by Cynthia Harrington
Software companies introduced Monte Carlo simulations to financial professionals over a decade ago.
The tool reached shooting-star status with the media when Nobel Prize winner William Sharpe launched Financial Engines, the retirement-planning Web site. Fears that the readily accessible Financial Engines would put advisors out of business have not come true -- and part of the reason is the future is not that easy to predict.
"We use Monte Carlo simulations as a tool but we don’t bet a client’s financial life on it," said Steven P. Kanaly, CFP, CTA, vice chairman of Kanaly Trust, in Houston. "It helps us get a handle on rates of return and the affect of compounding certain inputs, but after the report is printed we still have to apply common sense."
Before Monte Carlos, advisors projected a client’s financial future from a set of assumptions of expected returns, risk and correlation characteristics of the assets in the portfolio. Old programs would calculate results and produce one outcome. Monte Carlos generate thousands of possible outcomes based on different historical scenarios. The result is a range of possible outcomes.
"Before, I could show clients what would happen using a straight-line projection," said Benjamin Tobias, CFP, CPA, CIMA, of Tobias Financial Advisors, in Ft. Lauderdale, Fla. "I never liked that. I would show it to the client and tell them the one thing we could count on is that the future wasn’t going to look exactly like this."
Clients benefit by being able to see the affects of multiple changes in current actions, like saving and spending, as well as the impact of a wide range of possible market conditions. While they have no control over market conditions, the tool helps guide clients’ choice of amounts to save and how much they can spend. The range of outcomes produces a level of probability of the client meeting their retirement goals. "Most clients are less conservative than I am and will accept a 15 percent to 20 percent chance of not meeting their goals," said Tobias. "I want to keep it around 5 percent. Above that amount I think clients have to worry."
The difficulty in using this tool explains why investors still seek professional investment advice. While the process of choosing inputs and running the simulator is easy, interpreting the results is hard. "We try to put the projections on a grid of human reality," said Kanaly. "We get down to basics with the client’s spending choices for instance. We ask them if they’re really going to buy used cars and not new cars to be able to send their children to an Ivy League school over a local university."
Kanaly pointed out the variables missing from the results of the simulation tool. Changes in the client’s health or whether government will absorb future Social Security costs aren’t included. The software can’t calculate the affect of delayed inheritances or children moving back home. The problem of living too long isn’t considered. "The present value of the cost of an extended term of medical care at the end of life is a huge number," said Kanaly. "Monte Carlos can’t address that."
Other concerns with the output keep Mary Katherine Dean, CPA/PFS, in San Diego, Calif., from using the tool at all. Dean saw a demonstration of the software over a decade ago and was wary of the projections built into the models. "I don’t use Monte Carlos and never did," explains Dean. "I wasn’t sure that the numbers they were using for returns and correlations were going to continue. That was 1989 and now we know things didn’t turn out the way they were projecting."
Nor does Dean believe that all clients need retirement projections. For those that want to see what might happen she developed a spreadsheet program. She assumes a 1 percent return for the first decade because that was the return for the worst decade since World War II, from 1965 to 1975. "I start out with a worst-case scenario," she said "Then I go to an 8 percent return for the succeeding years."
Unlike the off-the-shelf Monte Carlos, Dean’s program is customized to each client. After the top line calculations, Dean does a line-by-line annual projection. The specialized service costs $3,000 extra. "Looking back, I wish I had required the retirement projection for every client that wanted to be all in equities in the late 1990s," said Dean. "The analysis of each year would have gotten them really thinking about the impact of stock market fluctuations."
Even adherents of Monte Carlos agree their use is more an art than a science. Technology delivers up a product, but the proper application depends on human thinking. "We can’t forget to think strategically about a client’s financial life," said Kanaly. "We still have to ask the client if a particular view of the future lets them sleep at night. We have to ask ourselves if the projections made economic or business sense. And we have to consider the tax impact of the recommended actions."
One way of customizing results using off-the-shelf programs is to run projections from several vendors. The "what-if" planning tool comes standard with many of the popular retirement planning packages, like Money Tree Software’s Silver Financial Planner, and Frontier Analytics AllocationMaster. Both Kanaly and Tobias use this strategy. "I look at different permutations using all the programs," said Tobias. "Usually my faith in the results is confirmed because they’re all similar."
"But sometimes they’re different," he continued. "I show these to clients as another reminder that we don’t know what’s going to happen."
Being able to put the results into a context demands deep understanding of the assumptions built into each software program, as well as the experience to understand when results don’t make sense. "The average individual can be educated to use the simulators, but still need professional advice," said Tobias. "You could hand someone something as simple as a hammer, but if they’ve never seen a hammer work they aren’t going to know what to do with it."
