"Estimating the Stock/Bond Risk Premium" (My Notes)

The folks at Hoisington Investment Management did some great research into the performances of stocks versus risk-free treasuries. Unfortunately I can not post a direct link, but below are my notes on their report, published in the 2002 Journal of Portfolio Management.


"This study sheds new light on the risk premium of stocks over US Treasury bonds, which indicates most research overstates the advantages of stocks over bonds. Our research also indicates long periods when bonds actually outperformed stocks and the conditions that produce these results."

With regard to the first point with the comparison of returns of stocks against bonds, the report goes all the way back to the creation of the S&P500 in 1871. From 1871 to 2001, stocks returned 9.3%,versus bond returns of 5.0%, much lower than most calculations on the matter.

More important for us is the second point- the scenarios under which bonds or stocks outperform. They concluded that the relative performances are most affected by three considerations: the inflation rate; dividend yield of stocks versus treasuries; and the P/E ratio. The following chart shows the 4 best 10 and 20 year periods for stocks and bonds:


The overriding factor has been price change, with deflation good for bonds and inflation good for stocks. But in periods where dividend yields have been excessively more than treasuries, stocks have tended to outperform. And in periods where treasury yields are higher than dividend yields, bonds tend to outperform. Also, a third factor has been the P/E yield. The report goes on to mention the implications of these findings in 2002's environment, and how they expected bonds to outperform .

And, they have been right so far. One problem I do have though is that in the examples they give, dividend yields are exceeding treasury yields. This is something that hasn't happened since the 1960's. But anyways, looking at these factors in today's environment:

1. Inflation rate:
This is the most important factor, and also the hardest for most people to predict. My opinion has been we have too much capacity if anything, because we have been requiring debt and assets to pay for our spending, rather than just income. And it's hard to see people spending considerably more from their already highly indebted levels. But these things are notoriously hard to predict, and who knows whether I'll be right or not.

2. Relative yields
Currently, the yield on a 10-year treasury is 4.57%. Meanwhile, the dividend yield of the S&P500 for 2006 was 1.77%, leaving a spread of 2.80% in favor of treasuries. As I mentioned before though, the dividend yield has not exceeded the treasury yield since the 1960's. It is difficult to say what to make of the fact that the dividend yield has been so low for so long. The common response is companies are reinvesting capital at favorable returns, or they have also been buying back shares. But dividends are money in the pocket today for investors, whereas the benefits of capital expenditures will only show over the long term and have usually been marginal at best.

3. P/E Ratio
The ratio for the S&P500 in 2006 was 17.3, above average but not by much. Still, it is nowhere near the low levels seen before the greatest stock advances.

The report concludes with the following:

We know that over very long terms stocks must outperform bonds, because investors must be rewarded for riskier assets, and we will experience again in the future conditions that warrant higher prospective returns. First, the baseline conditions must change, a process that may result in an extended period when bond returns will equal, or even exceed, returns on stocks.