For a trader, the lesson is simple – asset pricing models can help us understand how markets work, but where the model relies upon simplifying assumptions, the predictions from the models might not always be accurate. A good trader should understand both the predictions of the model and the limitations of the model. Without both, a trader will be vulnerable – even if this vulnerability takes years to be revealed.
Informed traders versus noise traders
A noise trader is simply a trader who holds no new information about a security. Any knowledge upon which he trades is assumed to be already imputed in the security’s price. Given this definition, it might seem that noise traders would be largely irrelevant to the functioning of markets. However, Gemmill and Thomas (2002) argue that the setting of prices in a market is determined through the interactions of arbitrageurs and noise traders. Furthermore, many models for understanding how security prices are set are based on the notion that a market comprises informed traders (those who know the fair value of a security) and noise traders (those who do not know).
Who are these noise traders?
Although rarely made explicit, noise traders are implicitly assumed to include non-professional traders (e.g. retail investors) – even though it is likely that at least some retail investors have better investing track records than some professionals. Noise traders might also include traders forced to trade because of a need for liquidity. Dow and Gorton (2006) argue that “noise traders play an important role in modern finance theory”, but state that their “identities, motivations and ability to persist” are not well understood.
In other words, we do not know much about a group of people that we believe plays an important role in the workings of markets. This is quite some confession!
Noise traders can have both benign and adverse effects on markets. Black (1986) argues that with more noise trading, markets will be more liquid, in the sense of having frequent trades that allow prices to be observed. However, security prices will reflect both the information upon which information-traders trade, but also the noise upon which noise-traders trade.
As noise trading increases, information trading becomes more profitable, because of the greater noise contained in prices. However, apparent ‘information’ may already be reflected in security prices, making it difficult to differentiate information from noise. Noise can create the opportunity for profitable trading, but simultaneously makes it difficult to trade profitably.
Even without short-sale constraints, the existence of noise trading means that securities need not be rationally priced, and arbitrage becomes risky. Information can give a trader an edge, but not a guaranteed profit. Consequently, informed traders will not take large enough (i.e. risky enough) positions to eliminate the noise.
Black surmises that it will be difficult to show that information-traders perform better than noise-traders, and argues:
there will always be a lot of ambiguity about who is an information trader and who is a noise trader.
Noise traders, through their uninformed trades, can set up mis-pricing opportunities for better-informed traders to exploit. But noise traders can also overwhelm informed traders, if their scale is large and they trade in a similar fashion to one another.
Noise traders should thus be applauded for creating opportunities for traders, but also feared when they move as a pack. Superior knowledge alone is not enough to guarantee success as a trader. We know from financial history that even well-informed arbitrageurs can be quite vulnerable.
Why smart arbitrageurs don’t always win…
Shleifer and Vishny (1997) describe one of the ways in which a well-informed arbitrageur can fail. A textbook description of arbitrage suggests that the process requires no capital, entails no risk and generates guaranteed and immediate profits. This kind of arbitrage would bring prices towards equilibrium and keep markets efficient. However, the authors argue that:
the textbook description does not describe realistic arbitrage trades and, moreover, the discrepancies become particularly important when arbitrageurs manage other people’s money.
Types of arbitrage that appear to be simple, such as that between two similar bond futures contracts traded on different exchanges, can take on the characteristics of risk arbitrage when considered fully. Even mechanically hedged arbitrage positions, such as long stock/short future, can result in financial distress if the arbitrageur earns paper profits on the stock leg but is unable to meet the cash requirements arising from losses on the futures leg. Risk arbitrage bears risk of loss and requires capital – an important distinction from the textbook definition of arbitrage.
The role of clients
Furthermore, the model of arbitrage assumed in many popular asset pricing models is inconsistent with how arbitrage is practised in financial markets. Instead of vast numbers of small arbitrageurs, arbitrage is conducted in practice by relatively few specialised professionals, who generally use outsiders’ money to take large positions. An agency relationship thus exists between the specialised arbitrageurs and their clients. Where a prospective client seeks to place money with a hedge fund but has a limited knowledge or experience of arbitrage, he might simply allocate capital to those funds with the strongest track records. Consequently, the size of funds under management becomes related to the past performance of the arbitrageur.
This dynamic can generate some interesting outcomes for markets. As an illustration, assume the existence of noise traders, so that securities need not be always rationally priced. Idiosyncratic risk (risk that cannot be hedged) can deter arbitrageurs. Consequently, securities with idiosyncratic risk can remain mis-priced for some time. With the existence of noise traders, arbitrage positions can widen and the arbitrageur loses money. Some clients might react to these losses by seeking to redeem their fund assets. However, if we assume that any market mis-pricings will eventually be corrected, the expected returns from arbitrage positions are high exactly when past returns are low. Thus, arbitrageurs can be forced to close positions that offer high expected returns, exacerbating deviations from equilibrium.
The poor performance of many classes of hedge fund during 2008 was followed by large client redemptions, and presumably the closure of some attractive arbitrage positions. In so far as this created deviations from equilibrium, those traders with capital to deploy and an ability to spot the mis-pricings would find such a trading environment very fertile. For those risk-arbitrageurs who suffered redemptions, the need to liquidate attractive positions must have been a galling experience.
Where the price of a security moves far away from an estimate of its fundamental value, one might expect it to revert at some future point. But simply identifying a mis-priced security is not enough. It could remain over-priced for some time or the mis-pricing could even grow, resulting in losses and ultimately redemptions for the arbitrageur. The path the security price takes is important, because some market players might be unable to hold onto positions that produce losses.
Hedge funds attempt to mitigate the risk of clients redeeming in response to losses by using devices such as ‘lock-in periods’ and ‘gates’ that impose contractual restrictions on clients seeking to withdraw funds. However, potential clients might fear being locked in to a poorly performing fund and so it could be more difficult to promote and market this type of fund. Only managers with strong track records are likely to be able to persuade clients to accept lengthy lock-in periods.
Educating clients about the need to hold on to attractive positions after losses is another important, albeit time-intensive, initiative to minimise redemption risk. Arguably, the best time to do this is when returns have been strong and the client can understand the principle of holding on to attractive positions that have experienced near-term losses. If education is left to the last minute, when the losses start appearing, there is a higher risk that the client’s