Why Use Monte Carlo Simulations? To Guard Against Tunnel Vision

 
Sam Savage, author of “The Flaw of Averages,” is a missionary for making business decisions using probability distributions instead of point estimates. I’m an avid disciple, which might surprise some who read the previous post.

The earlier post, “Modeling Expected Returns: The Future Is Not What It Used to Be” described a formula derived by Jacquier, Kane, and Marcus (JKM) to estimate the return over a future horizon given a sample return history. The result of that formula is a point estimate, not a probability distribution, so it might seem contradictory for me to celebrate that formula while advocating the use of probability distributions.

The power of the JKM formula is it empowers you to estimate a “back of the envelope” expected return given a set of historical returns. It won’t help you decide whether future asset returns will be higher or lower than in the past, but it can help you run a sanity check on your financial planning software if it bases its Monte Carlo simulations on historical returns.

The power of Monte Carlo simulations is they prompt you to imagine alternative future states of the world, and challenge you to come up with contingency plans in case of an unexpected or undesirable outcome. Even if you experience a state of the world for which you did not plan, the mindset of regularly considering alternative outcomes ought to expedite you arriving at an proper course of action.

So make use of Monte Carlo simulations. If you already use software that includes Monte Carlo simulations, great – just make sure someone you trust who knows what he’s doing verifies that the software works appropriately. If you don’t, you can easily build Monte Carlo simulations in Excel without too much effort. Your local Excel expert can help you use it appropriately.