The integral hasn't really been evaluated. As you guessed they recognised that the (1- the integral) is some multiple of the standard normal distribution function, and made appropriate substitutions.
This "trick" is quite common (maybe more so in CT6).
As far as I know that integral does not have a closed form solution (I think this is the right term).
The polar coordinate transformation works for integrating over all real numbers, and I don't think (although I'm a bit rusty) that it can be applied for integrals over a smaller range.
Since it can't be evaluated algebraically, the common approach (at least for us statisticians/actuaries) is to relate it to the standard normal distribution function (for which we have a convenient pre-calculated table of values, which were obtained via numerical methods). Generally, whenever you need to integrate exp(-x^2), you usually transform it to the standard normal distribution function.
Last edited by a moderator: Mar 18, 2010