Taxes and Corporate FinanceJohn R. GrahamFuqua School of Business, Duke University, Durham NC 27708-0120phone: (919) @: Capital Structure, Corporate Finance, Compensation, Dividends, Payout Policy, TaxesJEL: G30, G32, G34, G35, H25May 29, 2003This article will be published in “Taxes and Corporate Finance,” in B. E. Eckbo (ed.), Handbook of CorporateFinance – Empirical Corporate Finance, Elsevier Science, Amsterdam (2004).A more focused version of this article is available at The more focusedversion will be published in the Review of Financial Studies 16:4 (2003).I thank Roseanne Altshuler, Alan Auerbach, Alon Brav, Merle Erickson, Ben Esty, Mary Margaret Frank,Michelle Hanlon, Cam Harvey, Steve Huddart, Ravi Jagannathan, Mark Leary, Jennifer Koski, Alan Kraus,Ed Maydew, Bob McDonald, Roni Michaely, Lil Mills, Kaye Newberry, Jeff Pittman, Michael Roberts, DougShackelford, and Terry Shevlin for helpful comments. Maureen O’Hara (the editor) and an anonymousreferee made numerous helpful suggestions that improved the structure and content of the paper. I also thankTao Lin, Rujing Meng, and especially Vinny Eng and Krishna Narasimhan for excellent research apologize to those who feel that their research has been ignored or misrepresented. Any errors are mine. Thisresearch is partially funded by the Alfred P. Sloan Research
Abstract:This paper reviews tax research related to domestic and multinational capital structure, debt maturity, payoutpolicy, compensation policy, risk management, earnings management, leasing, pensions, R&D partnerships,tax shelters, transfer pricing, and organizational form. For each topic, the theoretical arguments explaininghow taxes can affect corporate decision-making and firm value are reviewed, followed by a summary of therelated empirical evidence and a discussion of unresolved issues. Tax research generally supports thehypothesis that high-tax rate firms pursue policies that provide tax benefits. Many issues remain unresolved,however, including understanding whether tax effects are of first-order importance, why firms do not pursuetax benefits more aggressively, and whether corporate actions are affected by investor-level
Modigliani and Miller (1958, 1961) demonstrate that corporate financial decisions are irrelevant ina perfect, frictionless world. To derive this result, Modigliani and Miller (MM) assume that capital marketsare perfect, which implies that there are no corporate or personal taxes, among other things. During the past45 years, research has focused on whether financial decisions become relevant if capital markets are notperfect. The research reviewed in this chapter investigates the consequences of allowing corporate andpersonal taxation, highlighting the role that corporate and investor taxes play in affecting corporate policies1and firm value. This role is potentially very important, given the sizable tax rates that many corporations andindividuals face (see Fig. 1). Modigliani and Miller argue that corporate financial policies do not add value in equilibrium, andtherefore firm value equals the present value of operating cash flows. Once imperfections are introduced,however, corporate financial policies can affect firm value, and firms should pursue a given policy until themarginal benefit of doing so equals the marginal cost. A common theme in tax research involves expressinghow various tax rules and regulations affect the marginal benefit of corporate actions. For example, when taxrules allow interest deductibility, a $1 interest deduction provides tax savings of $1xτ(.). τ(.) measuresCCcorporate marginal tax benefits and is a function of statutory tax rates, nondebt tax shields, the probabilityof experiencing a loss, international tax rules about dividend imputation and interest allocation, organizationalform, and various other tax rules. A common theme that runs throughout this paper is the demonstration ofhow various tax rules affect the τ(.) benefit function, and therefore how they affect corporate incentives andCdecisions. A second but less common theme in tax research is related to how market imperfections affectcosts. Given that this chapter reviews tax research, I emphasize research that describes how taxes affect costsand benefits – and only briefly discuss the influence of nontax are multiple avenues for taxes to affect corporate decisions. As outlined in Table 1, taxes canaffect capital structure decisions (both domestic (Section 1) and multinational (Section 2)), organizationalform and restructurings (Section 3), payout policy (Section 4), compensation policy (Section 5),risk1The interested reader can find excellent reviews of how taxes affect household investment decisions(Poterba (2001)) and the current state of tax research from the perspective of accountants (Shackelford and Shevlin(2001)) and public economists (Auerbach (2002)). Articles reviewing how nontax factors such as agency andinformational imperfections affect corporate financial decisions can be found in the other chapters of this
management (Section 6), and the use of tax shelters (Section 7). For each of these areas, the sections thatfollow provide a theoretical framework describing how taxes might affect corporate decisions, empiricalpredictions based on the theory, and summaries of the related empirical evidence. This approach is intendedto highlight important questions about how taxes affect corporate decisions, and to summarize and, in somecases, critique the answers that have been thus far provided. Each section concludes with a discussion ofunanswered questions and possible avenues for future research. Overall, substantial progress has been madeinvestigating if and how taxes affect corporate financial decisions – but much work remains to be 8 concludes and proposes directions for future . Taxes and Capital Structure – . Tax Theory and Empirical PredictionsThis section reviews capital structure research related to the “classical” tax system found in the United States.(Section 2 reviews multinational and imputation tax systems.) The key features of the classical system arethat corporate income is taxed at a rate τ, interest is deductible and so is paid out of income before taxes, andCequity payout is not deductible but is paid from the residual remaining after corporate taxation. In this taxsystem, interest, dividends, and capital gains income are taxed upon receipt by investors (at tax rates τ,Pτ=τ, and τ, respectively). Most of the research assumes that equity is the marginal source of funds anddivPG2that dividends are paid according to a fixed payout policy. To narrow the discussion, I assume thatregulations or transactions costs prevent investors from following the tax-avoidance schemes implied byMiller and Scholes (1978), in which investors borrow via insurance or other tax-free vehicles to avoidpersonal tax on interest or dividend this framework, the after-personal-tax value to investors of a corporation paying $1 of interest is$1(1-τ). In contrast, if that capital were instead returned as equity income, it would be subject to taxation atPboth the corporate and personal level, and the investor would receive $1(1-τ)(1-τ). The equity tax rate, τ,CEE2This assumption implies that retained earnings are not “trapped equity” that is implicitly taxed at thedividend tax rate, even while still retained. See Auerbach (2002) for more on the trapped equity or “new”
3is often modeled as a blended dividend and capital gains tax rate. The net tax advantage of $1 of debtpayout, relative to $1 of equity payout, is (1)If expression (1) is positive, debt interest is the tax-favored way to return capital to investors, once bothcorporate and individual taxation are considered. In this case, to maximize firm value, a company has a taxincentive to issue debt instead of . (1) captures the benefit of a firm paying out $1 as debt interest in the current period, relative topaying out $1 as equity income. If a firm has $D of debt with coupon rate r, the net benefit of using debtDrather than equity is (2)Given this expression, the value of a firm with debt can be written as(3)where the PV term measures the present value of all current and future interest deductions. Note that eq. (3)4implicitly assumes that using debt adds tax benefits but has no other effect on incentives, operations or (1958) is the seminal capital structure paper. If capital markets are perfect, τ, τ and τ all equalCPEzero, and it does not matter whether the firm finances with debt or equity (., Value=Value). Thatwith debtno debtis, the value of the firm equals the value of equity plus the value of debt – but total value is not affected bythe proportions of debt and equity. I use this implication as the null throughout the capital mid-2003 Congress passed a law that reduced the tax rate on both dividends and capital gains to 15%for individual investors, thereby simplifying and greatly reducing the level of equity taxation relative to are other approaches to modeling the tax benefits of debt that do not fit directly into this generalframework. For example, Goldstein, Ju, and Leland (2001) develop a dynamic contingent-claims model in whichfirms can restructure debt. They estimate that the tax benefits of debt should equal between eight and nine percent offirm value. See Goldstein et al. for references to other contingent-claims
Null hypotheses:Firms do not have optimal tax-driven capital value of a firm with debt is equal to the value of an identical firm without debt(., there is no net tax advantage to debt).In their “correction article,” MM (1963) consider corporate income taxation but continue to assumethat τ and τ equal zero. In this case, the second term in eq. (3) collapses to PV[τ rD]: Because interest isPECDdeductible, paying $rD of interest saves τrD in taxes each period relative to returning capital as (1963) assume that interest deductions are as risky as the debt that generates them and should be5discounted by r. With perpetual debt, MM (1963) argue that the value of a firm with debt financing is D(4)where the τD term represents the tax advantage of debt. Note that eq. (4) contains a term that captures theCtax benefit of using debt (τD) but no offsetting cost of debt term. Eq. (4) has two strong implications. First,Ccorporations should finance with 100% debt because the marginal benefit of debt is τ, which is oftenCassumed to be a positive constant. Second, if τ is constant, firm value increases (linearly) with D due to taxCbenefitsThe first implication was recognized as extreme, so researchers developed models that relax the MM(1958) assumptions and consider costs of debt. In the early models, firms trade-off the tax benefits of debtwith costs. The first cost proposed in the literature was the cost of bankruptcy, or more generally, costs offinancial distress. Kraus and Litzenberger (1973) show in a state-preference framework that firms shouldtrade-off bankruptcy costs with the tax benefits of debt to arrive at an optimal capital structure that involves5The assumption that debt should be discounted at r is controversial because it requires the amount of debtDto remain fixed. Miles and Ezzell (1985) demonstrate that if the dollar amount of debt is not fixed but instead is setto maintain a target debt-equity ratio, then interest deductions have equity risk and should be discounted with thereturn on assets, r, rather than r. (Miles and Ezzell (1985) allow first period financing to be fixed, which requiresADadjusting the discount rate by (1+r)/(1+r)). In contrast, Grinblatt and Titman (2002) argue that firms often payADdown debt when things are going well and stock returns are high, and do not alter debt when returns are low. Suchbehavior can produce a low or negative beta for debt and hence a low discount rate for the tax benefits of debt. Ineither the Miles and Ezzell or Grinblatt and Titman case, however, the value of a levered firm still equals the valueof the unlevered firm plus a “coefficient times debt” term – the discounting controversy only affects the
less than 100% debt. Scott (1976) shows the same with continuous variables. The bankruptcy cost solution6does not appear empirically to ex ante offset the benefits of debt. Therefore other papers have proposed non-bankruptcy costs that could be traded off against the tax benefits of debt. For example, Jensen and Meckling7(1976) introduce agency costs of equity and leverage-related deadweight costs. Myers (1977) introducesunderinvestment costs that can result from too much debt. Regardless of the type of cost, the basic trade-off implications remain similar to those in MM (1963):1) the incentive to finance with debt increases with the corporate tax rate, and 2) firm value increases withthe use of debt (up to the point where the marginal cost equals the marginal benefit of debt). Note also thatin these models, different firms can have different optimal debt ratios depending on the relative costs andbenefits of debt (., depending on differing firm characteristics).Prediction 1: All else constant, for taxable firms, value increases with the use of debt because of tax benefits(up to the point where the marginal cost equals the marginal benefit of debt).Prediction 2: Corporations have a tax incentive to finance with debt that increases with the corporate marginaltax rate. All else equal, this implies that firms have differing optimal debt ratios if their tax rates 1 is based directly on eq. (4), while Prediction 2 is based on the first derivative of eq. (4) withrespect to (1977) argues that personal taxes can eliminate the “100% debt” implication, without the needfor bankruptcy or agency costs. (Farrar and Selwyn (1967) took first steps in this direction.) Miller’sargument is that the marginal costs of debt and equity, net of the effects of personal and corporate taxes,should be equal in equilibrium, so firms are indifferent between the two financing sources. In essence, thecorporate tax savings from debt is offset by the personal tax disadvantage to investors from holding debt,6Warner (1977) shows that direct costs of bankruptcy average no more than % ex post in railroadbankruptcies. More recently, Andrade and Kaplan (1998) show that the ex post costs of distress brought about byfinancing choice amount to 20% of firm value for a group of industrial firms. Miller (1977) notes that firms chooseoptimal debt policy by considering ex ante costs of distress, indicating that the costs mentioned above need to bemultiplied by the conditional probability of distress to measure ex ante costs. Miller points out that ex ante costs offinancial distress appear to be very small compared to the apparently large tax benefits of and Weisbach (1999) use simulations to conclude that the agency costs of debt are too small tooffset the tax benefits, and Esty (1998) empirically examines the effects of agency costs on capital structure in thebanking
relative to holding equity. All else equal (including risk), this personal tax disadvantage causes investors todemand higher pretax returns on debt, relative to equity returns. From the firm’s perspective, paying thishigher pretax return wipes out the tax advantage of using debt . 2 illustrates Miller’s point. The horizontal line in Panel A depicts the supply curve for debt; theline is horizontal because Miller assumes that the benefit of debt for all firms equals a fixed constant τ. TheCdemand for debt curve is initially horizontal at zero, representing demand by tax-free investors, but eventuallyslopes upward because the return on debt must increase to attract investors with higher personal income taxrates. By making the simplifying assumption that τ=0, Miller’s equilibrium is reached when the marginalE*investor with τ =τ is attracted to purchase debt. In this equilibrium, the entire surplus (the area between thePC*supply and demand curves) accrues to investors subject to personal tax rates less than τ.PThere are several implications from Miller’s (1977) analysis. The first two are new:Prediction 3: High personal taxes on interest income (relative to personal taxes on equity income) create adisincentive for firms to use 4: The aggregate supply of debt is affected by relative corporate and personal other implications are consistent with the null hypotheses stated above: 1) there is no net tax advantageto debt at the corporate level (once one accounts for the higher debt yields investors demand because of therelatively high personal taxes associated with receiving interest), 2) though taxes affect the aggregate supplyof debt in equilibrium, they do not affect the optimal capital structure for any particular firm (., it does notmatter which particular firms issue debt, as long as aggregate supply equals aggregate demand), and 3) usingdebt does not increase firm general version of Miller’s argument (that does not assume τ=0) can be expressed in terms of (3). Once personal taxes are introduced into this framework, the appropriate discount rate is measured after-personal income taxes to capture the (after-personal-tax) opportunity cost of investing in debt. In this case,8the value of a firm using perpetual debt is:8See Sick (1990), Taggart (1991), or Benninga and Sarig (1997) for derivation of expressions like eq. (5)under various discounting assumptions. These expressions are of the form V = V + coefficient*D., withwith debtno debtthe coefficient an increasing (decreasing) function of corporate (personal income) tax
(5)If the investor-level tax on interest income (τ) is large relative to tax rates on corporate and equity incomeP(τ and τ), the net tax advantage of debt can be zero or even negative. Note that eq. (5) is identical to eq. (4)CEif there are no personal taxes, or if τ =τ.PEOne way that eq. (5) can be an equilibrium expression is for the right-most term in eq. (5) to equalzero in equilibrium (., (1-τ)=(1-τ)(1-τ) ), in which case the implications from Miller (1977) arePCEunchanged. Alternatively, the tax benefit term in eq. (5) can be positive and a separate cost term can beintroduced in the spirit of the trade-off models; in this case, the corporate incentive to issue debt and firmvalue both increase with [1-(1-τ)(1-τ)/(1-τ)] and firm-specific optimal debt ratios can exist. The bracketedCEPexpression specifies the degree to which personal taxes (Prediction 3) offset the corporate incentive to usedebt (Prediction 2). Recall that τ and τ are personal tax rates for the marginal investor(s), and therefore arePEdifficult to pin down empirically (more on this in Section ).DeAngelo and Masulis (1980; hereafter DM) broaden Miller’s (1977) model and put the focus onthe marginal tax benefit of debt, represented above by τ. DM argue that τ(.) is not constant and always equalCCto the statutory rate. Instead, τ(.) is a function that decreases in nondebt tax shields (., depreciation andCinvestment tax credits) because nondebt tax shields (NDTS) crowd out the tax benefit of interest. Further,Kim (1989) highlights that firms do not always benefit fully from incremental interest deductions becausethey are not taxed when taxable income is negative. This implies that τ(.) is a decreasing function of a firm’sCdebt usage because existing interest deductions crowd out the tax benefit of incremental τ(.) as a function has important implications because the supply of debt function canCbecome downward sloping (see Panel B in Fig. 2). This implies that there is a corporate advantage to usingdebt, as measured by the “firm surplus” of issuing debt (the area above the dotted line but below the supplycurve in Panel B). Moreover, high-tax-rate firms supply debt (., are on the portion of the supply curve to9
the left of its intersection with demand), which implies that there can exist tax-driven firm-specific optimaldebt ratios (as in Prediction 2), and that the tax benefits of debt add value for high-tax-rate firms (as inPrediction 1). The DeAngelo and Masulis (1980) approach leads to the following prediction, whichessentially expands Prediction 2:Prediction 2': All else equal, to the extent that they reduce τ(.), nondebt tax shields and/or interest deductionsCfrom already-existing debt reduce the tax incentive to use debt. Similarly, the tax incentive to use debtdecreases with the probability that a firm will experience nontaxable states of the Empirical Evidence on Whether the Tax Advantage of Debt Increases Firm ValuePrediction 1 indicates that the tax benefits of debt add τD (eq. (4)) or [1-(1-τ)(1-τ)/(1-τ)]D (eq. (5)) to firmCCEPvalue. If τ=40% and the debt ratio is 35%, eq. (4) indicates that the contribution of taxes to firm value equalsC14% (=τ x debt-to-value). This calculation is an upper bound, however, because it ignores costs andCother factors that reduce the corporate tax benefit of interest deductibility, such as personal taxes, nontax costsof debt, and the possibility that interest deductions are not fully valued in every state of the world. Thissection reviews empirical research that attempts to quantify the net tax benefits of debt. The first group ofpapers study market reactions to exchange offers, which should net out the various costs and benefits of remainder of the section reviews recent analyses based on large-sample regressions and concludes byexamining explicit benefit functions for interest Exchange OffersTo investigate whether the tax benefits of debt increase firm value (Prediction 1), Masulis (1980) examinesexchange offers made during the 1960s and 1970s. Because one security is issued and another simultaneouslyretired in an exchange offer, Masulis argues that exchanges hold investment policy relatively constant andare primarily changes in capital structure. Masulis’ tax hypothesis is that leverage-increasing (-decreasing)exchange offers increase (decrease) firm value because they increase (decrease) tax deductions. Note thatMasulis implicitly assumes that firms are underlevered. For a company already at its optimum, a movementin either direction (., increasing or decreasing debt) would decrease firm (1980) finds evidence consistent with his predictions: leverage-increasing exchange offers10
increase equity value by %, and leverage-decreasing transactions decrease value by %. Moreover, theexchange offers with the largest increases in tax deductions (debt-for-common and debt-for-preferred) havethe largest positive stock price reactions (% and %, respectively). Using a similar sample, Masulis(1983) regresses stock returns on the change in debt in exchange offers and finds a debt coefficient ofapproximately (which is statistically indistinguishable from the top statutory corporate tax rate at thatera). This is consistent with taxes increasing firm value as in eq. (4) (and also consistent with some alternativehypotheses discussed below) but it is surprising because such a large coefficient implies near-zero personaltax and non-tax costs to debt. That is, the debt coefficient in Masulis (1983) measures the average benefit ofdebt (averaged across firms and averaged over the incremental net benefit of each dollar of debt for a givenfirm) net of the costs. An average net benefit of requires that the costs are much smaller than the benefitsfor most dollars of debt. For the post-exchange offer capital structure to satisfy the MB=MC equilibriumcondition, the benefit or cost curves (or both) must be very steeply sloped near their (1984) and Cornett and Travlos (1989) argue that Masulis’ (1980) hypothesis is firms optimize, they should only adjust capital structure to move towards an optimal debt ratio, whetherthat involves increasing debt or equity. In other words, increasing debt will not always add to firm value, evenif interest reduces tax liabilities. Graham, Hughson, and Zender (1999) point out that if a firm starts at itsoptimal capital structure, it will only perform an exchange offer if something moves the firm out ofequilibrium. They derive conditions under which stock-price-maximizing exchanges are unrelated to marginaltax rates because market reactions aggregate tax and non-tax informational aspects of capital structurechanges. Therefore, nontax reactions might explain Masulis’ (1980) results. As described next, several papershave found evidence of non-tax factors affecting exchange offer market reactions. It is important to note thatthese post-Masulis papers do not prove that the tax interpretation is wrong – but they do offer , some papers find evidence of positive (negative) stock reactions to leverage-increasing(leverage-decreasing) events that are unrelated to tax deductions: Asquith and Mullins (1986), Masulis andKorwar (1986), and Mikkelson and Partch (1986) find negative stock price reactions to straight equityissuance, and Pinegar and Lease (1986) find positive stock price reactions to preferred-for-common11
exchanges. Second, Mikkelson and Partch (1986) and Eckbo (1986) find that straight debt issuance (withoutequity retirement) produces a stock price reaction that is indistinguishable from zero. Third, some papers findthat exchange offers convey non-tax information that affects security prices, perhaps due to asymmetricinformation problems along the lines suggested by Myers and Majluf (1984) or due to signaling (Ross (1977)and Leland and Pyle (1977)). For example, Shah (1994) correlates exchange offers with information aboutreduced future cash flows (for leverage-decreasing offers) and decreased risk (for leverage-increasing offers).Finally, Cornett and Travlos (1989) provide evidence that weakens Masulis’ (1983) conclusions. Cornett andTravlos regress event stock returns on the change in debt and two variables that control for information effects(the ex-post change in inside ownership and ex-post abnormal earnings). They find the coefficient on thechange in debt variable is insignificant while the coefficients on the other variables are significant, which9implies that the positive stock price reaction is related to positive information conveyed by the and Travlos conclude that equity-for-debt exchanges convey information about the future – but findno evidence of increased value due to tax recent papers examine the exchange of traditional preferred stock for monthly income preferredstock (MIPS). These two securities differ primarily in terms of their tax characteristics, so any market reactionshould have minimal non-tax explanations. MIPS interest is tax deductible for corporations (like debt interest)and preferred dividends are not. On the investor side, corporate investors enjoy a 70% dividends received10deduction (DRD) for preferred dividends but recipients of MIPS interest receive no parallel issuing MIPS to retire preferred, corporations gain the tax benefit of interest deductibility butexperience two costs: underwriting costs, and possibly an increased coupon due to the personal tax penalty(because investors are fully taxed on MIPS interest in contrast to corporate investors receiving the DRD onpreferred dividends). Engel, Erickson, and Maydew (1999) compare MIPS yields to preferred yields andconclude that the tax benefit of MIPS are approximately $ per dollar of face value, net of the9Cornett and Travlos do not report whether they get a significant positive tax coefficient (like Masulis(1983) did) when they exclude the information variables. Therefore, their results could be driven by their using adifferent sample than Masulis A 70% DRD means that a corporation that owns another firm’s stock only pays tax on 30% of thedividends received. Note that evidence in Erickson and Maydew (1998) implies that corporations are the marginalinvestor in preferred stock (see footnote 27).12
aforementioned costs. Irvine and Rosenfeld (2000) use abnormal announcement returns to estimate the valueat $. Given that MIPS and preferred are nearly identical in all legal and informational respects, thesestudies provide straight-forward evidence of the positive contribution of taxes to firm value, net ofunderwriting and personal tax Cross-Sectional RegressionsFama and French (1998; hereafter FF) attempt to estimate eq. (4) and Prediction 1 directly, by regressing VLon debt interest, dividends, and a proxy for V. They argue that a positive coefficient on interest is evidenceUof positive tax benefits of debt. FF measure V as the excess of market value over book assets. They proxyLV with a collection of control variables including current earnings, assets, R&D spending, as well as futureUchanges in these same variables. (All the variables in the regression are deflated by assets.) If these controlvariables adequately proxy for V, the regression coefficient on interest will measure the tax benefit of debtU(which is hypothesized to be positive). The main difficulty with this approach is that if the control variablesmeasure V with error, the regression coefficients can be biased. FF perform a series of regressions on a broadUcross-section of firms, using both level-form and first-difference specifications. In all cases, the coefficienton interest is either insignificant or negative. Fama and French interpret their results as being inconsistentwith debt tax benefits having a first-order effect on firm value. Instead, they argue that interest providesinformation about earnings that is not otherwise captured by their controls for V. In other words, V isUUmeasured with error, which results in the interest coefficient picking up a negative valuation effect relatedto financial distress or some other and Nissim (2002) attempt to circumvent this measurement problem. They perform a switchof variables, moving the earnings variable (which they assume proxies V with error) to the left-hand sideUof the regression and V to the right-side. Therefore, their regression tests the relation V =V - coeff* Kemsley and Nissim regress EBIT on V and debt, the debt coefficient is negative, which theyLinterpret as evidence that debt contributes to firm value. The coefficient also changes through time inconjunction with changes in statutory tax rates. In my opinion, the Kemsley and Nissim analysis should beinterpreted carefully. First, their regression specification can be interpreted as measuring the effect of debton earnings just as well as it can be interpreted as a switch-of-variables that fixes a measurement error13
