Investigating Herding Behavior On Toronto Stock Exchange
In this study, we employ an innovative new methodology inspired from the approach of Hwang and Salmon (2004) and based on the cross sectional dispersion of trading volume to examine the herding behavior on Toronto stock exchange. Our findings show that the herd phenomenon consists of three essential components: stationary herding which signals the existence of the phenomenon whatever the market conditions, intentional herding relative to the anticipations of the investors concerning the totality of assets, and the third component highlights that the current herding depends on the previous one which is the feedback herding.
Since the advent of behavioral finance in the 1980s, a considerable amount of research in finance has been devoted to the employment of psychological concepts in order to picture the evolution of stock prices on the grounds of various aspects of investor’s behavior. A substantial part of this research has focused upon the specific issue of herd behavior, which used to be confined traditionally within the realm of the popular Finance literature (Kindleberger, 1978; Soros, 1987; Galbraith, 1994). Academic interest on this issue has been notably intense during the last couple of decades and has led to the generation of a voluminous research output, as the reviews of Bikhchandani and Sharma (2000) and Hirshleifer and Teoh (2003) illustrate.
Herding in financial markets has been typically described as a behavioral tendency for an investor to follow the actions of others. Practitioners are interested in whether herding exists, because the reliance on collective information rather than private information may cause prices to deviate from fundamental value and present profitable trading opportunities. Herding has also attracted the attention of academic researchers, because the associated behavioral effects on stock price movements may affect their risk and return characteristics and thus have implications for asset pricing models.
Theoretical models of herding behavior have been developed by Bikhchandani, Hirshleifer and Welch (1992), Scharfstein and Stein (1990), and Devenow and Welch (1996). Empirical studies have mainly focused on detecting the existence of herding behavior among mutual fund managers (Lakonishok, Shleifer, and Vishny, 1992; Wermers, 1999) or financial analysts (Trueman, 1994; Graham, 1999; Welch, 2000; Hong, Kubik, and Solomon, 2000; Gleason and Lee, 2003; Clement and Tse, 2005).
Measuring herding empirically has proved challenging. Besides some special contexts or experimental settings, it is difficult to separate imitating behavior from clustering of trades. The empirical herding literature for the most part, therefore, uses herding as a synonym for systematic or clustered trading. Herding measures are, therefore, at best noisy proxies for imitative behavior. When herding is defined in a more general sense of clustered trading, specific forms of systematic trading patterns deriving from past returns, capital gain and loss position, and attention can also be interpreted as herding. However, when it comes to drawing conclusions on asset pricing, it is the overall clustering that is the primary concern.
The empirical study usually does not test a particular model of herding behavior described in the theoretical literature; instead, they gauge whether clustering of decisions, in purely statistical sense, is taking place in financial markets or within certain investor groups. Two streams of empirical literature have been developed to investigate the existence of herding in financial markets. The first stream analyzes the tendency of individuals or certain groups of investors to follow each other and trade an asset at the same time [Lakonishok et al. (1992) and Wermers (1995)]. These studies use the trading volume to detect herding in financial market. The second stream focuses on the market-wide herding, that is, the collective behavior of all participants towards the market views and therefore buying or selling particular asset at the same time [Christie and Huang (1995), Chang et al. (2000) and Huang and Salmon (2001, 2004, 2006)]. These measures are based on the cross-sectional dispersion of beta to detect herding toward the market index.
To improve the existent measures and to investigate the herding towards the market in major financial markets is the main purpose of our paper. There are two specific objectives to this study. Firstly, we intend to propose a new herd measure to detect the degree of herding in financial market. In constructing this measure, we take as our starting point the model of Huang and Salmon (2004), but we employ a proxy pioneered by Lakonishok, Shleifer, and Vishny (1992) which is the trading volume. Secondly, we shall apply our herd measure to detect herding behaviour in Toronto stock market. We use monthly data from January 2000 to December 2006.
This paper is divided into fore additional sections. In the second section we provide a review of the literature on the herding measurement. The third deals with methodological details and the presentation of our new measure of herding. The forth includes the data description and empirical evidence based on our new measure on Toronto stock exchange. Finally, the fifth section offers concluding remarks and discusses implications of our findings.
2. Literature review
Herd behavior is a term implying alignment to a mode of collective conduct and is expressed as a “similarity in behaviour” following the “interactive observation” of actions and payoffs (arising from those actions) among individuals (Hirshleifer and Teoh, 2003). In the stock market context, herding involves the intentional sidelining of investors’ private information in favor of the observable “consensus” (Bikhchandani and Sharma, 2000) irrespective of fundamentals (Hwang and Salmon, 2004) and the roots of such behavior can be traced to a series of factors be they of psychological or rational nature.
