Out[91]: 10.373513531207843
Compared to the case of a regular asset, the option case has somewhat different
characteristics. We only see a little more than 500 losses due to a default, although we
again have about 1,000 defaults. This results from the fact that the payoff of the option at
maturity has a high probability of being zero:
In [ 92 ]: np.count_nonzero(L * D * hT) # number of losses
Out[92]: 582
In [ 93 ]: np.count_nonzero(D) # number of defaults
Out[93]: 1031
In [ 94 ]: I - np.count_nonzero(hT) # zero payoff
Out[94]: 43995
Figure 10-22 shows that the CVaR for the option has a completely different frequency
distribution compared to the regular asset case:
In [ 95 ]: plt.hist(L * D * hT, bins= 50 )
plt.xlabel(‘loss’)
plt.ylabel(‘frequency’)
plt.grid(True)
plt.ylim(ymax= 350 )
Figure 10-22. Losses due to risk-neutrally expected default (call option)
