CMStatistics 2015: Start Registration
View Submission - CMStatistics
B0925
Title: Working capital and market risk Authors:  Alexandra Dias - University of York (United Kingdom) [presenting]
Silvia Pazzi - University of Leicester (United Kingdom)
Abstract: The estimation of market risk is essential in financial risk management. The measure of risk mostly used in practice still is Value-at-Risk (VaR). Most of the methodologies available to estimate VaR rely on modelling the price process disregarding other information concerning the firm. We investigate the relation between equity VaR and the liquidity of the firm. We use as a measure of the liquidity of a firm the Working Capital to Total Assets Ratio (WCR). The WCR enters in the well-known Z-score model for the probability of bankruptcy. We find that the dependence structure between VaR and WCR can be modelled by a t-copula. Hence, the relation between VaR and WCR is linear with tail dependence. We find that the correlation between VaR and WCR is positive although we would expect higher risk to be associated with lower liquidity. We find the explanation for this puzzling result based on the quality of the accrual components of the working capital. We show that the strong dependence found between VaR and WCR via a t-copula can be used to improve the estimation of market risk.