Retirement Monte Carlo

[rokid]

Attached is a Monte Carlo simulator created by Richard Pritz with assistance from Peter Vann and Peter Ponzo. The latter is a retired Canadian math professor known as "Gummy". His web site, http://gummy-stuff.org/, contains a lot of interesting, and mathematical, investment information.

Anyway, the Monte Carlo simulator allows you to determine the probability of having enough money in retirement given a specified number of years, withdrawal rate, inflation rate, equity/fixed income mix, and return/standard deviation/correlation estimates. Earlier Spaf proposed a question that assumed a 3.5% withdrawal rate and a 3.0% inflation rate.

Entering the average G Fund rate (6.63%) and standard deviation (1.49%) along with the approximately 12% equity rate of return and a plus 19% standard deviation, Spaf gets a 100% probability of success with a 3.5-4.0% withdrawal rate investing 100% in the G Fund - essentially 100% success with virtually no risk! However, adding equities (risk and return) nets him a lower probability of success - counter intuitive!

Consequently, in retirement, the highest risk/return doesn't necessarily provide the highest probability of success.

 

[EW_ret]

I too have been running estimates on achieving 95% probability of success on my money lasting 33 years using financial engines and another calculator, FIRECalc. I have reached the same conclusion regarding risk. I do not need more that 35%-40% equities for my money to last, at a 4.5% withdrawal rate. I used 8% average yearly return (37% Stock/63% Bond) with a Standard Deviation (SD) of 9%. If I increase the equity exposure to 60% the average return increases to 9% (60% Stock/40% Bond), but the probability of my money lasting decreases to below 80 percent because of the larger SD of 13%. The likelihood of getting a few bad years in row increases, and its impact is the greatest if it occurs during the first five years of retirement.

 

[rokid]

EWGuy,

Yes, the results really are surprising.