A Semiparametric Approach to Value-at-Risk, Expected Shortfall and Optimum Asset Allocation in Stock-Bond Portfolios Before and After the Global Financial Crisis
56 Pages Posted: 24 Aug 2010 Last revised: 19 Jul 2012
Date Written: July 9, 2012
This paper investigates a stock-bond portfolios tail risks such as value-at-risk (VaR) and expected shortfall (ES) and the optimum asset allocation, and the way in which these measures have been a¤ected by the recent global financial crisis (GFC). The semiparametric method is used to estimate bivariate copulas for modelling stock-bond returns' joint distributions for G7 countries and Australia. The t-copula is found to be adequate for modelling these joint distributions. Empirical results show that the weak (negative) dependence has increased notably for seven countries after the GFC, while it has decreased only for Italy. However, both VaR and ES have increased for all eight countries. In addition, before the GFC, the minimum portfolio VaR and ES were achieved at an interior solution only for the US, the UK, Australia, Canada and Italy, while at the corner solution for France, Germany and Japan. After the crisis, on the other hand, the corner solution was found to produce these minimum VaR and ES for all eigh t countries. The "flight to quality" phenomenon has become stronger, after the GFC. Based on a VaR forecasting exercise, Christoffersen's (1998) conditional coverage test's results reinforce the adequacy of semi-parametric t copula models for stock-bond joint distributions. The central focus of both the regulators and the Basel Committee is on the tail risk measures and the right level of capital requirements. For any level of bond weight in the stock-bond portfolios, the capital requirements have increased following the GFC.
Keywords: Dependence, Blanket tests, Semi-parametric method, copula, Investment decision, Value-at risk
JEL Classification: C14, C52, G11, F36, G15
Suggested Citation: Suggested Citation