Black (1986) defines noise trading as trading on noise as if it were information. In addition, he notes the importance of noise trading in capital markets: “Noise makes financial markets possible but also makes them imperfect.” That is, in a world without noise traders, all trading is motivated by informational advantages. Recognizing they will be trading against another informed investor, traders will be reluctant to transact. Noise traders provide the necessary liquidity to financial markets. In providing liquidity, however, they also provide noise.
Why Do Noise Traders Trade?
Noise trading may arise for various reasons. Some investors may simply enjoy trading or erroneously believe they have unique information or insights. In addition, some traders may trade on “sentiment.” Shiller (1984), for example, argues that evidence from social psychology, sociology, and marketing suggests that individual investor’s decisions are likely to be influenced by “fads” or “fashion.” Alternatively, Friedman (1984) suggests that institutional investors may be more inclined to trade on sentiment, due to the close-knit nature of the investment community, the importance of performance relative to other institutional investors, and the asymmetry of incentives. Similarly, Froot et al. (1992) develop a model in which rational short-horizon investors may trade on the same signal, but the signal need not be related to fundamental value (e.g. technical analysis). Trueman (1988) suggests that institutional investors may engage in noise trading because it provides an imperfect signal to clients that the manager is informed. In sum, noise trading may result from perceived information advantages, sentiment, trading appearing in the utility function, or agency problems.
The Impact of Noise Trading on Prices
Noise trading can explain excess volatility in security prices (i.e. price will be more volatile than value), temporal patterns in stock prices (e .g. momentum and/or mean-reversion) and the use of technical analysis and positive feedback trading (see Shleifer and Summers, 1990). The magnitude of noise traders’ impact on security prices will depend on both the degree of noise trading in the market and the systematic nature of noise trading. The greater the degree of noise trading, the greater the deviation between price and value. As the deviation between price and value increases, rational arbitrageurs should work to push prices toward fundamental value. In real markets, however, arbitrage is costly (e.g. short sale proceeds are not available for investment). Moreover, in a world with noise traders and finite horizons, arbitrage can be risky. For example, rational arbitrageurs with limited horizons may be forced to unwind their positions in a period when noise traders have pushed prices even further away from fundamental values (see DeLong et al., 1990).
If noise trading is cross-sectionally independent, then the impact of noise traders on a security’s price is likely to be small relative to a world in which noise trading is cross-sectionally correlated. That is, if orders from noise traders are equally likely to be buy or sell initiated at a point in time, then many noise traders’ orders will cancel out and the impact on price should be relatively small. Alternatively, if the noise traders’ orders generally come from the same direction (i.e. primarily buy initiated or primarily sell initiated), their impact on a security’s price is likely to be large. A similar argument holds for the expected impact of noise traders on the market. If noise traders’ orders are cross-sectionally correlated across securities, then they are likely to impact market averages. That is, if noise traders systematically enter (or exit) financial markets, market averages may be affected.
Empirical evidence regarding the impact of noise traders on security prices is mixed. Lee et al. (1991) argue that the systematic noise trading of individual investors influences both closed-end fund share discounts (since individual investors play a more important role in closed-end fund shares than in the market for the underlying assets of the funds) and the prices of small capitalization securities (that are also dominated by individual investors). Although there is evidence that there is some correlation between closed-end discounts and the returns of small capitalization securities, there is considerable debate regarding the statistical and economic significance of the correlation (see Chen et al., 1993).
Alternatively, recent investigations into the behavior of institutional investors suggests that noise trading by institutional investors may impact security prices. Wermers (1994) documents results consistent with the hypothesis that some mutual funds engage in positive feedback trading and that such trading moves prices. Assuming that previous returns do not indicate future fundamental values, this suggests that some institutional investors engage in systematic noise trading.
Can Noise Traders Survive?
Historically, the impact of noise trades has be en assumed to be minimal since noise traders should lose wealth (and therefore eventually become unimportant) when trading against rational “smart-money” arbitrageurs. Shiller (1984), however, argues that there is little reason to suspect that rational smart money speculators dominate financial markets. DeLong et al. (1990, 1991) develop formal models that allow for the survival of noise traders. In DeLong et al. (1991), noise traders systematically underestimate variances of risky assets and therefore invest a greater fraction of their wealth in the risky asset than would an otherwise equally risk-averse rational investor. Their excessive risk-taking may not only allow noise traders to survive, but they may come to dominate the market. Alternatively, in DeLong et al. (1990), the actions of noise traders are cross-sectionally correlated (systematic) and influence asset prices. Like any other systematic risk, the risk impounded by the random sentiments of noise traders should be priced. Thus, noise traders may be compensated for a risk that they create. Moreover, even though the model predicts th at noise traders will lose (on average) when trading against rational arbitrageurs, noise traders may garner higher rates of returns than sophisticated investors if they concentrate their holdings in assets that have a greater sensitivity to innovations in noise trader sentim ent.
Sias et al. (1995) examine the issue of whether investors are compensated for bearing noise trader risk. Specifically, DeLong et al. (1990) suggest that assets with greater sensitivity to noise-trader risk will tend to sell below fundamental values (reflecting the pricing of noise trader risk). They suggest that such a scenario can explain the fact that most closed-end funds sell at a discount to their underlying assets (assuming individual investors are noise traders). Specifically, the discount from fundamental values reflects the additional risk from the ownership structure: closed-end fund shares are held primarily by noise traders (individual investors), but noise traders play a less import ant role in the underlying assets of the funds. Thus, under these conditions, passive closedend fund shareholders should garner larger returns than passive investors of the underlyi ng assets as compensation for bearing noise trader risk. Sias et al. (1995) demonstrate that, despite selling at discounts, (passive) closed-end fund shareholders do not garner larger returns than the holders of the underlying assets. In fact, discounts are just large enough t o cover the expenses incurred by the funds. In addition, Sias et al. (1995) demonstrate that, holding capitalization constant, NYSE stocks with greater exposure to individual investors (and presumably greater exposure to noise trader risk) earn lower returns than stocks with greater exposure to institutional investors.
Unresolved Issues
Our understanding of noise traders is small relative to their likely importance in the market. Thus, noise traders represent a substantial and promising area for future research. Some of the key questions to be answered include: Who are the noise traders? Why do they trade? Is their trading independent or systematic? What is their impact on security prices? What is the relationship between informed traders and noise traders? Finally, how can noise traders survive?
