WARWICK BUSINESS SCHOOL
IB9CU0: Research Methods in Accounting
Research Assignment
2022
SUBMISSION DEADLINE
19th of April at 20:00
Please submit your research assignment report via my.wbs before the submission deadline.
Assessment Task
Accounting conservatism or the precept:
“anticipate no profit, but anticipate all losses” (Bliss, 1924)
is a prevalent feature of financial reporting and is usually interpreted as the requirement of a higher
degree of verification to recognize ‘good news’ as gains than to recognize ‘bad news’ as losses.
To illustrate this point, consider the case of a pharmaceutical firm that developed a vaccine
to prevent infection from a nasty virus. If the vaccine is discovered to have severe side effects
(e.g., increase the chances of heart failure), the ‘bad news’ would be recognized by writing off any
associated R&D assets and recognizing a provision for the firm’s best estimate of its expected future
losses (e.g., payments to those affected, expected litigation costs, sales returns, etc). However, if
there are some ‘good news’ (e.g., decreases the chances of heart failure) its recognition would
be delayed until the expected increased sales and cash-flows are verifiable. That is, bad news is
recognized in a more timely and complete manner than good news.
Interpreting accounting conservatism in this way enable different magnitudes of conservatism,
because “the greater the difference in degree of verification required for gains versus losses, the
greater the conservatism” (Watts, 2003). This form of conservatism is also known as ‘conditional
conservatism’.1
1This notion of conservatism should be distinguished from ‘unconditional conservatism’(or news independent).
Unconditional conservatism arise from systematically choosing accounting methods and estimates that produce rela-
tively lower book values of net assets. For instance, (a) expensing rather than capitalizing and amortizing R&D; (b)
overestimate allowance for doubtful accounts or sales returns; (c) using LIFO (instead of FIFO) when inventory costs
1
The work of Basu (1997) is the first empirical study that attempts to proxy the concept of
‘conditional conservatism’. Building on the ‘asymmetric verification’ intuition, Basu (1997) argues
that ‘conditional conservatism’ can be measured using a piece-wise cross-sectional regression model
of earnings on favourable returns (good news) and unfavourable returns (bad news). The generic
Basu (1997) model is:
Xi = β0 + β1 ∗D + β2 ∗Ri + β3 ∗D ∗Ri + i (1)
where:
ˆ Xi denotes a measure of accounting earnings.
ˆ Ri denotes a measure of stock market returns.
ˆ D is a dummy variable that takes value 1 when Ri < 0, and 0 otherwise.
Even though Basu (1997) never actually proposed a measure of the degree of conservatism, the
straightforward intuition and fairly simple way in which the Basu’s model can be estimated lead
to hundreds of empirical papers studying conditional conservatism to use Basu’s model to proxy
conditional conservatism.
Despite its acceptance in the past, many subsequent articles criticized the Basu’s measure of
conditional conservatism and questioned the validity of the inferences drawn from empirical studies
using the measure.
Required:
1. Estimate the Basu (1997) model using data from U.S. listed firms and discuss whether the
results obtained provide evidence on the financial reporting being ‘conditionally conservative’
or not.
ˆ Explanation of research choices involved in the estimation. [40 points]
ˆ Presentation and Interpretation of the Results. [10 points]
ˆ Discussion. [10 points]
2. Discuss two limitations of the Basu’s measure of conditional conservatism and its conse-
quences. [40 points]
are increasing; (d) accelerated depreciation methods (e.g., using short estimated asset lives) so that accounting dere-
ciation is larger than economic depreciaiton. This type of ‘conservatism’ creates ‘hidden’ (i.e., unrecorded reserves)
that provide managers with flexibility to report higher earnings in the future.
2
General Instructions
In this research assignment you are required to do an ‘empirical research task’. You are allowed
to discuss with your peers, but the assessment is individual you will be assessed only on your
own project reports and ‘Do-Files’2. Evidence on poor academic practice will result in lower
marks, while evidence of cheating/plagiarism/collusion would lead to a ‘plagiarism investigation’.
Please familiarize yourself with the guidelines regarding this on the ‘Masters Handbook’ (Link
to Handbook), and make sure that you properly reference the work of others (Link to Library
Referencing Resources).
The marking is based on how well you will address the assigned research task in your project report.
You have much room for developing your empirical research in a way that satisfactorily addresses
the research task (within the requirement of not exceeding 2500 words). This research
assignment accounts for 40% of the final mark, and the submission deadline is 19th of April
at 20:00, you should submit your research assignment report via my.wbs before the submission
deadline.
Format
The minimum (maximum) length of the report is 2000 (2500) words (exclusive of embedded refer-
ences or citations in the main body text, tables, diagrams, charts, figures, appendices, and bibli-
ography). The body text of the project report should be double-spaced and formatted in 12-point
font (Times New Roman). Footnotes, if any, should be single-spaced and formatted in 10.5-point
font. Margins should be one inch from top, bottom, and sides. Cites and references should be made
using the Harvard Referencing System/Style.
Practicalities
ˆ Before you start doing any analysis make sure you know the literature. It may not be obvious
but this research task will require that you do a fair amount of literature review.
ˆ You will have to download the data needed for the analysis on your own from WRDS.
ˆ You should use STATA for the analysis.
2Do-files are required for replicability and transparency purposes (i.e., to avoid plagiarism and poor academic
practices).
3
ˆ You have to submit the ‘Do-File’ containing the code that replicates the results in your
research report. For marking purposes both the ‘Do-File’ and the ‘Research Report’ will be
considered.
ˆ Make sure you provide enough explanations regarding your research choices both in the Do-
File and in the research report.
ˆ Don’t copy/paste from anywhere it will show up on the turnitin report, high levels of similarity
may result on lower marks and/or cheating/plagiarism investigations.
ˆ Use your own words. Avoid using Google Translator and/or direct translations from your
own language. This usually results in phrases that lack meaning.
ˆ Read proof your work before submitting.
ˆ Make sure you submit the correct version of your file and that the file is not corrupted.
ˆ Do not exceed the 2500 words limit.
Assessment Criteria
Your work will be assessed on the basis of how well you address the required research tasks. This
include:
- Explanation of the research choices related to the estimation of the Basu (1997) model.
[Comprehension]
- Estimation of the Basu (1997) model and presentation of results. [Technical Capability]
- Interpretation of your empirical results. [Analysis]
- Conclusions drawn from your empirical analysis. [Analysis, Critical Evaluation]
- Discussion of limitations and potential concerns related with the empirical analysis carried
out. [Critical Evaluation]
- Quality of your academic writing. [Academic Writing]
4
Marking Criteria
In evaluating all of the above the following elements will be considered:
ˆ Comprehension: Showing knowledge & understanding about the subject matter.
ˆ Analysis: Presenting logical arguments supported by evidence.
ˆ Critical Evaluation: Showing capacity for original thought by questioning relevant argu-
ments and/or identifying their strengths and weaknesses.
ˆ Academic writing: Presenting a clear and structured assignment; use of relevant litera-
ture;academic honesty;referencing and citation.
ˆ Technical Capability: Using data effectively to draw out conclusions and illustrate dis-
cussions, including the use of appropriate quantitative data manipulation techniques/use of
formulae etc.
References
Basu, S., 1997. The conservatism principle and the asymmetric timeliness of earnings. Journal of
Accounting and Economics 27, 3–37.
Bliss, J., 1924. Management through accounts. New York, NY: The Ronald Press Co. .
Watts, R. L., 2003. Conservatism in accounting part I: Explanations and implications. Accounting
Horizons 17, 207–221.
5