Print
E-MailDavid Hirshleifer is the Robert G. Kirby Chair in Behavioral Finance and Professor of Finance and Business Economics at the University of Southern California Marshall School of Business; Andrew W. Lo is the Charles E. and Susan T. Harris Professor at MIT Sloan School of Management; and Ruixun Zhang is Assistant Professor and Boya Young Fellow at Peking University School of Mathematical Sciences, Center for Statistical Science, and Laboratory for Mathematical Economics and Quantitative Finance. This post is based on their recent paper.
There is evidence of social contagion of investment behavior in financial markets that does not derive from rational information processing. Given recent developments in information technology and the growth of social networks, it is important to incorporate the influence of contagion via social interactions when studying economic and financial behavior. To better capture these dynamics, the Efficient Markets Hypothesis can be complemented by the Darwinian perspective of natural selection on investment strategies.
In a recent working paper, we model bias in the transmission of ideas among investors to analyze the evolutionary consequences of competing investment styles. We consider a market in which each investor has a propensity to invest in one of two investment styles. We refer to this propensity as the investor’s investment philosophy. Investors with higher realized returns produce more “offspring” in the next period of the model by transmitting their ideas to other investors via social interaction. Selection results in differential survival of investors’ behavioral traits, i.e., their investment philosophies.
A basic stylized fact about modern financial markets is that numerous competing investment styles coexist. Examples include value versus growth, momentum versus contrarianism, large-cap stocks versus small-cap stocks, diversification versus stock-picking, domestic versus global, technical versus fundamental, and so on. The rapidly evolving hedge fund sector is a good example of an industry with widely varying investment styles. We show that the survival of diversity is a consequence of general principles of evolution in the face of risk.
Consistent with this evidence, a key finding of our model is that heterogeneous investment styles are able to coexist in the long run, implying the survival of a more diverse set of strategies than occurs in traditional portfolio theories. As a simple benchmark for comparison, under the Capital Asset Pricing Model (CAPM), all investors hold the market portfolio, meaning that each investor holds a fraction of all available assets in proportion to their outstanding supply. So in this popular theoretical framework, all investors adopt the same investment strategy. The same lack of diversity exists with more sophisticated asset pricing models such as the Intertemporal CAPM, in which all investors hold the market and the same set of hedging portfolios, usually presumed to be small in number, implying only a very limited amount of diversity. In contrast to such homogeneity, our approach allows for considerably more variety and offers a possible explanation for the richness of investment strategies pursued by different stakeholders in actual capital markets.
We characterize the survival and popularity of styles in relation to the statistical distribution of security returns. These results lead to several testable implications. Under the CAPM, the quality of an investment style is often measured by its “alpha” with respect to the market, defined as the excess return above the market’s return at a given level of risk. This suggests that high-alpha strategies may tend to survive (at least to the extent that alpha persists over time). However, we find that the survival of an investment style is determined by several features, including its expected return, beta, and volatility. When determining a strategy’s survival with respect to these return characteristics, an investment style’s beta-scaled expected gross return—defined as the expected gross return of a style divided by its beta—plays a critical role. We call this return its scaled alpha.
Scaled alpha plays two roles in our model. The first is in determining the relationship between an investment style’s beta and its popularity and future survival. This relationship turns out to be non-monotonic. In particular, an investment style with low beta is promoted in market evolution only when its scaled alpha is comparable to that of the alternative style. In contrast, when a style has a much higher scaled alpha than its alternative, high beta can promote its popularity. This result implies that a style’s scaled alpha, not the traditional CAPM alpha, is a key determinant of the popularity of low-beta investment styles in a given population of investors.
The second role of scaled alpha is in determining a non-monotonic relationship between market volatility and the popularity and survival of an investment style. In particular, high market volatility promotes investment styles with high scaled alphas, and is opposed to investment styles with low scaled alphas. A high scaled alpha can therefore be understood as a defensive characteristic of an investment style, in the sense that investors will tend to allocate to styles with high scaled alpha in volatile markets. This can be empirically tested by examining shifts in investment style such as value versus growth, momentum versus contrarian, or fundamental versus quantitative as a function of market volatility.
In an extension of our model, we allow for important psychological forces that affect investor receptiveness toward the investment philosophies of others. The first is conformist transmission, the phenomenon that investors view others as being well-informed and therefore follow the choices of these others. We show that conformist preference slows down evolutionary convergence, potentially leading to oscillations and bubbles in certain financial environments.
The second psychological force we investigate is attention to novelty, the phenomenon that investors are more likely to pay attention to a novel investment philosophy if it is sufficiently different from the most popular philosophies. Attention to novelty acts in opposition to conformist preference, resulting in an even higher degree of diversity among investment philosophies in the long run. We also propose potential empirical tests for the survival of investment philosophies in relation to different proxies for attention in the empirical finance literature.
Our evolutionary model provides a systematic framework for understanding the general multi-period dynamics of the social contagion of competing investment styles. To explore additional implications of social contagion for investment styles and investor behaviors, our model can be extended to include resource constraints (which may generate strategic interactions), and autocorrelation in environments (which may generate intelligent behaviors with memory). Another possible extension is to model overlapping investors making decisions at different frequencies, such as high-frequency traders versus medium-term pension funds versus long-term private-equity investors. The existence of investors at different horizons can be a source of stock market bubbles and busts. Our model, and more generally, the evolutionary finance approach, offers a compelling set of explanations for the impact of social contagion on investor behavior, asset prices, and market efficiency.
The complete paper is available for download here.
