Measuring and managing market, credit and Op risk [88]

Go to: Summary | Previous | Next   
Bullet points include: For (multi-variate) Normally distributed random losses, VaR is coherent Proof, recall that and verify axioms Subadditivity holds because standard deviation of sum of two random variables is less than or equal to sum of the standard deviations Monotonicity holds since Normally distributed random variables have positive support on real line, so x1 > x2 in all states of the world if and only if x1 and x2 are perfectly correlated, when x1 = x2 + c, hence VaR(x1) = VaR(x2) + c for any alpha Homogeneity and translation invariance are trivially satisfied In fact, VaR coherent for a broader class of loss distributions, notably elliptic distributions (another special case of which is Student’s t-distribution)

Contents | Prev | Next | ERM Lecture Series

Desktop view | Switch to Mobile