problem in Fama and French (1998). Second, the debt coefficient has the correct sign for the full sample onlyin a nonlinear specification in which all the right-hand side variables are interacted with a crude measure ofthe discount rate. Finally, the coefficient that measures the net benefit of debt has an absolute value of consistent with Masulis (1983), such a large coefficient implies near-zero average debt costs and anear-zero effect of personal Marginal Benefit FunctionsUsing a different approach, Graham (2000, 2001) simulates interest deduction benefit functions and uses themto estimate the tax-reducing value of each incremental dollar of interest expense. For a given level of interestdeductions, Graham essentially integrates over possible states of the world (., both taxable and nontaxablestates) to determine a firm’s expected τ, which specifies the expected tax benefit of an incremental dollarCof interest deduction. Marginal tax benefits of debt decline as more debt is added because the probabilityincreases with each incremental dollar of interest that it will not be fully valued in every state of the simulation methods (described more fully in Section ) and various levels of interest deductions,Graham maps out firm-specific interest benefit functions analogous to the supply of debt curve in Panel Bof Fig. integrating under these benefit functions, Graham (2000) estimates that the tax benefit of debtequals approximately 9-10% of firm value during 1980-1994 (ignoring all costs). I update Graham’s estimatesand find that the tax benefit of debt is %, %, %, %, and % of firm value in 1995-1999,respectively (see Table 2). The fact that these figures are less than the 14% estimated (at the beginning ofSection 1) with the back of the envelope “τD” calculation reflects the reduced value of interest deductionsCin some states of the world. When personal taxes are considered, the tax benefit of debt falls to 7-8% of firmvalue during 1980-1994 (., this is Graham’s estimate of the “firm surplus” in Panel B of Fig. 2).Graham also estimates the “money left on the table” that firms could obtain if they levered up to thepoint where their last dollar of interest deduction is valued at the full statutory tax rate (., the “kink,” which14
11is the point just before incremental tax benefits begin to decline). I update the money left on the tablecalculations in Graham (2000, his Figure 2). If all firms lever up to operate at the kink in their benefitfunctions, they could add % to firm value over the 1995-1999 period (see Table 2). This number can beinterpreted either as a measure of the value loss due to conservative corporate debt policy, or as a lower boundfor the difficult-to-measure costs of debt that would occur if a company were to lever up to its kink. In theformer interpretation, these estimates imply that large tax benefits of debt appear to go unexploited, and thatlarge, profitable firms (which would seem to face the lowest costs of debt) are the most conservative in their12use of debt. In general, these implications are hard for a trade-off model to explain. Graham (2000),Lemmon and Zender (2001) and Minton and Wruck (2001) try to identify nontax costs that are large enoughin a trade-off sense that perhaps these firms are not in fact sum up, a fair amount of research has found evidence consistent with tax benefits adding to firmvalue. However, some of this evidence is ambiguous because non-tax explanations or econometric issuescloud interpretation. Additional research in three specific areas would be helpful. First, we need more market-based research along the lines of the MIPs exchanges, where tax effects are isolated from information andother factors and therefore the interpretation is fairly unambiguous. Second, additional cross-sectionalregression research that investigates the market value of the tax benefits of debt would be helpful in termsof clarifying or confirming the interpretation of existing cross-sectional regression analysis. Finally, if thetax benefits of debt do in fact add to firm value, an important unanswered question is why firms do not use13more debt, especially large, profitable firms. We need to better understand whether this implies that somefirms are not optimizing, or whether there are costs and other influences that have not been adequately11For example, if during 1995-1999 all firms levered up to just before the point of declining benefit,simulations performed for this paper indicate that the average company would have total tax benefits of debt ofaround 18% of firm value. That is, by levering up, the typical firm could add interest deductions with tax benefitequal to 10% of firm value, above and beyond their current level of tax (2002) argues that the prevalence of writing puts or purchasing calls on their own shares is alsoevidence that many firms pass up potential interest deductions. For example, writing a put (which involves implicitborrowing) can be replicated by explicitly borrowing today to purchase a share on the open market and repaying theloan in the future. The cash flows are identical in these two strategies but the latter results in the firm receiving a taxdeduction. The fact that many firms write puts is consistent with them passing up interest tax -Sunder and Myers (1999), Lemmon and Zender (2002), and related papers investigate whether thetrade-off model is the correct model of capital structure, which has implications towards interpreting these
modeled in previous Empirical Evidence on Whether Corporate Taxes Affect Debt vs. Equity PolicyTrade-off models imply that firms should issue debt as long as the marginal benefit of doing so (measuredby τ) is larger than the marginal cost. τ(.) is a decreasing function of nondebt tax shields, existing debt taxCCshields, and the probability of experiencing losses, so the incentive to use debt declines with these threefactors (Prediction 2'). In general, high-tax rate firms should use more debt than low-tax rate firms (Prediction2). The papers reviewed in this section generally use reduced-form cross-sectional or panel regressions to testthese predictions – and ignore personal taxes altogether. For expositional reasons, I start with tests ofPrediction 2'. Nondebt Tax Shields, Profitability, and the Use of DebtBradley, Jarrell and Kim (1984) perform one of the early regression tests for tax effects along the linessuggested by DeAngelo and Masulis (1980). Bradley et al. regress firm-specific debt-to-value ratios onnondebt tax shields (as measured by depreciation plus investment tax credits), R&D expense, the time-series14volatility of EBITDA, and industry dummies. The tax hypothesis is that nondebt tax shields are negativelyrelated to debt usage because they substitute for interest deductions (Prediction 2'). However, Bradley et that debt is positively related to nondebt tax shields, opposite the tax prediction. This surprising finding,and others like it, prompted Stewart Myers (1984) to state in his presidential address to the American FinanceAssociation, (p. 588) “I know of no study clearly demonstrating that a firm’s tax status has predictable,material effects on its debt policy. I think the wait for such a study will be protracted.”One problem with using nondebt tax shields, in the form of depreciation and investment tax credits,to explain debt policy is that nondebt tax shields are positively correlated with profitability and profitable (., high-tax rate) firms invest heavily and also borrow to fund this investment, this can inducea positive relation between debt and nondebt tax shields and overwhelm the tax substitution between interestand nondebt tax shields (Dammon and Senbet (1988)). Another issue is that nondebt tax shields (as well asexisting interest deductions or the probability of experiencing losses) should only affect debt decisions to the14An alternative test would be to match NDTS-intensive firms to companies that are similar in all waysexcept for their use of nondebt tax shields and examine whether the NDTS-intensive firms use less
extent that they affect a firm’s marginal tax rate. Only for modestly profitable firms is it likely that nondebt15tax shields have impact sufficient to affect the marginal tax rate and therefore debt -Mason (1990) and Dhaliwal, Trezevant, and Wang (1992) address these issues by interactingNDTS with a variable that identifies firms near “tax exhaustion,” at which point the substitution betweennondebt tax shields and interest is most important. Both papers find that tax-exhausted firms substitute away16from debt when nondebt tax shields are high. Even though these papers find a negative relation between theinteracted NDTS variable and debt usage, this solution is not ideal. For one thing, the definition of taxexhaustion is ad hoc. Moreover, Graham (1996a) shows that the interacted NDTS variable has low power todetect tax effects and that depreciation and investment tax credits (the usual components of nondebt taxshields) have a very small empirical effect on the marginal tax rate. Ideally, researchers should capture theeffects (if any) of nondebt tax shields, existing interest, and the probability of experiencing losses directlyin the estimated marginal tax rate, rather than including these factors as stand-alone similar issue exists with respect to using profitability as a measure of tax status. Profitable firmsusually have high tax rates and therefore some papers argue that the tax hypothesis implies they should usemore debt. Empirically, however, the use of debt declines with profitability, which is often interpreted asevidence against the tax hypothesis (., Myers (1993)). My view is that profitability should only affect the17tax incentive to use debt to the extent that it affects the corporate marginal tax rate; therefore, when testingfor tax effects, the effects (if any) of profitability should be captured directly in the estimated might also include a stand-alone profitability variable to control for potential nontax Directly Estimating the Marginal Tax RateOne of the problems that led to Myers’ capital structure puzzle is related to properly quantifying corporatetax rates and incentives. For example, many studies use static MTRs that ignore important dynamic features15The marginal tax rate for unprofitable firms will be close to zero whether or not the firm has NDTS. Thetax rate for highly profitable firms will be near the top statutory rate, unless a firm has a very large amount of (1995) finds the same for Swedish firms. Trezevant (1992) finds that Compustat PST firms mostlikely to be tax-exhausted decreased debt usage the most following the 1981 liberalization of tax laws that increasednondebt tax in mind that a marginal tax rate is bound between zero and the top statutory rate while profitabilityis not bounded, which can introduce difficulties into interpreting profitability as a proxy for the tax
of the tax code related to net operating losses carryback and carryforwards, investment tax credits and othernondebt tax shields, and the alternative minimum tax. Static MTRs miss the fact that a company might beprofitable today but expect to experience losses in the near future. This firm might erroneously be assigned18a high current-period tax rate even though its true economic tax rate is low. Conversely, an unprofitable firmmight have a large current economic marginal tax rate if it is expected to soon become and remain profitable(because extra income earned today increases taxes paid in the future: today’s extra income reduces lossesthat could be carried forward to delay future tax payments, thereby increasing present value tax liabilities).Shevlin (1987, 1990) uses simulation techniques to capture the dynamic features of the tax code19related to net operating loss carrybacks and carryforwards. The first step in simulating an MTR for a givenfirm-year involves calculating the historic mean and variance of the change in taxable income for each second step uses this historic information to forecast future income for each firm. These forecasts canbe generated with random draws from a normal distribution, with mean and variance equal to that gatheredin the first step; therefore, many different forecasts of the future can be generated for each firm. The third stepcalculates the present value tax liability along each of the income paths generated in the second step,accounting for the tax-loss carryback and carryforward features of the tax code. The fourth step adds $1 tocurrent year income and recalculates the present value tax liability along each path. The incremental taxliability calculated in the fourth step, minue that calculated in the third step, is the present value tax liabilityfrom earning an extra dollar today; in other words, the economic MTR. A separate marginal tax rate iscalculated along each of the forecasted income paths to capture the different tax situations a firm mightexperience in different future scenarios. The idea is to mimic the different planning scenarios that a managermight consider. The fifth step averages across the MTRs from the different scenarios to calculate the expectedeconomic marginal tax rate for a given firm-year. Note that these five steps produce the expected marginaltax rate for a single firm-year. The steps are replicated for each firm for each year, to produce a panel of firm-18Scholes and Wolfson (1992) define the economic marginal tax rate as the present value of current andfuture taxes owed on an extra dollar of income earned today, which accounts for the probability that taxes paid todaywill be refunded in the near future. 19 Auerbach and Poterba (1987) and Altshuler and Auerbach (1990) simulate tax rates using first-orderMarkov probabilities that weight the probability of transition between taxable and nontaxable
year MTRs. The marginal tax rates in this panel vary across firms and can also vary through time for a givenfirm. The end result is greater cross-sectional variation in corporate tax rates (and hence tax incentives) thanimplied by statutory difficulty with simulated tax rates is that they require a time-series of firm-specific , they are usually calculated using financial statement data, even though it would be preferable touse tax return data. With respect to the first problem, Graham (1996b) shows that an easy-to-calculatetrichotomous variable (equal to the top statutory rate if a firm has neither negative taxable income nor NOLcarryforwards, equal to one-half the statutory rate if it has one but not the other, and equal to zero if it hasboth), is a reasonable replacement for the simulated rate. With respect to the tax return issue, Plesko (2000)compares financial-statement-based simulated rates for 586 firms to a static tax variable calculated usingactual tax return data. He finds that simulated rates (based on financial statements) are highly correlated withtax variables based on tax return data. Plesko’s evidence implies that the simulated tax rates are a robustmeasure of corporate tax that by construction the simulated tax rates capture the influence of profitability on the corporatemarginal tax rate. Graham (1996a) extends the simulation approach to directly capture the effects of nondebttax shields, investment tax credits, and the alternative minimum tax. Graham (1996b) demonstrates thatsimulated tax rates are the best commonly available proxy for the “true” marginal tax rate (when “true” isdefined as the economic tax rate based on realized taxable income, rather than simulations of the future).Using the simulated corporate marginal tax rates, Graham (1996a) documents a positive relation between taxrates and changes in debt ratios (consistent with Prediction 2), as do Graham, Lemmon, and Schallheim(1998) and Graham (1999) for debt levels. Since that time, numerous other studies have also used simulatedtax rates to document tax effects in debt decisions. These results help to resolve Myers’ (1984) capitalstructure puzzle; when tax rates are properly measured, it is possible to link tax status with corporate Endogeneity of Corporate Tax StatusEven if measured with a very precise technique, tax rates are endogenous to debt policy, which can haveimportant effects on tax research. If a company issues debt, it reduces taxable income, which in turn can19
reduce its tax rate. The more debt issued, the greater the reduction in the marginal tax rate. Therefore, if oneregresses debt ratios on marginal tax rates, the endogeneity of corporate tax status can impose a negative biason the tax coefficient. This could explain the negative tax coefficient detected in some specifications (.,Hovakimian, Opler, and Titman (2001) and Barclay and Smith (1995b)). Note that endogeneity can affectall sorts of tax variables, including those based on NOLs or that use an average tax rate (., taxespaid/taxable income).There are two solutions to the endogeneity problem. MacKie-Mason (1990) proposed the firstsolution by looking at (0,1) debt versus equity issuance decisions (rather than the debt level) in his influentialexamination of 1,747 issuances from 1977-1987. Debt levels (such as debt ratios) are the culmination of manyhistorical decisions, which may obscure whether taxes influence current-period financing choice. Detectingtax effects in the incremental approach only requires that a firm make the appropriate debt-equity choice atthe time of security issuance, given its current position, and not necessarily that the firm rebalance to itsoptimal debt-equity ratio with each issuance (as is implicit in many debt level studies). To avoid theendogenous effect of debt decisions on the marginal tax rate, MacKie-Mason uses the lagged marginal tax20rate to explain current-period financing choice. He finds a positive relation between debt issuance and taxrates. Graham (1996a) follows a similar approach and examines the relation between changes in the debt ratio21and lagged simulated MTRs. He finds positive tax effects for a large sample of Compustat taxes exert a positive influence on each incremental financing decision, then the sum of theseincremental decisions should show up in an analysis of current debt levels – if one could fix the endogenous20Wang (2000) argues that firms do not consider the level of the marginal tax rate when making incrementaldecisions but rather consider how far the marginal tax rate is from the “optimal MTR.” Holding the level of the taxrate constant, Wang shows that companies with tax rates above the optimum are those that use the most debt (anaction which should endogenously reduce the marginal tax rate and move it closer to the optimum, essentiallyreducing MB until it equals MC). The difficulty with this approach is that Wang’s “optimal MTR” is ad hoc andbased on the probability of bankruptcy (as measured by Altman’s Z-score).21A number of other papers corroborate these results. For example, Shum (1996) finds similar evidence forCanadian firms. Alworth and Arachi (2000) show that lagged after-financing simulated tax rates are positivelyrelated to changes in debt for Italian firms. Henderson (2001) finds that changes in total liabilities and changes inlong-term debt are both positively related to simulated tax rates in a sample of . banks. Schulman et al. (1996)find that debt levels are positively correlated to tax rates in Canada and New
22negative effect on tax rates induced by cumulative debt usage. The second approach to fixing theendogeneity problem is to measure tax rates “but for” financing decisions. Graham, Lemmon, and Schallheim(1998) measure tax rates before financing (., based on income before interest is deducted). They find apositive relation between debt-to-value and (endogeneity-corrected) but-for tax rates. (They also find a“spurious” negative correlation in an experiment that uses an endogenously affected after-financing tax rate.)Examining changes in debt answers the question “are incremental decisions affected by tax status?”An alternative approach is to ask “if tax rates exogenously change, how will a firm alter debt usage?” TheTax Reform Act of 1986 greatly reduced corporate marginal tax rates (see Fig. 1), which in isolation impliesa reduction in the corporate use of debt. Givoly, Hahn, Ofer, and Sarig (1992) find that firms with high taxrates prior to tax reform (firms that therefore probably experienced the largest drop in their tax rate) reducedebt the most after tax reform. This finding is somewhat surprising because their corporate marginal tax ratesuffers from the negative endogeneity bias described above. Moreover, personal taxes are not modeled23directly, even though they fell by more than corporate tax rates after the 1986 tax reform. In a paper thatexamines international evidence during the same time period, Rajan and Zingales (1995) provide weakinternational evidence that taxes affect debt Time-Series and Small Firm Evidence of Tax EffectsThe empirical evidence described thus far confirms cross-sectionally that firms with high tax rates use moredebt than those with low tax rates. Presumably, there should also be time-series tax effects. For example, ifa firm starts public life with a low tax rate, one would expect increased debt usage if the tax rate increasesas the firm matures. I am not aware of any study that documents tax-related time-series effects in debt example, Graham (1999) uses panel data to document that cross-sectional variation in tax status affects22Dittmar (2002) studies corporate spin-offs, which potentially allows her to avoid the endogeneity problemby observing capital structures that are not the end result of a long history of accumulated debt policy , it is still the case that past decisions can influence the parent’s and/or spun-off unit’s new capital does not find evidence that corporate tax rates affect spin-off debt et al. (1992) include lagged dividend yield in their specification to control for personal tax effects,which might allow their tax variable to isolate corporate tax effects. Personal tax effects are examined more fully inSection
debt usage but he finds no evidence that time-series variation studying capital structure decisions among newly formed firms, one might be able to avoid long-lasting effects of past financing decisions. For example, Baker and Wurgler (2001) show that today’s market-to-book ratio and debt-equity issuance decisions continue to affect firm’s debt ratios for ten or more , Qureshi, and Olson (2000) describe various start-up financing issues including selecting a target debtratio, as well as how market conditions and collateralization affect the sequence of initial financing and Klassen (2001) examine capital structure in the years following IPO. They performannual (., years since IPO) cross-sectional regressions and find evidence that taxes have a positive effecton the use of debt in the early years of a firm’s public life – but this relation wanes as the firm ages. Pittmanand Klassen attribute this waning to an increase in refinancing transactions costs as firms age. Note that theirevidence is not time-series in terms of firms altering capital structure as tax rates change through time, thoughthey do link debt policy to firm age. Pittman and Klassen also find that firms use relatively more NDTS asthey all capital structure papers study Compustat companies. Ayers, Cloyd, and Robinson (2001)instead examine small companies with less than 500 employees that participated in the 1993 Federal ReserveNational Survey of Small Business Finances. 2,600 firms meet the Ayers et al. data requirements. The authorsregress interest expense divided by pre-interest pre-NDTS income on various variables including tax expensedivided by pre-interest income. They find a positive coefficient on the tax variable in both their outside andinside debt regressions (., interest owed to non-owners and owners, respectively). It is difficult to comparetheir results to Compusat-based research because Ayers et al. use a different dependent variable than moststudies, and they delete firms with a negative value for the dependent variable (which raises statistical issues).To summarize Section , once issues related to measuring debt policy and tax rates are addressed,researchers have supplied evidence in response to Myers’ (1984) challenge to show that corporate debt usageis positively affected by tax rates. These results are consistent with survey evidence that interest taxdeductibility is an important factor affecting debt policy decisions (ranking below only maintaining financialflexibility, credit ratings, and earnings volatility), and is especially important for large industrial firms(Graham and Harvey (2001)). Notwithstanding these empirical results, Myers is still not entirely convinced22
(Myers et al. (1998)); he argues that tax incentives are of “third order” importance in the hierarchy ofcorporate decisions. It would be helpful for future research to investigate whether the tax effects on debtversus equity choice are economically important, and if they are not, determine why not. Several other challenges remain. First, none of the papers cited above provide time-series evidencethat firm-specific changes in tax status affect debt policy. It would be quite helpful to examine whether a firmchanges its debt policy as it matures and presumably its tax status changes. Second, Fama and French (2001)point out that with few exceptions the panel data examinations do not use statistical techniques that accountfor cross-correlation in residuals, and therefore many papers do not allow for proper determination ofstatistical significance for the tax coefficients. Therefore, it is not clear if all of the tax effects documentedabove are robustly significant. Finally, most papers ignore the tax cost of receiving interest income from theinvestor’s perspective, an issue to which I now Empirical Evidence on Whether Personal Taxes Affect Corporate Debt vs. Equity PolicyMiller (1977) identifies a puzzle: the benefits of debt seem large relative to expected costs, and yet manyfirms appear to use debt conservatively. Miller proposes that the personal tax cost of interest income (relativeto the personal tax cost of equity) is large enough at the margin to completely offset the corporate taxadvantage of debt. The Miller Equilibrium is difficult to test empirically for several reasons, not the least ofwhich is that the identity and tax-status of the marginal investor(s) between debt and equity are , we can note that the tax rate on interest income (τ) was large relative to tax rates on corporatePand equity income (τ and τ) when Miller wrote his paper, so the Miller Equilibrium was , the statutory tax rates shown in imply that expression (1) has been positive since 1981, so24the strict form of the Miller Equilibrium is less plausible in the last two the corporate perspective, the relatively high investor-level taxation of interest leads to a“personal tax penalty” for debt: investors demand a higher risk-adjusted return on debt than on equity. By24If the statutory tax rates depicted in Fig. 1 are not representative of the tax rates applicable to the marginalinvestor(s), or if capital gains tax rates are effectively reduced through deferral and/or elimination at death, then theMiller Equilibrium is technically possible even in recent
