CONTENTS

Exercise: Computing ES and VaR for a binary payoff

Consider the P&L XX of a single defaultable zero-coupon bond with face value 100100. With probability 0.970.97 the bond pays 100100, and with probability 0.030.03 the bondholder loses the entire principal (so X=100X = -100).

Tasks

  1. Compute VaRα(X)\text{VaR}_\alpha(X) for α{0.01,0.03,0.05,0.10}\alpha \in \{0.01, 0.03, 0.05, 0.10\}.

  2. Compute ESα(X)\text{ES}_\alpha(X) for the same four values of α\alpha.

  3. For which values of α\alpha is VaRα=ESα\text{VaR}_\alpha = \text{ES}_\alpha? For which is ESα>VaRα\text{ES}_\alpha > \text{VaR}_\alpha? Interpret each case.

  4. Suppose the bond either pays 100100 (prob 0.970.97) or only 5050 (prob 0.030.03) — a haircut, not full default. How do VaR and ES at α=0.05\alpha = 0.05 change?

Hint

Write out the CDF as a step function, then read off the quantiles and integrate.