摘要：Despite of its many shortcomings, Pearson’s rho is often usedas an association measure for stock returns. A conditionalversion of Spearman’s rho is suggested as an alternativemeasure of association. This approach is purely nonparametricand avoids any kind of model misspecification. We derivehypothesis tests for the conditional rank-correlation coefficientsparticularly arising in bull and bear markets and studytheir finite-sample performance by Monte Carlo simulation.Further, the daily returns on stocks contained in the Germanstock index DAX 30 are analyzed. The empirical study revealssignificant differences in the dependence of stock returns inbull and bear markets.
关键词：Bear market; bootstrapping; bull market; conditional Spearman’s rho; copulas; Monte Carlo simulation; Pearson’s rho; stock returns
