Written by: Sheryl O’Connor | Income Conductor
We just love our probabilities in financial services. We rely on Monte Carlo engines to analyze and manage risk, make informed investment decisions, and optimize strategies. In fact, the vast majority of financial planning tools utilize Monte Carlo simulations to illustrate the ‘probability of success’ of a client meeting their financial goals in retirement and, accompanied by supporting data, refer to it as a ‘retirement income plan’.
There are several problems with using this approach:
1. Most people do not understand ‘probabilities of success’, or probability in general (see research by Amos Tversky and Daniel Kahnman). What does it mean when the probability of something happening decreases? What specific actions should be taken? Academics state that retirees just take less income, so it’s not a problem. Using the S&P as an historical benchmark, retirees could encounter six to eight annual losses during retirement. Expecting retirees to take less income during these periods while expenses like food, housing and healthcare continue to rise and also determine how much less is irresponsible and unfair to clients. And if they are fortunate enough to have some assets put aside for a legacy to their beneficiaries, how much can they draw from it in periods of high inflation and market downturns and still retain a buffer for the next round?
2. Even with the highest probabilities of success, clients focus on that probability of failure. They know that ‘failure’ means going broke before they die, so even the smallest of chances of that happening is frightening. Imagine getting on a plane and having the pilot announce there was a 75% chance of reaching your destination safely. Would you relax and look forward to the flight, or focus on the 25% probability of crashing? To statisticians, this fear may seem unreasonable. But to a retiree who is taking monthly distributions from a portfolio that has lost 40% of its value, fear of running out of money due to sequence of returns risk naturally takes over and can result in them jumping out of the market and locking in their losses. I saw this behavior firsthand with highly educated clients while overseeing a TAMP during 2008 and 2009.
3. Finally, a ‘probability of success’ is very different from a plan. The Merriam-Webster dictionary defines a plan as a method for achieving an end, a detailed formulation of a program of action, and an often customary method of doing something. A plan is a procedure, a step-by-step process, much like Google Maps provides detailed information for each step involved in getting from one location to another. Probabilities do not provide a procedure or a process. They merely provide an analysis and should only be viewed as a small part of a plan, not its replacement.
I am not suggesting that we do not utilize all the analytical tools we have available to us, including Monte Carlo simulations. What I am purposing is a significant shift in our thinking away from probabilities of success and the accompanying sequence of returns risk as an actual planning strategy. We also need to rethink what constitutes a financial plan and align it more closely to the Merriam-Webster definition. In Part 2 of this article, I’ll focus on these two areas and how these shifts can benefit both advisors and clients.
Related: Global Stock Markets Poised for Boost From China Property Revival Plans