There are a few common questions that many clients will eventually ask their financial adviser to answer.
How much will my portfolio be worth at retirement? Will I outlive my money? How would my plan be affected by a stock market downturn?
Chances are the adviser will not try to answer any of those questions before running the numbers through a computerized mathematical system that is practically unknown to the general public, yet widely used in the financial services industry. It is called the Monte Carlo method.
It has nothing to do with French Riviera, casinos, Grace Kelly or James Bond.
The Monte Carlo method used by financial advisers is a technology that analyzes the likelihood of a client’s portfolio being successful. The adviser inputs certain data, such as the client’s age and what assets are in the portfolio. The computer looks at thousands of portfolio return outcomes over multiple time periods and market conditions to arrive at the most likely — or highest probability — outcome.
Some form is used by just about every wealth management firm that has made an investment in financial planning tools and resources. But some industry players also warn that these simulations are only as good as the quality of the information provided, and that improper use can give users a false sense of security.
Robert Hapanowicz, president of Hapanowicz & Associates Financial Services in Pittsburgh, said his firm has been using this sort of analysis in financial and retirement planning for more than five years. “Monte Carlo enables us to analyze a client’s plan across many different life scenarios, thus providing a very effective way to stress test the plan and estimate the probability of success.”
One of the things advisers love so much about the method is its strong basis in mathematical statistics.
“With Monte Carlo, you can model thousands of potential life outcomes by using random variables — such as stock market returns, inflation rates, interest rates — to measure probability of success within a financial plan,” Hapanowicz said.
A Monte Carlo report will assign different probabilities to different outcomes. A 95 percent probability is considered the gold standard.
The method was named by mathematician Stan Ulam in 1946. The name refers to the casino in Monaco where his uncle would borrow money from relatives to gamble. According to mathematics website Motherboard, Ulam was inspired to develop the method with his partner John Von Neumann while playing solitaire and trying to calculate the likelihood of winning based on the initial layout of the cards.
Gambling and solitaire
The computational algorithms are used in a variety of fields outside of financial services, such as production management, engineering, physics and weather forecasting.
Robert Johnson, president and CEO of the American College of Financial Services in suburban Philadelphia, said those who rely too heavily on Monte Carlo reports should be reminded of previous instances where quantitative models have failed.
One of the greatest examples, he said, is the failure of Long Term Capital Management, a firm founded by some of the best minds in finance, including Nobel Prize winners. The Greenwich, Conn.-based hedge fund nearly collapsed the global financial system in 1998 as a result of high-risk arbitrage trading strategies.
“A more recent example is the failure of rating agencies to properly assess the risk of mortgage-backed securities failing,” Johnson noted.
A great tool
Nick Besh, investment director for PNC Bank, sees Monte Carlo analysis as a great tool to show probability of outcomes of a total portfolio of diversified assets. But it is not a forecaster of investment returns, nor is it used to evaluate individual stocks.
“Financial planning uses capital market assumptions — expected long-term returns and standard deviation — for each asset class used in a portfolio,” Besh said. “These assumptions are devised using historical data combined with projections of future returns.”
Besh said all projections PNC Bank shares with clients come with disclosures that past performance is not indicative of future results.
Tim Grant writes for the Pittsburgh Post-Gazette.