COURSE OUTLINE
Part 1:Research Methods in ACC/FIN
Part 2: Basic Theoretical Issues in ACC/FIN
Part 3: Special Topic in ACC/FIN
ACC/FIN
RESEARCH METHODS
INTRODUCTION TO RESEARCH
Cui Zhe
School of Business
Nantong University
OBJECTIVE
To ensure you have the skills necessary to produce an excellent dissertation
STRUCTURE
From developing the idea to presenting the completed piece of research
Developing the idea
Formulating testable hypotheses – traditional approach
Theoretical modelling
Developing a research plan
Models
Data
Published, existing, to be collected, qualitative and quantitative
Analysis
Writing, presentation, review
*
THE SCIENTIFIC RESEARCH PROCESS
STRUCTURE(2)
1 Searching the literature for ideas, previous experience and a plausible way forward
Week 1 information bases available
Project – select the broad area in which you plan to work
find 10 (or more) useful articles written on the topic
read them and set out their main features
this develops a knowledge base
Assignment – make a detailed appraisal of an article
How does it go about the research process
Critical appraisal - pluses/minuses, what do I learn from it for my research?
Designing the analysis
General issues - myself
Specific examples from subject areas Cahan/Berkman/Rouse
GETTING STARTED
Don’t rush
Choose something you are interested in
Something which is useful to the rest of your work or what you want to do next
It must be achievable, which normally means something narrow
Listen to advice
Look at some examples of successful work
Find a supervisor you think you can work with!
STEP 1 – GET A NOTEBOOK
A research ‘diary’
Record your thought process and your actions
Make a note of what you agreed with your supervisor – you might even want to send them a copy
Over-riding theme BE METHODICAL
When you get to the literature review take careful notes – set up some sort of filing system
It is your choice whether it is electronic, physical or a combination but Endnote is useful for references
When you have got your topic set up a research Plan – what do you need to do how long will it take?
A Gantt chart is helpful – record progress against plan
Good research is systematic, logical, objective and critical – enthusiasm is good, passion dangerous
A WARNING
Accounting and finance deal with human behaviour – it is not hard science
Findings will be tentative
Human beings are complex we can only model part of their behaviour
They are not consistent
They learn
Data are always difficult and expensive to obtain
Linearity is often misleading asymmetry, big vs small changes
Data may consistent with several theories
A RESEARCH PROBLEM
A specific problem within the research topic
Most important component of the research process
Ideal characteristics
Fills a gap within the research topic
Improves our understanding of the research topic
Relevance to firms / practitioners / regulators
“Researchable”
Theory driven
《 MONETARY POLICY ANNOUNCEMENTS AND STOCK REACTIONS 》
MOTIVATION
There is some evidence from the UK that financial markets’ responses to monetary policy have been completely different in the present crisis from the 10 years beforehand
We thought this surprising and wanted to see whether it applied elsewhere, particularly to the euro area and also to Australia and New Zealand, or whether it was due to something peculiar to the UK
FINDINGS
Although the UK finding is repeated for the euro area it is not found for Australia or New Zealand – we therefore conclude that it is the result of hitting the zero bound for interest rates and not an unusual change in behaviour
We show that behaviour is clearly asymmetric over the economic cycle with much stronger reactions in growth than in downturns
We show a second asymmetry in that markets do not respond in the same way to positive as they do to negative shocks
A PRELIMINARY
Financial markets respond to surprises not expected changes
Central banks operate monetary policy by setting short-run interest rates at regular intervals
We measure expectations by interest rate futures prices
We look at the behaviour of the Bank of England, the Eurosystem and the Reserve Banks of Australia and New Zealand for as long as they have been announcing decisions
WHAT DO WE EXPECT?
If interest rates are set higher than expected markets will think that stock price returns in the future will be lower than they previously thought (and symmetrically for a downward surprise)
. stock prices and surprises will be inversely related
ESTIMATED MODEL
(Bernanke and Kuttner, 2005),
rt = a + b ∆PRet + c ∆PRut + Xtd + εt (1)
where rt refers to the one-day return of a stock index on announcement day t, and PR refers to the policy rate, e denoting the expected change and u the unexpected change. X is a vector of all the other identifiable factors, other than policy rate changes, which affect the announcement day returns. a, b, c and d are parameters and ε is the residual.
ESTIMATED MODEL
The surprise component is calculated as the one-day change in the futures implied rate:
ΔPRut = fm, t – fm, t-1 (2)
where fm, t refers to the futures rate on the announcement day (month m, day t). The expected component of the rate change is then:
ΔPRet = ΔPRt – ΔPRut (3)
OUR ANALYSIS
We look at the behaviour of stock prices on the day of interest rate announcements by each central bank
We divide the data period in two pre and post crisis. We define the crisis as beginning with the collapse of Lehman Brothers in September 2008
We find that there is indeed a change for both the UK and the euro area but that it is better dated from when their short run interest rates effectively reached zero – the sign of the relationship is reversed
There is no change for Australia and New Zealand who did not reach zero
The relationship is disturbed over September-November 2008 in all countries
OUR ANALYSIS (2)
We divide the data into periods of economic growth and contraction
We find that the expected inverse relationship is clear in periods of economic expansion but not in periods of contraction (there are too few periods of contraction in Australia to make this judgement)
When we combine the crisis and this distinction between growth and recession there is no impact in Australia and New Zealand from the crisis
OUR ANALYSIS (3)
We distinguish positive from negative surprises and find that reactions are not the same
We are not able to say whether reactions to positive and negative surprises also vary across the economic cycle
A proviso – contrary to theory we find in some cases that expected changes have an effect on stock prices
The easiest reason for this is that we mismeasure expectations – this is bad news as it would mean that we also measure surprises wrongly
(Surprise = Actual – Expectation)
OUR ANALYSIS (4)
We also look at one aspect of New Zealand which is very interesting for monetary policy in the other areas
NZ offers a view on how interest rates will evolve in the future (as do Sweden and Norway)
Announcing that interest rates are likely to increase in future affects stock prices positively (it is symmetric for decreases)
This is difficult to rationalise as with the recession perversity – perhaps it is a reflection of the credibility of the central bank
IMPLICATIONS
Monetary policy surprises have more limited effect in downturns than when the economy is expanding
Here we are only looking at stock prices but there is clear evidence (Mayes and Viren, 2011) that monetary policy has only a trivial impact on inflation and output in downturns
The problem is even worse for the Bank of England and Eurosystem in the present crisis and may even be perverse because they have reached the zero interest rate bound
IMPLICATIONS(2)
People’s behaviour is clearly asymmetric. Not only do they react differently in contractions from expansions but they react differently to positive from negative surprises (stock prices are affected far less by positive than negative surprises)
Central banks need a sophisticated approach to setting policy
They need to react early and rapidly to stimulate action in a downturn to avoid being caught by the zero bound
Positive surprises may not achieve much in an upturn (surprises by definition cannot be repeated)
FINAL REMARK
The outlook for monetary policy in making a contribution to revitalising the European economies is poor
New Zealand and especially Australia have more scope for action should prospects worsen
FUTURE WORK
We might be able to tackle the zero bound problem by using a term structure equation, say of the Nelson-Siegel-Svensson variety. This would generate negative rates in the 3-month/overnight band and hence we would be able to observe differences on the announcement days despite actual zero rates (Kripner 2012)
*
