Moving average plays an important role in price dynamics. Future price is influenced by the current di erence between logarithmic price and its moving average. This di erence tends both to reduce and to enlarge, according to the examined market. The first open question is: this feature is a strict peculiarity of the market, or does it depends on the time frequency of data? In our cases, both daily datasets exhibit a repulsive moving average, at di erence with USD-DM high frequency dataset.
The answer could be that at di erent time scales, di erent moving average e ects are active, possibly attractive on short scales and repulsive on larger scales. The problem deserves further investigations. Another interesting point is to investigate if the moving average action is generated by a self-organized mechanism of traders reactions [19,20]. In other words, is it possible that traders, taking into account informations given by moving averages, make collectively induced financial choices producing, as a result, the observed phenomena? In our price dynamics model a random component is also present, which we do not have deeply investigated, being this out of the scope of this paper.
Nevertheless, our picture could help to determine the exact shape of the noise, since, in principle, we are now able to filter the deterministic contribution. Finally, in the light of our model, we have found a trading strategy that widely overcomes the intrinsic performance of the examined datasets. In this way, we have given a clear evidence that some inefficiency is present in financial markets. Next challenge is to find out if part of this inefficiency survives when applied by a real speculator, which lives in a real financial world, where transaction costs, unfortunately, are not omitted.
Acknowledgments
We thank Dietrich Stau er for many useful discussions and for a critical reading of the manuscript.
REFERENCES
[1] M.J. Pring, Technical Analysis Explained (McGraw-Hill, 1985).
[2] T.A. Meyers, The Technical Analysis Course (Probus Publishing, 1989).
[3] M. Ratner, R.P. Leal, Journal of Banking and Finance 23, 1887 (1999).
[4] K. Ojah, D. Karemera, The Financial Review 34, 57 (1999).
[5] D. Meyers, Futures 29, 60 (2000).
[6] A.M.R. Taylor, Oxford Bulletin of Economics and Statistics 62, 293 (2000).
[7] F.R. Johnston, J.E. Boyland, M. Meadows, E. Shale, The Journal of the Operational Research Society 50, 1267 (1999).
[8] R. Mantegna, H.E. Stanley, Nature 383, 587 (1996).
[9] R. Mantegna, H.E. Stanley, An Introduction to Econophysics: Correlations and Complexity in Finance (Cambridge University Press, 1999).
[10] M. Pasquini, M. Serva, Economics Letters 65, 275 (1999).
[11] M. Pasquini, M. Serva, Eur. Phys. J. B 16, 195 (2000).
[12] P.K. Clark, Econometrica 41, 135 (1973).
[13] B.B. Mandelbrot, J. Business 36, 394 (1963).
[14] R. Mantegna and H.E. Stanley, Nature 376, 46 (1995).
[15] O. Vasicek, Journal of Financial Economics 5, 177 (1977).
[16] M. Pasquini, M. Serva, Physica A 277, 228 (2000).
[17] Y.C. Zhang, Physica A 269, 30 (1999). L. Molgedey and W. Ebeling, Eur. Phys. J. B 15, 733 (2000). R. Baviera, M. Pasquini, M. Serva, D. Vergni, A. Vulpiani, Eur. Phys. J. B, to appear. R. Baviera, M. Pasquini, M. Serva, D. Vergni, A. Vulpiani, Quantitative Finance, submitted.
[18] J.L. Kelly Jr., Bell Syst. Tech. J. 35, 917 (1956).
[19] T. Halpin-Healy, Y.C. Zhang, Phys. Rep. 254 215 (1995).
[20] T. Palmer, Forecasting Financial Markets (Kagan Page, 1993).
By Prof. R. Baviera, Prof. M. Pasquini, Prof. J. Raboanary and Prof. M. Serva
Next: Books Related
Summary: Index