Do interest rate effects explain/predict the poor performance of value stocks over the past decade, and especially during 2017 through early 2020? In their May 2020 paper entitled “Value and Interest Rates: Are Rates to Blame for Value’s Torments?”, Thomas Maloney and Tobias Moskowitz investigate interactions between equity value factors and the interest rate environment. They first examine theoretical relationships and then explore relationships between several ways to measure the U.S. equity value premium and interest rates empirically, including interest rate level, change in short-term rates, change in long-term rates and slope of the yield curve. They look at subperiods and some international evidence. Finally, they assess ability of interest rate variables to predict the value premium and thereby inform factor timing strategies. Using U.S. interest rate and firm/stock data inputs for several ways of estimating the value premium as available since January 1954, and similar data for Japan, Germany and the UK since 1988, all through December 2019, they find that:
- Theoretical links between interest rate environment and the value premium are complex and make ambiguous predictions.
- Empirically:
- There is modest evidence of a positive relationship between the value premium and contemporaneous change in long-term bond yield.
- There is mild evidence of a negative relationship between the value premium and change in short-term interest rate.
- Combining these two patterns indicates a statistically stronger sensitivity of the value premium to change in the yield curve slope, but the effect: (1) is of small economic significance, and (2) varies for different versions of the value premium, subperiods and markets. During 2017-2019, flattening of the yield curve explains only a small part of steep losses in value stocks.
- Predictive relationships between interest rate environment and the value premium are even weaker than contemporaneous interactions, offering little hope that interest rate signals support value factor timing strategies.
In summary, evidence does not support belief that that value factor timing strategies based on interest rate signals work.
Lack of evidence, despite benefit of hindsight to develop hypotheses, supports belief that the value premium is simply risky, thereby justifying its long-term positive value.
Cautions regarding findings include:
- Value premium calculations are gross, not net. Accounting for frictions involved in periodic reformation of long-short value factor portfolios and ongoing shorting costs/constraints would reduce the premium.
- As implied in the paper, reliable exploitation of the value premium may require a very long holding period.
