sealPurdue News

August 2000

Past performance may indeed predict future investments results

WEST LAFAYETTE, Ind. – The standard mutual fund legal disclaimer notwithstanding, past performance may indeed predict future results, according to a study by two finance professors.

"The same professional investors and fund managers who sold their high-technology stocks when a judge ruled against Microsoft, which they knew was in the offing anyway, consistently underreact to important corporate events that have long-term effects on companies' financial future and investors' returns," says Raghu Rau, an assistant professor of management at Purdue University's Krannert Graduate School of Management.

Rau, whose area of research is corporate finance, is interested in the question of the stock markets' efficiency. The efficient-market hypothesis holds that over the long term, the sum total of stock buying and selling will determine the precisely correct value of stock shares. It's on the way to getting to that correct value that individual investors, whether they're Wall Street bankers, mutual fund managers and, ultimately, Main Street 401k participants make — and lose — money.

But what if markets aren't informationally efficient, ask Rau and Padma Kadiyala, an assistant professor of finance at the Cox School of Business at Southern Methodist University, in a recent study they undertook.

Here is the market dynamic they developed.

ABC Corp., rolling along with good sales, profits and market share, announces that it will earn 35 cents a share rather than the 30 cents analysts predicted. Investors should buy more shares. But they tend not to do that.

XYZ Corp., on the other hand, has seen things go badly in the past few years and announces it will lose 35 cents a share rather than the 30 cents analysts predicted. Investor should sell their shares. But they tend not to do that.

These are investor underreactions.

The professors came to their conclusions by analyzing four types of firms that announced major financial initiatives and comparing them to a benchmark portfolio of companies of similar size, financial fundamentals and stock movement (momentum).

The bottom line is that John Q. Market is overcorrecting when he should be holding steady to his financial course and not paying enough attention to the real blue skies and storms on the long-term investment horizon, Rau and Kadiyala concluded.

"There are policy implications for investors, whether they are on Wall Street or Main Street," Rau says. "Look at the past history of a company's stock. History tends to repeat itself. Past performance tends to be a good predictor of future performance."

So the research indicates that investors should buy shares of good companies when they announce good news and sell shares of bad companies when they announce bad news. And if investors follow the professors' advice, they may not validate the efficient market hypothesis, but they should be more successful, which is even more important.

Sources: Raghu Rau, (765) 494- 49658,

Padma Kadiyala, (214) 768-3155,

Writer: Mike Lillich, (765) 494-2077,

Purdue News Service: (765) 494-2096;

The Underreaction Phenomenon and the Long-Run Return Performance of Firms Conducting Corporate Events
Padma Kadiyala and P. Raghanvendra Rau

We examine the long-run abnormal returns to firms announcing three stock-based corporate events (seasoned equity offerings, share repurchases and stock-financed acquisitions) and one cash-based event (cash-financed acquisitions). We find that the long-run abnormal returns to firms announcing each of the four events can be explained by investor tendency to underreact to new information. Firms announcing corporate events after negative prior information earn significantly lower post-event abnormal returns than firms announcing these events after positive prior information released about the firm. We also document negative (positive) returns at earnings announcements for up to three years after the original event announcement for those firms that offer a negative (positive) earnings surprise. Our study suggests that earnings, management activities and managerial timing are not likely explanations for the long-run abnormal returns to corporate events.

* To the Purdue News and Photos Page