The combined size of the U.S. stock, bond, and derivatives markets may run into the trillions of dollars, but the underlying money flows on which these bets are based only represent 20-30% of our GDP. The other 70% -- representing mostly salaries and other income flows -- can't be traded against for the most part. And given how badly financial instruments performed in 2008, that may seem like a godsend.
Except to Robert Shiller. In a new paper with York University's Mark Kamstra, Shiller contends that it would be a benefit for society if we could trade against the other 70%. And the way to go about it is for the Treasury to start issuing new debt securities with coupons (i.e., interest payments) tied to the U.S.'s nominal (not inflation-adjusted) GDP:
A small-denomination GDP share might pay, for example, a coupon each year of one-trillionth of that year’s GDP, or about $14.40 at current levels. On this basis, we propose the name “Trill” be used to refer to this new security. Similar to shares issued by corporations paying a fraction of corporate earnings in dividends, the Trill would pay a fraction of the “earnings” of the U.S. Given the characteristics of GDP growth, our valuation of the Trill indicates its yield would be very attractive to the issuer, the U.S. government, and, for the same reasons, would be a useful new source of income to investors who want exposure to income growth and protection against inflation.
One way Trills would be different from ordinary Treasury bonds is that, while the latter payoff no matter what, their payoff has nothing do with the rate of economic growth. Trills, on the other hand, would be budget-stabilizing in that the interest payment would go up during a boom and down during a recession.
Others who could benefit from nominal-GDP linked bonds would be pension funds and pensioners, argue Shiller and Kamstra:
Shiller has been pitching this idea for at least a couple of years, and the U.S. wouldn't be the first country to issue GDP linked bonds (Argentina did so earlier this decade).
How good an investment would such a security be? Shiller and Kamstra say returns could be "as high as the S&P 500, with half the volatility," which, if true, would make for a very nice investment indeed. The pair also think that Trills could eventually make up a very significant 25% of a diversified portfolio. And when you look at correlations, there is a strong one between nominal GDP growth and income growth, but not between nominal GDP and stock-price growth, which makes such a security an attractive way to diversify. The first chart shows one-year percent change in GDP on the vertical axis versus average hourly earnings from 1964 to 2008, while the second shows GDP on the vertical axis versus S&P 500 growth over the same period:
Shiller's other forays into instrument-creation have been less than successful, but I hope the Treasury Department gives this idea serious thought.