Tuck School of Business Dartmouth College Contact Us Site Index Search Our Site
News and Events
Governance Sites

Spring 1999

The Inside Track Is Not Always Faster

Stock market trading is anonymous, so you never know if the other party in the trade is a corporate insider. Insiders often possess private information about the company's future prospects, so when trading with an insider, you're doomed to lose money since they will "buy low and sell high" at your expense. At least that's what traditional empirical research on insider trading would have you believe. But in a recent article in the Journal of Finance, Tuck Professor B. Espen Eckbo presents surprising systematic evidence to the contrary.

Global Governance
Center Research
Eckbo and his coauthor, Professor David Smith, perform a simple but powerful empirical experiment: Suppose you were allowed to replicate all insiders' trades in and out of all stocks over an extended period. If insiders truly are able to buy low and sell high, then this trading strategy should produce excess profits; that is, you should effectively "beat the market" after adjusting for risk.

Eckbo and Smith's laboratory for this experiment was the Oslo Stock Exchange (OSE), which has a reputation as an insider's market. Insider stock ownership averages close to 20 percent and stock prices are relatively volatile, both of which increase the value of inside information. The authors also studied insider transactions on the OSE during a period with lax enforcement of insider trading regulations (1985-1993). "If you should ever find evidence of excess profits to insiders, you should find it in these data," says Eckbo.

They subjected their data to the most powerful tests available in the modern theory of finance. They aggregated all insider trades into one portfolio, the insider fund, and examined whether you would have made excess profits had you invested in that fund. They eliminated the portion of the fund's return that is simply a normal reward for bearing market risk. "When we first found the insider fund did not produce turns in excess of what any investor could have expected given the risk, we were sure there was a bug in our computer programs," Eckbo recalls. "We beat the data to death, but the confession always came up the same: Outside investors would have been better off investing in the ordinary managed mutual funds on the OSE than in the insider portfolio!"

Anticipating that the academic finance profession would attribute the anomalous finding to use of the wrong model, Eckbo and Smith designed another test strategy that does not require any model specification, but nevertheless would reveal excess profits if any existed.

"Think about a strategy where you try to buy low and sell high," Eckbo says. "If you're successful, you should find that you actually increased your stock holdings in periods prior to unexpected price increases (bought low) and decreased your stock holdings just prior to unexpected price decreases (sold high). You should find a positive correlation between the weights in your stock portfolio and the portfolio's subsequent realized return. We found this correlation was plainly zero. That negates the presence of excess profits to insider trading, no matter what the risk of the insider fund."

B. Espen Eckbo is the Tuck Centennial Professor of Finance and director of the Center for Corporate Governance at the Tuck School of Business at Dartmouth. Professor Eckbo conducts research in corporate finance and capital markets, with emphasis on corporate governance, mergers and acquisitions, investment banking, portfolio management, and performance evaluation. His work has been published in the top international academic finance journals, including Journal of Financial Economics, Journal of Finance, and Review of Financial Studies. His email address is b.espen.eckbo@dartmouth.edu.

David C. Smith is associate professor of finance at the Norwegian School of Management in Oslo. This article originally appeared in Tuck Forum, IV(1), Spring 1999, published by the Office of Publications at Tuck.

<-- Back to Commentary main page