The most widely used herding measure is that invented by Lakonishok, Shleifer and Vishny (1992). This measure seeks to detect whether more investors are trading on either the buy or sell side of the market than would be expected if investors traded independently. Lakonishok, Shleifer and Vishny (1992) use the investment behavior of 769 U.S. tax-exempt equity funds managed by 341 different money mangers to empirically test for herd behavior. Lakonishok, Shleifer and Vishny (1992). conclude that money managers in their sample do not exhibit significant herding. There is some evidence of such behavior being relatively more prevalent in stocks of small companies compared to those of large company stocks. Lakonishok, Shleifer and Vishny (1992) explanation is that there is less public information on small stocks and hence money managers pay relatively greater attention to the actions of other players in making their own investment decisions regarding small stocks.
Wermers (1995) develops a new measure of herding that captures both the direction and intensity of trading by investors. This new measure, which he calls a portfolio- change measure (PCM) of correlated trading, overcomes the first drawback listed above. Intuitively, herding is measured by the extent to which portfolio weights assigned to the various stocks by different money managers move in the same direction. The intensity of beliefs is captured by the percent change of the fraction accounted for by a stock in a fund portfolio. Wermers (1995) finds a significant level of herding by mutual funds using the PCM measure.
Measuring the herding behavior on the basis of Lakonishok et al. (1992) has important limitations. First, this measure captures correlation in trades but does not, by itself, disentangle the determinants of herding. Second, this measure does not take in consideration whether the correlation trades results from imitation or merely reflects that traders use the same information. Finally, this measure is biased when there are limitations to short selling strategies.
Two studies that have proposed methods of detecting herding behavior using stock return data are Christie and Huang (1995) (hereafter referred to as CH) and Chang, Cheng, and Khorana (2000) (hereafter referred to as CCK). CH suggest that the investment decision-making process used by market participants depends on overall market conditions. They contend that during normal periods, rational asset pricing models predict that the dispersion in returns will increase with the absolute value of the market return, since individual investors are trading based on their own private information, which is diverse. However, during periods of extreme market movements, individuals tend to suppress their own beliefs, and their investment decisions are more likely based on the collective actions in the market. Individual stock returns under these conditions should tend to cluster around the overall market return. Thus, they argue that herding will be more prevalent during periods of market stress, which is defined as the occurrence of extreme returns on the market portfolio.
Demirer and Kutan (2006) apply the CH method to examine herding in Chinese equity markets. They use daily stock return data from 1999 to 2002 for 375 Chinese stocks and find no evidence of herding. One of the challenges associated with the approach described above is that it requires the definition of extreme returns. CH note that this definition is arbitrary, and they use values of one percent and five percent as the cutoff points to identify the upper and lower tails of the return distribution. In practice, investors may differ in their opinion as to what constitutes an extreme return, and the characteristics of the return distribution may change over time. In addition, herding behavior may occur to some extent over the entire return distribution, but become more pronounced during periods of market stress, and the CH method captures herding only during periods of extreme returns. Additional challenges arise when applying this method to Chinese stock market data because the relatively short history of these markets makes it difficult for investors to identify when extreme returns occur.
An alternative to the CH test for herding is that of Chang, Cheng, and Khorana (2000) (CCK). They examine several international stock markets, and find no evidence of herding in developed markets, such as the U.S. and Hong Kong. However, they do find evidence of herding in the emerging markets of South Korea and Taiwan. CCK note that the CH approach is a more stringent test, which requires “a far greater magnitude of non-linearity” in order to find evidence of herding.
Hwang and Salmon (2004) (hereafter HS) develop a new measure in their study of the US and South Korean markets. This model is price-based and measures herding on the basis of the cross-sectional dispersion of the factor sensitivity of assets. More specifically, HS (2004) argued that when investors are behaviourally biased, their perceptions of the risk-return relationship of assets may be distorted. If they do indeed herd towards the market consensus, then it is possible that as individual asset returns follow the direction of the market, so CAPM-betas will deviate from their equilibrium values. HS (2006) note that stock returns and herding are likely to be affected by fundamentals, at the level of the market or the individual firm. They use variables such as the dividend-price ratio, the Treasury bill rate, the term spread, and the default spread in their analysis of herding in the US, UK, and South Korean equity markets.
Need help with your literature review?
Our qualified researchers are here to help. Click on the button below to find out more:
In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: