1/24/2025 - By Matthew Wicker, CFA (Dimensional Fund Advisors)
Every year, many investment and consulting firms release new CMAs, or capital market assumptions. CMAs can be useful for setting risk and return expectations and for long-term financial planning. However, as the old saying goes, it’s difficult to make predictions, especially about the future.(1) Even well-constructed CMAs are not a crystal ball and, when used inappropriately, can lead to suboptimal outcomes for investors.(2)
While it will take some time until the accuracy of this season’s CMAs can be determined, we can compare previous asset class return forecasts to realized outcomes. In this analysis, we look at forecasts by 10 asset management and investment advisor firms for the 10-year period 2014 to 2023. Our analysis includes the major asset classes covered most often by forecasters. However, firms differ in the asset classes for which they issue forecasts, so the number of forecasts included for each asset class in our analysis varies.
Exhibit 1, Panel A presents annualized compound nominal return forecasts for US, developed ex US, and emerging markets equities (blue dots) against realized returns (light blue diamonds). We see wide dispersion across forecasts in all regions. For example, forecasts for US equities range from about 5.5% to 7.7% annualized, which translates to 10-year cumulative returns of about 70% to 110%. The realized return, however, lies well above the range at 11.5% annualized, or a nearly 198% cumulative return.3 For emerging markets equities, the forecast dispersion was even greater, although all forecasts turned out to be overly optimistic. Across all regions, equity forecasts differed from reality by 1.7 to 6.8 percentage points, with an average error of a little over 4 percentage points.
In Panel B, we show the results for subasset classes within US fixed income as well as inflation. Fixed income asset classes tend to have lower volatility compared to equities, and yet we still see meaningful dispersion across forecasts, on the order of 2 percentage points annualized. In terms of forecasting accuracy, all firms underestimated inflation, and most firms overestimated bond returns, with the realized return falling outside the entire forecast range for both government and investment-grade corporate bonds. Overall, forecasting errors range from 0.1 percentage points to 3 percentage points.
Panel C reports results for US REITs and commodities. REIT forecasts were off by approximately 1 percentage point on average, with an error range of 0.3 to 1.8 percentage points. Commodity forecasts exhibited the greatest degree of error in our study, with even the best forecast off by more than 3 percentage points and an average difference versus reality across forecasts of a little over 5 percentage points.
Past performance is not a guarantee of future results.
While some forecasts turned out to be closer to the actual outcomes than others, no firm’s forecasts stood above the rest as consistently more accurate. These results are consistent with those of other studies based on different time periods or CMA providers.4 Given the level of variability we see in the accuracy of CMAs, it is worth asking how much weight should be put on them when building portfolios.
Rather than relying on a crystal ball that does not exist, planning for the possible range of outcomes can better prepare investors for what’s ahead. When setting expectations about potential outcomes, we believe a sensible starting point is to look at the long-term historical performance of asset classes, to incorporate information from real-time market prices when relevant, and to consider a wide range of possible values.
Footnotes