Meanwhile, Alan Greenspan appeared on CNBC on Friday, and said with a straight face (I believe he has no other kind) that the "equity risk premium" on stocks was higher than it has been in 50 years. Evidently, this remark reflects the standard "Fed model" idea that the forward operating earnings yield on stocks can be usefully compared with the yield on 10-year Treasuries in order to estimate the attractiveness of stocks.
With all due respect, Mr. Greenspan - no less reckless than he was during the dot-com and housing bubbles - is out of his gourd.
The operational way to think about the equity risk premium to define it as the difference between the expected total return of the S&P 500 and the expected yield on Treasury securities. In order to estimate the equity risk premium, you had better have an empirically testable measure of long-term expected returns that actually works historically. On this basis, the forward operating earnings yield fails miserably (see Long-Term Evidence on the Fed Model and Forward Operating P/E Ratios). In contrast, the chart below presents the average of three models that I've detailed in previous comments - one based on forward operating earnings (adjusted for cyclical margins), one based on normalized earnings, and one driven by yields. At present, the average 10-year total return projection for the S&P 500 is just 3.76% annually. While it is true that this is higher than the 3.02% yield available on 10-year Treasury debt, it is also true that the average historical premium has been 4% (and even higher prior to 1995). When one subtracts the 10-year Treasury yield, the present equity risk premium is far from the highest in post-war data (an honor that goes to the premium of well over 15% in the late 1940's), but is in fact the lowest in history outside of the late-1990's stock market bubble over which Mr. Greenspan presided.
Technically, the equity risk premium should be measured over a horizon equal to the effective duration of stocks, which is currently over 50 years. Unfortunately, the longest horizon, liquid Treasury series is the 30-year bond, which has an effective duration of about 17 years. Since we have to work with the data we've got, the chart below presents our projection of the 17-year annual total return for the S&P 500, which presently stands at 5.61%. This compares with a yield of 4.32% available on similar duration Treasury bonds. Regardless of which long-term maturity one chooses, the implied equity risk premium is only slightly over 1% annually. Prior to the late 1990's market bubble, the average post-war equity risk premium was closer to 5%, and stocks had much shorter durations and therefore less risk in response to yield changes.
In the chart below, note that the bubble peak of 2000 is associated with a positive outlier for the actual return versus what would have been projected in 1983, but aside from this effect, valuations clearly explain a great deal of the variation in historical market returns.
Again, the likely premium for equity risk is presently among the lowest in history. In fact, the situation is even worse, because in return for just over 1% of additional expected return versus Treasury bonds, equity investors are accepting a duration (and accompanying volatility risk) that is three times as great as that of a 30-year Treasury.
Finally, last week, Jean Claude Trichet, the head of the European Central Bank, provided early indications that the ECB would be stepping up its buying peripheral government debt issued by Ireland, Greece, Portugal and Spain. Of course, the ECB prefers to "sterilize" these interventions, so it can be expected to sell the debt of stronger members such as Germany. Accordingly, yields dropped on the debt of credit-strained European countries, while German yields pushed to fresh yearly highs.
It doesn't take much thought to recognize that, like Bernanke's actions, the actions of the ECB are ultimately likely to represent not monetary policy but fiscal policy. When you buy the debt of countries that have a high likelihood of defaulting on this debt, or will avoid default only by the creation of currency that could have been issued to finance fiscal expenditures, it follows that you are engaging in fiscal policy without the authorization of elected governments.
This is a dangerous and cowardly road. We are allowing 99% of the world to accept budget cuts and austerity in order to defend bondholders from taking losses or having to accept debt restructuring. When bondholders lend money to a financial company or to a country, at a spread over the yield available safe debt, they are explicitly accepting the risk that the bet will not work out, and that they may lose money in the event of a restructuring. When government policy at every level focuses on making bondholders whole, then government policy at every level focuses equivalently on protecting the inefficient and dangerous misallocation of capital. The situation is even worse when unelected officials such as Ben Bernanke and Tim Geithner bend the clear meaning and intent of the law (such as section 13-3 of the Federal Reserve Act) in order to arbitrarily reward private interests. This will end badly. That doesn't mean that we need to persistently avoid market risk until it does, but investors who put these risks too far from their minds are likely to be blindsided as repeatedly as they have over the past decade.
Government Deficits and Potential GDP
As a final note on the U.S. budget deficit, the working group on deficit reduction failed to reach a consensus on Friday. My own impression is that the most efficient tax systems have a very broad base, coupled with flat rates and a large exclusion at low income levels, and tax income as close to their source as possible. Moreover, if we want an efficient tax system that encourages job creation, we have to deal with payroll taxes at the same time, which are presently extremely regressive (more than three of four families pay more in Social Security than in income taxes).
