Chapter 9 • The secondary capital market and its efficiency
by the price increasing by a small but significant amount. It seems that investors
who exploit the tendency to overreact, by, for example, buying shares immediately
following some ‘bad’ news, can make abnormally large returns as the overreaction is
corrected (see, for example, Dissanaike 1997 and 1999).
There is also evidence of a ‘weekend’ or ‘Monday’ effect, where there are signi-
ficantly higher returns from either buying shares on a Monday morning and selling
them on a Monday evening, or selling them in the morning and buying them back in
the evening, than the normal expected returns from the shares over one day (see, for
example, Mehdian and Perry 2001, Sun and Tong 2002, and Brusa, Liu and Schulman
2003 and 2005). The problem with generating abnormal returns from exploiting this
apparent weakness is that the effect seems inconsistent. Over some periods (and for
the shares of some sizes of business), returns are positive, but over other periods (and
for different sizes of business) they are negative. Although the Monday effect seems to
be an anomaly, it falls short of providing evidence of a lack of market efficiency. Picou
(2006) showed similar anomalies around holiday periods.
Fama (1998) makes the point that some of these apparent anomalies may result
from using a ‘bad model’. Where researchers are saying that abnormal returns result
from a particular investment technique, they are comparing returns from using that
technique with those that would be expectedfor the particular securities concerned.
Typically, the capital asset pricing model (CAPM) would be used to determine these
expected returns. As we discussed in Chapter 7, Fama and French (2004) showed that
CAPM is a flawed model.Conclusion on weak-form efficiency
The broad conclusion on weak-form tests must be that the evidence on capital mar-
kets, including the LSE, is consistent with weak-form efficiency. While there might be
minor inefficiencies, they are generally not of any economic significance since they
cease to exist when dealing charges are considered.
It is particularly important to note that randomness does not mean that prices are
set irrationally. On the contrary, since new information becomes available randomly,
its reflection in security prices should also be random if the market is efficient. After
a particular price movement it may well be possible to explain, by reference to real
events, why the movement took place. Randomness should not be confused with arbi-
trariness here. Prices moving in trends and repeating past patterns would point to
available information not being fully reflected in those prices, that is, to inefficiency
and arbitrary pricing.Tests of semi-strong-form efficiency
Tests of semi-strong-form efficiency have centred on questions of whether new infor-
mation, which could reasonably be expected to affect a security’s price, actually does
so, in the expected direction, by the expected amount, and with the expected rapidity.
A fertile area for testing to see whether security prices react rationally to new infor-
mation is where some action of management might superficially seem to indicate
something that, on closer examination, is not the case. If security prices seem to reflect
the superficial view of the action, and not the rational one, it would imply that the
market was not efficient (in the semi-strong form) owing to the naïvety of investors. In
other words, it would imply that the managers would be able to fool the investors by
window-dressing activities.