What's The Best Estimator of The Equity Risk Premium?

I'm teaching advanced corporate finance this semester. As usual, I spend a fairly large chunk of time on the Weighted Average Cost of Capital. One of the issues in the CAPM approach to estimating the cost of equity is what estimate to use for the market risk premium.

So of course, I has happy to find a paper by Aswath Damodaran titled "Equity Risk Premiums (ERP): Determinants, Estimation and Implications." Here's the abstract:
Equity risk premiums are a central component of every risk and return model in finance and are a key input into estimating costs of equity and capital in both corporate finance and valuation. Given their importance, it is surprising how haphazard the estimation of equity risk premiums remains in practice. In the standard approach to estimating equity risk premiums, historical returns are used, with the difference in annual returns on stocks versus bonds over a long time period comprising the expected risk premium. We note the limitations of this approach, even in markets like the United States, which have long periods of historical data available, and its complete failure in emerging markets, where the historical data tends to be limited and volatile. We look at two other approaches to estimating equity risk premiums - the survey approach, where investors and managers are asked to assess the risk premium and the implied approach, where a forward-looking estimate of the premium is estimated using either current equity prices or risk premiums in non-equity markets. We close the paper by examining why different approaches yield different values for the equity risk premium, and how to choose the "right" number to use in analysis. (In an addendum, we also look at equity risk premiums during the market crisis, starting on September 12, 2008 through October 16, 2008.)
It's not a quick read - it runs 77 pages with tables and all. If you have the time, you can read it here.

In case you don't want to plow through the whole thing, here are a few things I found interesting and/or useful:
  • He provides an excellent overview of the different approaches to estimating the ERP (surveys, historical ERPs, Implied (from valuation models) ERPs, Default Spreads, Option Pricing Model-based estimates, etc...), along with their strengths, weaknesses, and ways they can yield different results.
  • He examines determinants of ERPs (T-bill rates, the T Bill-TBond spread, etc...)
  • He examines how well various approaches predict future ERPs
On that last point, here's a table from the paper that lays it out pretty well:

So, it looks like premiums implied from DCF valuation models do the best job of predicting future ERPS. But even then, their predictive power seems pretty low - a correlation of 0.758 implies an R-squared of 0.57 for predicting the future year's ERP, and the correlation of 0.376 fbetween the current Implied ERP and the next 10 years' ERP works our to about a 14% R-squared.

Having said that, the paper will definitely make it's way in to my class this semester.