“Now, about that risk premium. How does your current assessment of a risk premium of 1.5% compare to history?”
– FSD
This is a good question, FSD, and gets to the heart of the matter. Long story short, since 1926, the 3-month treasury bill has averaged around 3%, the 10-year treasury around 6%, and stocks have returned around 11%.
So, historically – “historically” being the operative word – the risk premium for stocks over short-term treasuries has been around 8%/year and over long-term treasuries about 5% per year.
The question is what SHOULD the risk premium be. After all, at 1926, the typical starting point for a lot of the long-term studies, stocks were “cheap” by today’s standards. Consequently, the long-term historical returns are skewed by this fact (although this skewness declines each year). Pick a different starting point – say 1968 – and you’ll get a much smaller risk premium because stocks were more expensive. Many researchers have suggested that the long-term risk premium realized by stock investors was too high with hindsight and suggest that the “real” risk premium over the 10-year treasury “should” be around 3.0%-3.5% based on relative risk and volatility under a “rational expectations” framework. I agree with this approach.
So, here’s my opinion. 30-day treasuries should yield about the rate of inflation over the long term (you shouldn’t get much a premium over inflation for taking no risk); let’s use 3% inflation just for argument’s sake. The 10-year treasury should yield some premium over short-term treasuries due to duration risk. In my view, this number should also be around 3% – that’s what it has been historically. In other words, historically investors have said, “If I’m going to invest for 10 years I want a 3%/year return premium for doing so because I don’t know what’s going to happen in the interim.” I think that’s a rational expectation.
Likewise, I think it’s rational to expect to earn another 3% per year over the 10-year treasury for taking on stock market risk. Again, these numbers aren’t exact (obviously), but they’re good ballpark numbers in my view based on what I think a reasonable person should expect given the relative risks involved.
Clearly the risk premium shrinks and expands over time based on participants’ appetite for risk and judgements about the future. But I think the 3%/6%/9% (approximate) framework is what we’ll see over the very long term as a mean-reverting series. Why? Because in the absence of some other, more logical view, it just makes sense.
So, the problem right now is that I think the current high margins will mean revert over the next couple of years AND the risk premium will expand to 3% (or so). The combination of these things means stocks are likely headed down. Whether that’s 20% or 40% I don’t know. But it’s going to be a surprise to most folks and it’s going to hurt. That’s how markets work.
[One other thing: the chart you posted is essentially the so-called “Fed Model,” which has been thoroughly discredited as a tool for equity valuation from both theoretical and empirical standpoints. Google “problems with fed model” and you can read about the model’s shortcomings, which are many and varied.)
