A recent article in The Wall Street Journal may have a catchy headline — “Do I Get Rich, or Go Bust? These Tools Predicted My Financial Future” — but the consensus among wealth management experts is that it falls into a classic retirement planning trap that has the potential to mislead the public.
That is, the article fails to spotlight the nuance baked into retirement income projections based on Monte Carlo simulations. Commonly, such analyses fail to contextualize their failing scenarios, projecting financial destitution in retirement as more likely than is the reality.
David Blanchett, the retirement researcher and planning expert with PGIM, took to LinkedIn on Monday to offer some respectful but illuminating criticism of the article.
As Blanchett has previously told ThinkAdvisor, a merely binary discussion of retirement outcomes wrongly compares Monte Carlo failures with plane crashes. This overlooks how much control and flexibility people have when it comes to managing assets during life after work.
In the article, the author recounts his experience with using well-known, direct-to-consumer financial planning tools to run various retirement income simulations, concluding that he has a 17% chance of “going broke” in retirement.
It sounds scary, Blanchett noted, but it is vital to realize that each “failure” in a Monte Carlo simulation can look very different. Falling even one dollar short of hoped-for income over the retirement period reads as a “failure.” Clearly, that is a much milder miss than going bankrupt in the first few years after retirement.
The odds of truly going bust “are actually a lot lower and the implications of a shortfall are likely a lot less severe,” Blanchett wrote. “I’m guessing he has a decent chunk of guaranteed lifetime income, like Social Security, and he/they could always adjust spending over time.”
Context Is King
In an email to ThinkAdvisor, Blanchett said his problem was not with the Monte Carlo simulation technique. It was rather with an absence of context that can mislead people into being overly pessimistic about their retirement.
“To be clear, I really like Monte Carlo, since I think it’s a better way to think about retirement than a simple time value of money/deterministic model,” Blanchett wrote. “But, if you’re going to use success rates as your outcome metric … I don’t think the actual probability of success should be conveyed to the user/client.”
Each client, he added, is likely going to internalize “success rates” very differently (and likely not correctly).
“For example, a success rate of 0% could actually be totally fine for someone, but I’m guessing it would totally freak a retiree out,” Blanchett wrote. “Therefore, if you are going to run a Monte Carlo projection for a client with success rates as the outcome metric … keep the success rates to yourself!”
What a 0% Success Rate Really Means
Blanchett offered the example of a theoretical client with a retirement income goal of $100,000.
This person is assumed to have $95,000 of guaranteed lifetime income across a personal pension and Social Security, complemented by just $50,000 in savings for an anticipated 30-year retirement.
