Concerns are intensifying over elevated US stock valuations and the perceived disappearance of the equity risk premium. However, a Bloomberg analysis suggests that a widely used calculation method is flawed, and after reasonable adjustments, US stocks, while expensive, are not yet in a danger zone.
Bloomberg columnist Nir Kaissar notes that the S&P 500's dividend yield of just 1.1% is far below the 4.5% yield on 10-year US Treasury bonds, a stark gap causing many investors to worry the equity risk premium has vanished. The analyst argues this comparison is inherently flawed, as dividend yield fails to capture a company's full earnings power, making any risk assessment based on it potentially misleading.
A recalculation using a more appropriate blended earnings metric shows the S&P 500's equity risk premium is approximately 2.3%. While below its historical average, this level remains within a manageable range and does not yet constitute a systemic warning signal. For investors allocating assets based on the risk premium, the current environment still supports a modest overweight to stocks over bonds, with equities continuing to offer a positive risk-adjusted return.
Common Calculation Methods Are Misleading
When valuations are high, investors instinctively compare stock returns to bond yields. The theory is that stocks must offer a higher yield as compensation for their greater volatility; the disappearance of this premium is often seen as a red flag. The two most popular comparison methods, however, have significant shortcomings.
The first is a direct comparison of dividend yield to Treasury yield. Pitting the S&P 500's 1.1% dividend yield against a 4.5% Treasury yield yields a concerning conclusion. The problem is that dividends represent only a fraction of corporate profits. Companies today increasingly favor buybacks over dividends, making dividend yield a poor proxy for total shareholder return and a weak anchor for valuation.
The second method substitutes earnings yield. This, however, compares apples to oranges. Treasury yields are pure nominal rates, while earnings yield is a metric incorporating real growth expectations and is highly sensitive to inflation and price swings. Simply subtracting these two different measures lacks logical rigor.
Adjusting for Inflation Eases Valuation Pressure
A more reasonable framework aligns the earnings yield with the real Treasury yield. Currently, the 10-year breakeven inflation rate (the spread between regular Treasuries and TIPS) is about 2.2%. This implies a real Treasury yield of approximately 2.3% (4.5% nominal minus 2.2% inflation), not the headline 4.5%. This baseline adjustment significantly alleviates the apparent pressure on equity valuations.
While the measurement issue is addressed, the definition of earnings level presents a greater challenge. For the same company, using earnings from different periods leads to vastly different conclusions:
Forward 12-month expected earnings yield is around 5.0%. This is most current but relies on short-term extrapolation and can miss turning points.
Trailing 12-month actual earnings yield is around 4.0%. This is reliable but subject to high annual volatility and noise.
Cyclically Adjusted Earnings (CAPE), using a 10-year inflation-adjusted average, yields only about 2.6%, nearing the real Treasury yield (~2.3%). It smooths volatility but systematically underestimates current levels due to the long-term uptrend in earnings.
Each of these three measures has its use case and its blind spot. Choosing one involves a trade-off between timeliness, accuracy, and stability.
The Distortion of Cyclical Adjustment
When earnings growth significantly outpaces its historical trend, the CAPE metric's underestimation of current actual earnings is dramatically amplified. Long-term data from Yale professor Robert Shiller shows that since 1881, trailing 12-month earnings have averaged only about 11% higher than CAPE earnings. Currently, that gap has surged to over 60%.
This disparity stems from an exceptional surge in earnings growth. Since the post-pandemic recovery began in 2021, S&P 500 earnings have grown at a compound annual rate of about 13%—roughly double the long-term average since the 1950s and over triple the trend growth rate since the 1870s.
This high growth is directly reflected in the chasm between forward expectations and the cyclically adjusted average. According to Bloomberg data, the consensus forward 12-month earnings estimate is $365 per share, while the CAPE figure is only $188 per share—the forward estimate is nearly double. This gap is the largest on record since at least 1990, the earliest year for which analyst forecast data is available.
A Blended Approach: Expensive, But Not Dangerous
Both extreme measures are biased: CAPE is overly conservative due to the high profit growth of tech giants, while forward valuations may be overly optimistic due to the mean-reverting nature of earnings cycles.
A compromise is to use an eight-year average, blending five years of historical earnings with three years of forecasts. Based on this, the S&P 500's normalized earnings are $333 per share, implying an earnings yield of 4.5%. Subtracting the 2.2% breakeven inflation rate results in an equity risk premium of about 2.3 percentage points.
This level is below its historical mean but not excessively so. The conclusion: valuations are stretched, but no red light is flashing. For investors guided by the equity risk premium, stocks still offer positive compensation for risk.
Comments