Do econometric advances lead to more accurate VaR estimations when the legislation requirements are applied? The case of Turkish Lira FX Market.

Abstract

This paper empirically tests the benefits of econometric advances in the field of Value at Risk (VaR) within the existing legal framework that regulates the Banking (Basel III) and Asset Management industry (CESR (2010)). We test several VaR models, from the simplest and the most easily applied, i.e. the Historical VaR (HVaR), Variance Covariance VaR (VCVaR), and the Exponential Weighted Moving Average VaR (EWMAVaR), to advanced models such as GARCH(1,1).  We test these models by examining the extremely volatile FX market of the Turkish Lira (TRY) and we evaluate the models according to the criteria set by the legal framework on VaR. The empirical findings suggest that the HVaR model is a very reliable model if the goal of a risk manager is to satisfy the legislation requirements. However, we should mention that models that make use advances in econometrics lead to more accurate and more representative VaR estimations than the HVaR. The law should give incentives to adopt more representative models in the financial industry in order to take advantage of advances in financial modeling.

JEL Classification: C53; G15

Keywords: Value-at-risk; GARCH estimation; Historical VaR; Variance Covariance; Exponential Weighted Moving Average VaR; Back-testing; Legislation; FX Market