New Gamble in Predicting Returns

Securities brokers are getting a sweet deal on Valentine’s Day that may or may not endear them to their clients.

Beginning Monday they’ll be able to use investment analysis tools aimed at helping clients predict how their portfolios may fare over time. These tools have been used by institutional investors for years, but weren’t available to retail brokers because regulators had barred them from predicting investment returns.

Now, industry regulators have carved out an exception, allowing brokers to use mathematical models, such as the so-called Monte Carlo simulation, to show clients possible long-term portfolio results.

Computer-generated Monte Carlo simulations, named after the famed gambling resort, were first used to understand the behavior of neutrons during the development of the atomic bomb. In the financial business, the model has been adopted to illustrate the variability of investment returns.

The decision by the NASD, the brokerage industry’s self-regulatory group, to allow brokers to use Monte Carlo simulations is being greeted with a mixture of optimism and concern. The reason: These tools can be invaluable when used correctly, but they’re also complicated and fairly easy to manipulate to show a desired result -- which could be handy for a shady broker who’s trying to pitch a specific investment.

“Used properly, these are another tool to help inform customers about investment decisions,” said Marc Menchel, general counsel for the NASD. “But we are constantly concerned with the potential for abuse, whether it’s with investment analysis tools or the recommendation of an individual stock.”

Some experts also worry that individual investors could confuse a written analysis with a promise, rather than a projection of a possible outcome.

Here’s a look at the NASD’s rule change and what it means for brokers and their clients:

Question: What’s changing?

Answer: In the past, securities brokers were barred from providing any predictions of future investment returns, whether on an individual stock or a portfolio. As of Monday, they still will be barred from predicting that, say, XYZ stock will jump 30% this year. But brokers will be permitted to provide their customers with long-term portfolio growth projections based on mathematical models such as Monte Carlo simulations.

These projections are expected to help investors in a number of ways. For example, clients may better understand the likelihood of having sufficient assets at retirement based on their current saving and investing patterns. The projections also could give investors a better picture of the risks they shoulder with an individual stock or a particular mix of assets.

Q: How do Monte Carlo simulations work?

A: They use powerful computers to calculate thousands of possible investment outcomes for a portfolio, usually based on risk and return data gathered from past market performance.

In the end, the investor usually is presented with a simple figure that assesses the chance of coming out with the amount of money needed at a particular point in the future.

For example, an individual investor may find that, with his current investment strategy, he has a 20% chance of having the $100,000 he wants in 20 years.

The analysis also can be used to help determine how much a retiree can spend from his nest egg in any given year without facing a serious risk of running out of money before he dies.

Q: What is the potential downside of using these simulations?

A: The models all require a series of assumptions about investment returns and volatility. But those assumptions are not standardized, nor are they individually monitored by regulators.

If the assumptions a broker plugs into a simulation are inaccurate or unreasonable, they could make a miserable investment look smart. And although the assumptions must be disclosed to a client, unreasonable assumptions might not be obvious.

Q: Why not?

A: Because these analyses can be quite complicated and involve lots of moving parts.

Consider a broker who uses a Monte Carlo simulation to compare the benefits of a variable annuity with those of a mutual fund.

The broker knows that the benefits of the annuity are that taxes on any trading profits are deferred until the money is pulled out at retirement, and that an insurance component in the account can limit the investor’s losses.

But annuities also have drawbacks: When money is pulled out in retirement it’s taxed at ordinary income tax rates that can be twice as high as the capital gains rates that would be applied to long-term profits earned on a mutual fund.

Moreover, the annual management expenses of annuities often are far higher than what mutual funds charge.

To be accurate, a comparison of an annuity and a mutual fund over time must make fair assumptions about market returns, tax rates, the frequency that the investor would trade in his account, and the fees charged. Most valid analyses would show that it’s the rare investor who comes out ahead with the annuity because of the higher expenses and tax rates.

But if you assume that the investor trades frequently and never enjoys the benefit of a lower capital gains rate in the mutual fund -- or if you ignore the annuity’s higher annual fees -- the annuity might look like a lower-risk investment that’s sure to provide a higher return than the mutual fund over time.

Q: Why would a broker feed inaccurate assumptions into a portfolio simulation?

A: He could simply err. For instance, trading frequency is something that’s hard to predict and will vary widely from one investor to the next. Or the broker could be swayed by self-interest: The commissions paid to brokers on the sale of annuities are often far higher than those paid on mutual funds.

Q: How should I evaluate a simulation that a broker performs for my investments?

A: Look at the assumptions that are made and ask lots of questions. If, for example, you’re presented with a simulation comparing an annuity with a mutual fund, ask whether the assumption accounts for the difference in fees, and ask what it anticipates with respect to trading frequency and tax rates.

The broker should be able to explain all of the variables that go into a simulation and why he used particular assumptions.

A simple guideline to remember: Monte Carlo simulations should reflect a symmetry between risk and reward, said Joseph Grundfest, professor of law and business at Stanford Law School. If a projection shows the possibility of fabulous returns without a fairly high possibility of loss, it’s a warning that there’s something wrong in the mix.

“If someone is trying to explain that they can get you a big return without subjecting you to a material risk, run away,” Grundfest said. “It’s a fairy tale.”

Kathy M. Kristof can be reached at kathy.kristof @latimes.com.