A Non-random Walk down the Main Street: Impact of Price Trends on Trading Decisions of Individual Investors

We analyze the impact of price trends on trading decisions of more than 40, 000 households with accounts at a major discount brokerage house. Buying and selling decisions of investors in our sample are influenced by short-term (less than 3 months) price trends. By comparing the observed distributions of average trend before buys and average trend before sells with the average trend distributions (obtained using Monte Carlo simulations) when investors trade randomly, the null of non-trend motivated random trading is easily rejected (p-value < 0.002). Using Monte Carlo simulations again, we examine investor heterogeneity in trading based on prior returns and classify investors into (i) momentum buy (MB), (ii) momentum sell (MS), (iii) contrarian buy (CB) or (iv) contrarian sell (CS) category. Overall, approximately onethird of the investors are classified based on either the buy or the sell trades. A comparison of the portfolio characteristics and demographics of the identified investor segments reveal no significant differences. However, the trading characteristics of the segments show systematic differences, particularly in their response to reference points such as monthly high and low prices and in their strategies for selling losers. Contrarian buy investors are more likely to buy near monthly low prices while the contrarian sell investors tend to sell near the monthly high prices. The momentum investors do not exhibit such timing behavior. All four investor segments are reluctant to sell losers but the effect is the strongest for contrarian sell investors who expect price reversals and hence show a greater tendency to hold on to the losers. The effect is very weak for momentum sell investors who believe that a downward price trend is likely to continue and hence are more likely to realize their losses.