11/26/2008 10:43:53 AM

Andy has been hosting some very good sessions about communications and planning in the current economic crisis, and 2 of recent have focused around planning and Monte Carlo Simulations (MCS). To me, it seems the message by Bob Curtis was the advisor is using MCS incorrectly.

David Leoper has some very good white papers and hypothetical illustrations on what happens when MCS is run using year end market value of accounts and average annual returns. They equate to garbage-in, garbage-out.

MCS is a very powerful statistical application known as resampling. Done correctly, a sample size needs to be large enough to support the analysis, a minimum of 36 periods. Secondly, the sample should use a bootstrap technique which is resampling with replacement. You have to use the actual data, not its average and standard deviation.

The position of Bob Curtis seems to be, 'hey, planning just tells you how you can move forward based off your account value. Your investment manager is to blame for that.' That would be fine if the majority of the audience did not claim to be both financial planners and investment managers.

In his illustration he showed how the client's confidence level deteriorated from 73% to 44% based on a 20% decrease in market value. When I questioned this, he claimed it was because the period that was used to run the projections had increased to incorporate both 1973/74 crisis and today. However, if you examine the charts, the average rate of return and standard deviation were EXACTLY the same. What Bob failed to show us is with his calculations, the client would appear better off NOT following our advice.

The problem is that MoneyGuide Pro is using bad data (average annual returns and standard deviation) to answer the wrong question. The illustration attempts to solve the probability of past average annual returns repeating themselves, not if a client's portfolio can create $x in retirement income over the next 20 years.

MoneyGuide Pro is a tremendous application, but should be avoided because of the asset allocation assumptions used to run outcomes. With today's enabling technology as account aggregation, returns-based style analysis and downside risk optimizaton, MoneyGuide Pro could easily become a very powerful and meaningful application.

However, Bob is backed into a corner. He has 18000 users he must attempt to make beleive he did not send them over the falls without a lifevest. He also derives a great amount of income from broker-dealers and dually registered reps, who are subject to FINRA.

FINRA Interpretive Material IM-2210-06 is very clear on how Rule 2210 applies to software. Quoting from this letter, "IM-2210-6 defines an investment analysis tool as 'an interactive technological tool that produces simulations and statistical analyses that present the likelihood of various investment outcomes if certain investments are made or certain investment strategies or styles are undertaken, thereby serving as an additional resources to investors in the evaluation of the potential risks and returns of investment choices. The filing requirement does not apply to hypothetical illustrations of mathematical principles that do not predict or project the performance of an investment or investment strategy, such as Web site calculators that compute future returns based upon assumed variables, since Rule 2210(d)(1)(D) does not prohibit (and IM-2210-6 thus does not cover) such illustrations."
Endnote 3 elaborates on this by saying, "The 'hypothetical illustration' exception to the prohibition in Rule 2210(d)(1)(D) applies to tools that serve the function of a calculator that computes the mathematical outcome of certain assumed variables without predicting the likelihood of either the assumed variables or the outcome. For example, this exception would apply to a calculator that computes a net amount of savings that an investor would earn over an assumed period of time with assumed variables of rates of returns, frequency of compounding, and tax rates. On the other hand, this exception would not apply to a calculator that predicted the likelihood of achieving these assumed variables and outcomes."

Clearly MoneyGuide Pro is on the fence attempting to sell a one size fits all tool. Here in lies the problem - does the client know you were not actually projecting outcomes when you said they had a 73% chance if they transferred their account? Try to explain to an arbitration panel how a 20% drop in account value cuts their ability to achieve their goals in half.

Commented By Brent E. Bentrim