Comparison of risk measures: Value at risk versus expected shortfall

Abstract

Research objectives

Value at risk has become the standard risk measure of financial institutions during the past twenty years. After the unsuccessful revision of the regulative market risk framework, which was initiated after the financial crisis of 2007-2008, Basel Committee proposed switching the market risk measure from value at risk to expected shortfall.

The objective of this thesis is to compare the two different risk measures, value at risk and expected shortfall, with each other. The main research question is: should value at risk be replaced by expected shortfall in banking regulation? The backtesting procedure is probably the most important aspect of this comparison and will be highlighted in the thesis.

Research methods and data

Both theoretical properties and practical issues are reviewed. The theoretical properties are discussed in a literature review, which aims to compile a summary of the strengths and weaknesses of the risk measures and explain the highly technical details of the risk measures in plain language.

The practical issues are studied in an empirical research conducted with the daily closing values of Nokia's stock price and Euro Stoxx 50 index price. The daily risk estimates are calculated and the estimates are then backtested on yearly level. Two different years, 2005 with low market volatility and 2008 with high market volatility, are the backtested years.

Main findings

Expected shortfall has some theoretical properties which make it preferable to value at risk in perfect conditions. However, these conditions are seldom fulfilled in the markets. In practice, the robustness of value at risk makes it the more suitable option for banking regulation. Especially the superiorly effective backtesting of value at risk estimates will most likely prevent the move to expected shortfall.