Wednesday, May 28, 2008

Identifying Overvalued Equity

here's an interesting paper by Daniel Beneish of Indiana U. and Craig Nichols of Cornell titled "Identifying Overvalued Equity."

Here's the abstract:
Jensen (2005) argues that overvaluation changes the behavior of managers in ways that increase agency costs, but suggests that overvaluation is difficult to identify. We show that observable characteristics of changes in managers' accounting, operating, investing and financing decisions can be used to predict two likely consequences of overvalued equity: future stock price declines and overstatement of accounting earnings. In particular, we show that an overvaluation score (O-Score) that combines proxies for earnings overstatement, prior merger activity, excessive stock issuance, and the manipulation of real operating activities identifies firms with one-year-ahead abnormal price declines averaging -27%. We also estimate a model that integrates these various attributes to predict accounting restatements associated with fraud. In light of the costs associated with overvalued equity, the findings that firm characteristics can be used to identify overvalued equity should interest researchers who study overvaluation and professionals who oversee management on behalf of investors.
RTWT here.

It's an interesting paper, because it uses publicly available information to identify firms with high probabilities of negative returns. While it's probably not that applicable to individual investors, I can see their approach being of use to short-sellers (or those running long-short funds).