New research by CKGSB Professor Neng Wang, published in The Journal of Finance, sheds light on one of the most puzzling behaviors in financial markets.
Why do investors often sell winning stocks too early but hold on to losing stocks for too long? This well-known phenomenon—called the disposition effect—has puzzled economists and investors for decades. New research by CKGSB Dean’s Distinguished Chair Professor of Finance and Senior Associate Dean Neng Wang and his co-authors Min Dai (Hong Kong Polytechnic University) and Cong Qin (Shanghai University of Finance and Economics) provides fresh insights into why this behavior occurs and when investors finally decide to sell losing stocks.
Their paper, Dynamic Trading with Realization Utility, recently published in The Journal of Finance, develops a dynamic model of “realization utility”, in which investors has two layers of mental accounts that are interconnected but serve different purposes in their trading decisions. It shows that investor decisions are shaped not only by gains and losses on individual stocks, but also by how investors mentally organize their portfolios and manage risk over time.
The research begins from a simple observation: investors often evaluate each stock separately rather than thinking about their portfolios as a whole. At the same time, they still keep track of their overall investment budget.
Professor Wang explains that an investor effectively maintains two mental accounts. “For each investment episode, he uses a stock-level mental account to evaluate his utility burst based on the realized gains/losses,” noted Professor Wang. “In addition, he has a dynamic mental account that aggregates all of the individual stock investment episodes via a stochastically evolving budget at the trading account level.”
This dual accounting system plays a crucial role in determining trading behavior. Individual trades trigger emotional reactions to gains and losses, while the broader mental budget shapes risk-taking and future decisions.
One of the most robust findings in finance is that investors tend to sell winning stocks but keep losing ones. But when they choose to sell their losing stocks has rarely been studied.
As Professor Wang and his co-authors note, “One of the most robust findings about individual investors’ trading behaviors is the disposition effect… whereby an investor has greater propensity to sell a stock that has gone up in value since purchase than one that has gone down.”
The new research shows that this behavior depends heavily on how investors allocate their investment budgets between stocks and safe assets such as cash or bonds. The authors find that investors are more likely to sell their losing stocks when they set aside sufficient funds in risk-free accounts during downward stock-price jumps.
The model predicts two common patterns in real markets.
First, investors with sufficient savings may sell even severely losing stocks. “With sufficient savings, voluntarily selling a stock at a deep loss is a unique prediction of our model,” explain the authors.
Second, investors without enough financial cushion tend to wait until the stock recovers somewhat before selling. “Retail investors often sell their losing stocks after these stocks rebound a bit,” notes the study.
This explains why many investors hold declining stocks for long periods and only sell after partial recoveries—a pattern widely observed in brokerage data.
A key innovation of the study is showing how the use of leverage affects investors’ trading strategies.
According to Professor Neng Wang, “By using savings, the investor spreads his stock trades out over time… and becomes less sensitive to losses in his stock position.”
Investors who keep part of their funds in safer assets tend to trade more frequently and are more willing to realize losses. In contrast, “By using leverage the investor increases his trading size beyond his budget… [and] is more reluctant to realize losses, increasing the disposition effect,” explains Professor Wang.
This finding has practical implications for brokerage firms and regulators. Margin investors who borrow money to invest may be especially prone to holding losing stocks too long.
Interestingly, leverage limits can actually improve decision-making. According to the paper, “Binding leverage constraints make the investor realize losses sooner, mitigating his disposition effect.”
The research also predicts that investors will favor stocks with either very high or very low volatility.
As Professor Neng Wang explains, “Our model predicts that investors prefer stocks with either high or low volatility over stocks with intermediate volatility.”
High-volatility stocks encourage investors to keep more savings on the side, while low-volatility stocks encourage borrowing and larger positions.
The findings offer several important lessons for investors and policymakers.
First, investor behavior cannot be understood by looking at individual trades alone. Portfolio-level decisions and mental budgeting strongly influence trading behavior.
Second, leverage can significantly increase the tendency to hold losing investments, suggesting that risk management policies matter for investor outcomes.
Third, selling decisions are often driven by psychological reference points rather than purely rational calculations.
By combining psychological insights with rigorous financial modeling, Professor Wang and his co-authors provide a deeper understanding of investor behavior. Their research helps explain not only why investors systematically hold losing stocks, but also how savings, leverage, and market rules can influence those decisions.
