|
In a traditional framework, traders are rational and no frictions exist. As a
consequence, markets are efficient in the aggregation level. The price of a financial instrument
reflects all the information currently available. Only new information generates change in
asset prices. Since news is unpredictable, it is impossible to earn abnormal capital gains on
the basis of market information. For example, we may find auto correlations in stock
returns, but taking into account trading costs there is no evidence of profitable
trading opportunities. Also technical or fundamental analysis would not enable an investor
to achieve excess returns. If some strategies give excess returns with respect to the
buy-and-hold strategy, it is only by chance or by more risk taken. Also in efficient markets,
asset prices reflect their fundamental value, i.e., the discounted sum of future dividend if
the transversality condition is satisfied (LeRoy, 1989). Any deviation from this value
must trigger an immediate reaction from rational traders and the rapid disappearance
of mispricing. Empirical studies have shown some facts contrasting with the implications
of the efficient market theory. Predictability with economic gain has been identified
in longer-horizon stock returns. De Bondt and Thaler (1985) show reversal effects
while (Jegadeesh and Titman, 1993) illustrate momentum effects. On the other hand,
asset prices can deviate from no arbitrage values. Some stocks were overvalued or
undervalued in the past. An often cited example of overpricing is the internet stocks (Ofek
and Richardson, 2003). These anomalies sometimes persist in financial markets
before disappearing. The fundamental question is how can we explain the presence
and persistence of these anomalies. The first explanation supposes that all market
participants are not fully rational. For example, there are participants who use their belief or
sentiment to evaluate an asset instead of using only relevant information. These participants
are called noise traders. The others who use only relevant information are called
rational traders or arbitrageurs. The presence of noise traders can sometimes influence the
asset prices. A well-known objection to this point of view (see Friedman, 1953), is that
even if noise traders were to create such deviations, it would be for a small time. Indeed,
noise traders often buy overvalued and sell undervalued assets. So, they will make
less money and die out. They can also become by experience smart investors so that for a
bit, only rational agents remain. However, this argument is not shared by all. Figlewski
(1978) underlines that the process tending to lead the irrational towards failure may persist
before being realized. Even if a noise trader buys overvalued assets, by chance the realization
of the first asset returns may be higher than the average in the short-term. So, there is
not a systematic loss for this agent who continues to be influent in the market. This form
of risk is known as fundamental risk. Another incurred risk for arbitrageurs is
the unpredictability of the noise traders' behavior. The sentiment or belief that they have,
may diminish or augment in the short-term. This fact constitutes also a risk for an arbitrageur called Noise Trader Risk. More information on this concept may be found in Shleifer
and Summers (1990) work. Other facts which can limit arbitrage are synchronization risk
and transaction costs (Tuckman and Vila, 1992; and Abreu and Brunnermeier, 2002). All
concepts mentioned above give no riskless arbitrage so that arbitrageurs are reticent to
exploit them. Their risk is also accentuated by the fact that if their loss attains a certain
level, one can demand them to liquidate their portfolios in the case where they are not
owners. In general, it is the case as Shleifer and Vishny (1997) underlined it. To sum up,
anomalies are interpreted in this approach by noise traders' effect and limits to arbitrage which
avoid arbitrageurs to correct the mis-evaluation.
|
