Monte Carlo Simulations: Jackpot or Russian Roulette?

By: Investor Solutions
More and more financial advisors are using Monte Carlo simulations to help advise their clients regarding contribution and withdrawal rates before and during retirement, respectively. The inputs, results and interpretation of results may vary from advisor to advisor. However, does the reliance on this information ensure you’ll hit the jackpot or result in the equivalent of financial Russian roulette?
In technical terms, the simulation utilizes random draws of numbers from pools of results based on rates of return and risk. A single test is constructed from these random draws. The process is then repeated 1,000 times and the results are summarized. The summary quantifies the range and distribution of possible returns and assigns a probability of success.
In simple terms, a Monte Carlo simulation calculates the likelihood that an individual’s savings will enable him or her to reach his or her retirement objectives. Different variables come into play in these calculations. Mainly, current savings, expected contributions to the portfolio prior to retirement, expected return and risk for the portfolio and inflation. With this information, the simulator models a range of possible outcomes. It then assigns a probability of success in realizing each.
Let’s take a look at an example based on the following assumptions:

  • Husband and wife each 25 years old
  • Current portfolio value = $100
  • Expected return = 10.34%
  • Standard deviation (risk) = 10.43%
  • Pre-retirement contributions beginning in 1 year = $15,500 each
  • Retirement at age 60
  • Retirement withdrawals = $250,000
  • Age at passing = 95
  • Inflation = 3%

The results are as follow:
The overall result of the simulation, exemplified by the probabilities of success, indicates that our young couple will not have accumulated sufficient assets to see them through retirement at the current withdrawal rate of $250,000 annually. This is illustrated by the fact that the probabilities of success are low; that is, below the 90th percentile. The reason for the failure is the unsustainable withdrawal rate of nearly 6.5%*. A more realistic and therefore, sustainable withdrawal rate would be closer to 3% to 4%. If the couple would alter their withdrawal rates during retirement to $125,000 or 3.1%*, the results would be as follow:
This reduced withdrawal rate dramatically increases the probability the couple will be successful in having assets to see them through full retirement. They will, additionally, also have assets accumulated to pass on to heirs or charitable organizations. For purposes of calculating potential ending wealth, the 50% median provides us with a more realistic projection than do the other percentiles as the 0% and 100% may be skewed.
Clients and practitioners alike should realize that the simulations are only as good as the inputs used to generate results and practitioners should err on the side of conservative when deciding the variables. Additionally, when changes in an individual’s cash flow, investment horizon or risk tolerance arise, changes should be made to the Monte Carlo simulations to reevaluate the chances of clients achieving their goals.
It is important to realize that a Monte Carlo Simulation is a useful and appropriate tool for financial advisors to provide their clients throughout the financial and retirement planning process. Hopefully, the process will be more an affirmation that the client is on the right track to meeting their goals than a wake up call that their expectations are unrealistic.
*Withdrawal rates are based on the median assets accumulated after 35 years.

By | 2018-11-29T16:24:28+00:00 September 20th, 2012|Blog|

About the Author: