The Way the World Should Work

Corporate bonds have historically earned a 1% or so premium over government bonds. And stocks have historically earned a 3–4% premium over corporate bonds. These premia are compensation for bearing additional risk, and they are known as the credit premium and equity premium, respectively.

But the credit premium should depend on the discount that investors demand at a given time to own corporate credit versus government bonds. Likewise, the equity premium should depend on the discount that investors require to own stocks versus corporate bonds.

The way we think the world should work is that corporate bonds should offer a yield about 1% above the yield of government bonds to compensate for default risk. And we think stocks should have an earnings yield that is 3–4% above the yield on corporate bonds to compensate for equity risk and volatility. We can compare our simple model of the way the world should work to how markets actually price stocks and bonds.

We believe that, within the US market, this simple model is beautifully predictive. Companies with higher risk of default (as measured by credit rating) issue bonds at higher yields and have cheaper equity that trades at a higher earnings yield. Below, we show the bond yields and earnings yields of US companies. It appears there is a near perfect relationship between the two: the higher the bond yield, the higher the earnings yield—credit risk and equity risk scale in a nicely linear manner.

Figure 1: US Companies’ Bond Yields and Earnings Yields by Credit Rating

Source: Verdad bond database

Delightfully, this simple model also holds across countries. In each major developed market we can observe a credit risk premium (corporate bond yield – government yield) and an equity risk premium (earnings yield – corporate bond yield) that together account for the total risk premium for equities (earnings yield – government yield). The credit risk premium varies from 0.3% to 1.4%, and the equity risk premium varies from 1.8% to 6.1%, but both premia are observable in each major market. 
 
Figure 2: Yields by Country

Source: Wall Street Journal, Capital IQ, Star Global

Although we do see a positive credit risk premium and a positive equity risk premium across different markets, the size of these premia is wildly variable and seems to follow no particular pattern. We see no correlation between interest rates and equity valuations (contrary to popular mythology), nor do we see a correlation between the credit risk premium and the equity risk premium. For example, the US has a high government bond yield and a high credit risk premium but the lowest equity risk premium, whereas Japan has a very low government bond yield and a very small credit risk premium but the highest equity risk premium. There are, of course, many potential explanations for why the earnings yield does not vary with bond yields: differing growth and inflation expectations, different levels of historical price volatility, etc.

But since price and return are related, we believe the size of the equity risk premium offered today should end up having some predictive power about future equity returns, both within countries and across countries. The size of the risk premium realized should depend on the size of the risk discount at time of purchase. And, indeed, the data suggests that the earnings yield does have some predictive power.

First, we show below the US equity risk premium (calculated as the earnings yield minus the risk-free rate) versus subsequent 10-year returns on US equities for the last 100 years. The earnings yield explains 36% of the variance in subsequent 10-year returns over the period.

Figure 3: The Equity Risk Premium vs. Subsequent 10-Year Equity Returns

Source: Robert Shiller

Second, this relationship appears to hold across countries. Star Capital analyzed 17 countries back to 1979 and found that their calculation of the equity risk premium (using cyclically adjusted price-to-earnings ratios) explained 48% of the variance in subsequent 10-year returns, representing a 70% correlation between the equity risk premium offered at purchase and returns over the next decade.

Figure 4: Relationship between CAPE and Average Real Returns of the Subsequent 10 Years

Source: Star Capital

Paying high prices for stocks simply because interest rates are low is not sound logic. We believe investors shouldn’t be distracted by interest rates but should focus instead on what can be learned from the equity risk premium, which appears to have far more predictive power and is the far more variable number.

And if our simple model of the way the world should work is roughly correct, and today’s yields and risk premia in fact end up predicting future returns, then there are a few rather significant takeaways from the above data.

First, the US equity risk premium is far too low, probably by 2–3%. Based on our analysis, that should mean abnormally low US equity returns relative to bonds. Conversely, the equity risk premia in Japan, the Eurozone, and the United Kingdom look high relative to long-term historical averages, implying that investors could potentially earn abnormally high equity returns relative to bonds by buying stocks in those regions.

Second, we believe the US credit market remains the most attractive globally, with the highest government bond yields of any market and a robust 1.1% credit risk premium. US corporate bonds look attractive relative to US equities. Investors would probably only need to lever up BB corporate bonds 1.2x to achieve the same yield as the earnings yield of the US equity market, which, if those yields translated into future returns, could potentially imply a much smoother, less volatile and more secure way to achieve the same results.

The world is complex, and the future is unpredictable, but we believe very simple models can help guide us. And our very simple model of the way the world should work implies that US equities, the core holding of most US investors, are unattractive relative not only to international equities, but also to US corporate fixed income.