My article on Dr. Shiller’s idea of Trills was warmly received in some quarters, and not in others. Though I think Trills are a very bad idea for financing the US Government (though I think the US Government is not creative enough in its financing plans), there is one thing that I do agree with Dr. Shiller on — the price of Trills. He says they would trade near $1400 — a 1% initial yield. My view of a 1.06% initial yield largely agrees — the bond math is the bond math; we can argue about assumptions, but given the assumptions, there is only one answer. Comments about risk premium miss the point. Yes, we don’t know the future growth of the economy, but pricing assumes a growth path. Also, these are US Treasury securities — default risk is nonexistent in dollar terms, right? 
Why are Trills a bad idea? First, they seem cheap because of the low initial yield, but they aren’t cheap because the interest rate grows exponentially. Assuming 3.4% average nominal GDP growth, the coupon doubles every 21 years, forever. Suppose we issue a Trill in 2010, and we get a 1% coupon like Schiller suggests. What will future coupons be?
- 2031 — 2%
- 2052 — 4%
- 2073 — 8%
- 2094 — 16%
- 2115 — 32%
- 2136 — 64%
- 2157 — 128%
And so on. Unlike ordinary bonds, the price for calling the issue grows along with the coupon, so there is no escape. The best that can be said is that Trills would guarantee the default of the US Government if issued widely enough.
But now for the illusion of Trills. Because they pay one one-trillionth of GDP each year, some deem them to be the equivalent of owning one one-trillionth of the US economy. Oh no, they are far more valuable than that.
The US economy is subject to all manner of risks, risks that are higher than that of the US Government paying on its debts. Whether thinking of all the producers in aggregate, or all of the consumers in aggregate, the average cost of capital is a lot higher than what the US Government could borrow at in perpetuity. Rather than evaluate the Trills at the rate of the US Government at 4.4%, I think the proper rate is more like 7%, give or take a percent. You might think 7% is too high, but there are significant risks to growth in any economy from government intervention, plague, war, famine, etc. At 7%, the value of a trill drops from $1400 to $400. The difference is the guarantee of payment by the US government, rather than the economy as a whole, so long as the US Government is solvent.
And, that is another significant difference. If the US government is ever credit impaired, Trills will cease to be a share of one one-trillionth of the US economy.
I have suggested before that Dr. Shiller is more creative than the markets will allow. Derivative instruments are facile creatures — we can dream up anything, but what derivatives the cash markets will support is another matter. Shiller’s housing ETFs have gone away.
Trills are government bonds, should they ever be created. They are not shares in the US economy. That distinction is significant to the value of the securities, and the effect on the republic of the United States.
Happy new year! Dr. Shiller spells his name without the “c” in it, you should correct that
Done, thanks Jay.
The coupon rate goes up, but the yield stays, more or less, the same, which is what matters. Ability to pay tracks ability to raise revenue, which tracks GDP, which drives coupon rates. So payments will track ability to pay.
Can’t inflate away that debt? Sounds good to me, that’s one more thing to keep them less dishonest.
Robert — the alternative is debt repudiation/default, which is worse than inflation.
Government debt is a claim on the future earnings of the productive class (us and, hopefully, our heirs) made by the unproductive class (politicians and bureaucrats) and is therefore immoral. When the federal government issues a 30 year bond, they are making people who haven’t even been born yet a party to a contract with little prospect for the unborn to reap any benefit. I say repudiation is the preferred alternative.
Rothbard said it best: http://mises.org/daily/1423
I suppose you can justify your stance with the caveat “if issued widely enough”. However, I repeat what I said before, payments will tack ability to pay. How does this necessarily lead to default?