In a more complicated post-9/11 world, one in which highly unlikely things seem to occur more frequently, a growing number of advisors and wealth managers are modifying their retirement planning tools and techniques to adjust to the new realities today. Lately, sacred cows such as buy-and-hold investing have come under fire.
One popular tool getting a second look is Monte Carlo simulation, often used in financial planning to estimate the odds of a client reaching his financial goals in retirement. The current knock against Monte Carlo is that even though the tools typically run client portfolios through hundreds or thousands of potential market scenarios, they often assign minuscule odds to extreme market events, or so-called “black swan” events—for example, a market plunge of, say, 58%, or a terrorist attack. Now, some experts and academics are pushing for Monte Carlo to more clearly assimilate scarier scenarios.
Larry K. West, a CFP licensee in Huntsville, Ala., and a strong believer in Monte Carlo simulation techniques, has always used them to test for unpleasant surprises, such as a client’s spouse dying or the unanticipated need for long-term care. But more recently he has begun to test for worst-case scenarios as well—for example, a major market drop that can drag assets down significantly. Ten years ago, he might not have taken this step.
Another financial advisor, Jeff Cedarholm, a CFP in Gadsden, Ala., is also modifying is also modifying his Monte Carlo simulations.
“During the ’90s,” says Cedarholm, “you could just pick any well-diversified mutual fund and get a great return and the plans would work. But over the last ten years, return patterns in the market have made planning very difficult.”
At least one mutual fund company has reacted, too—Charles Schwab Corp., which uses Monte Carlo software to create its target-date fund models. Schwab has modified these funds to reduce their risk quotient, though so far some of its peers have not followed suit.
“The whole methodology of traditional Monte Carlo analysis is bogus from the perspective of economics,” maintains Laurence Kotlikoff, a Boston University economics professor who developed the DSPlanner financial planning software. “It assumes that people will spend the same amount of money year after year in retirement no matter what happens in the market. No one in his right mind acts this way, so they’re assimilating a behavior that is mathematically convenient but economically ridiculous.”
The problem with Monte Carlo and similar retirement planning tools, say critics, is that extreme market events fall towards the “tail” of the bell curve, while the sunny probability of 7%-8% stock market returns year after year falls in line with “normal” events, and these assumptions lull people into a false sense of security. Facing this more hostile environment, a growing number of advisors and wealth managers are tweaking the simulations—or, in industry lingo, “fattening the tail” of the bell curve by taking into account more worst-case scenarios than before, things that would have fallen outside the bell curve before along the edges or tail.
The criticisms have thrown a spotlight on target-date (or “lifestyle”) funds in recent years, funds that have grown more popular as an “all-in-one” retirement planning vehicle and which often use Monte Carlo simulation patterns. Typically, target-date funds’ moving asset allocations, or “glide paths,” invest more aggressively as the investor approaches retirement, and then the asset mix shifts to something more conservative in the years immediately preceding and following retirement. But critics say these funds concentrated too much on longevity risk during the 1990s—which increased rather than lowered equity allocations as people neared retirement—because reports indicated many investors could outlive their assets.
There are nearly 380 such funds—including 100 or so that came on stream last year amid the worst of the financial crisis. The median loss last year for all of them including those in existence for all of 2008, was 33%, according to Morningstar Inc., which says the funds have to do a better job communicating their risks to investors.
Throwing The Dice
The Monte Carlo technique is not new. The name was first coined in the 1940s by physicists working on nuclear weapons projects at the Los Alamos National Laboratory. There is no single Monte Carlo method. Instead, the term describes a large and widely used class of approaches.
The tool allows some flexibility in entering key variables such as an investor’s current age, the expected age at retirement, the spouse’s age (if applicable), current investable assets and the amount of anticipated costs through retirement. According to Eagle Asset Management in St. Petersburg, Fla., a firm that uses Monte Carlo in crafting its retirement models, the calculator will typically program in assumptions related to returns, standard deviations, correlations and inflation, and it will then generate an output from perhaps hundreds or thousands of inputs, according to a series of calculations run in the background, which show the odds or likelihood of particular outcomes. More traditional or more linear scenarios would typically arrive at a single number as opposed to a range and likelihood of outcomes.
“The best and simplest benefit of using Monte Carlo simulations is showing clients that achieving any financial goal, like retirement, is not a sure thing,” points out Christopher J. Cordaro, chief investment officer at RegentAtlantic Inc., Morristown, N.J. “The flaw in MC simulations is that the tails of the distribution of investment returns are fatter than a normal distribution implies; bad outcomes happen more frequently than we expect. That’s where we need to improve our simulation analysis.”
Professor Moshe Milevsky, an associate finance professor at York University’s Schulich School of Business in Toronto and an authority on Monte Carlo, says these simulations are the “personal finance equivalent of bank stress tests.” In a recent paper, Milevsky proposed that users compute the sustainability ratio using two sets of assumptions. One would take into account pre-existing data and require no new mathematical tools or other inputs. The second would factor in an unusual event – something with a “one-in-100” chance of occurring within three years. “A ratio greater than 1% should set off alarm bells, while anything above 2% should set off piercing sirens,” Milevsky said.
