Stress-Testing A Modest Retirement Portfolio

by: Lowell Herr

On January 7 of this year I posted a "Modest Portfolio"; below is the same portfolio with a stress test included. Since this is a middle-of-the-road portfolio, it is an excellent example to use for a stress test. The portfolio holds only nine ETFs, and there are times when one will want to be out of SDS. Now is such a time, but I included it in this portfolio so as to hold down the volatility. Note the low (12.0%) standard deviation. That low percentage will help in this stress test.

If the market behaved as projected, this 50-year-old, who plans to retire at age 66, should not have a problem with $500,000 already saved and a monthly savings plan of $1,000 per month. With this portfolio and these assumptions, there is only a 10% chance of running out of money by age 84. One more assumption: This individual plans to withdrawal $40,000 per year from the portfolio.

Retirement planning does not follow projected plans as laid out in this scenario. Instead, we experience three Sigma events, like the one in 2008 and early 2009. What impact does a severe bear market have on this investor should she/he experience another "black swan" market?

Below are the Monte Carlo projections following a three-Sigma market decline. With a standard deviation of 12.0%, a negative three-Sigma event results in a 36% hit on this portfolio. The savings is lowered from $500,000 to $320,000. If this happened at age 50, how are the retirement projections altered?

Instead of a 10% chance of running out of money at age 84, the bear market moves this age down to 78. Not catastrophic, but certainly bordering on the uncomfortable. Before the three-Sigma decline, this investor had only a 50% chance of running out of money at age 105. But after the big hit, the 50% probability drops to age 91. While this loss is survivable, one does need to seriously think about the risk involved when constructing a portfolio. Avoiding a three-Sigma loss is one of the primary reasons we are using the ITA Timing Model.