Empirical Study of Weak form Efficient Market Hypothesis
Fama[1970］ says a market in which prices always ”fully reflect” available information is
called ”efficient”, that is, successive price changes (or more usually, successive one-period
returns) are independent. In other words, if the flow of information is unimpeded and information
is immediately reflected in stock prices, then tomorrows’s price change will reflect
only to tomorrow’s news and will be independent of the price changes today. But news is, by
definition, unpredictable and, thus, resulting price changes must be unpredictable and random.
Weak form Efficiency Market Hypothesis is one of the type of Efficient Market Hypothesis
and implies that it is impossible to predict future price changes by using information about
historical changes. This study examines whether or not this hypothesis holds. This study
examines the efficiency in a foreign exchange market, using linear regressive model. When
using time series data like exchange rate, whether or not stationarity of the data must be
confirmed. Henceforth in chapter 2, I confirm the stationarity of the process I use. In chapter
3, basing on chapter 2, I estimate regressive model. In chapter 4 and 5, tests relating to estimated
parameters of the regressive model are done, and I examine the efficiency in a foreign
exchange market. In chapter 5, structual change is confirmed.
2 Stationarity of process
2.1 Unit root test
It is very important to confirm stationarity when using time series data like in this study.The
reason for which we confirm stationarity is we occasionally obtain statistically meaningful t
value even if the model is completely meaningless. Stationarity which this paper implies is
especially weak stationarity, which satisfies the properties as follows:
Let Xt be stochastic variable at point t; and Xt satisfies
E(Xt) = ¹ (1)
var(Xt) = ¾2 (2)
cov(Xt;Xt¡s) = °s (3)
That is, equation (1) implies mean is constant, equation (2) implies variance is constant and
equation (3) implies auto-covariance depends on only time lag s, and are uncorrelated to time.
If exchange rate is random walk, then it does not satisfy stationarity condition as above. But
if, differeciating one order lag, the process is sational, it is said that the process has one unit
root. This study use DF(Dicky Fuller)test to confirm unit root.
Empirical Study of Weak form Eﬃcient Market Hypothesis
Fama[1970］says a market in which prices always ”fully reﬂect” available information
is called ”eﬃcient”, that is, successive price changes (or more usually, successive one-
period returns) are independent. In other words, if the ﬂow of information is unimpeded
and information is immediately reﬂected in stock prices, then tomorrows’s price change
will reﬂect only to tomorrow’s news and will be independent of the price changes today.