Fat Tail Distributions

I received an email newsletter today from a vendor that sells financial planning software to advisory firms like Cascade Wealth. This particular company is Money Tree Software and it is located in Corvallis, Oregon. While we are not currently using their software, I have in the past.

This email caught my attention, because it mentioned Monte Carlo Analysis. You may recall from one of my prior blog postings, this is the methodology used to test the probability of outcomes. Financial Planners often use Monte Carlo simulations to determine how likely clients are to reach their financial objectives.

The recent stock market crash has caused some advisors to question the reliability of traditional Monte Carlo Analysis. The problem is that these projections rarely show major market declines. The reality, as well all know, is that dramatic market drops are not only possible, but they actually occur. We have witnessed two of these periods in the last decade alone in the tech crash in the early 2000’s and the financial sector meltdown in 2008.

Money Tree plans to enhance its software by incorporating wider distribution patterns in their Monte Carlo simulations. This means that events considered statistically very unlikely to occur will be included in their simulations. If you have ever taken a statistics class, you may remember that a normal distribution curve looks like the graph below. It displays the outcomes that occur when the pattern of distributions is symmetrical with a single, central peak at the mean (or average) of the data. Half of the outcomes fall on each side of the mean.

The spread or dispersion of the outcomes is dictated by the standard deviation. In the graph below of a normal distribution, we can expect that 68.2% of the events will fall between -1 and 1. Only 1.5% of events will fall below at -2.5 or lower.
What would this graph look like if we assumed that more outcomes occurred towards the ends of the graph and fewer near the middle or mean? The tails would become thicker. The midrange would become thinner. This is sometimes described as a “Fat Tailed” distribution.

By giving greater statistical weight to the possibility that market returns could fall outside the traditional range, Money Tree’s software will reduce the probability that a given client scenario will succeed. This should motivate the planner and the client to make more conservative assumptions.