Reading 18 Q13

[broadex]

Q. Recommend a statistical method that ILIC should use to obtain information about a proposed asset allocation’s performance over time, given its forecasts concerning liabilities and its capital market expectations.
The answer is Monte carlo but im not sure whether Monte Carlo is the only possible solution..and if yes why not the mean–variance, resampled efficient frontier, Black–Litterman, exprienced based methods Or ALM???

 

[ink]

i think the reason is because “over-time” implies path dependency, and this is best modeled with monte-carlo.

 

[Galli]

You can eliminate experienced based or ALM because they are not statistical methods.
If you thinking about the differences in the remaining statistical methods one could argue that monte carlo actually incorporates all of the other methodologies to some degree. Therefore it’s best suited to meet all of the information you’re looking to obtain.