rearranging Equation (1), the net tax advantage of debt can be represented as (6)where τ is the corporate income tax rate, τ is the personal tax rate on equity income, and τ is the personalCEPtax rate on interest income. The bracketed term in Eq. (6) accounts for the personal tax penalty: τ - (1-τ)τ.PCETo quantify the effect of personal taxes in expression (5), Gordon and MacKie-Mason (1990) andothers implicitly assume that investors form clienteles based on firm-specific dividend payout ratios, andtherefore that τ is a weighted combination of the tax rates on dividend payout and capital gains income:Eτ=(payout)τ + (1-payout)τ. This and related papers use historic averages to estimate dividend payoutEdivcap gainsand measure τ as equaling τ, where τ is implicitly estimated using the difference between the yield ondivPPtaxable and tax-free government bonds. τ is often assumed to equal a fraction of the statutory capitalcap gainsgains tax rate (to capture the benefit of reduced effective tax rates due to deferral of equity taxation and25omission of equity tax at death).Given these assumptions, Gordon and MacKie-Mason (1990) estimate that the tax advantage of debt,net of the personal tax penalty, increased following the Tax Reform Act of 1986. Recall that Miller (1977)implies that the aggregate supply of debt is determined by relative corporate and personal tax rates. Gordonand MacKie-Mason document that aggregate corporate debt ratios increased slightly in response to tax reform(consistent with Prediction 4). This is the only research of which I am aware that investigates this aggregateprediction. Note that Gordon and MacKie-Mason focus on a single point in time, while the Miller Equilibriumhas implications for any point in time. Also note that if the marginal investor is taxable at rates like those25Green and Hollifield (2003) simulate an economy to investigate the degree to which capital gains deferralreduces the effective tax rate on equity income (and therefore, from the company’s perspective, increases thepersonal tax penalty for debt relative to equity). Green and Hollifield find that the ability to defer taxation reducesthe implicit tax on capital gains by about 60%. If they were to factor in deferral at death and the lower tax rate oncapital gains relative to the rate on dividends and interest, it would reduce the implicit tax rate on capital gains evenfurther. (On the other hand, their calculations ignore the high turnover frequently observed for common stocks andmutual funds, which increases the effective tax rate on equity.) Overall, their evidence suggests that there is ameasurable personal tax disadvantage to debt but it does not appear large enough to offset the corporate tax benefitsof debt. However, Green and Hollifield find that when coupled with fairly small costs of bankruptcy (., realizedbankruptcy costs equal to 3% of pretax firm value), the personal tax penalty is sufficient to offset the corporate taxadvantage to debt at the margin and lead to interior optimal debt
reflected in Fig. 1, then the 2003 reduction in dividend and capital gains tax rates to τ=τ=15% shoulddivPreduce the aggregate amount of debt used in the . (1999) tests similar predictions using firm-specific data. He finds that the net tax advantage26of the first dollar of interest averaged between 140 and 650 basis points between 1980 and 1994. He findsthat the firms for which the net advantage is largest use the most debt in virtually every year. Graham alsoseparately identifies a positive (negative) relation between the corporate tax rate (personal tax penalty) anddebt usage. These results are consistent with Predictions 2 and (2001) assumes that a given firm’s debt and equity are held by a particular clientele ofinvestors (with the clienteles based on investor tax rates). He investigates the capital structure response to thelarge reduction in personal taxes (relative to the smaller reduction in corporate tax rates) after the Tax ReformAct of 1986. Campello finds that zero-dividend firms (which presumably have high-tax-rate investors andtherefore experienced the largest reduction in the personal tax penalty) increased debt ratios in response totax reform, while high-dividend payout firms (which presumably have low-tax-rate investors and thereforeexperienced a small reduction in the personal tax penalty) reduced debt usage relative to peer Market-Based Evidence on How Personal Taxes Affect Security ReturnsWhile consistent with personal taxes affecting corporate financing decisions in the manner suggested byPrediction 3, the papers cited above are not closely tied to market-based evidence about the tax characteristicsof the marginal investor between debt and equity. Instead, these papers assume that dividend clienteles exist,and also make assumptions about the personal tax characteristics of these clienteles based on a firm’s payoutpolicy. For example, these papers implicitly assume that there is a certain marginal investor who owns bothequity and debt and (to estimate τ) that this same investor sets prices between taxable and tax-free truth is that we know very little about the identity or tax-status of the marginal investor(s) between anytwo sets of securities, and deducing this information is update Graham’s (1999) annual tax regressions from his Table 5, Panel B. The tax variable is the taxadvantage of debt net of personal taxes, as expressed in Eq. (5), with the personal tax penalty based on firm-specificdividend payout ratios. The dependent variable is debt-to-value. The estimated tax coefficients for 1995-1999 , , , , , respectively, indicating that debt ratios are positively related to net tax the tax coefficients are significant at a 1% level, except in 1996 when the p-value is
For example, assume that munis yield 7%, Treasuries 10%, and equities 8% (and assume that thisequity return has been adjusted to make its risk equivalent to the risk of munis and Treasuries). In a Gordon/MacKie-Mason/Graham type of equilibrium, r = r(1- τ) = r(1- τ) = 7%, which implies thatmuniTreasuryPequityequityτ=30% and τ=%. This in turn implies that a large portion of equity returns are expected to come fromPequitycapital gains (because τ is so much lower than τ). However, things are rarely so simple. First, it isequityP27difficult to determine the risk-adjusted equity return. Second, if there are frictions or transactions costslimiting arbitrage between pairs of markets (or if risk adjustments are not perfect), one could observe, say,munis yielding 7%, Treasuries 10%, and equities 12%. In this case, it is not clear which pair of securitiesshould be used to deduce τ. If Treasuries and equities are used, the implicit τ could be negative. ForPPexample, assume that dividend payout is 15%, that τ=5%, and that τ is modeled as a weightedeffective cap gainsequityaverage between dividends and retained earnings: τ=(1-τ) + (1-τ), where τ=τ.equitydiveffective cap gainsdivPTo ensure that r(1- τ) = r(1- τ), in this example τ= -30%; clearly, market frictions drive relativeTreasuryPequityequityPreturns in this example, so the usual approach can not be used to deduce the personal tax characteristics ofthe marginal investor(s). Williams (2000) points out that when there are more than two assets, different pairs of assets can bearbitraged by different investors, so prices might reflect a mixture of tax characteristics. It is difficult to knowwhich assets are directly benchmarked to each other by the marginal investor(s) and which are “indirectlyarbitraged,” and it is even difficult to know whether capital gains or income tax rates are priced into would be helpful if future research could quantify the relative importance of personal taxes onsecurity prices, with an eye towards feedback into capital structure decisions. One area where there has beena fair amount of research along these lines involves determining the investor tax rate implicit betweenmunicipals and taxable government bonds. Poterba (1989) finds that the yield difference between high-gradeone-year munis and government bonds approximates the top statutory personal tax rate, implying that the27Gordon and MacKie-Mason (1990) Graham (1999) avoid the issue of adjusting the equity return. Instead,they assume that τ equals the τ implicit between munis and Treasuries, that τ= τ,divPeffective cap gainsstatutory capital gainsand weight these two pieces by the portion of earnings returned as dividends and retained, respectively, to deduceτ. It would be informative if future research could calibrate this approach to market-driven estimates of τ.equityequity26
marginal investor between these two securities is a highly taxed individual. However, even this experimentis not without difficulty. First, returns on long-term munis and taxables imply a tax rate for the marginalinvestor that is approximately half that implied by the short-term securities. Chalmers (1998) shows that thisholds even when the muni interest payments are prefunded by T-bonds held in “defeasement,” and thereforedifferences in risk between munis and T-bonds do not explain this conundrum. Green (1993) proposes thattaxable bonds might not be “fully taxable” because a portion of their return can come from capital gains(especially for long-term bonds) and also because to some degree the interest income can be offset byinvestment interest deductions. Mankiw and Poterba (1996) suggest that munis might be benchmarked toequities by one clientele of investors and taxable bonds might be benchmarked to equities by anotherclientele. In this case, munis and taxables might not be directly benchmarked to each other, which couldexplain the unusual implicit tax rate that is sometimes observed between the two an example of trying to link the effects of personal taxes to capital structure issues, consider the28implications from Engel et al. (1999) and Irvine and Rosenfeld (2000) about the personal tax that corporations are the marginal investors in preferred stock but not in the similarityof the securities, in equilibrium we expect their after-investor-tax returns to be equal, within transactions cost28Recall that these authors investigate MIPS for preferred exchanges. These two securities are similar inmost respects, except that MIPS interest is tax deductible for issuing corporations and preferred dividends are the investor side, corporate investors can take the 70% dividends received deduction (DRD) for preferreddividends but recipients of MIPS interest receive no parallel and Maydew (1998) provide evidence that corporations are the marginal investors in preferredstock, though they do not precisely identify the numeric value of the marginal investor’s tax rate. They study themarket reaction to the announced (but never implemented) change in the dividends received deduction (DRD). TheDRD allows corporations to deduct a portion of the dividends they receive from other corporations to attenuate“triple taxation” of equity income. Individual investors do not receive the DRD. When the Treasury made a surpriseannouncement in December, 1995 that they were planning to reduce the deduction from 70% to 50%, the typicalpreferred stock experienced a statistically significant -1% abnormal return, while there was no reaction amongcommon stocks. This implies that corporations are the marginal investors (., price-setters) in preferred stocks butnot in common stocks. One advantage of the Erickson and Maydew study is that they are able to control for riskwhen examining abnormal returns because they compare a security to itself before and after the exogenousannouncement. The authors are unable to precisely deduce the tax rate of the marginal (corporate) investor, however,because they can not pinpoint the probability assigned by the market that the Treasury would actually implement Erickson and Maydew (1998) find no evidence that corporations are the marginal investors incommon stocks, Geisler (1999) shows that common stock holdings by insurance companies vary positively with theallocation of the dividends received deduction among insurance companies. (The allocation of DRD can vary acrossinsurance companies for regulatory reasons.) Geisler’s evidence is consistent with clienteles: insurance companiesrespond to tax incentives to hold common stocks when their tax rate is low (., when their DRD allocation is high).27
bounds: r(1-τ) = r(1-τ). Plugging in r=% and r=% from Engel et al.’s TablepreferredDRDMIPSPpreferredMIPS4, and assuming that the marginal corporate investor is taxed at 35% so that τ=%, we can back outDRDthe personal tax rate associated with interest income: () = (1-τ) implies that τ=13%.PPIf I ignore the 30 basis point “yield premium” on MIPS imputed by Engel et al. and use r=%, τ=16%.MIPSPTo the extent that results based on MIPS interest carry over to debt interest, finding τ=16% for thePmarginal debt investor is intriguing. First note that the mean after-financing corporate tax rate in 1993-1999is approximately 18% (see Table 2), which is a rough estimate of the tax benefit of the last dollar of interestdeduction (ignoring all costs). If we make Miller’s (1977) assumptions that τ=0 and that all firms face theEsame 18% marginal benefit of debt, then τ should equal 18% (., MC should equal MB), quite close to thePτ=16% MIPS estimate. As argued by Green and Hollifield (2003), it would only take fairly small costs ofPbankruptcy to equalize the costs and benefits of debt, creating a environment conducive to an equilibriumwith internal optimal debt ratios.. However, τ is most likely not zero for the marginal investor in (Green and Hollifield argue that deferral reduces effective τ to about half its statutory level.) Another issueEis that the estimated MIPS costs and benefits are average, not marginal. Even if the marginal costs andbenefits are equal in an equilibrium like that depicted in Figure 2a, there is a firm surplus/benefit to usingdebt. Therefore, even if personal tax costs are large enough at the margin to equal marginal benefits, thereappear to be tax-driven preferred capital structures for some firms – presumably the incremental benefitwould be near $ per dollar for high-tax-rate firms, while the personal tax cost is only half that if the nontax costs of debt are large for these high-tax rate firms could a Miller-type equilibrium hold,in which the benefits of debt are zero for all firms in sum, the implicit personal tax costs estimated here suggest that at the margin the tax costs and taxbenefits might be of similar magnitude. However, they do not explain cross-sectionally why someinframarginal firms (with large tax benefits of interest) do not use more debt. More on this in Section . Oneother place where there has been a fair amount of success (though not unambiguously so) in deducingmarginal investor tax characteristics is related to ex-day dividend returns. I defer this discussion to Section3, where I explore how taxes affect corporate dividend
In the most general sense, any research that shows that personal tax rates affect security returns shedslight on Miller’s (1977) claims. Using the CAPM-with-taxes specification, Auerbach (1983) finds evidencethat tax-related preferences result in clienteles of investors that purchase stocks based on firm-specificdividend-price ratios. Constantinides (1983) and Dammon, Spatt, and Zhang (2001) investigate how favorablecapital gains taxation affects investment and consumption choices. Seida and Wempe (2000) show thatindividual investors accelerated recognizing capital gains (and delayed losses) in anticipation of the increasein capital gains tax rates associated with the 1986 tax act. See Poterba (2001) for a review of articles relatedto how personal taxation affects the timing and value of asset sales and capitalization: Another group of papers investigates tax capitalization. These papers argue that personaltaxes are capitalized into share prices via retained earnings. This in turn affects the relative tax advantage todebt because retained earnings are assumed to be the marginal source of funding. Harris and Kemsley (1999),Collins and Kemsley (2000), and related papers assume that all earnings are eventually paid out as taxabledividends (and none via repurchases or liquidating dividends), which is consistent with the “new view” of30the effects of dividend taxation. They argue that (nearly) full dividend taxation is impounded into shareprices and therefore, there is no incremental personal tax penalty when a firm pays a dividend. Therefore,personal taxes are large on interest income and small on equity income, and the personal tax penalty to debtfinancing is and Kemsley (1999) regress stock price on variables including retained earnings, and they inferthat retained earnings are penalized at a dividend tax rate of approximately 47%. Collins and Kemsley (2000)argue that reinvesting current earnings leads to investor capital gains taxation when shares are sold, on topof the already impounded dividend taxation. This implies that there is no personal tax penalty to dividendpayments (it is already impounded into share prices and therefore paying a dividend does not lead to furthervaluation effects). In fact, this leads to the counterintuitive argument that paying dividends leads to areduction in future capital gains payments and therefore, dividend payments are tax advantageous. Thisimplication only holds if arbitrage by tax-free investors is restricted to the point that personal investors are30See Auerbach (2002) for cites. The “new view” or “trapped equity” assumptions are in contrast to theassumptions I made at the beginning of Section 1 that “equity is the marginal source of funds” and that “dividendsare paid out according to a fixed payout policy.”29
the marginal price-setters in stocks. Collins and Kemsley find empirical evidence that they interpret as beingconsistent with their hypotheses. An untested implication of their argument is that there should be a largevalue gain in deals that result in firms returning capital to investors in any form other than taxable dividends(such as mergers). Research into this area could be than dividend taxes, an alternative argument is that capital gains taxes on future earnings areimpounded into share prices. Consider a shareholder in a nondividend-paying firm and assume that the firmis expected to pay dividends at some point in the distant future. If the market expects that low-tax investorsare likely to be the dominant owners of this company when the dividend payments are initiated, the only(future) tax that current investors face is capital gains. In support of this argument, Lang and Shackelford(2000) show that upon announcement that capital gains tax rates were going to decline, stock prices increasedmost among firms for which capital gains are most important (., firms with the lowest dividend yield). Thisis opposite the reaction predicted by lock-in models like Klein (2001), in which returns fall when capital gainsrates fall, for firms with substantial accrued retained earnings, because the required return declines along withthe tax rate. See Shackelford and Shevlin (2001) for further discussion of the tax capitalization , the tax status of the marginal investor and therefore the empirical magnitude of the personaltax penalty is an open empirical question. This is an important issue. For one thing, failing to control forpersonal tax considerations can result in an omitted variable bias. For example, personal tax considerationscould cause clientele behavior that is correlated with dividend-payout ratios. In a regression that omitspersonal tax considerations, the dividend-payout coefficient might erroneously be interpreted as supportinga nontax hypothesis. As another example, business students are often taught that the tax advantage of debtis captured by τD (see eq. (4)), which ignores personal tax effects. If it can be demonstrated that personalCtax effects are not particularly important, this simplified view of the world might be justified. In contrast, ifinvestor taxes affect security returns in important ways, more care needs to be taken in modeling these effectsin corporate finance research. Investigations of personal tax effects face several challenges, not the least ofwhich is that risk-differences between securities must be properly controlled to allow one to deduce implicittax rates from market return
Beyond Debt vs. EquityLeasing The discussion thus far has considered the debt versus equity choice; however, it can be extendedto leasing arrangements. In certain circumstances, a high-tax rate firm can have a tax incentive to borrow topurchase an asset, even if it allows another firm to lease and use the asset. With true leases (as defined by theIRS) the lessor purchases an asset, and deducts depreciation and (if it borrows to buy) interest from taxableincome. The lessee, in turn, obtains use of the asset but can not deduct interest or depreciation. Thedepreciation effect therefore encourages low-tax rate firms to lease assets from high-tax-rate lessors. Thisoccurs because the lessee effectively “sells” the depreciation (and associated tax deduction) to the lessor, whovalues it more highly (assuming that the lessee has a lower tax rate than the lessor). This incentive for low-taxrate firms to lease is magnified when depreciation is accelerated, relative to straight line depreciation. Further,the alternative minimum tax (AMT) system can provide an additional incentive for a lessee to lease, in orderto remove some depreciation from its books and stay out of AMT status are other tax effects that can reinforce or offset the incentive for low tax rate firms to ’s with relatively large tax rates receive a relatively large tax benefit of debt, which provides anadditional incentive (to borrow to) buy an asset and lease it to the lessee. Moreover, tax incentives providedby investment tax credits (which have existed at various times but are not currently on the books in the .)associated with asset purchsaes are also relatively beneficial to high tax-rate lessors. In contrast, the relativelyhigh taxes that the lessor must pay on lease income provide a tax disincentive for firms with high tax ratesto be lessors (and similarly the relatively small tax benefit that a low tax rate firm obtains from deductinglease expense works against the incentive for low tax rate firms to lease rather than buy). The traditionalargument is that low tax rate firms have a tax incentive to lease from high tax rate lessors, though thisimplication is only true for some combinations of tax rules (., depreciation rules, range of corporate taxrates, existence of investment tax credits or AMT) and leasing arrangements (., structure of leasepayments). See Smith and Wakeman (1985) for details on how nontax effects can also influence the 5: All else equal, the traditional argument is that low tax-rate firms should lease assets from high-tax rate lessors, though this implication is conditional on specifics of the tax code and leasing
There are several complications associated with investigating whether firms lease in response to taxincentives. First, because leasing expense is tax deductible, leasing endogenously reduces a lessee’s effectivetax rate, which can bias an experiment in favor of detecting tax effects. Likewise, lessor tax rates could beendogenously increased from the effects of lease income. Second, financial statement definitions of leasingare not one-to-one with IRS definitions, making it difficult to use Compustat data to test Prediction 5. Usingendogenously-affected tax variables, Barclay and Smith (1995b) and Sharpe and Nguyen (1995) find thatlow-tax-rate firms use relatively many capital leases. However, capital leases do not meet the IRS definitionof true leases (instead they are likely a mixture of true leases and conditional sales contracts, the latter ofwhich are treated like debt so the lessee deducts interest and depreciation), and therefore the documentednegative relation between capital leases and taxes is hard to interpret because it might be , Lemmon, and Schallheim (1998) address the first issue by measuring tax incentives “but-forfinancing decisions,” ., calculating tax rates using income before debt interest and the implicit interestportion of lease payments are deducted. They address the second issue by focusing on operating leases, whichare defined in a manner similar to the IRS definition of true leases. Graham et al. (1998) find that the use ofoperating leases is negatively related to before-financing tax rates, consistent with Prediction 5, and thatcapital leases are unrelated to before-financing tax rates. Graham et al. also show that erroneously using anafter-financing tax rate would double the magnitude of the negative tax coefficient for operating leases, andspuriously assign a negative tax coefficient to capital lease and Marston (2001) find that lessors tend to be high-tax rate firms (consistent with Prediction5). Finally, O’Malley (1996) finds no evidence that firms systematically lease in response to tax incentivesimposed by the AMT. We need research investigating whether the tax benefit of leasing adds to firm jury is also still out on whether debt and leasing are substitutes for the lessee (as they might be in aDeAngelo and Masulis (1980) sense because both lead to tax deductions).Pensions: Black (1980) assumes that pension plans and the overall company are a single economic entity thatshould have an integrated financing and investment strategy. Due to interest tax deductions, the cost ofcorporate borrowing is the after-tax cost of debt. Because they are tax-free entities, defined benefit pensionplans (DBs) earn the before-tax rate of interest on bond holdings. Therefore, Black suggests that DBs should32
increase (decrease) bond (equity) holdings, while the rest of the firm should do the reverse. This action shouldnot increase firm risk because the increase in corporate debt offerings is offset by the increase in bonds heldin the pension plan. In a M&M (1963) world, the net effect is that the company earns τ times the amount ofCbonds held, as in Eq. (4). Tepper (1981) argues that there can be a tax advantage to the strategy of corporateborrowing and DBs investing in bonds, even in a Miller (1977) world. In this case, the benefit occurs whenthe DB is an inframarginal investor in bonds, thereby earning the “extra” return necessary to compensateindividual investors for the personal tax penalty associated with interest income (., DBs capture some ofthe investor surplus depicted in Figure 2). The Tepper incentive for DBs to hold bonds increases with thedifference between personal tax rates on interest and equity 6: Defined benefit pension plans have an incentive to hold bonds (equity) that increases (decreases)in the corporate tax rate, while the rest of the firm has the reverse (2001) finds evidence consistent with the Black (1980) case: she finds that DB bond holdingsincrease with a simulated corporate marginal tax rate. She does not find evidence consistent with the Tepperargument. In a less direct test of the same incentives, Thomas (1988) finds time-series evidence that firmsdecrease DB contributions when their tax rate is falling, and cross-sectional evidence that high-tax firms havelarger DB funding levels. Clinch and Shibano (1996) study pension reversions, which occur when a firm terminates anoverfunded pension, settles its liabilities, and reverts the excess assets to the firm, all in one year. The revertedassets are taxable in the reversion year. Clinch and Shibano find that firms with the largest tax benefit ofreverting do so, and also that firms time reversion decisions to occur in years with particularly large taxbenefits. One nice thing about the Clinch and Shibano experiment is that their tax variable equals the taxconsequence of reverting relative to the tax consequence associated with the next best alternative (.,31amortizing the excess assets over several years).31Chaplinsky and Niehaus (1990) describe the potential tax benefits of Employee Stock Ownership Plans, aform of defined contribution benefit plan. ESOPs offer deferred compensation to employees and a deductibleexpense to employers. ESOPs are designed to allow firms to borrow to purchase own-company stock on employees’behalf, which provides an interest deduction to the firm. Moreover, half of the interest income received by thelenders is tax-free. Shackelford (1991) finds that lenders keep only 20-30% of the tax benefit associated with thisinterest, with the remainder being passed along to the ESOP in the form of a lower interest rate on the loan. In late1989, tax rules changed to restrict the interest exclusion to loans where the ESOP own more than 50% of the stock,33
Debt maturity: In the spirit of Modiglinani and Miller (1958), Lewis (1990) derives an irrelevance nullhypothesis for debt maturity. If corporate taxes are the only market imperfection, Lewis shows that theoptimal firm-specific debt policy (., optimal level of promised interest payments) can be achieved byvarious combinations of short- and long-term debt. This implies that firm value is unaffected by debt maturitystructure and that capital market imperfections beyond corporate taxes, like costs to restructuring debt orunderinvestment, are needed for debt maturity to than modeling the simultaneous choice of debt level and maturity structure as in Lewis (1990),Brick and Ravid (1985) assume that firms choose debt level before debt maturity. If the expectations theoryof interest rates holds, firms pay the same present value of interest in the long run regardless of debt maturity;however, issuing long-term debt accelerates interest payments, thus maximizing the present value of theinterest tax shield. Brick and Ravid (1985) use this logic to argue that debt maturity should increase with theslope in the yield curve. Prediction 7: Debt maturity increases in the slope in the yield curve. Most empirical evidence does not support their prediction. Barclay and Smith (1995a) and Stohs andMauer (1996) include a stand-alone yield curve variable that is either insignificant or has the wrong and Opler (1996) argue that the slope of the yield curve should only affect firms with a positive taxrate, and therefore interact the yield curve variable with the corporate marginal tax rate. Neither Guedes andOpler (using a crude measure of the corporate tax rate), nor Harwood and Manzon (1998, using a simulatedcorporate tax rate) find a significant coefficient on the yield curve variable. The one exception is Newberryand Novack (1999), who use a dummy variable equal to one during 1992 and 1993 (when the term premiumwas relatively high) and equal to zero for all other years 1987-1995. Newberry and Novack find a positivecoefficient on the yield curve dummy in their public debt regression but not in their private debt , Marcus, and McDonald (1985) determine optimal debt maturity in a model that trades offcorporate tax benefits with personal tax, bankruptcy, and flotation costs. The implications of their model arethat debt maturity decreases with the corporate MTR and increases with the personal tax rate: long maturityimplies less frequent recapitalization and relatively low transactions costs, so long-term debt can be desireablewhich effectively killed the interest exclusion except for a few very unusual
