Brad DeLong (2011): The Interest Rate That Did Not Bark in the Night: The Surge in U.S. Treasury Debt and the Non-Reaction of Rates: Tuesday Hoisted from Archives Weblogging
At the very start of the 2000s in the years of the Clinton budget surpluses--remember those?--the U.S. government was repaying its debt at the rate of $60 billion a quarter: each quarter saw $60 billion less of U.S. Treasury debt out there in the private market for savers to hold.
George W. Bush--with assistance from Alan Greenspan and the entire rest of the Republican Party--quickly fixed that: from 2002 to 2007 the average quarter saw net Treasury issues of some $70 billion. Each quarter saw the U.S. Treasury having to find extra buyers for another $70 billion of bonds.
A bunch of us (definitely including me) feared that this shift from prudent to feckless fiscal policy would put substantial upward pressure on interest rates. We were wrong. A weak economy lowered private issues of bonds to fund investment. The desire of China and other emerging economies to keep their currency values low made them eager to soak up every dollar earned by their businesses' exports that they could find and invest it in U.S. Treasury debt. Add to that the emerging private rich abroad for whom U.S. Treasuries became more and more desirable as a hedge, and late-2007 saw the 10-year U.S. Treasury rate exactly where it was when the Clinton surpluses had come to an end at the end of 2001. The market took six years of this swing in bond issues and ate it, without a burp.
Then came the recession. Revenues collapsed. Spending on unemployment insurance and other social insurance expenditures rose. The Recovery Act added an extra $600 billion of debt on top of that. And Treasury interventions in financial markets required debt issues as well. A U.S. government that had been paying back $60 billion a quarter at the start of the 2000s and issuing a net of $70 billion a quarter in the mid-2000s was suddenly issuing $380 billion a quarter of bonds.
Back in the third quarter of 2008 I wondered: "Who is going to buy all of these things?" There were nearly $5.3 trillion of Treasuries held by the public in mid-2008. But we were about to start adding to that at a pace of $1.4 trillion a year: $6.7 trillion by mid-2009, $8 trillion by mid-2010, $9.4 trillion by mid-2011, and we are on target to have $10.7 trillion outstanding by mid-2011--doubling U.S. Treasury debt held by the public in four short years.
Who would buy these things?
And at what prices would they buy and hold them?
It is astonishing. By next summer [i.e., the summer of 2012] the U.S. Treasury will have expanded its marketable debt liabilities by $5.4 trillion--an amount equal to more than half all equities in America, an amount equal to at least one-fifth of all traded dollar-denominated securities. And yet the market has swallowed all these past issues and is looking forward to swallow the next tranches without a single burp. Supply-and-demand are supposed to rule--and a sharp increase in Treasury borrowings is supposed to carry a sharp increase in interest rates along with it to crowd out other forms of interest sensitive spending. What has happened to them?
And the interesting thing is that I almost believed that what we have seen was going to be what would happen. Almost, but not quite. You see, I had read John Hicks (1937). And I had almost believed him.
Let me give you the Hicksian argument about what happens in a financial crisis--a sudden flight to safety that greatly raises interest rate spreads, and as a result diminishes firms' desires to sell bonds to raise capital for expansion and at the same time leads individuals to wish to save more and spend less on consumer goods as they, too, try to hunker down.
In Hicks's model, the immediate consequence is an excess demand for (safe) bonds in the hands of investment banks: bond prices rise, and interest rates falls. As interest rates fall, (a) firms see that they can get capital on more attractive terms adn so seek to issue more bonds, and (b) households see the interest rate they can get on their savings fall, and so lose some of their desire to save. The market heads toward equilibrium. But as the market heads toward equilibrium, something else happens as well: the fall in interest rates and the rise in savings is accompanied by a greater desire on the part of households and businesses to hold more of their wealth safely--in pure cash. And so the speed with which cash turns over in the economy, the velocity of money, falls. And as the velocity of money falls, total spending falls, and workers are fired, and as workers are fired and lose their incomes their saving goes from positive to negative.
Thus the process of return to equilibrium takes two forms:
- interest rates fall, boosting investment and curbing savings.
- spending and thus employment and production fall, further curbing savings.
In normal times, the correct policy response is for the Federal Reserve to inject more money into the economy: through open-market operations it should buy bonds for cash, thus increasing the amount of cash so that even at the lower velocity we still have the same volume of spending, and thus transform the adjustment process from a fall in interest rates, spending, employment, and production to a fall in interest rates alone.
A little thought, however, will lead us to the conclusion that such open-market operations may fail. In them, the Federal Reserve is buying bonds, shrinking the supply of bonds out there--and thus pushing up their price and pushing down interest rates. For each amount that the Federal Reserve expands the money stock, therefore, it puts downward pressure on interest rates and thus on monetary velocity. In the limit where interest rates are so low that people don't really see a difference between cash and short-term government bonds like Treasury bills, open-market operations have no effect because they simply swap one zero-yielding government asset for another.
It is in this situation that we want a government deficit--the government to print and issue the safe bonds that private investors really want to hold. As these bonds hit the market, people who otherwise would have socked their money away in cash--thus diminishing monetary velocity and slowing spending--buy the bonds instead. A large and timely government deficit thus short-circuits the adjustment mechanism, and avoids the collapse in monetary velocity that was the source of all the trouble. And as long as output is depressed--as long as monetary velocity is low and there is slack in the economy--printing more and more bonds will have next to no effect increasing interest rates.
I will never doubt John Hicks again.
But how much longer can this go on?
UPDATE: And here it is, twenty-one months later. Things are still the same--only more so:
Real and nominal Treasury rates are both substantially lower than they were 21 months ago. By all market indicators, the U.S. is still further from reaching the limits of its debt capacity now than it was then: