Michael Woodford is one of the world's preeminent monetary theorists, which is why you should read the thorough profile written about him by my colleagues at Bloomberg Markets. Buried inside the article is the revelation that Woodford wants the Federal Reserve to cut back on its asset purchases as soon as possible. That wasn't what I was expecting from a known advocate of "unconventional" monetary stimulus, so I asked him to explain his views more thoroughly.

First, a little background. Among central bankers and those who watch them, Woodford is most associated with something called "forward guidance." The idea comes from a 2003 paper he wrote with Gauti Eggertsson, then an economist at the International Monetary Fund. They were trying to come up with a way to stimulate weak economies after short rates had already fallen to zero percent. (Rates can theoretically go below zero but that would probably make things worse.) Their insight was that central banks can theoretically suppress long-term interest rates by promising to keep short rates low for longer than traders expect. All of the big rich-country central banks have since adopted this approach.

Last year, Woodford made a big splash by arguing that the Fed hadn't done nearly enough to boost the economy. He thinks that the U.S. central bank should commit to keeping short-term interest rates close to zero percent until economic output, expressed in current dollars, returns to its pre-crisis trend path. This output gap is currently about $2 trillion per year and rising. Most Fed officials, by contrast, generally expect to start raising rates sometime in 2015, if not earlier. When it comes to interest rates, in other words, Woodford supports much more aggressive stimulus than the people currently in charge of monetary policy.

Woodford also argued that asset purchases were relatively ineffective except insofar as they corroborated the Fed's promises about interest rates. Purchases targeted at particular sectors of the economy, such as housing, might also help somewhat, according to Woodford, although those operations would be closer to fiscal policy than monetary policy. Woodford's agnosticism on asset purchases accords with recent research by Northwestern's Arvind Krishnamurthy and Berkeley's Annette Vissing-Jorgensen, which I discussed in a previous post.

None of this explains why Woodford "welcomes the Fed's intention to taper off its purchases of Treasury securities and mortgage debt." If the economy is still depressed and asset purchases are at worse useless and at best mildly beneficial, why not keep doing them? I e-mailed Woodford, who is currently traveling, and got a response that I think helps explain things.

For Woodford, the most important point is that the Fed's balance sheet cannot keep growing without imposing costs on the financial system and broader economy -- even when inflation is low and unemployment is high. While Woodford didn't explicitly tell me what those costs were, a possible explanation can be found in this brief passage from the paper he presented at last year's Jackson Hole Economic Symposium:

An increase in the safety premium obtained by making “safe assets” (in the relevant sense) more scarce would in itself be welfare-reducing. If Treasuries provide a convenience yield not available from other assets (including bank reserves), then reducing the quantity of Treasuries in the hands of the public reduces the benefits obtained from this service flow.

In other words, Treasury bonds are uniquely useful for savers. When the Fed makes these securities more expensive -- or restricts their supply through asset purchases -- the central bank harms regular savers without doing much to boost the broader economy. Moreover, the relative scarcity of newly-issued Treasury bonds has been causing havoc in the repo markets.

Woodford suspects that the Fed agrees with him. In fact, he thinks that the pace of tapering will (and should) be determined almost exclusively by the size of the balance sheet rather than the health of the economy:

This explains, in my view, how it was possible for Fed officials to indicate that it would likely be time to begin slowing the rate of purchases later in the year, even while admitting that it was not yet time for the tapering to begin last spring. The point was not so much that they felt confident that they could already predict labor market conditions in the remainder of the year, but rather that they could already predict how large the balance sheet would have gotten by later in the year --- and they knew that, barring substantial unexpected developments with regard to economic conditions, they would be concerned by then about allowing the growth of the balance sheet to continue too much further.

Hopefully this explains why someone known as a monetary "dove" can support tapering without being inconsistent.

(Matthew C. Klein is a writer for Bloomberg View. Follow him on Twitter.)