In this paper we test the efficiency hypothesis in financial market. A market is called efficient if the price variations "fully reflect" relevant information, i.e. a speculator cannot make a profit out of it. A currency exchange market is a natural candidate to check efficiency because of its high liquidity. We perform a statistical study of weak efficiency in Deutschemark/US dollar exchange rates using high frequency data. In the weak form of efficiency the information can only come from historical prices. The presence of correlations in the returns sequence implies the possibility of a statistical prevision of market behavior. We show the existence of correlations by means two statistical tools. A first analysis has been performed using structure functions. This approach gives an indication on the returns distributions at different lags \tau. We have also computed the generalized correlation functions of the return absolute values; roughly speaking this is a test of the independence of the fluctuations of fixed size. In both cases we have obtained a clear evidence of long term return anomalies. This implies a failure of the usual "random walk" model of the returns; nevertheless the presence of long term correlations does not directly imply the fault of the weak efficiency hypothesis : it is not obvious how to use time correlation to make a profit in a realistic investment. Then we show how this information is relevant for a speculator. First we introduce a measure of the available information relevant from a financial point of view, with a technique which reminds the Kolmogorov e-entropy. Second in the case of no transaction costs, we propose a simple investment strategy which leads to an exponential growth rate of the capital related to the available information. We have performed two kind of information analysis in the return series. We show that the available information is practically zero if the speculator wants to change his portfolio systematically after a fixed lag \tau : for him the market is efficient. Instead, a finite available information is observed by a patient investor who cares only of fluctuation of given size \Delta. This is the first case, as far as we know, in which the available information obtained by a suitable data analysis is directly linked to the possible earnings of a speculator who follows a particular trading rule

Weak efficiency and information in foreign exchange markets

SERVA, Maurizio;
1999-01-01

Abstract

In this paper we test the efficiency hypothesis in financial market. A market is called efficient if the price variations "fully reflect" relevant information, i.e. a speculator cannot make a profit out of it. A currency exchange market is a natural candidate to check efficiency because of its high liquidity. We perform a statistical study of weak efficiency in Deutschemark/US dollar exchange rates using high frequency data. In the weak form of efficiency the information can only come from historical prices. The presence of correlations in the returns sequence implies the possibility of a statistical prevision of market behavior. We show the existence of correlations by means two statistical tools. A first analysis has been performed using structure functions. This approach gives an indication on the returns distributions at different lags \tau. We have also computed the generalized correlation functions of the return absolute values; roughly speaking this is a test of the independence of the fluctuations of fixed size. In both cases we have obtained a clear evidence of long term return anomalies. This implies a failure of the usual "random walk" model of the returns; nevertheless the presence of long term correlations does not directly imply the fault of the weak efficiency hypothesis : it is not obvious how to use time correlation to make a profit in a realistic investment. Then we show how this information is relevant for a speculator. First we introduce a measure of the available information relevant from a financial point of view, with a technique which reminds the Kolmogorov e-entropy. Second in the case of no transaction costs, we propose a simple investment strategy which leads to an exponential growth rate of the capital related to the available information. We have performed two kind of information analysis in the return series. We show that the available information is practically zero if the speculator wants to change his portfolio systematically after a fixed lag \tau : for him the market is efficient. Instead, a finite available information is observed by a patient investor who cares only of fluctuation of given size \Delta. This is the first case, as far as we know, in which the available information obtained by a suitable data analysis is directly linked to the possible earnings of a speculator who follows a particular trading rule
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11697/41076
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact