Earning surprises and information asymmetry

Research by Stephen Brown

Information asymmetry is one of the factors that drives the spread between the buying and selling price of a stock and pushes up the cost of capital for firms, so there is real economic significance to understanding what drives the activity of privately informed investors and uninformed investors. Meeting or beating analyst’s earnings forecasts has also become important, not just to firms but to investors whose behavior is driven by them. But while it is clear that earnings benchmarks are important to investors, and that earnings surprises have a short-term effect on investor’s pricing associations and short-term trading reactions, little has been known about how earnings surprises affected a stock’s investment visibility and trading activity over the long term.

Stephen Brown, assistant professor of accounting and information assurance, examined the effect of earnings surprises on information asymmetry between informed and uninformed investors. With co-authors Stephen A. Hillegeist, INSEAD, and Kin Lo, University of British Columbia, Brown conducted time-series tests of the association between earnings surprises and the change in information asymmetry before and after the earnings announcement.

Brown and his co-authors considered the trading activity level of informed investors, those with some kind of private knowledge that would influence their trading decisions, and uninformed investors, i.e. those who rely solely on publicly available information or who trade for liquidity reasons. This latter group includes both individual investors and certain professional investors such as index funds, that maintain balanced portfolios.

They found that information asymmetry decreased in the quarter following a positive earnings surprise, when a firm beat earnings forecasts, and increased during the quarter following a negative earnings surprise, when firms failed to meet an earnings forecast.

Earnings surprises attract media attention, and that extra attention raises awareness of the firms in the minds of investors. When the surprise is positive, there is increased trading activity in the stock. Brown expects this is because newly-aware investors believe that a firm that beat its earnings forecasts represent an attractive buying opportunity.

There are thousands of stocks for investors to choose from, so publicity and media attention brings a stock to the awareness and attention of uninformed investors. Good attention in the form of beating earnings expectations result in a decrease in information asymmetry. Why?

“Beating expectations stop the private information events from occurring,” says Brown. “Firms that are doing well are generally happy to share that good news. You always hear about people getting engaged before they actually get married, but you rarely hear about people getting divorced until it has already happened. We humans are eager to share good news and reluctant to share bad news.”

The increase in uninformed trading provides additional camouflage for informed trading, because it is harder for the market maker to tell that any particular trade is based on private information. After a positive surprise, the level of informed trading appears to increase commensurately with the level of uninformed trading on those days when informed traders have private information. However, there are fewer such days on which they have such information and so the overall level of information asymmetry is reduced.

Negative earnings surprises cause a reduction in the level of trading by both uninformed and informed traders, despite the fact that the surprise may garner just as much media attention. This may be because uninformed investors, particularly individual investors, aren’t in a position to sell the stock short. However, the reduction in trading by uninformed investors is disproportionately greater than the reduction in trading by the informed and the overall level of information asymmetry increases.

A company that doesn’t meet its earnings forecasts in a big way, particularly when that follows a string of positive earnings surprises, is more likely to cause greater media attention, which will then cause greater asymmetry.

These changes in information asymmetry persisted for at least a year following the initial earnings surprise, whether it was positive or negative.

While it is commonly held belief that investors place importance on earnings benchmarks, this study is valuable because it provides evidence for what investors actually do, rather than what executives believe investors do.

“The effect of earnings surprises on information asymmetry,” is forthcoming from theJournal of Accounting and Economics. For more information about this research, contact sbrown@rhsmith.umd.edu.