Investor Confidence and Returns Following Large One-Day Price Changes

-- Ray R Sturm

In this study, I hypothesize that post-event price behavior following large one-day price shocks is related to pre-event price and firm-fundamental characteristics and that these characteristics proxy for investor confidence. Several theories of behavior suggest how investors form their expectations and I suggest four investor confidence hypotheses based on these theories. In addition to documenting further evidence of investor overreaction, my findings indicate that investors respond differently to negative price shocks than to positive price shocks. In particular, large decreases in price generally appear to drive positive post-event abnormal returns while large increases in price do not drive positive or negative abnormal returns. However, the main finding of this study is that this relationship is altered when pre-event return and firm characteristics are introduced. This suggests that certain pre-event characteristics influence investor confidence which in turn influences buying and selling decisions thereby driving post-event returns. However, investors' confidence appears to be lessened by a price shock effect. Many studies have investigated and successfully documented the presence of stock price overreactions in the US stock market. All of these studies detect overreactions by examining the market's response to a large price shock. The presence of a subsequent abnormal reversal in prices following such a shock has been presented as evidence to argue the overreaction hypothesis. The purpose of this study is to determine if the market's response to a price shock differs depending upon certain pre-event price and firm characteristics. In particular, I examine whether post-event abnormal returns are related to the price shock's direction, pre-event returns for the periods [-260, -10], [-135, -10] and [-30, -10] or three-year average changes in earnings per share, book value per share and debt-to-equity ratio. If so, then this suggests that these characteristics contain information about investors' confidence¹ in the market's ability or inability to provide future returns. In other words, I hypothesize that the post-event price behavior will differ based upon investors' perceptions of the market and that these perceptions will be at least in part formed by pre-event price and firm characteristics. My findings extend the current literature in five ways. First, I provide further and more current evidence that price behavior is asymmetric² and that investors do appear to overreact. Also, although others have tested the relationship between long-term (short-term) pre-event returns and long-term (short-term) post-event returns, I test the relationship between long-term pre-event returns and short-term post-event abnormal returns. In addition, certain pre-event characteristics appear to cause a different response by investors given similar price shocks, suggesting that the characteristics tested herein may be valid proxies for investor confidence. Finally, although the size of large price changes has been shown to be inversely related to the subsequent reversal, I provide evidence that the price shock is actually positively related to post-event abnormal returns thereby mitigating the reversal. I refer to this as the price shock effect. The paper is organized as follows: Section I presents the motivation and variable selection method, Section II is a review of related literature, Section III describes the data and methodology, Section IV presents the results and Section V gives a summary and conclusion.

© 2003 Lawrence Erlbaum Associates, Inc. (www.leaonline.com).This article was published earlier in Journal of Behavioral Finance, Vol. 4, No. 4, pp. 201-216. Reprinted with permission.

Investors' Activity and Trading Behavior

-- Petri Kyröläinen and Jukka Perttunen

Utilizing a comprehensive data set on daily holdings of Finnish stocks, this paper examines momentum trading and herding of active investors vs. passive investors during the Information Technology (IT) stock bubble period of 1997-2000. Modern stock markets are characterized with tremendous amount of trading. Still the market participants are highly heterogeneous in terms of their trading activity. Employing the number of trading days as a measure of trading activity, we segregate investor population into ten activity categories. After aggregating these categories into larger investor groups, our aim is to compare trading styles of these contrasting activity groups. In theoretical literature, it is often found that momentum trading and herding can potentially destabilize asset markets, therefore, our focus is on these specific trading styles. We find that particularly large active investors engage in momentum trading. Active investors as a group also tend to herd in their trading decisions. Furthermore, their herding tendency is increasing monotonically, year on year. Buy pressures of active investors are positively associated with contemporaneous daily returns, implying either price pressures caused by their trading or intraday momentum trading. Passive investors' and small active investors' trading styles, on the other hand, exhibit the contrarian fashion. Moreover, the passive investors' herding tendency is very strong over the sample period. Their buy pressures are negatively associated with contemporaneous returns. Finally, neither trading of active investors nor trading of passive investors seem to have predictive ability on returns. Our results are consistent with large active investors being contributors to the recent price bubble. Thus, active trading might not have only positive effects on the efficiency of asset markets.

© 2003 Petri Kyröläinen and Jukka Perttunen (www.ssrm.com). Reprinted with permission.

Sentiment Strategies

-- Xuewu Wang

This paper documents the profitability of the sentiment strategies. Using the aggregate closed-end fund discount as a proxy for investor sentiment, a simple sentiment strategy is constructed on the basis of the exposure of stock returns to the closed-end fund discount. The sentiment strategies buy stocks with highest exposure to closed-end fund discount and sell stocks with lowest exposure to closed-end fund discount in the past 36 months. It is shown that such a strategy can lead to an annualized profit of 11%. The source of the profitability is explored and it is found that neither market risk nor momentum anomaly can account for the profitability. However, the traditional four factor asset pricing model when augmented with an additional sentiment factor can account for the profit. This finding is interpreted as supportive evidence to the fact that the pricing of the investor sentiment risk may be a potentially useful explanation for profitability.

© 2004 Xuewu Wang (www.ssrn.com). Reprinted with permission.