The authors analyze stock transactions of individual investors to identify and explain patterns in their repurchasing of previously owned stocks. Repurchase behavior is examined at both the aggregate and individual level. Consideration is given to factors that are potentially motivating this repurchasing behavior as well as to the role of investor sophistication.

The fact that investors prefer to repurchase stocks that have been previously sold for a gain or stocks that have declined in price since being sold is evident by a statistically significant shorter time from sale until repurchase under these scenarios. This behavior may be explained by regret avoidance. Stocks responsible for prior losses are avoided because they are associated with the unpleasant emotion of regret, whereas prior winners are sought out because of the pleasant emotion of pride. The avoidance of stocks that have appreciated in value since a prior sale is associated with regret because of the missed opportunity of selling at a higher price, whereas a drop in stock price since a sale may cause investors to overestimate their ability to understand and profit from the stock. Sophisticated investors are found to be significantly less prone to this behavior.

The alternative motives tested but determined not to drive repurchase behavior include public information regarding timing of dividends, tax considerations, private information, and contrarian strategy implementation. Some repurchases fit a pattern that suggests they may be motivated by dividend dates or tax considerations, but the authors’ results demonstrate that the time to repurchase is accelerated when the stock was sold for a gain or when the stock price declined following the sale.

Trading on private information or implementing a contrarian strategy conflicts with the patterns present in the data. Trading on private information implies that the advantage of insider information should produce superior returns when, in fact, repurchased stocks are outperformed by initial stock purchases. Focusing on investors who repurchased stocks whose prices have declined since the prior sale to determine whether the behavior is part of a contrarian strategy, the authors find that the initial purchases by these investors have high market-adjusted returns prior to purchase, suggesting that they prefer momentum strategies to contrarian strategies.

This research serves as a precautionary reminder for investment advisers working with retail clients and for investors themselves to be aware of how emotions influence financial decisions. Keeping these emotions in check may improve portfolio diversification and performance.

A European online brokerage firm supplies the trading activity of 84,500 individual investors from 1999 to 2006, which yields a subset of 34,129 investors who repurchased a stock during this time period—a total of more than 6 million trades. Prices for stocks traded on the Euronext are provided by EUROFIDAI, and price data for stocks traded elsewhere are from Bloomberg.

For the purposes of this study, the prior sale must have occurred within one year in order for a buy transaction to be considered a repurchase. Historical ownership beyond one year is disregarded by assuming the trade was forgotten by the investor and irrelevant to current investment decisions. Investors are classified as sophisticated when they are in the top quintile of the number of foreign stocks held, are in the top quintile of total stocks held, or use multiple trading account types for tax benefits. Sophisticated investors in this study also have larger portfolios and a greater tendency to trade options, both of which traditionally support the claim of sophistication.

Repurchase behavior is examined at the aggregate level using repurchase rates of prior winners and losers as well as the repurchase rates of stocks that have gained or lost value since the prior sale.

At the individual level, the authors use survival analysis. The duration prior to the hazard event—in this case, the stock repurchase—is regressed on independent variables called “covariates,” and the authors estimate a Weibull distribution for the survival function. The alternative motives are tested by including dummy covariates in the regression model.

Two of the three criteria used by the authors to classify investors as being sophisticated are that they be in the top quintile among all investors in terms of the number of stocks owned and in the top quintile in terms of the average number of stocks held from different countries. Such a high standard seems somewhat exclusive. However, 47% of all investors satisfy the third criteria of having both a taxable and tax-free account.

Further research could expand the time period for determining repurchase versus initial purchase beyond stocks sold within one year. Additionally, it may be of interest to examine how the magnitude of a gain or loss on prior holdings affects repurchase behavior—the assumption being that large gains or losses may have a greater impact on behavior than small ones.