even if the net tax benefit is low. Maturity also decreases with the volatility of firm valuebecause volatilefirms are more likely to restructure debt. Prediction 8: Debt maturity decreases with the corporate MTR and the volatility of firm value and increaseswith the personal tax and Mauer (1996) find support for the latter prediction: volatile firms generally use shorter termdebt. The evidence is weaker related to the tax rate prediction. Stohs and Mauer find that debt maturitydecreases with corporate tax rates – but their MTR variable is very crude (equal to income tax expensedivided by pretax income when this ratio is between zero and one, and equal to zero otherwise). Opler andGuedes (1996) find a negative coefficient on a tax expense divided by assets variable but the wrong sign onan NOL-based tax variable. Finally, Harwood and Manzon (1998) and Newberry and Novack (1999) finda positive relation between a simulated tax rate variable and debt maturity, opposite the Kane et . A positive coefficient makes sense if large simulated MTRs identify firms that use long-termdebt cause they are relatively likely to be able to deduct interest in current and future , debt maturity can affect the tax-timing option for firms to opportunely retire debt (.,Emery, Lewellen, and Mauer (1988)). If the corporate tax function is convex, the expected present value taxbenefit of short-term debt declines with interest rate volatility, while the tax deductions with long-term debtare fixed. Therefore, long-term debt is preferred when interest rates are volatile. Long-term debt also increasesthe value of the timing option for investors to tax-trade securities (Kim, Mauer, and Stohs (1995)) becauseoption value increases with security maturity and long-term bond prices are more sensitive to changes ininterest rates. Prediction 9: Debt maturity increases with interest rate et al. (1995) find that debt maturity increases with interest rate volatility but Guedes and Opler32Harwood and Manzon’s variable equals the Graham (1996a) simulated tax rate divided by the topstatutory tax rate. This variable has a large value for firms that do not currently have NOLs and that do not expect toexperience a loss in the near future. Harwood and Manzon predict a positive relation between this tax variable anddebt maturity. They argue that firms with large values for the tax variable are likely to fully utilize tax deductions inthe future, and therefore lock into long-term debt now. In new analysis for this chapter, I perform a more direct teston the hypothesis that uncertainty about future tax-paying status reduces the use of long-term debt. I use the standarddeviation of the simulated marginal tax rate to measure uncertainty about tax-paying status, with the standarddeviation calculated across the simulated scenarios for any given firm-year. I do not find any relation between debtmaturity and uncertainty about tax-paying
(1996) do not. Nor do Guedes and Opler find significance for a second variable that interacts interest ratevolatility with a corporate MTR evidence linking tax incentives to debt maturity is mixed. One thing that makes it difficult todraw general conclusions is that debt maturity is defined differently in various papers. Barclay and Smith(1995a) use a dependent variable measuring the portion of outstanding debt that matures in four or moreyears, Guedes and Opler (1996) use the log of the term to maturity for new debt issues, Stohs and Mauer(1996) use the book value weighted-average of the maturity of a firm’s outstanding debt, Newberry andNovack use the same for new issues, and Harwood and Manzon (1998) use the portion of outstanding debtthat is long-term. Another issue that might affect inference about tax variables is the apparently nonlinearrelation between debt maturity and nontax influences (Guedes and Opler (1996)). Unless the nonlinearity ofthe overall specification is properly controlled, it might adversely affect the ability to detect tax , the yield curve was never inverted during the periods studied by most of these papers, so the testsof Brick and Ravid (1985) focus on the steepness of the yield curve, rather than the . Taxes and Capital Structure – International Tax IssuesSection 1 reviews capital structure choice in the context of a domestic-only firm operating in a classical taxsystem (in which interest is tax deductible but equity payments are not). While much academic researchfocuses on this paradigm, international tax issues have become more important in recent years. This sectionreviews how international tax law can affect corporate financing decisions in a multinational firm. Theperspective is generally for a firm headquartered in the but many of the implications hold if the firm isheadquartered elsewhere. The general framework is still based on taxes affecting firm value via an expression like V =with debtV + τ(.)*D. The research in this section demonstrates that multinational tax rules can affect the τ(.)no debtCCfunction and therefore the incentive to use both domestic and foreign debt. So as not to bog down ininternational tax law, this section only sketches the effects of multinational tax incentives. To focus on thecentral factors that affect multinational firms, I make several simplifying assumptions (described below). Fora more detailed description of international tax law, see Hines (1996) or Scholes, Wolfson, Erickson,36
Maydew, and Shevlin (2002) and the references Tax Incentives and Financial Policy in Multinational Firms: Theory and Tax RulesA multinational corporation can finance its foreign operations with internal equity (., an equityinfusion from a parent or subsidiary to an affiliated subsidiary), internal debt (., a loan from the parent toa subsidiary), external funding, or earnings retained by the foreign subsidiary. If internal equity is used, theparent receives its return on equity when the subsidiary repatriates dividends back to the home repatriations based on active operating earnings can usually be deferred indefinitely, until the parent33needs an infusion of cash, or to optimize the worldwide tax situation of the firm. In contrast, interest frominternal debt is paid according to a fixed schedule. Like a repatriated dividend, interest counts as “world-wideincome” on the . tax return of the parent. Unlike a repatriated dividend, the interest is often deductibleon the foreign tax return, allowing for a foreign tax deduction analogous to the tax benefit of debt described34in Section important items affect the financing choices of .-based multinational firms: foreign taxcredits and interest allocation rules. The . government taxes individuals and corporations on the basis ofresidence or place of incorporation, meaning that they are taxed because they are from the ., regardlessof where they earn income. (Note that the . only taxes “active foreign source income” at the time ofrepatriation to the . parent.) At the same time, the government recognizes that income earned abroad isusually taxed by a foreign entity, so the . offers foreign tax credits to offset taxes paid abroad. If the not offer such credits, the foreign operations of . corporations would face double taxation andtherefore have a tough time competing with foreign corporations. For the purposes of this analysis, think ofthe foreign tax rate (τ) as a weighted average of tax rates the firm pays in the various countries in whichFor33To illustrate the potential economic importance of repatriations and taxes on such transfers, note that in2003 the Bush administration proposed reducing the tax on all repatriated income to 5%. The goal was to spur areturn of capital to . domiciled firms in hopes that these firms would productively invest the funds and stimulatethe . economy. This provision was eliminated during negotiations with Congress over the tax are restrictions to shifting interest deductions abroad by lending from the domestic parent to theforeign subsidiary: thin capitalization rules (., limits on the magnitude of foreign debt ratios), withholding taxesimposed by the foreign government on interest payments and other repatriations, and netting rules that restrict theeffect of interest payments on the determination of foreign source income (Newberry and Dhaliwal (2001) andScholes et al. (2002)). For example, withholding taxes are above and beyond foreign income taxes and are collectedby foreign governments on remittances to parent
it earns foreign income, with the weights being the relative share of active (., non-passive) foreign sourceincome repatriated from a particular simplest terms, if the foreign tax rate is smaller than the . corporate income tax rate (τ), a firmUSreceives credit for foreign taxes paid but still must remit to the . government taxes equal to (τ -USτ)*(foreign source income). Such a firm is called deficit credit because it lacks sufficient foreign tax creditsFor(FTCs) to avoid all . taxes. For example, if repatriated foreign earnings are $200, τ=15%, and τ =ForUS35%, the firm must pay $40 in tax to the . In contrast, if τ> τ, the firm does not have to pay . tax because it receives foreign tax creditsForUSproportional to τ. For example, if τ=45% and τ = 35% and repatriated earnings are $200, the firm paysForForUS$90 in foreign tax; however, the firm’s foreign tax credits are limited to FTC=$70 (=min[$200allowτ,$200τ]), which is just enough to shield it from . tax obligation. The $20 in unused FTCs can beUSForcarried back up to two years or carried forward up to five years to offset taxes on repatriated income (or theycan be deducted rather than used as a credit). This firm is excess credit because it has more FTCs than it isallowed to use in the current year and accumulates the excess tax credits to potentially shield income inanother year. The tax benefit of debt, τ(.), can be modeled as a decreasing function of accumulated FTCsC35because FTCs can act as nondebt tax shields that are substituted for interest 10: All else equal, the incentive τ(.) to finance with domestic debt decreases with accumulatedC36foreign tax credits for deficit credit can affect tax incentives to use debt in a manner that is not reflected in a one-period model. Assumethat a multinational firm has accumulated unused FTCs that it has carried forward to the present (or assume that itanticipates receiving excess FTCs sometime in the next two years). If a firm has carried forward FTCs from previousyears it very likely was excess credit, and therefore subject to τ> τ at some point in the past. For the most part, aForUSfirm can use these accumulated FTCs only if the foreign tax rate becomes smaller than the . corporate income taxrate. This can occur if there is an exogenous shift in relative tax rates (τand τ) or if a firm repatriates moreFor USforeign source income from low-tax countries, thereby reducing the average τ (., the latter case is an example ofFora firm endogenously reducing its τ). If a firm expects to use accumulated FTCs to reduce taxes, the FTCs competeForwith interest deductions in a DeAngelo-and-Masulis sense and reduce the incentive to finance with a firm with $1 in pre-tax foreign earnings that it will repatriate back to the . to pay that the firm has $ in accumulated FTCs, τ=, τ=, τ=, and the . corporate tax rate isPEForτ=. Ignoring foreign considerations, τ= and eq. (1) equals , so it appears that the firm should financeUSCwith domestic debt. However, τ= once the effect of FTCs is considered (the firm pays $ in foreign tax andCno . tax because the FTCs offset any potential tax owed to the .); therefore, eq. (1) equals – and the firmshould finance with equity. This implication holds for deficit credit firms but not for excess credit firms (because an38
Prediction 10 is a static prediction. Considering the dynamic carryback and carryforward features of the taxcode, a dynamic prediction is that the tax incentive to finance with debt decreases with the probability of afirm being deficit credit and the probability of accumulating second important tax principle affecting multinational corporate financing decisions is theallocation of debt interest between domestic and foreign operations. Via the allocation of domestic interest,the . limits allowable foreign tax credits, thereby possibly reducing the tax benefit of domestic debt. (. does this to limit tax deductions on debt that might possibly be used to finance foreign operations andproduce foreign profits.) To implement this policy, the . allocates domestic interest to foreign operationsbased on the proportion of total assets that are in foreign subsidiaries. In rough terms, if two-thirds of acompany’s worldwide assets are held by foreign subsidiaries, then two-thirds of domestic interest deductionsare allocated to foreign income when determining the allowable-FTC calculation. Note that this is a ruling and does not mean that foreign governments recognize the allocated interest as a deductionagainst foreign income. Also note that the allocation of a portion of domestic interest abroad technicallyaffects only the allowable-FTC calculation; that is, ignoring FTC, domestic interest deductions are notdirectly interest allocation procedure can reduce the tax incentive for . firms to use domestic debtbecause τ(.) also declines with the degree of interest allocation. When a firm is excess credit (., τ < τ)CUSForand taxable on both foreign and domestic operations, the interest allocation procedure reduces the tax benefitof domestic interest deductions by setting τ(.) equal to τ(.)*[domestic assets/worldwide assets]. Thus, forCUSexcess credit firms the incentive to finance with domestic debt decreases with the proportion of assets held37abroad. One implication of the interest allocation rules is that debt policy research can not assume thatfinancial statement (or Compustat) “domestic interest expense” is fully beneficial to . credit firm would not pay . tax at repatriation, regardless of whether they have accumulated FTCs).37If a . multinational is deficit credit (., τ > τ) and taxable both in the . and overseas, τ(.)=τUSForCUSand the incentive to use domestic debt is not affected by interest allocation rules. The interest allocation rules limitthe amount of deductions a firm is allowed to use to offset repatriated foreign income. When a firm is deficit credit,it pays tax at the rate τ regardless of the amount of FTCs applied to foreign-source income, so reducing allowableUSFTCs via interest allocation does not affect the current-year tax
Table 3 summarizes the tax incentives to use external domestic or foreign debt in a one-period table is intended to be self-explanatory so I emphasize only the main points in the text. The modelignores personal taxes, carryforwards and carrybacks, and assumes that all foreign income is repatriated eachyear. The worldwide tax liability (Tax) is equal to the sum of . tax on worldwide income (Tax) andWorldUSforeign tax on foreign income (Tax), less allowable FTCs. The table shows the change in Tax thatForWorldoccurs, for various tax credit and interest allocation situations, when an additional dollar of domestic or38foreign interest is the most part, the results in Table 3 are what you would expect without thinking too deeply aboutthe complexities of foreign taxes. If Tax is zero (rows (1) and (3)) or domestic income is negative (row (6)),USthere is no tax benefit from issuing domestic debt; there is, however, a benefit of τto deducting $1 ofFor 39foreign interest when foreign income is positive (rows (1) and (6)). If foreign income is negative butdomestic income is positive (row (2)), there is no tax incentive to issue foreign debt but an incremental dollarof domestic interest provides a benefit of τ. USTwo situations are more subtle. If a . multinational is deficit credit (., τ is greater than τ)USForand profitable both in the . and overseas (row (4)), a dollar of domestic or foreign interest produces a taxbenefit of τ. To see how foreign interest produces a tax benefit proportional to τ, consider a case in whichUSUSa multinational earns $2 of income in a country with τ=45% and $4 of income in a country with τ=25%,ForForand assume that τ=35%. The $2 of high-tax foreign income produces Tax=$. The firm receivesUSForFTC=$ on this income and has $ of unused FTCs. The $4 of low-tax foreign income producesallowTax=$. As a stand-alone item, this income produces $ of . tax at repatriation ($4*(35%-25%));Forhowever, the $ of extra FTC offsets half of this . tax liability. On net the firm pays the . $ intax on foreign earnings and has a total tax liability of Tax=$ ($= $ in high-tax country, $ 38This model ignores many techniques by which firms can minimize worldwide taxes. See Scholes et al.(2002) for more information on these alternative there is a positive probability that tax-losses will be used if carried backward or forward, the tax benefitcan be positive even in rows (1), (3), or (6). Conversely, if there is a positive probability that losses will occur and becarried back from the future, positive tax benefits might be smaller than those shown in the table. Also, in a morecomplicated model, one could also net out the personal tax costs associated with interest income. Finally, seeAltshuler and Newlon (1993) for the marginal tax costs of repatriations when there are also withholding
40in low-tax country, and $ on income repatriated from low-tax country). If this firm deducts $1 ofinterest in the low-tax country, it reduces its tax bill by $ ($ reduction in Tax and $ reductionForin . tax owed on that dollar). If the firm uses $1 of interest in the high-tax country, it reduces its tax billby $ ($ reduction in Tax but $ less FTC available to offset taxes owed on the incomeForrepatriated from the low-tax country.) Either way, the tax benefit of deducting $1 of foreign interest is τUSwhen a firm is deficit credit and profitable both in the . and second subtle situation involves the tax benefit of deducting domestic interest when a firm isexcess credit and Tax and Tax are both positive (row (5)). In this case, a portion of domestic interest isUSForallocated to foreign source income, thereby reducing the benefit of a dollar of interest by the ratio of foreignassets to worldwide assets. (Recall that this allocated interest will not reduce Tax.) The allocation ofFordomestic interest reduces the incentive of an excess credit firm to issue domestic debt, especially when thefirm has substantial foreign assets. Altshuler and Mintz (1995) note that more than 60% of firms were excesscredit during the late 1980s, so interest allocation is potentially important. Prediction 11: Due to interest allocation, the tax benefit of domestic interest deductions declines with theprobability that a firm will operate as excess credit and with the proportion of assets held in foreignsubsidiaries. The analysis can be modified to examine the tax incentives associated with the parent supplying theforeign subsidiary with internal debt. The incentive is similar to that for external foreign debt shown in therightmost column in Table 3, with one difference: with internal debt, the interest is taxable to the parent atrate τwhen Tax>0, so in some cases τshould be added in the rightmost column. Specifically, if the debtUS USUS is internal rather than external, the entries in the rightmost column are -τ,+τ, 0, 0, τ - τ, and τ - τForUSUSForUSForin rows (1)-(6), respectively. (Recall that a negative term means tax savings.) First consider the deficit creditcase (row (4)) where the tax incentive to fund a foreign subsidiary with internal debt is nil: there is no taxincentive to use internal debt because the net benefit of deducting in the foreign country is exactly offset by40In most situations the income from the high- and low-tax country would be summed and treated as incomefrom one “basket”, with τ=(4x25% + 2x45%)/6=%. I treat the countries separately in this example toForhighlight how income from one country can lead to FTCs that shield income repatriated from another
the increased tax in the home country. In the excess credit case (rows (5) and (6)), the net tax benefit is τ-For τ. For these rows, there is a tax incentive to issue debt increases with τ but it is offset by taxes owed byUSForthe domestic parent. In row (2), when Tax=0 and Tax>0, there is a tax disincentive of τ per dollar ofForUSUSinternal interest: the extra foreign interest does not further reduce Tax and yet there is a positive tax liabilityForof τ on the remitted interest. In contrast, when Tax=0 (rows (1) and (3)) using internal rather than externalUSUSdebt does not change the entries in Table 3: there is no tax on the interest received by the parent because thefirm otherwise has domestic 12: The tax incentive to fund a foreign subsidiary with internal debt generally increases with τ;Forhowever, this incentive is offset in several situations as shown in Table 13: The tax incentive to issue external foreign debt increases with τ, although this incentive canForbe affected by the relative taxation of interest and equity income at the investor that the incentive to save on foreign taxes might be tempered by investor-level taxes along the linessuggested in Miller (1977).Other than in this paragraph, the results in this section are derived for the case where the domesticparent operated under a classical tax system in which interest is tax deductible but equity payments are instead there is an imputation or integrated tax system (like in the ., France, or many other countries),equity holders receive a credit for taxes paid at the corporate level, which partially or fully eliminates thedouble taxation of equity income. This at least partially reduces the net tax advantage to debt. For example,ignoring personal taxes, Cooper and Nyborg (1999) show that the value of a levered firm in an imputationtax system equals(7)where τ is the rate of imputation tax. In a full imputation tax system, dividend recipients receive a tax creditIfor income taxed at the corporate level, which they can use to offset their personal tax liability. If imputationresults in a full tax credit at the corporate rate, then τ = τ in Eq. (7) and there is no tax advantage to a partial imputation system, stockholders only receive a partial credit for taxes paid at the corporate level,42
which is analogous to making equity (at least partially) tax deductible, which in turn reduces the net taxadvantage of debt. I am not aware of research that investigates the following prediction. Prediction 14: The tax incentive to issue debt decreases with the degree of dividend imputation dictated bythe tax law under which a company Empirical Evidence Related to Multinational Tax Incentives to Use DebtTesting multinational tax hypotheses is difficult because the data are hard to obtain and noisy. Mostof the international capital structure tests are based on implications found in row (4) and especially row (5)of Table 3. Table 4 summarizes some empirical evidence related to multinational debt respect to Prediction 11 (due to interest allocation, the tax benefit of domestic interestdeductions declines with the probability that a firm will operate as excess credit and with the proportion ofassets held in foreign subsidiaries), Froot and Hines (1995) find that debt usage is reduced for excess creditfirms, with the reduction proportional to the fraction of assets that are foreign. Altshuler and Mintz (1995)also show that the use of foreign debt increases with the proportion of assets held overseas (presumablybecause domestic interest would be allocated abroad). Newberry (1998) and Newberry and Dhaliwal (2001)find that the likelihood of issuing domestic debt is highest when a firm is not excess credit and when lessinterest is allocated abroad. A related prediction is that firms shift away from debt financing when interestis allocated abroad. Collins and Shackelford (1992) show that firms increase their use of preferred stock whendomestic interest allocation is unfavorable. Froot and Hines (1995) point out that, unlike interest, leasepayments are not allocable, and show that excess credit firms rely more heavily on papers provide evidence with respect to Prediction 12 (the tax incentive to fund a foreignsubsidiary with internal debt increases with τ) and Prediction 13 (the tax incentive to issue external foreignFordebt increases with τ, although this incentive can be affected by the relative taxation of interest and equityForincome at the investor level). Examining a cross-section of countries with differing foreign tax rates, Desai(1997) indicates that the net internal debt infusion into foreign subsidiaries increases with τ (Prediction 12).ForNewberry and Dhaliwahl (2001) find that the propensity to issue bonds in foreign markets increases in τFor(Prediction 13). Hines (1995) demonstrates that royalty payments increase when they are a cheaper form of43
repatriation than are dividends. Finally, Grubert (1998) finds that an increase in the price of one form ofremittance does not reduce total payments. Firms hold the total constant and substitute between differentforms of remittance, such as dividends, interest, or respect to Prediction 10 (the incentive τ(.) to finance with domestic debt decreases withCaccumulated foreign tax credits for deficit credit firms) and Prediction 14 (the tax incentive to issue debtdecreases with the degree of dividend imputation), I am not aware of any research that explicitly investigatesthese than Altshuler and Mintz (1995), most papers use very general specifications to test for foreigntax effects or the influence of interest allocation. For example, when they are considered at all, separate termsindicating excess credit status, τ, or the ratio of foreign to worldwide assets are used, rather than interactingForthe variables in the manner suggested by the theory. Also, the sharper predictions are often ignored. Finally,I am not aware of any multinational tax research that directly links the tax benefits of debt to firm value. Tothe extent that data are available, variation across countries in tax rules and incentives provides a rich andunder-researched environment within which to investigate how variation in tax rules affect τ(.) and,Ctherefore, the financing decisions of multinational firms. Other Predictions and Evidence About Multinational Tax IncentivesInterest allocation can be avoided altogether if the domestic borrowing is performed by a domesticsubsidiary that is less then 80% owned by the parent (although this subsidiary must allocate interest on itsown books). I am unaware of any systematic research investigating this issue. Scholes et al. (2002) presentan example describing how Ford Motor Co. implemented this directly altering where and whether it issues debt, there are many related mechanisms bywhich a firm might respond to multinational tax law. A company might alter its transfer prices (the prices atwhich goods and services are transferred between related entities) to shift income from the high-tax to thelow-tax affiliate. Though transfer prices are supposed to be “arms-length prices,” the rules are vague enoughto allow wiggle-room. Properly designed, transfer pricing allows for tax-free dividend repatriation. Consistentwith this means of reducing overall taxes, Lall (1973) finds that multinational firms overinvoice their low-taxColumbian subsidiaries. Mills and Newberry (2000) find that shifting income to foreign operations increases44
