ABSTRACT
We now come to the most important and useful technique in risk modelling. Indeed some people refer to risk modelling as ‘doing a Monte Carlo’.
The probability theory in the previous chapter was pretty useless for calculating risk models. The short section on functions of random variables looked at sums and products and even this was hedged around with assumptions of independence. The exact calculation of a percentile of a random variable – which is a complex, non-linear function of a set of probabilistic inputs – is impossible almost all of the time. What we need is an approximate method, preferably an easy one, and luckily one exists.
