Stats › Backtesting and evidence › Predicting the premium
Eighteen of twenty equity premium predictors have a negative out of sample R squared
Out of sample R squared for annual equity premium forecasts against a rolling historical mean benchmark, S&P 500 data to 2005. A negative value means the predictor forecast worse than simply using the average return so far. The kitchen sink model combining everything scores minus 139.03% and is omitted from the chart for scale.
Of twenty annual predictors of the equity premium, eighteen have a negative out of sample R squared. The 2024 update tests twenty nine newer variables and finds more than a third no longer significant even in sample, and half the rest poor out of sample.
Members only
The rest of this page is for members
Below this point there are 5 sections and 4 named sources, roughly 800 words of it. Every figure carries the source it came from and the date the data is from.
You can keep browsing every statistic in the library for free. The intro and the summary are always open.