in the difference between the . tax rate and the global tax rate. Alternatively, multinational firms can use“triangle schemes” in which one subsidiary is capitalized by or invested in by another affiliate subsidiary(Altshuler and Grubert (2000)). These schemes allow firms to optimally mix remittances from high- and low-tax subsidiaries in ways that reduce domestic taxes on foreign source generally, firms can time dividend repatriation to coincide with low overall tax cost to theparent and subsidiary. In particular, deficit credit firms owe . tax when they repatriate dividends, so theyhave incentive to delay repatriation. In contrast, excess credit firms often do not owe additional tax uponrepatriation. Taking debt versus equity choices as given, Hines and Hubbard (1990) find that excess creditfirms repatriate more than do deficit credit firms, and that repatriation by deficit credit firms is inverselyrelated to the tax cost of doing so. Altshuler and Newlon (1993) show that most repatriated dividends are“cross-credited”; that is, the parent firm simultaneously receives payments from both high- and low-foreign-tax subsidiaries, and can use the extra credits from one source to offset potential domestic taxes from another. 3. Taxes, LBOs, Corporate Restructuring, and Organizational Theory and PredictionsUnder perfect capital markets, an MM analysis implies a null hypothesis that organizational form andrestructurings are irrelevant to firm value. However, imperfections in the tax, legal, and informationenvironments can create situations in which the form of the organization or restructuring can Leveraged BuyoutsThere is a tax incentive for corporations to use substantial leverage in the management buyout process. Thisflows directly from the predictions in Section 1 that high tax-rate firms have incentive to use debt and thatthe associated tax benefits add to firm value. LBOs are particularly interesting because they lead to a muchlarger increase in leverage than do most debt issuances. LBOs also can provide an opportunity to mark assetsto market, thereby increasing depreciation and the associated tax 15: All else equal, the tax incentive to perform a highly-levered buyout increases with the firm’sexpected post-deal tax rate, τ(.). Distressed Reorganizations and Chapter 1145
Tax incentives can affect distressed reorganizations. Distressed firms with substantial accumulated netoperating losses (NOLs) have incentive to file Chapter 11 because it facilitates reducing debt ratios (Gilson(1997)). Chapter 11 allows the firm that emerges from bankruptcy to have unlimited use of the pre-filingNOLs to shield future income, as long as there is no change in ownership (., a large change in the ensuingtwo years in who owns the firm’s equity). Reducing the debt ratio during reorganization preserves debtcapacity and decreases the likelihood of precipitating an ownership change by future equity 16: The tax incentive for a firm to file Chapter 11 (versus a work-out), to better facilitate reducingits debt ratio in reorganization, increases with the firm’s accumulated NOL carryforwards and its expectedpost-deal tax C-corporations vs. S-corporationsTaxes affect organizational form in general, not just reorganizations. When an entity operates as a common“C Corporation,” revenues returned to investors as equity are taxed at both the firm and investor levels. Thefirm level taxation is at the corporate income tax rate, and the investor taxation is at the personal equity taxrate. The equity rate is often relatively low because equity income can be deferred or taxed at the relativelylow capital gains rate. In contrast, partnership income is passed-through and taxed only at the investor level,at ordinary income tax rates. The tax burden is often disadvantageous to corporate form. For example, atcurrent maximum statutory federal tax rates (Figure 1), in 2002 an investor would receive $ inpartnership income; in contrast, corporate equity payments would return only approximately $ (assumingequity is taxed at a 20% capital gains tax rate). There are, however, nontax benefits to corporate form thatoutweigh the tax costs for many firms. Gordon and MacKie-Mason (1994) argue that these nontax benefitsare large, annually equaling about 4% of equity value. See Scholes et al. (2002) and Gordon and MacKie-Mason (1997) for details about nontax costs and benefits of corporate form. See Shelley, Omer, and Atwood(1998) for discussion of the costs Prediction 17: All else equal, the tax incentive to operate as a C-corporation (versus a partnership or S-corp)increases in [(1-τ) - (1-τ)(1-τ)]. Divestitures and Asset SalesTax incentives can also affect the valuation, purchase and sale of assets. Alford and Berger (1998) argue that46
spinoffs are preferred by high tax rate firms when they shed assets that lead to taxable gains because spinoffscan be structured to avoid taxes to both the seller and buyer. In contrast, all else equal, sales are preferredwhen the transaction results in a loss because this loss can be deducted against corporate income. Moreover,when a firm sells an asset, the deal can be structured to benefit the seller or purchaser, possibly by financingthe deal with debt (Erickson (1998)).Prediction 18: There is a tax incentive for high-tax firms to shed assets in spin-offs when the deal is profitableand via sales when the deal is not profitable. When a firm acquires assets, high-tax firms have incentive touse “taxable deals” financed with R&D PartnershipsLeasing allows a low-tax-rate firm to “sell” tax deductions to high-tax-rate lessors. Analogously,research and development limited partnerships (RDLPs) allow low-tax firms to sell start-up costs and lossesto high-tax-rate investing partners. Prediction 19: All else equal, low tax rate R&D firms should form research partnerships with high tax Empirical EvidenceKaplan (1989) and others investigate tax benefits in leveraged buyouts. LBOs provide large interesttax deductions and also can provide an opportunity for asset value to be stepped up to market value. Note thatthe tax benefit of $1 of interest does not necessarily equal the top statutory tax rate. The net benefit is lessthan the top rate if all of the LBO interest expense can not be deducted in the current year, if there is apersonal tax penalty on interest income, or if there are nontax costs to debt. If he assumes that the net taxbenefit of $1 of interest is $ and that LBO debt is retired in eight years, Kaplan estimates that the tax41benefit of interest deductions equals 21% of the premium paid to LBO target shareholders. Kaplan alsoestimates that among firms electing to step up asset value, the incremental depreciation tax benefit equals 28%of the premium. I am not aware of any research that explicitly investigates whether the probability of41Graham (2000) accounts for the declining marginal benefit of incremental interest deductions andestimates that the gross tax benefit of debt equaled approximately one-fourth of firm value in the mid-1980s RJRNabisco and Safeway
choosing a highly levered form of reorganization increases with the expected post-deal MTR (Prediction 15).Gilson (1997) shows that firms in Chapter 11 reduce their debt ratios more when pre-filing NOLs arelarge (Prediction 16). He concludes that firms file Chapter 11 (versus a workout) in part because of taxincentives: Chapter 11 status offers smaller transactions costs to reducing the debt ratio, thereby minimizingthe chance of an ownership change that would result in the loss of pre-filing centered on tax reforms has linked taxes with organizational form. The Tax Reform Actof 1986 (TRA86) set corporate tax rates above personal income tax rates, and also equalized capital gains andordinary tax rates, providing a natural environment to test Prediction 17. These tax rate changes madepartnerships attractive by greatly increasing the tax disadvantage to operating as a corporation. Scholes etal. (2002) point out that there was a huge increase in formation of S-corporations (which are taxed aspartnerships) following TRA86. Gordon and MacKie-Mason (1997) show that the increased corporate taxdisadvantage due to TRA86 resulted in a reduction in the portion of aggregate profits paid via (and assets heldin) corporate form; however, the economic importance of this reduction was modest. Finally, Guenther (1992)investigates how corporations responded to the 1981 Economic Recovery Tax Act reduction in personalincome tax rates, which increased the tax disadvantage for corporations. He finds that firms altered policiesthat contribute to the double taxation of equity payout: firms reduced dividends and instead returned capitalby increasing the use of debt, share repurchases, and payments in mergers (which are often taxed as capitalgains).Ayers, Cloyd, and Robinson (1996) study small firms and find that entities choose to operate as S-corps, rather than C-corps, when they experience losses in their early years of operation. These losses canimmediately be passed through to S-corp investors, while C-corps must carry losses forward to offset futurecorporate income. The experiment of studying small firms is especially telling because small firms cangenerally choose between S- or C-corp form with little difference in cost or nontax considerations; therefore,the choice highlights tax incentives. Interpreting this result as strong tax evidence is somewhat clouded,however, because Ayers et al. do not find that the choice between C-corp and proprietorship/partnership formis affected by tax losses (though nontax considerations can affect this choice). Erickson and Wang (2002)argue that S-corps can be sold for more than C-corps because of favorable tax treatment. Finally, Hodder,48
McAnally, and Weaver (2001) conclude that banks convert to S-corp status to eliminate double taxation ofdividends and to reduce the onerous burden of the AMT. Research investigating organizational form choicesusing micro firm- and owner-specific tax information would be helpful. Such papers would most likelyrequire accessing confidential tax and Wolfson (1990) describe tax incentives that encouraged merger and acquisition activityin the early 1980s (following the 1981 tax act) and discouraged these activities after TRA86. They provideaggregate evidence that M&A activity surged in the early 1980s, and declined in 1987, consistent with taxincentives. See Scholes et al. (2002) for details of how acquisitions vary along the tax dimension dependingon whether the deal involves C- or S-corporations, subsidiaries, spin-offs, carve-outs, and Berger (1998) show that firms trade-off tax and nontax considerations when choosingbetween spinoffs and asset sales (Prediction 18). The authors estimate tax benefits as a means of determiningthe size and nature of nontax costs and argue that adverse selection, moral hazard, and agency costs are alltraded-off against tax benefits to influence how firms structure their deals. Erickson (1998) also demonstratesthat the structure of deals is affected by tax concerns. He shows that the probability that a sale is structuredas a “taxable deal,” financed with tax-deductible debt, increases with the acquirer’s tax rate; however, he findsno evidence that seller tax characteristics affect deal structure. Erickson and Wang (2000) find that the priceof subsidiary sales can be affected by tax considerations. These authors show that premiums (and sellerabnormal stock returns) increase when the sale is structured to allow a step-up in subsidiary basis, so that theacquiring firm receives additional depreciation tax benefits. Thus, contrary to a Modigliani and Miller prefectmarkets null hypothesis, tax considerations affect both the pricing and structure of asset taxes appear to affect the structure and price of some deals, the tax-minimizing form is notalways selected. Hand and Skantz (1998) argue that issuing new shares in equity carve-outs can avoid taxliabilities that occur when a firm issues secondary shares (at a price above the firm’s tax basis in the shares).The authors determine that, relative to issuing new shares, secondary carve-outs increase tax liabilities by anamount equal to 11% of the carve-out IPO proceeds. Hand and Skantz are not able to identify benefitsassociated with secondary carve-outs that are large enough to offset the increased tax payment. Maydew,Schipper, and Vincent (1999) find that incremental taxes incurred when firms perform taxable sales (rather49
than tax-free spinoffs) amount to 8% of the value of divested assets. The authors argue that firms incur thesetax costs 1) because they are smaller than the financial reporting benefits (., larger financial statementearnings), and 2) when selling firms are cash-constrained (sales provide a cash inflow; swaps do not).Shevlin (1987) investigates whether firms that perform R&D via partnerships have lower tax ratesthan firms that do R&D in-house (Prediction 19). Two notable features of Shevlin’s careful experimentaldesign are his use of simulated tax rates, and his specification of many explanatory variables in “as-if” form(., defining right-hand-side variables for all firms as if they funded R&D in-house, to avoid the endogenouschoice of in-house versus RDLP possibly affecting the variables’ values). Shevlin finds that tax rates exerta significant, negative influence on the probability of choosing an RDLP in two out of three as-if an NOL dummy to measure tax incentives, Beatty, Berger, and Magliolo (1995) find that low-tax firmsare more likely to finance R&D via a financing organization both before and after Research and Experimentation Tax Credit has also influenced corporate R&D spending. In hiseconomically-weighted regressions, Berger (1993) finds a positive market reaction to announcementsaffirming the tax credit. His regression coefficients indicate that three-fourths of the benefit of the creditaccrues to shareholders, with the remaining one-fourth increasing product price and therefore flowing toemployees or suppliers. This latter finding implies that the tax credit creates an implicit tax in the form ofhigher prices for tax-favored R&D activity and that this implicit tax offsets some of the intended benefit fromthe credit (in other words, some of the R&D tax credit is passed along in the form of higher prices to suppliersof R&D inputs). Berger also detects a negative market reaction among firms that do not use the creditthemselves but compete with firms that do. Swenson (1992) finds evidence consistent with low-tax-rate firmspursuing firm-specific R&D tax credits less aggressively than they are pursued by high-tax-rate , this research indicates that tax considerations affect the structure and pricing of research anddevelopment activity in the United States. The cited papers investigate R&D spending associated with pre-TRA86 tax rules. I am unaware of any similar research that investigates the influence of the tax credit onR&D activity based on post-TRA86 rules (under which the credit is based on the R&D-to-sales ratio, ratherthan on nominal R&D spending). Moreover, the R&D tax credit has temporarily expired several times since1986. It would be interesting to know whether these expirations have affected real R&D
4. Taxes and Payout PolicyModern dividend research began with Lintner’s (1956) field interviews with 28 firms. Lintner found thatdividends are stable, appear to adjust towards an earnings-payout target, and are rarely reduced. Miller andModigliani (1961) provide the theoretical foundation of payout policy and conclude that dividend policy isirrelevant in a frictionless world with perfect capital markets. Research since that time has explored howmarket imperfections create an environment in which payout policy affects firm value. This section highlightstax incentives related to corporate payout policy. For brevity, I narrow the discussion to payout issues thatparallel those in Section 1 or that shed light on unresolved capital structure issues (., whether personaltaxes affect security prices). See Allen and Michaely (1995, 2001) and Poterba (2001) for broad reviews ofthe various tax and nontax imperfections that can lead to payout policy affecting firm value and Theory and Empirical PredictionsMiller and Modigliani (1961) argue that in a perfect economic environment, firm value is determined byoperating cash flows, not by whether a company retains or pays out profits, nor by the form of payout. Thisline of reasoning produces the null hypotheses for this hypotheses: Firm value is not affected by payout policy. Taxes do not affect corporate payout and Michaely (2001) show that the null can also hold if different classes of investors are taxeddifferently and firms have differing payout policies, as long as the marginal price-setter is , firms can have a tax incentive to return equity capital via share repurchases rather thandividends if dividends are taxed more heavily than are capital gains for the marginal investor(s). Statementsby financial executives that repurchases are a “tax efficient means of returning capital to investors” supportthis point of view (though Brav, Graham, Harvey, and Michaely (2003) conclude that taxes only play asecond-order role in the choice between returning capital as dividends or repurchases). If dividends are taxed more heavily than repurchases, there can be a negative valuation of dividends(relative to repurchased shares) (., the CAPM with corporate and investor taxation in Brennan (1970) orAuerbach and King (1983)). All else equal, if a firm were to increase dividends, the pre-tax return on its stock51
would need to increase so that after-tax returns do not change. This effect increases as dividend taxationincreases relative to capital gains 20: All else equal, tax effects imply that firm value is negatively related to 1) the portion of payoutdedicated to dividends, and 2) dividend taxation relative to capital gains taxation. Analogously, required pre-tax stock returns increase with dividend payout and relative dividend are non-tax factors that also can lead to negative (., reduced funds to pursue positive NPV projects)or positive (., signaling or agency alleviation) dividend valuation (see Allen and Michaely (2001)). Note that dividend clienteles, in which high-tax rate investors own stocks with low dividend payouts,can occur under the null or Prediction 20. Under the null, firms can have different payout policies that do notaffect value, even if some investors are taxed more heavily on dividends (capital gains) and have a taxpreference for capital gain (dividend) income. Similar clienteles can form under Prediction 20, based on therelative taxation of dividends and capital gains for different groups of the extent that transactions are not costless, clientele tax characteristics can affect security example, the price of a stock changes from P to P as the stock goes ex-dividend. If the firm issues acumexdividend Div, its investors receive Div(1-τ) but simultaneously avoid capital gains taxes of the amount (Pdivcum- P)τ. With risk neutrality, continuous prices, and no transactions costs, and clienteles that do not varyexcap gainbefore and after ex-days, Elton and Gruber (1970) show that (P - P)(1-τ) =Div(1-τ) in equilibrium,cumexcap gaindivand therefore (8)where (P - P)/Div is referred to as the ex-day 21: The ex-day premium reflects the relative taxation of dividends and capital gains for a givenstock’s clientele of and Michaely (2001) call dividend clienteles “static” if they do not vary through , if there are advantages to trade among differentially-taxed investors, dividend clienteles mightbe dynamic, which can lead to changes in the composition of the clientele around certain dates. Dynamic52
clienteles might lead to abnormally high volume around ex-days. For example, low-dividend-tax investorsmight buy stocks just before ex-day, capture the dividend, then sell the stock after it goes this route, taxes might lead to ex-day behavior that produces trading volume but where the ex-daypremium is close to one. Thus, Prediction 21 is a joint prediction about clienteles being static as well as effects should vary with the tax rules of the country under consideration. For example,assuming static clienteles, the ex-day premium should increase with the degree of dividend imputation in agiven country (because a tax refund for corporate taxes paid is attached to dividends in imputation countries,which reduces the effect of dividend taxation). The premium can be greater than one if imputation makesdividends tax-favored relative to capital gains (Bell and Jenkinson (2001)). Empirical Evidence on Whether Firm Value is Negatively Affected by Dividend PaymentsBlack and Scholes (1974) test Prediction 20 by adding dividend-yield as a right-hand side variable in themarket model. They conclude that firm value is not related to dividends. In contrast, Litzenberger andRamaswamy (1979) find a significant, positive dividend-yield coefficient. Kalay and Michaely (2000)emphasize that the positive dividend effect should show up in cross-sectional (because of cross-firm variationin dividend-payout) long-run returns (., returns for stocks held long enough to qualify for capital gainstreatment). They point out that Litzenberger and Ramaswamy (1979) use monthly returns, and allow high-dividend yield firms to be considered zero-dividend in non-dividend months. Kalay and Michaely (2000) donot find cross-sectional or long-run return evidence that high-dividend stocks earn a tax premium. Kalay andMichaely’s findings imply that the effect identified by Litzenberger and Ramaswamy occurs for short-runreturns, perhaps only during the ex-dividend week. Fama and French (1998) test Prediction 20 by regressing (changes in) firm value on (changes in)42dividends and “firm value if no dividends.” If personal taxes reduce the value of dividends, and one could42As discussed in Section , FF regress the excess of market value over book assets on dividends,interest and a collection of variables that proxy for V, with all variables deflated by assets. The variables that proxyUfor V include current earnings, assets, R&D spending, and interest, as well as future changes in earnings, assets,UR&D, interest, and firm value. V is probably measured with error, which clouds interpretation of FF’s
design a clean statistical experiment that isolates tax effects, there should be a negative coefficient on thedividend variable in this specification. In contrast, Fama and French find a positive coefficient, whichprobably occurs because either their proxy for “firm value if no dividends” is measured with error and/or non-tax effects overwhelm the tax influence of dividends. For example, if firms use dividends to signal quality,dividend payments might be positively correlated with firm value. Or, if dividends are priced by tax-freeinvestors, one would not expect a negative influence of dividends on firm value. Fama and French conductthe only study of which I am aware that directly regresses firm value on dividend variables in an attempt to43determine the tax effect of Evidence on Whether Ex-day Stock Returns and Payout Policy are Affected by Investor Dividend ClientelesPrediction 21 is based on the existence of static dividend clienteles, so I start by reviewing dividend clienteleresearch. Blume, Crocket, and Friend (1974), Pettit (1977), and Chaplinsky and Seyhun (1987) provide weakevidence that investors hold stocks such that dividend yield is inversely related to personal tax rates;Lewellen, Stanley, Lease, and Schlarbaum (1978) find no such evidence. However, these studies have poormeasures for tax, risk, and wealth effects and therefore are hard to interpret. Auerbach (1983) concludes thattax-related preferences result in clienteles of investors that purchase stocks based on firm-specific dividend-price ratios. In perhaps the cleanest evidence on the subject, Scholz (1992) uses self-reported data from the1983 Survey of Consumer Finances. This survey has information on retail investor stock holdings, asophisticated measure of the investor’s relative dividend and capital gains tax rates, household wealth, andself-declared risk preferences. Scholz finds a negative relation between the dividend yield for an investor’sstock holdings and the relative taxation of dividends, which is consistent with a general preference fordividends by low-tax investors. In contrast, however, Allen and Michaely (2001) show that rich investorsreceive the bulk of dividend income (though perhaps not proportionally so, relative to wealth).Strickland (1996) finds that mutual funds and money managers hold low-dividend yield portfolios,43Another approach to study whether personal taxes affect asset prices investigates tax capitalization. SeeSection
while untaxed institutions such as pension funds show no preference. Dhaliwal, Erickson, and Trezevant(1999) find that the percentage of shares owned by institutional investors increases by about 600 basis points44in the year after a firm initiates paying a dividend. Overall, there is weak evidence that the preference fordividends decreases with income tax rates – but no direct evidence that this preference leads to static tax-based papers link corporate actions to the (assumed) tax characteristics of their investors. Pérez-González (2000) classifies firms by whether their largest shareholder is an individual or an institution andfinds that the former pay 30% fewer dividends than the latter. He also finds that when tax reform increases(decreases) the taxation of dividends relative to capital gains, firms with large retail shareholders decrease(increase) dividend payout. Poterba and Summers (1985) find a similar result for aggregate dividend behaviorin the . from 1950-1983. Lie and Lie (1999) also conclude that investor-level taxes affect payout find that firms with low-dividend payout (and presumably high-tax-rate investors) use self-tender-offershare repurchases more often than they use special dividends, and these firms also use open-marketrepurchases more often than they increase regular and Michaely (1995) point out that the trading volume around ex-days provides evidence aboutwhether clienteles are static (which would imply that trading only occurs between investors in the same taxbracket, who always hold stocks with the same dividend characteristics) or dynamic (in which case theremight be advantages to trade among differentially-taxed investors, potentially involving dividend-capture orarbitrage by low-dividend-tax investors). In the static case, there should be no abnormal volume because thereare no abnormal advantages to trade around the ex-day. Grundy (1985), Lakonoshok and Vermaelen (1986),and Michaely and Vila (1996) find evidence of abnormal trading volume on the ex-day, which is consistent45with dynamic tax-related trading on the Ex-day Premia and Returns44See Del Guercio (1996) and Brav and Heaton (1997) for evidence that institutional investors favor high-dividend stocks for nontax reasons like prudent-man and Michaely (2000) find that abnormal volume can be quite large on ex-days due to non-taxactivity. In their case, Japanese insurance companies captured dividends for regulatory reasons, using nonstandardsettlement procedures that allowed them to buy just before and sell just after the ex-day. Note, however, that thisform of nonstandard settlement ended in 1989, so it can not explain abnormal ex-day volume in recent
Elton and Gruber (1970) find that the ex-day premium was on average in the 1960s, which they interpretto imply that dividends are priced at a 22% disadvantage relative to capital gains (Prediction 21). Moreover,the premium ranged from (for the lowest dividend-yield decile of stocks) to (for the highest decile),which is consistent with the highest (lowest) tax-rate investors purchasing the lowest (highest) dividend-yieldstocks. The Elton and Gruber evidence is consistent with personal taxes affecting stock prices via dividendpayout and dividend clienteles. Their findings are strengthened by Barclay’s (1987) evidence that thepremium was in the early 1900s, before the advent of personal income are several complications to interpreting the ex-day phenomenon. Kalay (1982) points out thatabsent transactions costs and risk, arbitrage by tax-free investors should push the premium to . Kalayargues that transactions costs are too large for individual investors to be the marginal price-setters but insteadzero-tax-rate institutions might fulfill that role at ex-day. Kalay’s findings suggest that inferring tax rates fromex-day returns is complicated by transactions costs and the effect of institutional traders. Consistent with thisview, Michaely (1991) finds that the mean premium equaled approximately in both 1986 (when capitalgains tax rates were much lower than dividend tax rates for wealthy individuals) and in 1987-88 (whenstatutory dividend and capital gains tax rates were nearly equal), and was relatively invariant across dividendyield deciles during these years. Michaely’s evidence is not consistent with retail investor taxation affecting46stock prices, suggesting that prices might have been set by institutional investors in the and Hite (1998) argue that discrete stock prices lead to patterns consistent with those observedby Elton and Gruber (1970). Suppose a $ dividend is paid and, during the era when stock prices weredivisible by one-eighth, the stock price drops by the largest increment less than the dividend: $. Thisimplies an ex-day premium of , which occurs in the absence of personal tax effects. Moreover, thiseffect is strongest for low dividend stocks. Bali and Hite’s (1998) argument might explain some of theobserved ex-day phenomenon, however, it does not explain abnormal volume on the ex day, which Michaelyand Vila (1996) argue is evidence of tax-motivated discussion ignores the effect of risk (see Michaely and Vila (1995)) and transactions costs (see Boydand Jagannathan (1994), Michaely and Vila (1996), and Michaely, Vila, and Wang (1996)) on ex-day behavior. Forexample, Boyd and Jagannathan (1994) regress capital return on dividend yield and find a slope coefficient of oneand a negative intercept. They interpret the negative intercept as a measure of transactions
