You can’t predict the future, of course, but that doesn’t stop some financial professionals from trying.
Of the many methods devised to anticipate different possible futures in financial planning, Monte Carlo simulations are among the most popular. Named after the storied gambling mecca in Monaco, they are a mathematical technique for modeling the different possible outcomes of any action or decision.
Yet as helpful as they may be for calculating and comparing degrees of risk, some experts say their value can be overestimated and their limitations and shortcomings misunderstood—with disastrous consequences.
A Product Of The Computer Age
Steve Parrish, St. Augustine, Fla.-based adjunct professor of advanced planning at the American College of Financial Services, where he is also co-director of the Center for Retirement Income, calls it, “a powerful tool brought to us by fast-processing computers.”
He explained that the simulations take key financial indicators—such as historical returns and correlations among asset classes—and generate thousands of random results for different scenarios.
They cannot, however, “account for all the random events that occur in markets, and in life, for that matter,” he said, “and they can’t predict the consequences of these random events.” Rather, Monte Carlo simulations are good for “cutting through the fog of current events and avoiding the danger of static assumptions,” he says.
They are useful “to test strategies,” he notes. “But they should not be the final judge of strategies.”
How They Work
To understand how these simulations work, consider the “random walk theory,” which essentially holds that asset prices move unpredictably.
Richard Van Kuren, director of research at LVW Advisors in Pittsford, N.Y., puts it this way: “What happens this year is not related to what happened last year or what will happen next year.” The Monte Carlo simulation, though, allows you to take that random walk thousands of times, generating different results for random variables that can affect markets.
“It can paint a picture of the probability of success many years into the future,” says Dean Catino, president and co-founder of Monument Wealth Management in Alexandria, Va.
That does not mean the results hold forever. “This is not a set-it-and-forget-it analytical tool,” he clarifies. “We run annual Monte Carlo simulations.”
Each one is “a snapshot in time,” says Kristina Mello, senior financial advisor and director of financial planning at StrategicPoint Investment Advisors in Providence, R.I. “Clients need to understand that underlying assumptions can always change, and this will have an impact on the results of the analysis.”
Besides shifts in market assumptions—such as whether interest rates are headed up or down—the simulation can’t predict how clients’ lives may change.
“Monte Carlo does not plan for Armageddon or even Covid,” says Harold Evensky, founder of Evensky & Katz/Foldes Wealth Management in Lubbock, Texas.
How They Should Be Used
Still, many advisors see the simulations as tools for guiding adjustments in a client’s financial plan.
“The advantage of Monte Carlo simulations is that they provide boundaries that can be used to measure a plan’s progress,” says John Bochniak, a principal at Homrich Berg in Atlanta. “Simulation results are helpful markers.”
But they should not be the only markers, he stresses. Bochniak recommends supplementing Monte Carlo simulations with key metrics such as a client’s debt-to-income ratio and withdrawal rate to better gauge how on track each client’s financial plan truly is.
“From the standpoint of an advisor who is creating a long-term financial plan for a client, using the outputs from a Monte Carlo simulation to facilitate productive and understandable conversations can be extremely useful,” says Derek Schug, head of portfolio management at Kestra Investment Management in Austin, Texas, adding, “as long as the advisor and client understand the limitations and/or imperfections.”
Some Caveats
Those limitations and imperfections are many and varied.
One frequently cited problem is that the simulations are predicated on what economists call “efficient” markets, in which asset prices are reflective of all available information. Real-world markets don’t usually work that way.
In addition, most Monte Carlo simulations assume “the projected distribution of returns will be similar to a bell curve,” says LVW’s Van Kuren, referring to a statistical term for a symmetrically balanced array of results, also known as “normal distribution.” “In reality,” he says, “investment returns are often not ‘normal.’ … The extremes—both positive and negative—tend to occur more in reality than they will in simulations.”
Given these imperfections, he says, Monte Carlo modeling is “best used as part of a mosaic.” He and his team often combine the simulations with “other portfolio stress-testing techniques to build what we believe is a more complete picture,” he says.
Another shortcoming is that Monte Carlo simulations “rely on the data being input by the advisor,” says Matt Marini, associate director of planning at Coastal Bridge Advisors in Westport, Conn. “If accurate and comprehensive data isn’t used, the analysis is meaningless.”
Moreover, the simulations work off of historical market information. This “rarely reflects the variabilities of markets and human behavior,” says Michael Lent, a founding principal, CIO and CFO of Veris Wealth Partners in New York City.
Realistic Expectations
Nevertheless, he says, the simulations “can provide some insight into what a range of outcomes for a client’s portfolio could be over time.” But he stresses that these insights should be “used for client education, not as a prediction.”
If that’s misunderstood, he warns, clients can get a false sense of security.
Working with an advisor to make regular adjustments to a portfolio “can improve outcomes, relative to the potentially misleading comfort sometimes conveyed by a Monte Carlo simulation,” he says. “Maintaining a disciplined investment approach in the long term is likely to generate the best returns.”
Indeed, Monte Carlo simulations don’t adequately factor in an investor’s ability to change course.
“The ‘hole’ in the Monte Carlo is not in the assumptions it makes about investors’ erratic behavior. It’s in the assumptions it makes that the client will not adjust his or her behavior accordingly,” says Joe Maier, director of wealth strategy at Johnson Financial Group in Milwaukee. “It assumes set-it-and-forget-it spending, [which] has virtually nothing to do with reality.”
Factors No Model Can Foresee
Not that Monte Carlo simulations should be ignored. “You would be a fool not to use the tools that are available to help make decisions,” says Greg O’Donnell, founder and CEO of O’Donnell Financial Group in San Rafael, Calif. “At the end of the day, however, someone has to make a decision no matter what the Monte Carlo simulation says.”
So, in sum, Monte Carlo simulations may not be perfect, but what is? “We haven’t really come up with anything better in long-term planning,” says Charles Lewis Sizemore of Sizemore Capital in Dallas. “Perhaps the best rule here is to simply take all simulations with a grain of salt and always allow for a wide margin of safety.”
He’s not alone in this view. “There are still some unknowable factors that no model can foresee,” says John Bautista, a managing director at UHY Advisors in New York City.