Journal of the Italian Statistical Society

, Volume 5, Issue 3, pp 335–360

A monte carlo analysis of two spectral tests of the martingale hypothesis

  • Claudio Lupi
Article

DOI: 10.1007/BF02589095

Cite this article as:
Lupi, C. J. It. Statist. Soc. (1996) 5: 335. doi:10.1007/BF02589095

Summary

The size, power, and robustness properties of the Kolmogorov-Smirnov and Cramér-von Mises spectral tests of the martingale (difference) hypothesis are investigated by Monte Carlo methods. The results highlight a marked superiority of the Cramér-von Mises with respect to the Kolmogorov-Smirnov test. The paper also shows that the Cramér-von Mises test is simple to compute, more general and more powerful than other converntionally used tests.

Keywords and phrases

Monte Carlo Spectral Tests Martingales Hypothesis Testing 

Copyright information

© Società Italiana di Statistica 1996

Authors and Affiliations

  • Claudio Lupi
    • 1
  1. 1.ISPE-Instituto di Studi per la programmazione economicaRomaItaly

Personalised recommendations