Central banking is a weird business: Small groups of unelected officials affect the fates of billions. One particular concern is that policies meant to benefit everyone might help some more than others. It's a topic that occupied a good deal of the discussion this morning at the annual confab of academics, traders, and central bankers in Jackson, Wyoming, hosted by the Federal Reserve Bank of Kansas City.
Monetary policy makers usually don't worry about the distributional impact of their decisions. Instead, they prefer to focus on big headline numbers. Just a year ago, Federal Reserve Chairman Ben Bernanke was arguing that the Fed's bond-buying had lowered borrowing costs enough to boost the economic growth by 3 percent and add 2 million jobs. His analysis was based on a paper by, among others, the president of the Federal Reserve Bank of San Francisco.
Two weeks ago, the San Francisco Fed published a new paper refuting their earlier study. (Go figure.) Nowadays, the thinking is that asset purchases don't matter. The latest thinking is that the real punch comes from the Fed's so-called forward guidance on the future path of short-term interest rates. This is because long-term interest rates are always roughly equal to the weighted average of expected future short-term rates. Central banks can therefore reduce the cost of borrowing by promising to keep short-term rates lower for longer. At least, that's the theory.
Both points of view are simplistic. For one thing, interest rates come in many flavors depending on who is trying to borrow from whom. A small business in need of working capital isn't exposed to the same set of interest rates as a prime mortgage borrower or a large technology company or the U.S. government. There is no reason to think that measures that are supposed to lower rates for one of those entities would have any significant impact on the others. Moreover, interest rates are only one constraint on the cost of financing.
These criticisms were raised at this morning's symposium. Hyun Song Shin focused on the Fed's failure to reach small businesses, while Arvind Krishnamurthy and Annette Vissing-Jorgensen argued that the effectiveness of asset purchases depends on what the Fed actually buys.
Shin, a professor at Princeton University, made the charts below, which compare borrowing by big corporations and small businesses. As you can see, large enterprises have issued trillions of dollars of bonds at low interest rates while small business are borrowing less than they were four years ago.
This is a serious problem because businesses with fewer than 20 employees create the most jobs. Unfortunately, those companies are dependent on banks for their financing. The decline in real estate prices, which reduced the amount of collateral available for borrowing, and the stubborn persistence of relatively high interest rates on business loans help explain why the smallest companies have increased their hiring the least since the start of the recovery. That in turn helps explain why the Fed has consistently overestimated the potency of its stimulus measures.
Shin's concern with the details was reinforced by Krishnamurthy, a professor at Northwestern University, and Vissing-Jorgensen, a professor at the University of California, Berkeley. Their paper is worth reading in full, but the gist is that the Fed buying government-guaranteed mortgage bonds boosted the economy but buying Treasury debt wasn’t helpful.
Mortgage-backed security purchases helped because they freed savers and banks to buy other risky assets, like corporate bonds. This effect was most pronounced during the asset purchases at the end of 2008 and the beginning of 2009. It has been less potent since then as the risk of default has receded.
Over the past two years, the Fed pushed mortgage rates lower and encouraged additional lending by banks by gobbling up about half of newly issued mortgage bonds. The focus on new issues has been crucial because mortgage lenders only look at those bonds when deciding how to set mortgage rates. This is why Krishnamurthy and Vissing-Jorgensen recommend that the Fed keeps buying as many fresh mortgage bonds as possible even when it sells off the rest of its portfolio.
By contrast, Krishnamurthy and Vissing-Jorgensen find that Treasury bond purchases lower borrowing costs for the government but don't do much for interest rates in the private sector. Even worse, buying Treasury bonds "may actually reduce welfare" because those instruments "offer unique convenience services" to savers. That’s because those purchases reduced income for bondholders without lowering borrowing rates for anyone else.
We should be troubled by these two analyses. Together, the message is that the tools available to central banks force them to favor certain groups over others. If only there were a way around this problem.