Wednesday, September 22, 2004

Accounting fraud leads to higher management turnover--Go figure!

Our trip to the FMA meetings continues with a look at Accounting Fraud and Management Turnover By Jayaraman, Mulford, and Wedge.

Jayaraman, Mulford, and Wedge ask the question “What impact does accounting fraud have on the turnover of top management?” Their answer really should not surprise anyone who has been paying attention in recent years: fraud leads to higher turnover. However, just because the outcome is known, it does not mean the paper is not worth reading!

Short Version:
Accounting fraud leads to a higher likelihood of upper management turnover in the 5 year event window surrounding the inclusion in the SEC Accounting and Auditing Enforcement database

Longer version:

“Accounting fraud and scandals have been occupying a central stage in the public policy debate in the recent years. This paper studies firms’ management turnover as a result of accounting fraud.”

Existing literature on the topic is surprisingly mixed. For example, if you presume firms are more likely to cheat when times are bad, there is several papers that suggest that CEO turnover increases as performance is bad. However, in a paper that is closer to this one, Agrawal, Jaffe and Karpoff (1999) “do not find systematic evidence of unusually high turnover among senior managers and directors. Even for financial fraud firms, they do not find higher top management and director turnover rates than control firms.”

It is in this context that Jayaraman, Mulford, and Wedge begin their paper.

Their sample is composed of 291 firms that were the target of SEC action for accounting fraud from 1990 to 2000. The authors examine management changes over the 5 years (+/- 2 years) from the first SEC Accounting and Auditing Enforcement Release (AAER). This event window is noteworthy both because of its length (5 years) but also because it is for 2 years prior to the SEC announcement. This is because there is often a long time period between when the fraud was committed and when the SEC announces their investigation. For many firms it is likely that the news of the fraud, and possibly the pending investigation, is already incorporated by the market prior to the official announcement.

Over this 5 year time period, the authors “identify top management turnover in 180 of the 291 firms.” This is significantly more than found at control firms. Again to quote the authors:

“In all models, top management is significantly positively related with AAER
dummy at 1% to 5% level, after controlling for size, growth opportunities, returns and board information.”

Why the difference between this and previous work?

The authors respond to this question: “Our result is different from Agrawal et al (1999) where they do not find significant relationship between fraud and top management turnover. There are several reasons for our stronger findings: First, they do not document the turnover prior to the event when it is more likely to happen as a correspondence to the fraud investigation. Second, their sample includes frauds of all kinds while our sample primarily focuses on accounting fraud….Third, they collect data from Wall Street Journal, which is more likely to document fraud of larger companies. Our sample includes all firms in AAER releases and does not have the same size bias. One can argue that fraud has less direct impact on management turnover in bigger and more complicated firms.”

Overall it is an interesting paper that, while seemingly relatively early in the publication process, does fit what we would expect to see.

No comments: