Monte Carlo forecasts prove anything is possible — or not

With simulations showing extreme projected portfolio outcomes, some financial advisers won't even use Monte Carlo modeling for their younger clients for fear of looking silly.

Rick Ferri, the tenacious burr under the saddle of asset-based financial planning business models, is at it again.

Concurrent with his ongoing battle to push the financial advice industry toward lower and less-conflicted fee models, the feisty owner of Ferri Investment Solutions is calling out the industry for its generally haphazard use of Monte Carlo simulations to forecast potential portfolio outcomes for clients.

His biggest gripe is with the reliance on simulation projections, which are typically presented at such extremes they become useless for anything but making financial advisers look technologically savvy.

“I prefer to come up with reasonable expectations of what stock and bond markets will deliver, and whether that gets you to where you need to be down the road,” he said.

But, while Ferri might be among the more dogged critics of the popular portfolio analysis tool that is embedded in many of the most popular financial technology platforms, he isn’t alone in at least acknowledging the downside of a modeling technique that factors chance and random outcomes through formulas often loaded with assumptions.

“I just don’t see the point of pointing out to a client there is a 5% probability they could be living under a bridge eating dog food someday,” said Ferri, in a swipe on that severe ranges of potential outcomes that Monte Carlo forecasts produce.

“I’m not sure it’s the role of the adviser to give clients the whole spread of what could happen,” he added. “What’s the point in telling a client they have a 5% chance of running out of money and 5% chance their children will inherit $10 million?”

Marcio Silveira, a financial adviser at Toler Financial Group, said the range of potential outcomes is typically so wide that he avoids doing Monte Carlo forecasts for younger clients because the longer-term projections can appear comically unrealistic.

“If I have a 30-year-old client saving 30% of their $200,000 income year after year, the simulation shows they may end up with $40 million at age 90, and that happens on the downside as well, but it’s not as embarrassing to show because it’s less money on the downside,” he said. “The numbers are so disconnected from reality they make me look silly and I risk losing credibility.”

Silveira said a fundamental problem with the simulation methodology, originally named after a gambling destination in Monaco, is that it assumes independence of annual returns as if the performance of one year doesn’t affect future years.

He compares the Monte Carlo formula to flipping a coin.

“It could come up heads 10 consecutive times, even though that’s not likely, because each toss is independent,” he said. “But that’s not how it works in the financial markets because returns are related.”

Silveira does use Monte Carlo simulations for clients closer to retirement when the shorter time horizon doesn’t allow the projected compounding to produce such extreme outcomes.

“The numbers are out of whack with all clients, but with someone in their 60s there’s not enough of an opportunity for the compounding to take place,” he said.

Alec Quaid, a financial planner at American Portfolios Denver, recognizes the “output issues” of Monte Carlo forecasts that are “only as good as the assumptions made in the financial plan.”

But he also appreciates the way the simulations show clients “where the plan breaks.”

“The Monte Carlo simulation will produce those sequence of returns in a thousand different ways,” Quaid said. “You show the client the best and worst possible scenarios; that they could run out of money five years into retirement or they could retire with $20 million.”

Ferri compares presenting clients with such extreme potential outcomes to a doctor telling a patient they have a 2% chance of dying.

“You’re just scaring the shit out of the patient,” he said. “I don’t think all those what-ifs and squiggly lines do anything, so why add all that in there?”

The ‘why,’ as Ferri sees it, comes down to some financial advisers marketing themselves as brilliant conductors of a formula that he believes the majority of advisers don’t even understand.

Beyond the ability to “bedazzle clients” with the range of potential outcomes, Ferri questions the real value of Monte Carlo simulation in client portfolio construction.

“Monte Carlo was originally designed for science, where you have known facts, but what you don’t know is one particular variable, so you run the simulation to find the probability of what you don’t know,” he said. “But in finance, everything is an assumption, and a lot of it never happened before. And I don’t think anyone has ever gone back and checked the accuracy of all those simulated projections.”