The Risk of Common Investments

I'm frustrated by how we economists have failed to incorporate most of wealth into our theory of asset pricing. The CAPM says that a stock needs a higher expected return if its return is more correlated with the return of the stock market as a whole. That's a good start. It is true, too, that it is possible to hold a diversified portfolio of public stocks, whereas other assets such as private business stock can't be held by everybody.

I worry about other assets. How about bonds? Surely they should be in the CAPM, since they are public and easily diversified into.

How about housing? That's the asset most people hold. And it isn't valued well by economists. It is a hedge against housing consumption risk. If rents rise, then if I own my house, I am insulated. But when I sell my house, I do face risk. Also, if rents rise, maybe I'd like to consume less housing. The optimal house ownership contract isn't what we normally observe. It would involve some insurance against resale capital gains and losses, and adjustment for desired amount of house consumption.

How about labor income? Labor is our greatest wealth-- human capital, and just plain labor endowment. It's risky, and the risk is correlated with the stock market. I'd like a stock that does well when my salary goes down.

The Consumption CAPM is an advance. It notes that we want to have higher stock returns when we want higher consumption. That's odd, though, because consumption is endogenous. Really, we ought to look for the correlation between a stock's return and the return on wealth as a whole, which would be the ratio of GNP to wealth. We should look for the correlation between GNP and stock return, not consumption and stock return. But no, that's not right. Ideally, we'd want the change in total wealth, including labor wealth, which is not the flow of GNP, but a change in a capitalized value of future GNP. I'd like a stock which is uncorrelated with other stocks, and uncorrelated with changes in the value of my labor.

There's another problem. THe price of stocks is determined by the kind of people who buy them. Poor people don't. The only labor wealth that matters to stock prices is that of the people who buy stocks. So what we want to measure is the value of public stocks, bonds, private companies, housing of people who hold stocks, and labor wealth of people who hold stocks.