In risk management, it is important to find an effective model for the evaluation of the risk of the financial asset. Asset returns are mainly assumed to be normally distributed, but empirical financial data have shown that assets returns have distributions with fat tails and are often showed. In this study, the families of Partial Distribution (PD) and Normal Inverse Gaussian Distribution models are compared for the returns for the returns of financial data collected for End-of-Year exchange rates of different currencies (Japanese Yen, Swiss Franc and the Deutsche Mark) to US dollar from 1982-2002 sourced from IMF International Financial Statistics using the Vuong test. The list showed that the NIG distribution is better, considering two results in its favour.
Loading....