After the Capital Asset Pricing Model (CAPM) was developed within the Nineteen Sixties and Seventies, financial researchers began to check how well this theoretical model actually worked in the true world.
With increasing computing power and higher access to data, the Nineteen Eighties became an important era for assessing the validity of the CAPM as analysts examined the effectiveness of the beta model in predicting future returns.
Surprisingly, there was a general consensus that the predictive power of the beta factor for predicting returns was moderately low.
In the roughly 60 years because the CAPM was introduced, how well have the model and beta predicted returns over the many years? To discover, we analyzed every company traded on the NYSE and NASDAQ and constructed portfolios of firms based on their systematic risk (beta) using monthly returns and a 12-month rolling calculation.
If an organization had a beta below 0.5, it was assigned to the low-beta portfolio. Companies with a beta above 1.5 were assigned to the corresponding high-beta portfolio.
Using these groupings, we examined how the portfolios performed over the subsequent 12 months, each on a median and market-cap weighted basis. The portfolios were then reconstructed annually using recent beta calculations.
| Average return of a high-beta portfolio | Average return of a low-beta portfolio | Market weighted return of a high beta portfolio | Market-weighted return of a low-beta portfolio | Percent of years in accordance with CAPM | |
| Seventies | 14.9% | 2.5% | 14.3% | 3.5% | 80% |
| Nineteen Eighties | 13.0% | 14.4% | 12.1% | 18.1% | 40% |
| Nineties | 18.7% | 12.6% | 22.6% | 13.4% | 70% |
| 2000s | 15.2% | 8.9% | 10.7% | 5.2% | 80% |
| 2010s | 14.7% | 9.0% | 13.3% | 12.5% | 91% |
It seems that the Nineteen Eighties were a terrible time for beta. On an annualized basis, a low-beta portfolio outperformed its high-beta counterpart by a mean of 6 percentage points over the last decade, returning 18.14% versus 12.12%
We then checked out the proportion of years that reflected the CAPM forecasts on an ordinal basis over the last decade. In only 4 of the ten years did the CAPM accurately predict returns. That is, years with positive market returns ought to be related to high beta outperforming low-beta portfolios, and years with negative market returns with low beta outperforming high-beta portfolios. This implies that the CAPM underperformed a random walk over this era, and explains why researchers of the time were so skeptical of the model.
But the Nineteen Eighties were something of an outlier. As the many years went by, beta and CAPM became higher predictors. From 2010 to 2020, the CAPM was correct in 10 of the 11 years.
In fact, in every decade because the Nineteen Eighties, a high-beta portfolio has returned just over 5 percentage points greater than its low-beta counterpart on an annualized basis. That is, the high-beta portfolio returned 15.53% on average, while the low-beta portfolio returned 10.34%.
Overall, the outcomes show that beta just isn’t as bad a predictor of future returns as is usually assumed. The Nineteen Eighties were a terrible time for beta and CAPM, but since that decade, beta has been predictor of future returns.
If you liked this post, don’t forget to subscribe.
Photo credit: ©Getty Images / Eskay Lim / EyeEm
