How Changes In Volatility Can Impact Retirement Plans
However, there are still static variables when retirement plans are run through Monte Carlo. These are the standard deviations used for volatility and the correlations among the assets. Today I want to focus on how big of an impact a change in volatility has on Monte Carlo results. What would happen if volatility increased to the levels we saw in 2008 and 2009? What would happen if volatility decreased from here?
I ran an analysis in the WealthTrace Retirement Planner to see what kind of impact changes in volatility might have. I used a plan for a couple that is 50 years old and plans on retiring at age 65. Here were my starting assumptions:
Inflation (CPI) |
3.00% |
Current Age of Both People |
50 |
Age Of Retirement |
65 |
Age When Both People Have Passed Away |
95 |
Social Security at age 67 (combined) |
$45,000 per year |
Average Savings Rate |
6% on Income of $100,000 |
Total Investment Balance Today (all in IRAs) |
$500,000 |
Recurring Annual Expenses in Retirement |
$50,000 |
Investment Mix |
60% U.S. Value Stocks, 15% Emerging |
Investment Location |
50% in taxable accounts, 50% in IRAs |
Return Assumption Value Stocks |
6% per year |
Standard Deviation Value Stocks |
16.20% |
Return Assumption Emerging Mkt Stocks |
9% per year |
Standard Deviation Emerging Mkt Stocks |
25.80% |
Return Assumption Treasuries |
1.5% per year |
Standard Deviation Treasuries |
7.20% |
The standard deviations used are those derived by taking the standard deviation of monthly returns over the last 10 years. After generating their plan I found that the probability of them meeting all of their spending goals and never running out of money is 85%. That is a pretty good result and can give a couple confidence that they will meet all of their goals.
But what if volatility increases? What if it decreases? I ran several scenarios moving the volatility of equity returns up and down. Here are the results I found:
Scenario |
Probability Of Meeting |
Base Case |
85% |
Volatility Up 2% |
79% |
Volatility Up 4% |
71% |
Volatility Up 6% |
65% |
Volatility Down 2% |
90% |
Volatility Down 4% |
95% |
Volatility Down 6% |
97% |
These are interesting results for sure. If volatility increases by 2% (and stays there) this cuts 6 percentage points off the probability that this couple can meet all of their goals. By contrast, if volatility decreases by 2% it adds 5 percentage points to the probability. The effect is nearly linear in terms of the impact on the probability as we change the volatility levels further.
As I mentioned, volatility isn’t the only dynamic input into Monte Carlo. The correlations of assets tend to move closer to 1.0 in times of serious distress. This is something else that can be analyzed, and I will do this in a future article. There is also the idea of changing distributions. Our Monte Carlo simulator has a default assumption (though it can be changed by the user) that returns are normally distributed. It is also worthwhile looking at changing the type of distribution to one that has “fatter tails” and gives a higher probability of extreme events. This has a fairly large impact on the Monte Carlo results as well and I will look at this in a future article.
We live in a very dynamic world today. We have powerful tools to help us with this, but for a complete picture of how somebody’s retirement plan might play out, it is best to cover all the bases.