Frank and Jaganathan (1998) argue that dividends are a nuisance, and that market makers are well-situated to handle their collection and reinvestment. Therefore, investors unload the stock cum-dividend tomarket makers, who are compensated for handling the dividend by the dividend itself. This is especially truefor low-dividend stocks, for which the nuisance remains relatively the same but for which the reward forhandling the dividend is smallest. The implication is that prices should fall by less than the dividend in partbecause transactions are at the bid when the market-maker buys the stock on the cum date and are at the askwhen the market-maker sells the stock ex-dividend, and in part due to reduced demand on the cum-date. Theyfind evidence consistent with their arguments on the Hong Kong exchange, where the average premium isapproximately one-half during 1980-1993, even though dividends and capital gains are not taxed at thepersonal level. Kadapakkam (2000) strengthens this argument by showing that when the nuisance of handlingdividends (., cumbersome physical settlement procedures) was greatly reduced with the onset of electronicsettlement, the premium in Hong Kong became indistinguishable from , Michaely, and Roberts (2003) cast doubt on price discreteness (Bali and Hite) or bid-askbounce (Frank and Jagannathan) explaining ex-day pricing in the . These authors note that pricediscreteness and bid-ask bounce were greatly reduced as pricing increments changed from 1/8ths to 1/16ths(in 1997) to decimals (in 2001) on the NYSE. According to the price discreteness and bid-ask bouncehypotheses, the ex-day premium should have moved closer to one as the pricing grid became finer. Incontrast, the ex-day premium got smaller (further from one), which is inconsistent with the price discretenessand bid-ask bounce hypotheses. Graham et al. do find evidence consistent with the original Elton and Grubertax hypothesis, however. They find that the ex-day premium fell in conjunction with the 1997 reduction incapital gains tax , it is not possible to unambiguously interpret the ex-dividend day evidence in terms ofpersonal taxes. Green and Rydqvist (1999), however, provide convincing evidence of personal taxes beingimpounded into asset prices. Swedish lottery bonds are taxed like common stock with tax-free dividends (.,the coupon is tax-free and capital gains are taxed). Therefore, one would expect prices to be bid up cum-coupon (as high-tax rate investors purchase the bonds) and drop after the coupon is paid (with the dropleading to a capital loss deduction, which reduces taxes in proportion to the capital gains rate). Because the57
coupon is tax-free, the ratio of price drop to coupon should be greater than , reflecting the personal taxadvantage of the coupon. Moreover, regulations prohibit coupon capture or arbitrage of the type that mightbe expected to force the ratio to , and unlike the case of stock dividends, frictions and price discretenesswork in the opposite direction of the proposed tax effect. Green and Rydqvist (1999) find that the ratio ofprice drop to coupon averages for Swedish lottery bonds, implying that the relative tax-advantage ofcoupons relative to capital gains is impounded into bond prices. They also find that this implicit tax ratedeclined as tax reform reduced the top statutory personal tax rate during the 1980s and 1990s. Florentsen andRydqvist (2002) find that the ratio averages about for similar lottery bonds in (2001) investigates ex-day behavior in Germany, where the dividend imputation taxsystem attached to most dividends a tax credit for corporate taxes (until this feature was repealed in late2000). This tax credit means that dividends are more valuable to German investors than are capital gains, allelse equal. McDonald shows that tax considerations imply that the ex-day premium should be underthese conditions. In his empirical work, McDonald finds that the average premium is , indicating thatabout 60% of the dividend tax credit is impounded into the ex-day price. He also finds that 55% (35%) of thetax credit is reflected in futures (options) prices. Finally, McDonald demonstrates that there is abnormalvolume for the six days leading up to and including the ex-day, and that abnormal volume increases in thedividend yield. This is consistent with foreigners, who do not enjoy the German tax credit, selling the stockjust before ex-day. Bell and Jenkinson (2002) investigate the effects of a July, 1997 tax reform in the . Prior to 1997,the imputation tax credit attached to dividends was such that tax-free investors received a full tax refund, eventhough they did not pay taxes on the dividend. In other words, a $1 dividend was worth more than $1 to tax-free investors. The tax reform eliminated imputation credits for tax-free investors, implying that a $1 dividendis now worth only $1 to these investors. Bell and Jenkinson show that tax-free institutions like pension fundsown the majority of . equities and argue that they therefore are the marginal price-setters. Bell andJenkinson find that dividend valuation decreased after the tax reform effectively reduced the imputation taxbenefit of dividends. It would be helpful if there were more research like this that exploits the rich variationin tax codes around the
Overall, some ex-day papers provide clear evidence that personal taxes affect asset prices. Thisconclusion is not unambiguous across all papers, however, because of potential non-tax explanations forabnormal ex-day returns. If these alternative hypotheses completely explain ex-day returns, then in thesecircumstances personal taxes are not impounded into stock prices. However, even if tax rates do not appearto affect stock returns directly, tax considerations might still affect financial markets if they increase payout results have implications for capital structure research. If the marginal investor in equitiesis tax-free but the debt price-setter is not, then the personal tax penalty to using debt might be quite large. Ifthe marginal investor in equities and debt is tax-free, there is no personal tax penalty associated with debtfinancing. Finally, if the marginal-price-setter for equities is taxable and his tax rate is impounded into stockreturns, this reduces the personal tax penalty on debt relative to the Miller (1977) scenario. Understandingthe tax characteristics of the marginal price-setter(s) in various securities is an important issue for futureresearch. 5. Taxes and Compensation Theory and Empirical PredictionsAn MM perfect capital markets analysis would lead to a null hypothesis that compensation policy does notaffect firm value absent market imperfections. There has been a great deal of research investigating howagency costs and informational asymmetry can drive a wedge between employee objectives and shareholder47wealth, as well as how compensation policy can improve the group of papers investigateshow the tax code can affect the choice of when and how to pay employees. Analogous to Miller’s (1977)arguments about capital structure, Scholes et al. (2002) argue that to understand compensation policy, onemust consider the tax implications for both the employer and employee. Scholes et al. show how differenttax rates for the firm and its employees, or changing tax rates for either party, produce trade-offs between47See Murphy (1999) for a broad review of compensation research including pay-performance sensitivityand linking salary, bonuses, and stock compensation to firm performance. See Core, Guay, and Larcker (2002) for areview that focuses on using equity compensation to align executive and shareholder
salary and bonuses, deferred compensation, compensatory loans, pension contributions, fringe benefits, andstock option compensation. This section reviews compensation research most closely linked to taxes andcorporate finance: the choice of salary versus equity compensation, the choice between incentive stockoptions (ISO) and nonqualified stock options (NQO), and, linking back to Section 1 of this paper, the trade-off between compensation deductions and debt tax first tax issue is straightforward. Salary payments lead to an immediate deduction that reducestax liabilities, while employee stock options lead to a corporate deduction only when the options areeventually exercised (if then – see below). Ignoring incentives and other nontax issues, the first compensationprediction isPrediction 22: All else equal, the tax preference of paying salary compensation instead of optioncompensation increases with the corporation’s tax rate because salary expense is deducted immediately andoption expense is second tax issue involves the choice between paying employees with incentive versusnonqualified stock options. ISOs and NQOs are similar in most respects other than tax treatment, allowing48researchers to isolate how tax imperfections affect corporate compensation decisions. With ISOs, the firmnever gets a tax deduction, and the employee pays capital gains tax on the amount the share price exceeds thegrant price when the stock is eventually sold (assuming that the option is exercised at least 12 months aftergrant and the share of stock is sold at least 12 months after exercise). With NQOs, on the exercise date thefirm gets a deduction equal to the amount by which the price upon exercise exceeds the grant price, and theemployee pays ordinary income taxes on this same amount. The tax trade-off between incentive andnonqualified stock options amounts to comparing the relatively light burden of the employee paying capitalgains taxes for ISOs to the net NQO benefit (., the corporate deduction less the cost to the employee ofpaying taxes sooner and at a higher rate with NQOs).Prediction 23: All else equal, when the corporation is taxed at a higher rate than the employee on ordinaryincome, nonqualified options are preferred to incentive stock options because they lead to lower “all parties”48Stock appreciation rights are similar except the net benefit is paid in cash, not shares of stock. With SARs,the employee pays tax at ordinary personal tax rates on the cash benefit when it is paid, and the firmcontemporaneously deducts the
taxation of option compensation. Incentive stock options are generally preferred if the corporation has a lowtax rate relative to the final tax issue reviewed in this section investigates whether deductions from employee stockoptions serve as nondebt tax shields that substitute for the use of interest tax deductions by and Masulis (1980) argue that firms with substantial nondebt tax shields will use less debt. Amongpapers investigating this hypothesis, most find weak or no evidence that the traditional measure of nondebttax shields (depreciation) crowds out debt tax shields (see Section 1). Section reviews recent research thatexamines whether option deductions might serve the role of nondebt tax shields as laid out in DeAngelo andMasulis. I do not formally state this as a prediction because it is already stated in Prediction 2'. Empirical EvidenceSeveral papers investigate whether corporate and employee tax status affect compensation choice, and drawmixed conclusions. Hall and Liebman (2000) assume that all firms pay the top statutory tax rate and findsome evidence that the use of executive options increased as the corporate tax rate declined from the 1970sto the 1980s. This is consistent with Prediction 22 (the tax benefit of options increases as corporate tax ratesfall because the forgone opportunity to deduct salary expense immediately is less important). However, whenthey allow for cross-sectional differences in tax rates and annual fixed effects, the Hall and Liebman taxcoefficient becomes insignificant. In contrast, Core and Guay (2001) examine stock-option plans for employees other than the top fiveexecutives. Non-executives hold two-thirds of outstanding compensation options. Core and Guay find thathigh tax rate firms grant fewer options, consistent with Prediction 22; however, they do not find evidence thatlow-tax rate firms grant more options. Finally, Klassen and Mawani (2000) find that option use decreaseswith the corporate marginal tax rate, as in Prediction 22 (note that option compensation is not deductible forCanadian corporations, which only strengthens the incentive to deduct salary expense immediately). Overall,then, the evidence is weakly consistent with Prediction papers investigate whether corporate and employee tax status affect the choice betweenincentive and nonqualified options. Austin, Gaver, and Gaver (1998) assume that executives are taxed at the61
highest statutory rate and investigate whether high tax rate firms use NQOs. They use five different variablesto measure the corporate tax rate and find that none of them are statistically related to the form of option conclusion is generally consistent with the finding by Madeo and Omer (1994) that low- rather thanhigh-tax-rate firms switched from ISOs to NQOs following the 1969 Tax Act, opposite the tax am not aware of research that provides evidence unambiguously consistent with Prediction with personal tax incentives, Huddart (1998) finds that some employees accelerated NQOoption exercise in 1992, prior to the anticipated 1993 increase in upper income personal tax rates (from 31%to %); however, he concludes that only one-in-five employees took this action, indicating that nontaxfactors more than offset personal tax incentives in many situations. Goolsbee (1999) finds that in aggregatean abnormally large number of options were exercised in 1992, prior to the tax increase. Hall and Liebman(2000) note that Goolsbee defines abnormal based on a linear trend in exercise activity. When they insteadconsider the number of vested options and recent changes in stock prices, Hall and Liebman do not find thatemployees accelerated options exercise in anticipation the personal tax rate increase; nor do they find a delayin exercise in anticipation of personal tax rate reductions in the 1981 and 1986 tax , Shevlin, and Shores (1992) conclude that tax factors affect the disqualification of ISO plan is disqualified (., treated as an NQO plan for tax purposes) if an employee sells her stock lessthan 12 months after exercising incentive stock options. A company might want to disqualify an ISO planto receive the corporate deduction associated with NQOs if the corporate tax rate increases relative to thepersonal tax rate and/or if the ordinary personal rate falls relative to the capital gains tax rate, both of whichhappened after the 1986 tax reform. Matsunga et al. perform a careful “all parties” tax analysis and concludethat firms with the largest net benefit of disqualification were the firms most likely to , there is only modest evidence that taxes are a driving factor affecting corporate or employeecompensation decisions. This is perhaps surprising because popular press articles indicate that the size of thecorporate deduction provided by NQOs is huge, completely eliminating corporate taxes for many large,49profitable firms in the late 1990s (., NY Times, June 13, 2000).49I thank Bob McDonald for making me aware of the corporate tax deductibility of employee stock and Shevlin (2002) present evidence about options deductions for NASDAQ 100 firms. Hanlon and Shevlinprovide an excellent summary of the accounting issues related to options
Financial statement footnote information can help us understand whether the magnitude of optioncompensation deductions is sufficient to affect overall corporate tax planning and also to determine whetherthese deductions are inversely correlated with interest deductions (and therefore might explain why somefirms use little debt). I gather information on the exercise and grant prices for all options exercised from 1996to 1998 by employees of Fortune 500 firms (see Table 5). Assuming that all of the options are nonqualifiedimplies that the corporate options deduction equals the difference between the exercise and grant prices ofthe exercised options. Note that these deductions appear on tax returns and reduce taxes owed to the50government; they do not appear as a deduction on financial statements, nor are they collected by could multiply these deductions by τ to estimate their tax average (median) Fortune 500 firm had $85 ($16) million of annual deductions resulting fromemployees exercising stock options during 1996-1998 (Panel A of Table 5). These numbers are skewed: theththfirm at the 90 (95) percentile had $185 ($379) million in deductions. As a percentage of financial statementtax expense, the deductions average 50%. As a percentage of the amount of interest it would take to lever afirm up until there are declining benefits associated with incremental deductions (., levering up to the kinkin the Graham (2000) benefit functions discussed in Section ), the option deductions average 49%. PanelB of Table 5 shows the numbers for some specific firms. In the years shown in the table, option deductionsare larger than interest deductions for Dell Computer, Intel, Dollar General Corporation, General Motors, andCircuit City. Moreover, options deductions are larger than tax expense for Intel, GM, and Circuit , the magnitude of the compensation deductions are large for some firms; however, they aremoderate for many companies and therefore do not appear to provide the final answer to the puzzle of why50Option deductions do not reduce financial statement tax expense because the deductions are not treated asa permanent expense. Instead, the deductions are added to stockholders’ that the discussion in this section apply to the vast majority of firms because they elect to present theirstock option information using the intrinsic value method and therefore do not expense options and reduce netincome or earnings per share, but instead present the information in the financial statement footnotes (and neverexpense the option compensation to reduce net income). In 2002 and 2003 several dozen firms announced that theywould begin expensing option costs on financial statements (which has the effect of reducing net income on financialstatements). At press time most other firms appear to be in a holding pattern, waiting to see whether accountingstandards change with respect to reporting option compensation, so it is not possible to say whether more firms willelect to expense options in their financial statements. Note that the tax rules have not changed regarding options –this footnote simply discusses whether firms report options expense as a net income reducing item on
some firms appear to be underlevered. Nonetheless, Panel C of Table 5 reveals that compensation deductionsappear to substitute for interest deductions, and so at least partially address the puzzle. The Pearson(Spearman) correlation coefficient between the magnitude of option deductions and the degree to which afirm appears to be conservatively levered (as measured by amount of interest it would take to lever up to the51kink in the benefit function) is ().Two recent papers investigate whether option deductions displace the use of debt along the lines ofsuggested in DeAngelo and Masulis (1980), that is, to explore whether option deductions serve as a form ofnondebt tax shield that might substitute for interest deductions (Section ).. Graham, Lang, and Shackelford(2002) find that the magnitude of option deductions is large enough to reduce the median MTR for Nasdaq100 and S&P 100 firms from 34% (when option deductions are ignored) to 26% (when option deductions are52considered) in 2000. Documenting a reduction in MTRs is important because, as argued in Section ,nondebt tax shields should reduce the use of debt to the extent that the NDTS alter the marginal tax et al. find that debt ratios are positively related to tax rates and negatively related to the amount bywhich option deductions reduce marginal tax rates (consistent with Prediction 2'). Similarly, Kahle andShastri (2002) find that long- and short-term debt ratios are negatively related to the size of tax benefits fromoption exercise. Finally, Graham et al. show that firms that appear conservative when option deductions areignored appear significantly less underlevered when options are considered. Overall, the evidence is consistent with managers substituting away from debt when their firm hassubstantial option deductions. It would be interesting for future research to investigate whether other nondebttax shields play this role (., R&D tax credits or foreign tax credits), especially in eras during which optiondeductions were less prevalent. One “secretive” source of such deductions is tax shelters, which areinvestigated in Section shortcoming of this analysis is that I measure the tax benefit of realized compensation deductions,which are not necessarily the same as the deductions that managers expect ex ante, when they plan their capitalstructure. Nor do I distinguish between ISOs and NQOs, although Hall and Liebman (2000) note that NQOs accountfor 95% of option grants. Future research should address these light of the large reduction in tax rates for some firms, it is surprising that (for tax reasons) some ofthese firms do not use more incentive and fewer nonqualified stock options. One reason might be restrictions on thetotal amount of incentive stock options that can be granted in a given
6. Taxes, Corporate Risk Management, and Earnings ManagementIf capital markets were perfect, there would be no benefit to corporate hedging because investors would beable to achieve the same outcome by hedging on personal account. The null hypotheses is therefore thatcorporate hedging does not increase firm value. And yet, the corporate use of derivatives (presumably) tohedge has increased enormously in the past decade. For example, OTC swaps increased from $11 trillion in1994 to $69 trillion by 2001 ( A large number of corporate finance researchpapers investigate which market imperfections create situations that can make corporate hedgingadvantageous. Theory suggests that hedging to reduce volatility can reduce expected costs of bankruptcy (Smith andStulz (1985)), reduce underinvestment costs by shifting funds into states where they would otherwise bescarce (Froot, Scharfstein and Stein (1993)), help offset conservative decision-making that results fromemployee risk-aversion (Tufano (1996)), and reduce the effects of information asymmetry between managers,investors, and the labor market (DeMarzo and Duffie (1991) and Breeden and Viswanathan (1998)). Thoughnarrower in scope, taxes can also provide an incentive to hedge. In this section, I review imperfections in thetax code that can lead to corporate hedging being beneficial and also explore how similar imperfections canprovide an incentive to manage earnings. Theory and Empirical PredictionsSmith and Stulz (1985) show that if the function that maps taxable income into tax liabilities is convex, a firmcan reduce its expected tax liability by hedging to reduce income volatility. The tax function is generallyconvex because corporate income tax rates are progressive, though the degree of progressivity for positiveincome is small. The main form of progressivity occurs because profits are immediately taxed at a positiverate, while tax reducing effect of losses is effectively spread through time via tax-loss carrybacks andcarryforwards and is only valuable in states in which the firm is profitable. Due to the time value of money,therefore, the tax function is convex because the present value tax benefit of $1 in losses is less than the tax65
53cost of $1 in profits. With a convex tax function, firms have incentive to use derivatives to shift taxableincome from good to bad states to reduce volatility and expected tax 24: All else equal, the corporate incentive to hedge increases with the degree of tax second tax incentive to hedge involves increasing debt capacity by reducing income the extent that increased debt capacity leads to greater debt usage, it also leads to greater tax benefits andfirm value. Alternatively, increased debt capacity might go unexploited, thereby reducing expectedbankruptcy costs (Smith and Stulz (1985)). Ross (1997) and Leland (1998) argue that the former effectdominates and therefore that hedging increases firm value via the tax benefits of 25: There is a tax incentive to hedge because it increases debt capacity. When firms use this extradebt capacity, the tax benefits of debt with derivatives transfers income across states within a given time-period. In contrast,earnings management is usually thought of as smoothing income through time. Like the hedging case, tax54function convexity can provide an incentive to smooth income. However, tax incentives to smooth are moreunidirectional: All else equal, companies prefer to delay paying taxes due to the time value of , if tax rates are expected to fall, tax incentives to delay income are strengthened. The followingprediction summarizes three conditions that can lead to a convexity-like incentive to smooth that worksagainst the incentive to delay income recognition:Prediction 26: Unless one or more of the following conditions are met, there exists a tax incentive to delayrecognition of taxable income: 1) the tax function is progressive, 2) net operating loss carryforwards and otherdeductions are less than fully valued due to limitations on use and the time value of money, and/or 3) tax rates53 The logic is that the government effectively holds a call option on corporate tax liabilities and writes aput on corporate tax refunds, the net value of which can be reduced by reducing volatility. That is, present valueconsiderations from delayed tax refunds reduce the value of the government’s written put, so on net thegovernmental call option on tax collections is more valuable. Reducing volatility reduces the value of the call andthe government’s claim on corporate are numerous nontax explanations for earnings management. Schipper’s (1989) review notes thatfirms might manage earnings to reduce required returns by lowering earnings variance, to impress outside investorswho value stock via earnings multiples, because executive compensation is tied to accounting numbers, and becauseinsiders can not credibly convey private information via other
are expected to Empirical EvidenceMany empirical papers measure tax function convexity using variables based on the existence of NOL or taxcredit carryforwards (Prediction 24). These papers regress corporate derivative usage on a proxy for convexityand several nontax right-hand-side variables, and generally do not find evidence that convexity affects thecorporate use of derivatives (., Nance et al. (1993) for Fortune 500 types of firms or Tufano (1996) forgold-mining firms). Rather than proxying for convexity, Graham and Smith (1999) explicitly map out taxfunctions and find that they are convex for about half of Compustat firms, and that the average among thesefirms could save approximately $125,000 in expected tax liabilities by reducing income volatility by 5%.Graham and Rogers (2001) compare this explicit measure of tax function convexity to derivatives usage fora broad cross-section of firms and find no evidence that firms hedge in response to tax function contrast, Dionne and Garand (2000) use regression coefficients from Graham and Smith (1999) to estimateconvexity and find that hedging among gold-mining firms is positively related to estimated and Rogers (2001) use simultaneous equations to investigate the joint hedging/capitalstructure decision, to determine whether firms hedge to increase debt capacity (Prediction 25). In one equationthey regress derivatives usage on variables, including debt ratios, that explain corporate hedging and in theother equation they regress debt ratios on variables, including derivatives usage, that explain debt and Rogers find that hedging leads to greater debt usage. For the average firm, hedging withderivatives increases the debt ratio by 3% and adds tax shields equal to % of firm , the empirical evidence suggests that the tax incentive to hedge because of tax functionconvexity is weak at best. The statistical evidence is stronger that the tax incentive to increase debt capacityleads to greater hedging – though the economic importance of this effect appears to be only terms of earnings management, very little research directly investigates the conditions that can leadto a tax incentive to smooth earnings, particularly with respect to the three conditions in Prediction , Wilson, and Wolfson (1992) find that firms delayed recognizing income in 1986 in anticipation oflower future tax rates. Barton (2001) regresses a measure of earnings management (., discretionary67
