Don't Buy Suze Orman's 12% Return Projection, Retirement Experts Say

Why Geometric Returns Matter More in Wealth Management

Blanchett’s initial post continues: “In my opinion, the geometric/compounded return should always be the return assumption when talking about building wealth over multiple periods, not the simple (arithmetic) average. In my review, this is a mistake Dave Ramsey (among others) consistently makes as well.” (See “Supernerds Unite Against Dave Ramsey’s 8% Safe Withdrawal Rate Guidance.”)

Additionally, Blanchett writes, it’s “kind of cheating” to not mention the bite of inflation, especially over a 40-year time horizon. A 2.5% inflation rate over 40 year results in $1 eventually being worth around 37 cents.

“In reality, that $1 million she mentions in the article (saving $100 a month for 40 years) is only likely to be around $250,000, using a more realistic geometric real return of 7% versus the 12% noted,” Blanchett concludes. “I get that $250,000 doesn’t sound as cool as $1 million, but at least the $250,000 is actually a (more) probable outcome!”

Deeper Insight Into Average Returns

Asked by ThinkAdvisor to expand on this discussion, Blanchett wrote via email that “there’s not much of a debate here” regarding the superior planning methodology, though confusion does arise from the fact that arithmetic returns are important in the Monte Carlo simulation process — but that is another matter entirely.

“The arithmetic return is only the appropriate input/assumption in a Monte Carlo projection because the realized return by the client will be the geometric return (incorporating the volatility within the forecast),” Blanchett explains. “If you’re doing any kind of future/present value calculation you should always use a geometric return, because that’s the return that’s going to be realized by the investor.”

Blanchett again pointed to the importance of incorporating inflation, “because obviously $1 today should be worth less in 40 years.” He also points out that, the higher the volatility of an investment, the greater the difference is going to be between the geometric and arithmetic return, as shown in the graphic below, which is sourced from the Jordà-Schularick-Taylor Macrohistory Database.

The chart contains the simple (i.e., arithmetic) and compound (i.e., geometric) returns for the a few of the key asset classes in the Ibboston SBBI series from 1926 to 2023, using calendar year returns.

“If you want to approximate the impact moving from the arithmetic average to the geometric average you would just subtract half the variance,” Blanchett says. “So, the long-term standard deviation for U.S. stocks has been about 20%. Variance is the square root of standard deviation, [which gives us] 4.47%. Take half of that and you get 2.236%. The actual difference in geometric and arithmetic returns has been 1.885%, so that’s a pretty good approximation.”

Pictured: Suze Orman. (Photo: Marc Royce)