A simple approach for building a financial plan is to decide on a rate of return for the investment portfolio and to plug that value into a spreadsheet to represent assumed asset growth. Historical data may be used to calculate historical average returns for different asset classes, which are then combined to create the overall portfolio return. This approach is also known as deterministic modeling, as there is no randomness in the future outcome. The same return is obtained each year without variability.
Deterministic approaches are overly simplified because they do not account for volatility and therefore miss the impact of sequence-of-returns risk. The basic approach of assuming a fixed return reflecting the best guess about future market returns leads to a retirement plan with only a 50 percent chance to work. The outcomes are too optimistic and could lead a retiree down an unsustainable path.
Monte Carlo simulations provide an alternative that is now widely used in financial planning software. Simulations are used to develop sequences of random market returns fitting predetermined characteristics, in order to test how financial plans will perform in a wider variety of good and bad market environments. The use of Monte Carlo tools has increased considerably over the past decade, which can likely be attributed to lower computing costs, increased recognition that returns are random, and desires to provide more robust financial plans. A thousand or more simulations could be created to test the robustness of a retirement plan in many market environments.
Monte Carlo simulations can be created for different asset classes or for an overall portfolio. With the asset class approach, one defines the arithmetic average return, the standard deviation for that return, and the correlations with other asset classes. Random draws are then taken from statistical distributions sharing these characteristics. By combining the arithmetic mean with volatility, the resulting simulated returns will display the appropriate compounded return over time. Historical data is commonly used to set these input characteristics. Most financial planning software works in this way.
With Monte Carlo based financial planning software, retirees generally focus on building a plan that achieves a high probability of success, such as 80 or 90 percent. This implicitly means the underlying assumed return is below average. But when thinking in terms of a fixed return assumption, we usually consider what we view as the best guess for future returns. Again, the best guess only implies a 50 percent chance for success. Half of the time, the realized return will be higher and half the time it will be less. In order to have a conservative fixed return assumption, we must further scale down from our best guess estimate. This is a point which many investment management professionals have not internalized into their thinking, as they are conditioned to using their idea about average returns as the input.
Implied fixed investment returns are usually not shown with Monte Carlo simulation output in financial planning software, but they do exist underneath the hood. We can reverse engineer their values. So which implied portfolio fixed return supports a 90 percent chance for success? The implied return will be lower than the average return input for the simulation, and I find support for appropriate portfolio return assumptions in the postretirement period to be more conservative than in the preretirement period.
This is an excerpt from Wade Pfau’s book, Safety-First Retirement Planning: An Integrated Approach for a Worry-Free Retirement. (The Retirement Researcher’s Guide Series), available now on Amazon.