accruals) on a crude convexity variable. Barton’s measure of convexity is the excess of a firm’s marginal taxrate over its average tax rate (., tax expense divided by taxable income); a positive number indicates aprogressive tax function. Barton finds that the absolute value of discretionary accruals is positively relatedto this measure of convexity, which he interprets as evidence of income smoothing in response to taxincentives. Similarly, using a NOL based convexity variable, Pincus and Rajgopal (2002) find that profitableoil and gas firms use derivatives to smooth income in response to tax incentives. As with the corporatehedging evidence, tax incentives appear to be a second-order consideration, rather than a dominant influenceon earnings . Tax SheltersTax shelters offer a means of reducing taxes that may displace traditional sources of corporate taxdeductions. Three common characteristics of shelters are that they reduce tax liability without greatly alteringfinancial statement information, they are shrouded in secrecy, and they are often shut down once detected bythe Treasury. Tax shelters can take many different forms, and the current “hot product” is always usually exploit glitches in the tax system such as asymmetric domestic and foreign tax treatment or asituation in which income is allocated beyond economic income. In the short-run, before detection, shelterscan create a money pump for some firms, with benefits far exceeding transactions costs and the probability-weighted cost of audit/detection. One could imagine a long-run equilibrium in which the benefits of sheltersare competed away or greatly reduced but, as a class, their secretive nature and the proliferation of newproducts appears to make “short-run” benefits continue unabated for those who type of shelter, the high-basis low-value variety, involves an untaxed foreign investor and ataxable domestic corporation both participating in a deal. The untaxed investor is allocated a large portionof the income from the deal and then exits the transaction in a manner that leaves a large economic loss. Thecorporation can deduct the loss against taxable income. To get a feel for the magnitude of the benefit,Bankman (1999) presents an example in which the corporation contributes $11 million to a deal and receives$10 million in property and a $40 million deductible loss. Therefore, the company effectively pays $1 million(plus maybe $3 million in transactions costs and a small expected cost of being caught) for a tax benefit of68
$40τ forms of shelters, such as the tax deductible preferred stock (MIPS) discussed in Section 1,receive positive rulings from the Treasury and go on to become accepted financial transactions. Furtherdiscussion of tax shelters is beyond the scope of this chapter. The interested reader is directed towardsBankman (1999), the source for much of the tax shelter discussion in this . Summary and Suggestions for Future ResearchThis paper reviews research related to how taxes affect corporate activities. The research often finds that taxesaffect corporate financial decisions – but the magnitude of the effect is not always large. With respect to capital structure, there is cross-sectional regression evidence that high-tax rate firmsuse debt more intensively than do low tax rate firms. There is also exchange offer evidence that indicates thatdebt tax benefits add to firm value. However, there is room for much additional research to improve ourunderstanding of capital structure tax effects. One gap is the lack of time-series evidence about whether firm-specific changes in tax status affectdebt policy. Another important area for future research is to isolate the market value of the tax benefits of debtfor the broad cross-section of firms. Additional research is also needed to explain the apparently conservativedebt policy of many firms. Such analysis might investigate whether non-debt tax shields substitute for interestdeductions – and help solve the “conservative leverage puzzle.” Two such nondebt tax shields includeemployee stock option deductions and accumulated foreign tax credits. Recent research indicates that theformer help to (partially) explain apparent underleverage in some firms. Keep in mind, though, that nondebttax shields should only affect tax incentives to the extent that they affect the corporate marginal tax have also only scratched the surface regarding tax-related leasing research. There is currently notmuch analysis of whether taxes affect the pricing and structure of lease (or other financial) contracts, analysisof whether leases and debt are substitutes for the lessee, or evidence about how lessor tax rates affect is also little research into the effect of relative corporate and personal taxes on the aggregate demandand supply of debt. Unambiguous evidence about whether taxes affect debt maturity choices is also , all of this research should emphasize robust statistical treatment of standard errors and the economic69
importance of tax effects, in light of Myers et al. (1998) statement that taxes are of third-order importancein the hierarchy of corporate intriguing in theory, the profession has made only modest progress documenting whetherinvestor taxes affect asset prices and in turn affect the costs and benefits of corporate policies. There is strongevidence that personal taxes drive a wedge between corporate and municipal bond yields. There is alsoplausible evidence that the personal tax penalty on MIPs interest income is only modest, which might implythat the personal tax penalty on debt is only modest (relative to using equity) – but this implication needs tobe verified. Several papers assume that companies have clienteles of investors that have similar taxcharacteristics, and then link these companies’ policies to the assumed investor tax rates; however, it wouldbe helpful to make these linkages more direct. In general, we need more market evidence about theimportance of personal taxes affecting asset prices, the effective equity tax rate for the marginal investor(s),and information related to the identity of the marginal investor(s) between different securities. One leveldeeper, we also need evidence that corporate policies are altered in response to these investor tax influenceson security prices. Some of this evidence will be hard to come by and might require access to confidentialinformation or data from countries with unique data or institutional has been made relating multinational tax considerations to corporate financing decisions,especially in terms of the use of debt by affiliated foreign entities when foreign tax rates are high. However,there is a need for research that highlights capital structure comparisons between classical and other taxsystems and direct tests of multinational tax incentives, including the interaction of explanatory variableswhen appropriate (., excess credit status interacted with interest allocation considerations). It would behelpful if excess (or deficit) credit tax position were measured more precisely than simply using current-period average tax studies link corporate payout policy to tax considerations. In particular, the ex-day stockreturn and volume evidence is consistent with investor tax considerations influencing asset markets. TheGreen and Rydqvist (1999) study of Swedish lottery bonds stands out in terms of presenting clean ex-dayevidence documenting personal tax effects and serves as a model for future research that isolates tax insights into some payout issues might be provided by comparing payout policy in classical versus70
other tax systems. In addition, there currently is no convincing evidence that the interaction of investor taxcharacteristics and payout policy affect firm value and stock returns. Finally, there is a need for directevidence that tax-based investor clienteles exist (., that investors hold certain securities because of theinvestor’s tax status and the form of payout) – because many of the payout hypotheses implicitly assume thatsuch clienteles is some recent evidence documenting tax motivated compensation payments (., the choicebetween salary and options paid to non-executive employees), risk management (., hedging to increase debtcapacity and the tax benefits of debt), and earnings management. However, we need more “all parties, alldeductions” research in these areas, as well as analysis of whether these forms of non-debt tax shields aresubstitutes for each other or for debt interest. We also need compensation studies based on firm- andemployee-specific tax rates and the choice between ISO and NQO plans. Finally, to date there have been fewdirect tests of whether earnings management is related to progressive tax schedules, less than full valuationof accumulated NOLs and other deductions, and/or expectations of changes in future tax studies have documented that firms choose organizational form based on relative corporate andpersonal tax rates, that asset sales are structured in response to tax considerations, and that corporatebankruptcy and highly levered restructurings have tax implications. However, we need more evidence aboutthe choice of corporate form using firm-specific data, evidence that firms choose ex ante to perform highlyleveraged buyouts in response to tax incentives, and, in general, more evidence about tax incentives affectingcorporate reorganizations, spinoffs, and other forms of , while it is convenient for academic research to investigate these tax issues one by one, thereis potential for large gains from investigating how these various policies and tax incentives interact from theperspective of a corporate financial manager or tax planner. Along these lines, explicit studies integrating themurky world of tax shelters into the overall tax planning environment would be helpful, though I suspect thatmuch of this research might end up being case studies. Overall, there are numerous important areas in whichcareful research can contribute to our understanding of how the imperfections created by taxes affectcorporate decisions and firm
ReferencesAlford, A. and P. Berger, 1998, "The role of taxes, financial reporting, and other market imperfections in structuring divisivereorganizations," working paper, Wharton School, University of PennsylvaniaAllen, F. and Michaely, R., 1995, "Dividend Policy," in R. Jarrow, V. Maksimovic, and . Ziemba (eds.), HandbooksIn Operations Research and Management Science, Finance 9, Elsevier North-Holland, , F. and Michaely, R., 2001, "Payout Policy," working paper, Cornell , R. and A. J. Auerbach, 1990, "The Significance of Tax Law Asymmetries: An Empirical Investigation," QuarterlyJournal of Economics, 105,1, , R. and H. Grubert, 2000, "Balance Sheets, Multinational Financial Policy, and the Cost of Capital at Home andAbroad", working paper, Rutgers University J Pub Econ, 2003!!Altshuler, R. and J. Mintz, 1995, ". Interest-Allocation Rules: Effects and Policy,” International Tax and Public Finance,2, , R. and T. S. Newlon, 1993, "The Effects of . Tax Policy on the Income Repatriation Patterns of Corporations", in Alberto Giovannini, R. Glenn Hubbard, and Joel Slemrod (eds.), Studies in InternationalTaxation, University of Chicago , J and G. Arachi, 2000, "The Effect of Taxes on Corporate Financing Decisions: Evidence from a Panel of ItalianFirms,"International Tax and Public Finance 8, , G. and S. N. Kaplan, 1998, "How Costly Is Financial (Not Economic) Distress? Evidence From Highly LeveredTransactions That Became Distressed", Journal of Finance, 53, , P. and D. Mullins, 1986, "Equity Issues and Offering Dilution,” Journal of Financial Economics, 15, , A. J., 1983, "Stockholder Tax Rates and Firm Attributes,” Journal of Public Economics, 21, , A. J., 2002, "Taxation and Corporate Financial Policy,” A. Auerbach and M. Feldstein (eds.), Handbook of PublicEconomics, vol. 3, , A. J. and J. M. Poterba, 1987, "Tax-Loss Carry forwards and Corporate Tax Incentives,” in Martin Feldstein (ed.),The Effects of Taxation on Capital Accumulation, University of Chicago Press, , A. J. and M. A. King, 1983, "Taxation, Portfolio Choice, and Debt-Equity Ratios: A General EquilibriumModel,” Quarterly Journal of Economics, 98, , J., J. Gaver, and K. Gaver, 1998, "The Choice Of Incentive Stock Options vs. Nonqualified Options: A MarginalTax Rate Perspective," Journal of the American Taxation Association, 20 (Fall), , B., C. Cloyd, and J. Robinson, 1996, "Organizational Form And Taxes: An Empirical Analysis Of Small Businesses",Journal of the American Taxation Society 18 (Supplement), , B., C. Cloyd, and J. Robinson, 2001, “ The Influence of Income Taxes on Use Of Inside and Outside Debt by SmallBusiness,” The National Tax Journal, 54, , R. and G. L. Hite, 1998, "Ex-Dividend Day Stock Price Behavior: Discreteness or Tax-Induced Clienteles?" Journalof Financial Economics, 47, , M. and J. Wurgler, 2001, “Market Timing and Capital Structure”, Journal of Finance , J., 1999, The New Market for Corporate Tax Shelters, Working Paper, Stanford , M. J, 1987, "Dividends, Taxes, and Common Stock Prices: The Ex-Dividend Day Behavior of Common Stock
Prices Before the Income Tax", Journal of Financial Economics, 19, , M. J. and C. W. Smith Jr., 1995a, "The Maturity Structure of Corporate Debt", Journal of Finance, 50, , M. J. and C. W. Smith Jr., 1995b, “The Priority Structure of Corporate Liabilities,” Journal of Finance, 50, , J., 2001, "Does the Use of Financial Derivatives Affect Earnings Management Decisions?" forthcoming in TheAccounting , Anne, Philip G. Berger, and Joseph Maglio, 1995, “Motives for Forming Research and Development FinancingOrganizations,” Journal of Accounting and Economics, 19, , L. and T. Jenkinson, “New Evidence on the Impact of Dividend Taxation and on the Identity of the Marginal Investor,”Journal of Finance 57, , S., and O. H. Sarig, 1997, "Corporate Finance: A Valuation Approach,” McGraw Hill, New , Philip, 1993, “Explicit and Implicit Tax Effects of the R & D Tax Credit,” Journal of Accounting Research, 31, 2,, F., 1980, "The Tax Consequences of Long-run Pension Policy", Financial Analysts Journal, 14, No. 7, , F. and M. Scholes, 1974, "The Effects of Dividend Yield and Dividend Policy on Common Stock Prices andReturns,” Journal of Financial Economics, 1, , M., J. Crockett and I. Friend, 1974, "Stock Ownership in the United States: Characteristics And Trends,” Surveyof Current Business, 54, 11, , J., and R. Jagannathan, 1994, “Ex-Dividend Price Behavior of Common Stocks,” Review of Financial Studies, 7, , M., G. Jarrell, and E. H. Kim, 1984, "On the Existence of an Optimal Capital Structure: Theory and Evidence,”Journal of Finance, 39, 857-878stBrav, A., J. R. Graham, C. R. Harvey, and R. Michaely, 2003, “Payout policy in the 21 century,” working paper, , A. and J. B. Heaton, 1997, "The Economic Effects of Prudent Man Laws: Empirical Evidence from Stock OwnershipDynamics,” working paper, Duke , D. and S. Viswanathan, 1998, “Why do Firms Hedge? An Asymmetric Information Model,” working paper, , M. J., 1970, "Taxes, Market Valuation and Corporate Financial Policy,” National Tax Journal, 23, , I. and S. A. Ravid, 1985, "On the Relevance of Debt Maturity Structure", Journal of Finance, 40, , M., 2001, "Taxes and Capital Structure: Do Investors’ Taxes Matter? Evidence from the Tax Reform Act of1986", working Paper, Arizona State , J., 1998, "Default Risk Cannot Explain The Muni Puzzle: Evidence from Municipal Bonds that Are Secured . Treasury Obligations,” Review of Financial Studies, 11, , S. and Gregory Niehaus, 1990, “The Tax and Distributional Effects of Levereged ESOPs,” FinancialManagement, 19, 29-38.
Chaplinsky, S. and H. N. Seyhun, 1987, "Tax Rationality and The Demand for Dividends,” working paper, University , G. and T. Shibano, 1996, "Differential Tax Benefits and the Pension Reversion Decision", Journal of Accountingand Economics, 21, 1, , J. and D. Kemsley, 2000, "Capital Gains and Dividend Capitalization in Firm Valuation: Evidence of TripleTaxation", Accounting Review, 75: , J. and D. Shackelford, 1992, "Foreign Tax Credit Limitations and Preferred Stock Issuances", Journal of AccountingResearch, (Supplement 1992), , G., 1983, "Capital Market Equilibrium with Personal Tax", Econometrica, 51, Clearing House, various years, “Standard Federal Tax Reporter,” Chicago, , I. A. and K. G. Nyborg , 1999, "Discount Rates and Tax", working paper, London Business SchoolCore, J. and W. Guay, 2001, "Stock Options Plans for Non-Executive Employees," Journal of Financial Economics, 61,, J., W. Guay, and D. Larcker, 2002, “Executive Compensation and Incentives: A Survey,” forthcoming in EconomicPolicy Review, Federal Reserve Bank of New , M. and N. Travlos, 1989, "Information Effects Associated with Debt for Equity and Equity for Debt ExchangeOffers,” Journal of Finance, 44, , R. and L. Senbet, 1988, "The Effect of Taxes and Depreciation on Corporate Investment and Financial Leverage,”Journal of Finance, 43, , R. C. Spatt, and H. Zhang, 2001, "Optimal Consumption and Investment with Capital Gain Taxes,” Review ofFinancial Studies, 14, , H. and R. W. Masulis, 1980, "Optimal capital structure under corporate and personal taxation,” Journal ofFinancial Economics 8, Guercio, D., 1996, "The Distorting Effect of the Prudent-Man Laws on Institutional Equity Investments,” Journal ofFinancial Economics, 40, , P. and D. Duffie, 1991, “Corporate Financial Hedging with Proprietary Information,” Journal of EconomicTheory, 53, , M. A., 1997, "A Multinational Perspective on Capital Structure Choice and Internal Capital Markets,” working paper,Harvard , D., M. Erickson, and R. Trezevant, 1999, "A Test of the Theory of Tax Clienteles for Dividend Policies,” NationalTax Journal, 52, , D., R. Trezevant, and S. Wang, 1992, "Taxes, Investment-Related Tax Shields And Capital Structure,” Journalof the American Taxation Association, 14, , G., and M. Garand, 2000, “Risk Management Determinants Affecting Firms' Values in the Gold Mining Industry:New Empirical Results,” HEC working , A. K. , 2002, "An Analysis of Capital Structure and Investment Policies: Evidence from Corporate Divestitures,”forthcoming in Journal of , K., and F. Marston, 2001, “Incentives for Leasing: Evidence from the Largest . Lessees and Lessors,” working
paper, University of , E., 1986, "Valuation Effects of Corporate Debt Offerings,” Journal of Financial Economics, 15, , E., 1995, “Taxation and Corporate Financial Policy,” working Ppaper, Uppsala University, , E. J. and M. J. Gruber, 1970, "Marginal Stockholder Tax Rates and the Clientele Effect,” Review of Economics andStatistics, 52, , D., W. Lewellen, and D. Mauer, 1988, "Tax-Timing Options, Leverage, and the Choice of Corporate Form", Journalof Financial Research, 11, , E. M. Erickson, and E. Maydew, 1999, “Debt-Equity Hybrid Securities,” Journal of Accounting Research, 37, , M., 1998, "The Effect of Taxes on the Structure of Corporate Acquisitions,” Journal of Accounting Research, 36,, M. and E. Maydew, 1998, "Implicit Taxes in High Dividend Yield Stocks,” Accounting Review, 73, , M. and S. Wang, 2000, "The Effect of Transaction Structure on Price: Evidence from Subsidiary Sales,” Journalof Accounting and Economics, 30, , M. and S. Wang, 2002, "The Effect of Organization Form on Acquisition Price”, Working Paper, University ofChicago. Esty, B., 1998, The Impact of Contingent Liability on Commercial Bank Risk Taking, Journal of Financial Economics 47,, B., F. Qureshi, and W. Olson, 2000, “Iridium LLC.”, Harvard Business School case study No. 200-039, Rev. 3/27/, E. F. and K. R. French, 1998, "Taxes, Financing Decisions, and Firm Value,” Journal of Finance, 53, 3, , E. F. and K. R. French, 2001, "Testing Trade-off and Pecking-Order Predictions About Dividends and Debt,” Reviewof Financial Studies, 15, , D. and L. Selwyn, 1967, "Taxes, Corporate Policy, and Return to Investors,” National Tax Journal, 20, , B., and K. Rydqvist, 2002, “Ex-Day Behavior when Investors and Professional Traders Assume Reverse Roles–The Case of Danish Lottery Bonds,” Journal of Financial Intermediation 11, , M. and R. Jaganathan, 1998, “Why do Stock Prices Drop by Less Than the Value of the Dividend?: Evidence Froma Country Without Taxes,” Journal of Financial Economics, 47, , K. and J. Hines, Jr., 1995, "Interest Allocation Rules, Financing Patterns, and the Operations of . Multinationals",in Martin Feldstein, James Hines Jr., and R. Glenn Hubbard, (eds.), The Effects of Taxation on Multinational Corporations,University of Chicago , K., D. Scharfstein, and J. Stein, 1993, “Risk Management: Coordinating Corporate Investment and FinancingPolicies,” Journal of Finance, 48, , F., 1999, “Equity Security Investments: Evidence on Tax-Induced Dividend Clienteles,” Journal of AmericanTaxation Association, 22, 1, , Stuart C., 1997, Transactions Costs and Capital Structure Choice: Evidence from Financially Distressed Firm,”Journal of Finance, 52, 111-133.
Givoly, D., C. Hahn, A. Ofer, and O. H. Sarig, 1992, "Taxes and Capital Structure: Evidence From Firms' Response to theTax Reform Act of 1986,” Review of Financial Studies, 5, , R., N. Ju, and H. Leland, 2001, "An EBIT-Based Model of Dynamic Capital Structure,” Journal of Business 74,, A., 1999, "What Happens When You Tax The Rich? Evidence Form Executive Compensation"Gordon, R. H., and MacKie-Mason, J. K., 1990, Effects of the Tax Reform Act of 1986 on Corporate Financial Policy andOrganizational For,” In: J. Slemrod, (ed.), Do Taxes Matter? (MIT Press, Cambridge), , R. H. and J. MacKie-Mason, 1994, "Tax Distortions to the Choice of Organizational Form,” Journal of PublicEconomics, 55, , R. H. and J. MacKie-Mason, 1997, "How Much Do Taxes Discourage Incorporation?" Journal of Finance, 52, 2,477-505Graham, J. R., 1996a, "Debt and the Marginal Tax Rate,” Journal of Financial Economics, 41, , J. R., 1996b, "Proxies for the Marginal Tax Rate", Journal of Financial Economics, 42, , J. R., 1999 ,"Do Personal Taxes Affect Corporate Financing Decisions?" Journal of Public Economics, 73, , J. R., 2000, "How Big Are the Tax Benefits Of Debt?” Journal of Finance, 55, , John R., 2001, “Estimating the Tax Benefits of Debt,”, Journal of Applied Corporate Finance 14, 42-54,Graham, J. R., 2004, “Taxes and Corporate Finance,” in B. E. Eckbo (ed.), Handbook of Corporate Finance – EmpiricalCorporate Finance, Elsevier Science, Amsterdam (2004) , J. R., and C. Harvey, 2001, “The Theory and Practice of Corporate Finance: Evidence from the Field,” Journalof Financial Economics, 60, , J. R., E. Hughson, and J. Zender, 1999, "Stock Market Reactions to Capital Structure Changes: Theory andEvidence", working Paper, Duke UniversityGraham, J. R., M. Lang, and D. Shackelford, 2002, “Employee Stock Options, Corporate Taxes and Debt Policy,” workingpaper, Duke , J. R., M. Lemmon, and J. Schallheim, 1998, "Debt, Leases, Taxes, and the Endogeneity of Corporate Tax Status,”Journal of Finance, 53, 131-162Graham, J. R., R. Michaely, and M. Roberts, 2003, “Do Price Discreteness and Transactions Costs Affect Stock Returns?Comparing Ex-Dividend Pricing Before and After Decimalization,” Journal of Finance , J. R. and D. Rogers, 2001, "Do Firms Hedge in Response to Tax Incentives?", Journal of Finance, 57, , J. R. and C. W. Smith, Jr., 1999, "Tax Incentives to Hedge", The Journal of Finance, 54, 2241- , R. C., 1993, "A Simple Model of the Taxable and Tax-Exempt Yield Curves,” Review of Financial Studies, 6,, R. C. and B. Hollifield, 2003, "The Personal-Tax Advantages of Equity,” Journal of Financial Economics, 67, , R. C. and K. Rydqvist, 1999, "Ex-Day Behavior with Dividend Preference and Limitations to Short-Term Arbitrage:The Case of Swedish Lottery Bonds,” Journal of Financial Economics, 53, 145-187.
Grinblatt, M. and S. Titman, 2002, Financial Markets and Corporate Strategy (Irwin McGraw-Hill, Boston, MA). Grubert, H., 1998, "Taxes and the Division of Foreign Operating Income Among Royalties, Interest , Dividends andRetained Earnings", Journal of Public Economics, 68, 2, , B. D., 1985, "Trading Volume and Stock Returns Around Ex-Dividend Dates,” working paper, University , J. and T. Opler, 1996, "The Determinants of the Maturity Structure of Corporate Debt Issues", Journal of Finance,51, , J. A., 1992, "Taxes And Organizational Form: A Comparison of Corporations and Master Limited Partnerships",The Accounting Review, 67, 17Hall, B. and J. Liebman, 2000, "The Taxation of Executive Compensation", in James Poterba (ed.), Tax Policy and theEconomy, 14, MIT Press, Cambridge, of Corporate Finance, 2004, B. Espen Eckbo (ed.), Elsevier (North Holland).Hand, J. R. M. and T. R. Skantz, 1998, “Market Timing Through Equity Carve-Outs”, working paper, UNC Chapel , M. and T. Shevlin, 2002, “Accounting for Tax Benefits of Employee Stock Options and Implications for Research,”Accounting Horizons 16, , T. and D. Kemsley, 1999, “Dividend Taxation in Firm Valuation: New Evidence,” Journal of Accounting Research37, , M. and A. Raviv, 1990, “The Theory of Capital Structure,” Journal of Finance 46, E. and G. B. Manzon Jr., 1998, “Tax Clienteles and Debt Maturity”, working paper, Boston , C. B., 2001, "The Financing Choices of Large Commercial Banks", working paper, University of Texas, , J. R., 1995, “Taxes, Technology, Transfer, and the R & D Activities of Multinational Firms,” in Martin Feldstein,James R. Hines Jr., and R. Glenn Hubbard (eds.). The Effect of Taxation on Multinational Corporations, University ofChicago Press, Chicago, , J. R., 1996, "Altered States: Taxes and the Location of FDI in America,” American Economic Review, 86, , J. R. and R. G. Hubbard, 1990, "Coming Home to America: Dividend Repatriations by . Multinationals", in AssafRazin and Joel Slemrod (eds.), Taxation in the Global Economy, The University of Chicago , L., . McAnally, and C. D. Weaver, 2001, “The Influence of Tax and Non-Tax Factors on the OrganizationalForm of Banks”, Working Paper, University of Texas , H., T. Opler, and S. Titman, 2001, “The Debt-Equity Choice,” Journal of Financial and Quantitative Analysis,36, , S., 1998, "Patterns of Stock Option Exercise in the United States", in J. Carpenter and D. Yermack (eds.),Executive Compensation and Shareholder Value, NYU and Kluwer Academic Publishers, Norwell, , S., 1998, "Tax Planning and the Exercise of Employee Stock Options", Contemporary Accounting Research, 15,2, 203-216Irvine, P. and J. Rosenfeld, 2000, "Raising Capital Using Monthly-Income Preferred Stock: Market Reaction andImplications for Capital-Structure Theory,” Financial Management, 29, 2, Summer , M. and W. Meckling, 1976, "Theory of the Firm: Managerial Behavior, Agency Costs, and Ownership Structure,”
Journal of Financial Economics, 7, , P., 2000, "Reduction of Constraints on Arbitrage Trading and Market Efficiency: An Examination of Ex-DayReturns in Hong Kong after Introduction of Electronic Settlement,” The Journal of Finance, 55, , K. and K. Shastri, 2002, “Firm Performance, Capital Structure, and the Tax Benefits of Employee Stock Options,”Working Paper, University of , A., 1982, "The Ex-Dividend Behavior Of Stock Prices: A Reexamination of the Clientele Effect,” Journal ofFinance, 37, , A. and Michaely, R.,2000,"Dividends and Taxes: A Re-Examination,” Financial Management, 29, , A., A. J. Marcus, and R. L. McDonald, 1985, “Debt Policy and the Rate of Return Premium to Leverage,” Journalof Financial and Quantitative Analysis, 20, , S., 1989, "Management Buyouts: Evidence on Taxes as a Source of Value", Journal of Finance, 44, 3, , D. and D. Nissim, 2002, "Valuation of the Debt-Tax Shield", Journal of Finance, 57, , E. H., 1989, "Optimal Capital Structure in Miller’s Equilibrium,” in S. Bhattacharya and G. Constantinides (eds.),Financial Markets and Incomplete Information, Rowman and Littlefield, (Totowa, NJ), C. S., D. C. Maurer, and M. H. Stohs, 1995, “Corporate Debt Maturity Policy and Investor Tax-Timing Option: Theoryand Evidence,” Financial Management, 24, , K., and A. Mawani, 2000, “The Impact of Financial and Tax Reporting Incentives on Option Grants to CanadianCEOs,” Contemporary Accounting Research 17, , P., 2001, "The Capital Gain Lock-In Effect and Long-Horizon Return Reversal", Journal of Financial Economics,59, 1, 33-62Koski, J. L. and R Michaely, 2000, "Prices, Liquidity, and the Information Content of Trades,” The Review of FinancialStudies, 13, , A. and R. H. Litzenberger , 1973, "A State-Preference Model of Optimal Financial Leverage,” Journal of Finance,28, , J., and T. Vermaelen, 1986, "Tax-Induced Trading Around the Ex-Day,” Journal of Financial Economics, 16,, S., 1973, "Transfer Pricing by Multinational Manufacturing Firms", Oxford Bulletin of Economics and Statistics, 35,, M. and D. Shackelford, 2000, "Capitalization of Capital Gains Taxes: Evidence from Stock Price Reactions to the1997 Rate Reductions,” Journal of Public Economics, 76, , H. E., 1998, "Agency Cost, Risk Management and Capital Structure,” Journal of Finance, 53, , ., and D. H. Pyle, 1977, "Informational Asymmetric, Financial Structure, and Financial Intermediation,” Journalof Finance, 32, , M. and J. Zender, 2001, "Looking Under the Lamppost: An Empirical Examination of the Determinants of CapitalStructure", working paper, University of , M. and J. Zender, 2002, "Debt Capacity and Tests of Capital Structure Theories", Working paper, University ofUtah.
