2

In retirement planning one of the inputs to a forecast is an assumption about return on investments... and there are a number of rules of thumb thrown around. 6%, 8%, something that goes from higher to lower the closer you get to retirement age, etc.

What I want to do is look at something like the S&P 500, for which data is readily available, not only to get a compound annual growth rate (CAGR) value for the annual return input, but also to look at volatility.

But what's a good way to incorporate volatility into this?

Since I have approximately 35 years left to invest for retirement, one thought I had is to take rolling 35 year slices of the index and calculate the CAGR for each slice. Then for my forecast use the median rate for my primary assumption on return, but then also forecast using the 25th and 75th percentile rates to give a sense of what the next 40 years could actually look like.

Is there any common way of doing something like this when doing retirement planning?

2

Monte Carlo methods are used given various pieces of data about securities for one idea. However, Mean Variance Optimization can sometimes be analyzed from various angles.

Course there is something to be said for the model you build in terms of how much income you need, how long does it have to last, what tax rates will be, etc. Some may use estimates and depending on what investing options one has there are other factors as if one has pensions or annuities these could pay out differently under various scenarios.

  • Some really good ideas in that article on MVO. I especially like the author's critique of complicated modeling techniques as potentially misleading. – mlibby Dec 24 '15 at 14:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.