Monte Carlo supporters say that it’s not the methodology itself that’s the problem. “It’s as good as the assumptions you put in,” says Cooper Abbott, a senior vice president at Eagle Asset Management. “Getting the assumptions right is the key to the quality of outcomes.”
Jerry A. Miccolis, a CFP in Madison, N.J., and co-author of Asset Allocation for Dummies (Wiley Publishing Inc.), says he has provided another layer of disaster scenario planning as a result of the current market slide. “We weren’t entirely comforted by the fact that our tails were already fat enough,” he says.
“We’re making some slight modifications in what we do with Monte Carlo, trying to get a more meaningful range of outcomes,” says Mike Palmer, principal of the Trust Company of the South, Raleigh, N.C.
Likewise, says Cheryl A. Costa, managing director of AFW Wealth Advisors in Purchase, N.Y. Costa says she has “begun overriding the settings that represent the ‘95% probable’ range on my software to include a wider range of possible outcomes. …I want clients to walk away knowing there are a wide range of possible outcomes. I don’t want them to walk away saying, ‘The planning software says I will run out of money at age 94.’”
Target-Date Funds Targeted
As Monte Carlo assumptions have drawn more scrutiny, Charles Schwab has cut the equity exposure in is target-date funds with a time horizon of ten years before the investor retires, and increased it in the longer years past retirement. For example, the equity allocation of the firm’s 2010 target-date fund is currently 43% two years from retirement, down from 52% before the investment change, according to Dan Kern, who manages the firm’s $514 million in target-date funds. The Schwab Target 2010 Fund, with $71.7 million in assets, lost 24% in 2008. The equity exposure in the Schwab 2040 Fund is now 91%, up from the previous exposure of 79%.
“In the years approaching the target date, clients are most vulnerable to a downturn in equity markets,” says Kern, explaining the changes. “They have the most money at risk; they have limited earnings power, and they typically have the most anxiety around their finances. So it made sense to us to reduce the risk in those years approaching retirement.
“Conversely, investors further away from their retirement target date have time on their side. Their earnings power is higher and in most cases growing, and they can ride out the bumps in the market. Consequently, that’s the time to build assets for retirement and invest more of those assets in equity.”
But Fidelity Investments, T. Rowe Price and MFS Investment Management currently have no plans to radically change the glide paths of their target-date funds. Fidelity, the largest target-date fund provider with $66.6 billion in assets in its Freedom Fund series as of March 31, says its 12 funds are meant for long-term investors.
According to Jonathan Shelon, who manages the funds, the Fidelity Freedom 2050 Fund is the most aggressive, being meant for the youngest participants, and has 90% of its assets in equities. Fidelity Freedom Income, the most conservative, is meant for investors in their 70s and 80s and has 20% of its assets in equities, while the Fidelity Freedom 2010, designed for those people closest to retirement, has 50% of its assets in stocks. The company uses several different Monte Carlo approaches in setting its funds’ models, Shelon says.
T. Rowe Price’s ten target-date funds have equity exposures that vary according to one’s age on a sliding scale, from 90% for those 25 or more years away from retirement to 20% for those who are 30 or more years past retiring at age 65. The company says that after some retesting, it ended up reaffirming its current glides.
“Many people right now are focused on this extreme bear market, which, while devastating, is a short-term snapshot within decades of retirement saving and withdrawals,” says Jerome Clark, lead portfolio manager of the T.Rowe Price retirement funds. “Just to focus on that is to take a shortsighted view and is not the best way to go about structuring your assets with the goal of generating an income stream throughout a long retirement.”
Joe Flaherty, a portfolio manager for MFS Investment Management’s asset allocation funds, says his firm adopted a conservative approach with respect to glide paths well ahead of the meltdown and is not likely to change it.
Morningstar fund analyst Greg Carlson maintains that fund companies with target-date funds “need to do a better job communicating the risks of these funds—and that they’re designed to be held for a long time after their target date. In most cases, they’re not going to get a very low-risk vehicle at the time of the target date because a lot of shops are focusing on longevity risk.”
Morningstar last year tweaked the asset allocation software it offers to institutional investors, allowing them to create either a normal bell curve or a “fat-tailed” distribution. In late May, the company announced it will launch ratings and in-depth research reports for the target-date funds it covers.
Though Russell Investments uses Monte Carlo simulations in much of its research on retirement investing, Richard Fullmer, a senior portfolio analyst at the firm, acknowledges the lack of uniformity in results in some retirement planning models like Monte Carlo. Fullmer chairs the methodologies committee of the Retirement Income Industry Association (RIAA), which is looking at standards for models like Monte Carlo.
“A lot of these tools are available to advisors and they give significantly different results. It’s hard for the public to know why. Typically, it’s because the assumptions are different, but there is no baseline to compare the assumptions with,” says Fullmer.
Some advisors, meanwhile, are making greater use of alternative retirement planning methodologies. Instead of Monte Carlo, planner Howard Cadwell of Brentwood, N.H., uses what he calls “rolling historical periods going back a century” to look at retirement and pre-retirement scenarios.
For his “mass affluent” clients, there’s a bigger payoff, he says. “We can tell them that the historical record involves all sorts of things that would be hard to model honestly with a bell curve: war, depression, runaway inflation, stagflation, bubbles and what have you. These severe events don’t just have random results; there are serially correlated consequences that resonate, or ebb and flow, over time.”