Lewellen, W., K. Stanley, R. Lease, and G. Schlarbaum, 1978, "Some Direct Evidence on the Dividend ClienteleHypothesis,” Journal of Finance, 33, , C., 1990, "A Multiperiod Theory of Corporate Financial Policy Under Taxation", Journal of Financial andQuantitative Analysis, 25, , E. and H. Lie, 1999, "The Role of Personal Taxes in Corporate Decisions: An Empirical Analysis of Share Repurchasesand Dividends,” Journal of Financial and Quantitative Analysis, 34, , John, 1956, “Distribution of Incomes of Corporations Among Dividends, Retained Earnings, and Taxes,”AmericanEconomic Review, 46, 2,, R. H. and K. Ramaswamy, 1979, "The Effect of Personal Taxes and Dividends on Capital Asset Prices:Theory and Empirical Evidence,” Journal of Financial Economics, 7, 163-700. MacKie-Mason, J. K., 1990, "Do Taxes Affect Corporate Financing Decisions?" Journal of Finance, 45, 5, 1990, , R., 2001, “Cross-Border Investing with Tax Arbitrage: The Case of German Dividend Tax Credits,” Review ofFinancial Studies 14, , R. L., 2002, "The Tax (Dis)Advantage of a Firm Issuing Options on Its Own Stock,” forthcoming in Journalof Public , S. and T. Omer, 1994, "The Effect of Taxes on Switching Stock Option Plans: Evidence From the Tax Reform Actof 1969,” Journal of the American Taxation Association, 16 (Fall), , N. Gregory and James M. Poterba, 1996, “Stock Market Yields and the Pricing of Municipal Bounds,” NBERWorking Paper , R. W., 1980, "Stock Repurchase by Tender Offer: An Analysis of the Causes of Common Stock Price Changes,”Journal of Finance, 35, , R. W., 1983, "The Impact of Capital Structure Change on Firm Value: Some Estimates,” Journal of Finance, 38,, R. W. and A. Korwar, 1986, "Seasoned Equity Offerings: An Empirical Investigation,” Journal of FinancialEconomics, 15, , S., T. Shevlin, and D. Shores, 1992, "Disqualifying Dispositions of Incentive Stock Options: Tax BenefitsVersus Financial Reporting Costs", Journal of Accounting Research, 30 (Supplement), , E., K. S. Schipper, and L. Vincent, 1999, "The Impact of Taxes on the Choice of Divestiture Method,” Journalof Accounting and Economics 28, , R., 1991, "Ex-Dividend Day Stock Price Behavior: The Case of the 1986 Tax Reform Act,” Journal of Finance,46, 3, , R. and J. Vila, 1995, "Investors’ Heterogeneity, and Volume Around the Ex-Dividend Day,” Journal of Financeand Quantitative Analysis, 30, 2, 171-198Michaely, Roni and Jean-Luc Vila, 1996, "Trading Volume with Private Valuations: Evidence from the Ex-dividend Day,”Review of Financial Studies, 9, , R., J. Vila, and J. Wang, 1996, "A Model of Trading Volume with Tax-Induced Heterogeneous Valuation andTransaction Costs,” Journal of Financial Intermediation, 5, 340n371.
Mikkelson, W. and M. Partch, 1986, "Valuation Effects on Security Offerings and the Issuance Process,” Journal ofFinancial Economics, 15, , James, and John Ezzell, 1985, “Reformulating Tax Shield Valuation: A Note,” Journal of Finance, 40, , M. H., 1977, "Debt and Taxes,” Journal of Finance, 32, , M. and F. Modigliani, 1961, "Dividend Policy, Growth and the Value of the Firm,” Journal of Business, 34, 411-433Miller, M. H. and M. Scholes, 1978, "Dividends and Taxes,” Journal of Financial Economics, 6, , L. and K. Newberry, 2000, “ The Influence of Tax and Non-Tax Costs on Book-Tax Reporting Differences: Publicand Private firms,” working paper, University of Arizona at , B. and K. H. Wruck, 2001, "Financial policy, Investment Opportunities and the Demand for External Financing:Evidence from Low Debt Firms,” working paper, Ohio State , F. and M. H. Miller, 1958, "The Cost of Capital, Corporation Finance and the Theory of Investment,” TheAmerican Economic Review, 48, , F. and M. H. Miller, 1963, "Corporate Income Taxes and the Cost of Capital: A Correction (inCommunications),” The American Economic Review, 53, , Kevin, 1999, "Executive Compensation,” in O. Ashefelter and D. Card (eds.). Handbook of Labor Economics, , North-Holland, , M. M., 2001, "The Impact of Taxes on Corporate Defined Benefit Plan Asset Allocation", Journal of AccountingResearch, 40, , S. C., 1977, "Determinants of Corporate Borrowing,” Journal of Financial Economics, 3, , S. C., 1984, "The Capital Structure Puzzle,” Journal of Finance, 39, , S. C., 1993, “Still Searching for the Optimal Capital Structure,” Journal of Applied Corporate Finance, 6, , S. C. and Nicholas S. Majluf, 1984, “Corporate Financing and Investment Decisions When Firms Have InformationThat Investors Do Not Have,” Journal of Financial Economics, 13, , Stewart, John McConnell, Alice Peterson, Dennis Soter, and Joel Stern, 1998, Vanderbilt University roundtable onthe capital structure puzzle, Journal of Applied Corporate Finance, 11, , D. R., C. W. Smith, Jr., and C. W. Smithson, 1993, "On the Determinants of Corporate Hedging,” Journal ofFinance, 48, , K., 1998, "Foreign Tax Credit Limitations and Capital Structure Decisions," Journal of Accounting Research,36, , K., and D. S. Dhaliwal, 2001,"Cross-Jurisdictional Income Shifting by . Multinationals: Evidence fromInternational Bond Offerings,” Journal of Accounting Research, 39, K. J. and G. F. Novack, 1999, "The Effect of Taxes on Corporate Debt Maturity Decisions: An Analysis of Publicand Private Bond Offerings", Journal of the American Taxation Association, 21, 2O'Malley, Michael P., 1996, The effects of taxes on leasing decisions: Evidence from panel data, Working paper, Board ofGoverners of the Federal Reserve , Robert, and Michael Weisbach, 1999, Measuring Investment Distortions arising from Stockholder-BondholderConflicts, Journal of Financial Economics 53, 3-42.
Pérez-González, F., 2000, "Large Shareholders and Dividends: Evidence from . Tax Reforms", working Paper, HarvardUniversityPettit, R. R., 1977, "Taxes, Transaction Costs and the Clientele Effect of Taxes", Journal of Financial Economics, 8, , M. and R. Rajgopal, 2002, “The Interaction between Accrual Management and Hedging: Evidence from Oil andGas Firms,” The Accounting Review, 77, M. J. and R. C. Lease, 1986, "The Impact of Preferred-for-Common Exchange Offers on Firm Value,” Journal ofFinance, 41, , J. and K. Klassen, 2001, “The Influence of Firm Maturation on Firms’ Rate of Adjustment to Their Optimal CapitalStructures,” Journal of the American Taxation Association 23, , G. A., 2000, "An Evaluation of Alternative Measures of Corporate Tax Rates,” forthcoming in the Journal ofAccounting and , J. M., 1989, "Lifetime Incidence and the Distributional Burden of Excise Taxes,” American Economic Review, 79,, J. M., 2001, "Taxation and Portfolio Structure: Issues and Implications,” in Luigi Guiso, Michael Haliassos, andTullio Jappelli (eds.), Household Portfolios, MIT Press, Cambridge, , J. M. and L. Summers, 1985, "The Economic Effects of Dividend Taxation,” in Altman and Subramanyam (eds.),Recent Advances in Corporate Finance, Richard , R. G. and L. Zingales, 1995, "What Do We Know About Capital Structure Choice? Some Evidence FromInternational Data" ,Journal of Finance, 50, , M. P., 1997, "Corporate Hedging: What, Why and How?" working paper, University of California, , S., 1977, "The Determination of Financial Structure: The Incentive Signaling Approach,” Bell Journal of Economics,8, , K., 1989, “Commentary on Earnings Management,” Accounting Horizons, 3, , M. and M. Wolfson, 1990, "The Effects of Changes in Tax Laws on Corporate Reorganization Activity,” Journalof Business 63, 141-164Scholes, M. and M. Wolfson, 1992, "Taxes and Business Strategy: A Planning Approach,” Prentice-Hall, Englewood Cliffs,, M., M. Wolfson, M. Erickson, E. Maydew, and T Shevlin, 2002, "Taxes and Business Strategy,” 2 edition,Prentice-Hall: Englewood Cliffs, , M., P. Wilson, M. Wolfson, 1992, "Firms Responses to Anticipated Reductions in Tax Rates: The Tax Reform Actof 1986,” Journal of Accounting Research, 30 (Supplement), , J., 1992, "A Direct Examination of the Dividend Clientele Hypothesis,” Journal of Public Economics, 49, , Craig, Deborah Thomas, Keith Sellers, and Duane Kennedy, 1996, “Effects of Tax Integration and Capital GainsTax on Corporate Leverage,” National Tax Journal, 49, , J. H., Jr., 1976, "A Theory of Optimal Capital Structure,” Bell Journal of Economics, 7, J. and W. Wempe, 2000, "Do Capital Gain Tax Rate Increases Affect Individual Investors’ Trading Decisions?", TheJournal of Accounting & Economics, 30, 1, 33-57.
Schackelford, D., 1991, “The Market for Tax Benefits: Evidence from Leveraged ESOPs,” Journal of Accounting andEconomics, 14, , D. and T. Shevlin, 2001, "Empirical Tax Research in Accounting,” Journal of Accounting and Economics, 31,, K., 1994, "The Nature of Information Conveyed by Pure Capital Structure Changes,” Journal of Financial Economics,36, , S. and H. Nguyen, 1995, Capital market imperfection and the incentive to lease, Journal of Financial Economics39, , M., T. Omer, and T. J. Atwood, 1998, “Capital Restructuring and Accounting Compliance Costs: The Case ofPublicly Traded Partnerships,” Journal of Accounting Research 36, , T., 1987, "Taxes and Off-Balance Sheet Financing: Research and Development Limited Partnerships,” TheAccounting Review, 62, , T., 1990, "Estimating Corporate Marginal Tax Rates with Asymmetric Tax Treatment of Gains and Losses,” TheJournal of the American Taxation Association, 12, , P. M., 1996, "Taxes and Corporate Debt Policy in Canada: An Empirical Investigation,” Canadian Journal ofEconomics, 29, 556-572Shyum-Sunder L., and S. Myers, 1999, “Testing Static Tradeoff Against Pecking Order Models of Capital Structure,”Journal of Financial Economics 51, , G. A., 1990, "Tax-Adjusted Discount Rates,” Management Science, 36, , ., Jr. and R. M. Stulz, 1985, "The Determinants of Firms' Hedging Policies,” Journal of Financial andQuantitative Analysis, 20, , Clifford and L. MacDonald Wakeman, 1985, Determinants of corporate leasing policy, Journal of Finance 40, , M., and D. Mauer, 1996, "The Determinants of Corporate Debt Maturity Structure", Journal of Business, 69 (July),, D., 1996, "Determinants of Institutional Ownership: Implications for Dividend Clienteles,” working Paper, OhioState , C. W., 1992, “Some Tests of the Incentive Effects of the Research and Experimentation Tax Credit,” Journal ofPublic Economics, 49, , R., 1991, "Consistent Valuation and Cost of Capital Expressions with Corporate and Personal Taxes,” FinancialManagement, 20, 8n20Tepper, I., 1981, "Taxation and Corporate Pension Funding Policy", Journal of Financial Economics, June, , J. K., 1988, "Corporate Taxes and Defined Benefit Pension Plans", Journal of Accounting and Economics, 10, , R., 1992, "Debt Financing and Tax Status: Tests of the Substitution Effect and the Tax Exhaustion HypothesisUsing Firms' Responses to the Economic Recovery Tax Act of 1981,” Journal of Finance, 47, 4, , P., 1996, "Who Manages Risk? An Empirical Examination of Risk Management Practices in the Gold MiningIndustry,” Journal of Finance, 51, 1097-1137.
Wang, S., 2000, “Optimal Expected Corporate Marginal Tax Rate and Leverage Adjustments,” working Paper, , Jerold B., 1977, “Bankruptcy Costs: Some Evidence,” Journal of Finance, 32, , M., 2000, “Tax Clienteles, Implicit Taxes, and Tax Capitalization: Interaction of Multiple Clienteles in FinancialMarkets,” working Paper, University of California, Los Angeles.
Table 1 Table of Contents1. Taxes and Capital Structure............................................................. Theory and Empirical Predictions................................................. Empirical Evidence on Whether the Tax Advantage of Debt Increases Firm Value......... Exchange Offers..................................................... Cross-Sectional Regressions............................................ Marginal Benefit Functions............................................ Empirical Evidence on Whether Corporate Taxes Affect Debt vs. Equity Policy........... Nondebt Tax Shields, Profitability, and the Use of Debt...................... Directly Estimating the Marginal Tax Rate................................ Endogeneity of Corporate Tax Status.................................... Time-Series and Small Firm Evidence of Tax Effects........................ Empirical Evidence on Whether Personal Taxes Affect Corporate Debt vs. Equity Policy.... Market-Based Evidence on How Personal Taxes Affect Security Returns........26Tax capitalization.................................................. Beyond Debt vs. Equity.......................................................31Leasing.................................................................31Pensions................................................................33Debt maturity............................................................342. Taxes and Capital Structure – International Tax Issues....................................... Tax Incentives and Financial Policy in Multinational Firms: Theory and Tax Rules........ Empirical Evidence Related to Multinational Tax Incentives to Use Debt................ Other Predictions and Evidence About Multinational Tax Incentives....................453. Taxes, LBOs, Corporate Restructuring, and Organizational Form.............................. Theory and Predictions........................................................ Leveraged Buyouts................................................... Distressed Reorganizations and Chapter 11................................ C-corporations vs. S-corporations....................................... Divestitures and Asset Sales............................................ R&D Partnerships.................................................... Empirical Evidence...........................................................484. Taxes and Payout Policy.............................................................. Theory and Empirical Predictions................................................ Empirical Evidence on Whether Firm Value is Negatively Affected by Dividend Payments.. Evidence on Whether Ex-day Stock Returns and Payout Policy are Affected by Investor Taxes....................................................................... Dividend Clienteles.................................................. Ex-day Premia and Returns............................................575. Taxes and Compensation Policy........................................................ Theory and Empirical Predictions................................................ Empirical Evidence...........................................................636. Taxes, Corporate Risk Management, and Earnings Management............................... Theory and Empirical Predictions................................................ Empirical Evidence...........................................................687. Tax Shelters........................................................................708. Summary and Suggestions for Future Research............................................71
Table 2Annual calculations of the mean benefits of debt and degree of debt conservatismBefore-financing MTR is the mean Graham (1996) simulated corporate marginal tax rate based on earnings before interest deductions, and after-financing MTR is the same based on earnings after interest deductions. Kink is the multiple by which interest payments could increase without a firmexperiencing reduced marginal benefit on incremental deductions (., the amount of interest at the point at which a firm’s marginal benefit functionbecomes downward sloping, divided by actual interest expense) as in Graham (2000). The tax benefit of debt is the reduction in corporate and statetax liabilities occurring because interest expense is tax deductible, expressed as a percentage of firm value. Money left on the table is the additionaltax benefit that could be obtained, ignoring all costs, if firms with kink greater than one increased their interest deductions in proportion with -financing MTRAfter-financing MTRKinkTax benefit of debtMoney left on
Table 3Tax incentive to use debt in a . multinational firm with foreign tax credits and allocable domestic interestAssume that a . multinational firm currently returns $1 of pre-corporate-tax earnings to its marginal investor as domestic equity. The one-periodmodel in this table shows the tax effect of instead paying the $1 as foreign interest (rightmost column in each panel below) or as $1 of domesticinterest (the second-to-rightmost column). The model is adapted from Collins and Shackelford (1992) and assumes that all foreign income (Inc) isForrepatriated every year and that tax rules are the same worldwide, except that only the . allocates interest. The model ignores the AMT, carrybacksand carryforwards, personal taxes, and allocable items other than interest. Because the real-world tax-code is dynamic (., allows for carrybacks andcarryforwards), the one-period nature of this model might overstate (understate) the largest (smallest) tax benefits. Note that foreign losses (., IncFor- Int < 0) can not be repatriated as losses back to the . FTC is allowable foreign tax credit (sometimes referred to as FTC), FA isForallowlimitationforeign assets net of foreign debt, WA is worldwide assets net of foreign debt, and FSI is foreign source income, which equals Inc - Int - = Tax + Tax - FTC = (Inc - Int + Inc - Int)τ + (Inc - Int)τ - FTC , whereWorldUSForallowUSUSForForUSForForForallowFTC = Max{0, Min [Tax, FSIτ, Tax] } = Max{0, Min [(Inc - Int)τ, (Inc - Int - Int)τ, (Inc - Int + Inc - Int)τ]}allowForUSUSForForForForForUSUSUSUSForForUSIf Taxand Taxthen FTC=and Tax =USForallowWorld(1)=0>00 *0(Inc - Int)τ-τForForForFor(2)>0=000(Inc - Int)τ-τUSUSUSUS(3)=0=00000Otherwise, if Tax>0 and Tax>0 andUSForif Inc-Intthen FTC =referred to asand τUSUSallowUS(4)>0> Tax /FSIdeficit credit(Inc - Int)τ-τ-τ -τ+ τ= ForForForForUSUSFor For-τUS(5)>0excess credit *< τ(Inc - Int - Int)τ-τ (1 - )-τ -τ + τ= ForForForUSUSUSUSForUS-τFor(6)<0, and < Inc - Int innot(Inc - Int + Inc -domestic losses but0-τ -τ+ τ= ForForUSUSForUSFor USabsolute value, so someapplicableworldwide profitsInt)τ-τForUSFortaxes paid(excess credit) **In a multi-period model, FTCs above the allowable amount could be carried back or accumulated and carried forward. For example, in the excesscredit case with interest allocation (row (5)), of unused FTCs accumulate per incremental dollar of domestic interest.
Table 4Summary of predictions and empirical evidence for multinational capital structurePredictionEmpirical EvidenceFirm uses less debt when it has accumulated FTCsNoneExcess credit firms should have less incentive than deficitDebt usage declines when firm is excess credit. The reduction iscredit firms to use domestic debt. increasing in the fraction of assets that are foreign (Froot andHines (1995)).The incentive for excess credit firms to use domestic debtdeclines with the proportion of assets that are of issuing domestic debt is highest when deficitcredit and decreases as FTC limitations increase (NewberryThe incentive to use foreign debt increases in the foreign(1998) and Newberry and Dhaliwal (2001)).tax credit firms’ use of foreign debt increases in τ and inForthe share of foreign assets (Altshuler and Mintz (1995)).Debt ratios of foreign affiliates increase in τ (Desai (1997)Forand Altshuler and Grubert (2000))If domestic losses, use foreign . multinationals borrow in foreign subsidiary when theyhave domestic NOL carryforwards (Newberry and Dhaliwal(2001)).Use a different financing source than domestic debt,Weak evidence that excess credit firms lease more than otherespecially when foreign assets are substantial. Forfirms (Froot and Hines (1995)).example, use leases instead of debt because leasepayments are not allocated to foreign . firms’ incentive to finance with preferred stock rather thandebt increases with proportion foreign assets (Collins andShackelford, 1992) and Newberry (1998)). Use internal debt infusion rather than internal equity toNet internal borrowing by subsidiary from parent increases infinance foreign subsidiary, especially when τ is high.τ (Desai (1997)).ForForSimilarly, finance via royalty agreement rather than withIncrease royalty payments when cheaper than (Hines (1995))Use transfer pricing to increase (decrease) cashflow toMultinationals overinvoice low-tax affiliates (Lall (1973)).low (high) tax -controlled . firms’ . tax expense is inverselyrelated to difference between the . and global tax rate (Millsand Newberry (2000) ).Repatriate dividends when excess credit firms repatriate more than deficit credit firms, andrepatriation by deficit credit firms is inversely related to the costRepatriation for deficit credit firms negatively related toof doing so. (Hines and Hubbard (1990))τ - τ.USForRemMost repatriated dividends are “cross-credited” (Altshuler andit dividends from high- and low-foreign-tax firmsNewlon (1993))simultaneously, to reduce potential domestic via . subsidiary that is less than 80% ownedExample: Ford Motor Co. set up domestic financing subsidiaryby multinational which it owned 75% (Scholes et al. (2002)).Use triangle arrangements between subsidiaries inLow-foreign tax subsidiaries invest in high-tax affiliateforeign jurisdictions with different tax burdens to reducesubsidiaries, who in turn remit funds to . parent at low ordomestic taxes owed on domestic tax liability; or low-foreign tax subsidiaries arecapitalized by high-tax affiliate subsidiary, so repatriations fromhigh-tax subsidiary are assigned a foreign tax rate that is amixture of the low and high tax rates (Altshuler and Grubert(2000)).
Table 5Corporate Tax Deductions Resulting from Option Compensation for Fortune 500 firms, 1996-1998Panel AAnnual OptionDeduction/Deduction/Fortune 500DeductionInterest ExpenseTax Expense1996-1998 ($ million) B Option Interest Tax Deduction/ Deduction/ ExpenseExpense Interest Tax ExpenseSpecific FirmsDeductionExpense ($million)($million)($million)Dell Comp. (1997) Corp. (1998) Gen. (1997) (1998) City (1998) Deduction is the dollar amount of option compensation expense that a firm can deduct from its taxableincome in a given year, which is calculated as the number of options exercised in a given year times the differencebetween the weighted average exercise price and the weighted average grant price. This calculation treats allexercised options as if they are nonqualified options. Deduction/Interest is the option compensation deductiondivided by interest expense, where interest expense is from financial statements. Deduction/Tax Expense is thecompensation deduction divided by tax expense, where tax expense is from financial statements.
Figure 1Corporate and Personal Income Tax RatesThe highest tax bracket statutory rates are shown for individuals and C corporations. The corporate capital gains taxrate (not shown) was equal to the corporate income tax rate every year after 1987 and equal to 28% every yearbefore 1988. Source: Commerce Clearing House, annual publications. Figure 2Equilibrium Supply and Demand Curves for Corporate DebtThe supply curve shows the expected tax rate (and therefore the tax benefit of a dollar of interest) for the firms thatissue debt. The demand curve shows the tax rate (and therefore the tax cost of a dollar of interest) for the investorsthat purchase debt. The tax rates for the marginal supplier of and investor in debt are determined by the intersectionof the two curves. In the Miller Equilibrium (panel A), all firms have the same tax rate in every state of nature, sothe supply curve is flat. The demand curve slopes upward because tax-free investors are the initial purchasers ofcorporate bonds, followed by low-tax-rate investors, and eventually followed by high tax-rate-investors. In theMiller equilibrium, all investors with tax rate less than the marginal investor's (., investors with tax rates of 33%or less in Panel A) are inframarginal and enjoy an "investor surplus" in the form of an after-tax return on debt higherthan their reservation return. In Panel B, the supply curve is downward sloping because firms differ in terms of theprobability that they can fully utilize interest deductions (or have varying amounts of nondebt tax shields), andtherefore have differing benefits of interest deductibility. Firms with tax rates higher than that for the marginalsupplier of debt (., firms with tax rates greater than 28% in Panel B) are inframarginal and enjoy "firm surplus"because the benefit of interest deductibility is larger than the personal tax cost implicit in the debt interest rate.
20020002599109915891089157910791Figure 1Corporate and Personal Tax RatesTax RatePersonal Ordinary Income70%60%Corporate Income50%40%Personal Capital Gains30%20%10%0%
Figure 2Debt Supply and Demand Curves Cost or benefit per dollar of interestPanel A40%Miller (1977) Supply Curve (based on τ)C30%Investor Surplus20%Demand Curve (based on τ)P10%0%Amount of InterestCost or benefit per Panel Bdollar of interest40%DeAngelo and Masulis (1980) Supply Curve30%Firm SurplusInvestor Surplus20%Demand Curve10%0%Amount of Interest
