To say that the risk premium is the difference between the return on stocks and the return on bonds is assuming that markets are constantly in equilibrium. This is not always true. In fact (to rehash an old handout) in the NYT from 2006, there’s an article by Daniel Altman which says,
For decades, the returns on stocks have usually been much higher, relative to bonds, than risk alone would seem to justify — perhaps as much as six or seven percentage points higher. If risk were the only explanation, the difference would suggest that investors were extremely risk-averse, to the point that they would never leave the house for fear of having to cross the street. [link]
Nevertheless, despite short-term blips (both, up and down) in the risk premium, long-term studies have shown that it is mean-reverting or constantly tending towards a steady average. Later, in that same article,
According to research by William N. Goetzmann and Robert G. Ibbotson, two finance professors at Yale, that premium has stayed fairly constant over long periods through virtually all of American history. For lack of a better reason, there may just be something special about American capital markets, so that a high equity premium would tend to revert to some sort of long-run average. In other words, the equity premium may be a partial predictor of future stock returns and even the future growth of the economy. [link]
[I’ve uploaded the paper to my server, if you’re interested in venturing into the realms of technical econ reading]
I blogged about this earlier, but interestingly enough, because the current (narrowly averted?) financial crisis had so much to do with debt and debt-related instruments (like bonds), it put a whole new spin on the risk premium – these instruments became riskier and less liquid than stocks. So what do you think happens to the risk premium when the return on bonds falls below the return on stocks? When bonds become riskier than stocks? Read more.