To verify whether an empirical distribution has a specific theoretical distribution, several tests have been used, for example: Kolmogorov-Smirnov and Kuiper. These tests try to analyze if all parts of the empirical distribution has a specific theoretical shape. But, in a Risk Management framework, the focus of analysis is on the tails of the distributions, since we are interested on the extreme returns of financial assets. This paper proposes a new goodness-of-fit hypothesis test with focus on the tails of the distribution. The new test is based on the
Conditional Value at Risk measure. Three major exchange rates (JPY/USD, GBP/USD and CHF/USD) are used as examples of a practical application of the test proposed. The new test, the Kolmogorov-Smirnov and Kuiper tests were applied to verify if the empirical data has a Normal, Scaled-t, Hyperbolic, NIG or GH distribution. For JPY currency, the Normal, Hyperbolic and scaled-t distributions were rejected by the new test. For the CHF and GBP, only Normality was rejected. Results are the same for CHF and GBP when using the other two tests. But for the JPY, the Scaled-t and the Hyperbolic distributions are rejected on the new test, and not rejected for the other two tests. We conclude that, for overall finance applications, we can use Scaled-t and Hyperbolic distributions for the JPY, but for Risk Management applications, they are not adequate.
[Authors: Farias, Aquiles; Haas Ornelas, Jose R.; Fajardo, Jose]