When the short-term policy rate hits zero, as was the case in most developed countries not too long ago and is still the case for Japan now, how can the central bank further loosen its monetary policy? One option is to purchase government bonds at longer maturities, and hope that this will push their rates down.
Economists and practitioners call the menu of government bond rates the yield curve. Lining up from the very short to the very long, the curve is usually upward sloping, meaning that the bond yields of longer maturities are typically higher. What the Federal Reserve, the Bank of England, and the Bank of Japan (BOJ), among many others, have been trying to do is to bend and shift their own yield curve.
Make no mistake, while the BOJ recently announced the plan of “yield curve control,” it does not mean that the BOJ is just starting to purchase longer government bonds. The BOJ has been doing that since the start of Abenomics in 2012 (not to mention the many similar efforts in the 1990s), and the BOJ is only changing the policy target from the quantity of bonds purchased to the price of the bonds purchased. In particular, the BOJ aims to keep the 10-year bond yield at around zero.
Will it work? Theoretically, there is a convincing reason that it will. Empirically, the evidence that it will is somewhat weaker.
The Yield Curve
Let’s start with a crash course on the yield curve. How is it determined? Why does it take a particular shape on a particular day? The standard answer is arbitrage. Just like a USD 100 bill will not be left lying around on the street for too long, bond yields that are “inconsistent” will be eliminated by investors. Suppose a particular bond is too cheap (or its yield is too high) — investors will be tempted to buy this bond and short the others. These greedy investors will make sure that the bond yields are in line.
According to this view, the yield curve is a combination of two things: the expectations of future short yields and compensation for the risk-taking of buying a particular bond. For example, the 10-year yield tells us something about what the short-term rate is going to be in the next 10 years, but part of it is also the reward of buying a risky (as its price does fluctuate) asset. There is less of a consensus on how to distinguish the two though.
Now go back to the bond purchase program. In 2011, the Fed announced a plan of selling shorter government bonds and buying longer ones simultaneously. Dubbed “Operation Twist,” the policy aimed at raising the shorter rates and reducing the longer ones, or, to put it more graphically, flattening the yield curve. If you prefer an older example, go back half a century earlier and we have another “Operation Twist” under the Kennedy administration: raising the shorter yields — hoping to strengthen the USD and reduce the trade deficit — and reducing the longer yields, hoping to encourage more private consumption and investment. The Bank of England and the European Central Bank have both purchased a large amount of long-term government bonds, and now we have the BOJ trying to stabilize the 10-year yield.
With arbitrage and the two components mentioned above held constant, it is unlikely that the effort of the central bank will have much effect. Suppose the central bank buys a large amount of 10-year bonds — what will happen? Seeing that the 10-year yield is going down, investors will take the opportunity to sell the same bond to the market, counteracting the action of the central bank.
It is unclear how this shift from targeting the quantity to the price of the bonds is going help with raising actual and expected inflation.
But a recent revival of an old idea provides a different answer. Back in the 1950s, there was a popular theory of the yield curve called the “preferred-habitat model.” This theory with a strange name is actually quite simple: some investors have a strong preference for government bonds at some maturities, and their behavior is able to move the yields of those bonds. One major example is pension funds that are required to hold bonds of long enough maturities. The modern version of the theory argues that, when both arbitragers and preferred-habitat investors exist in the market, the behavior of the latter can have a large impact: arbitragers, though greedy, may find it too risky to bet against these preferred-habitat investors.1
The central bank is one major preferred-habitat investor, and its sizable purchase of specific government bonds (like what BOJ is doing to the 10-year bond) will put a dent on the yield curve.
The new theory of preferred habitat is clever, but is it supported by empirical evidence? Do these eye-catching bond purchase programs work?
The most direct evidence is found by looking at the yield curve right before and after the announcement and implementation of purchase programs. This method, called event analysis, is relatively well accepted by economists. Over a short period (say a couple of trading days), it is more acceptable to assume that not many things are going on in the bond market, and we also have a good idea of what those things are.
Looking the episodes of bond purchases in the US, the UK, and the Eurozone, there is strong evidence that the relevant yields went down after the announcements (for about a tenth to half of a percentage point).2 Reactions on the days of implementation were much weaker, unless there was a surprise element in the program (e.g. an unexpected purchase at certain maturities). Response from other related markets, like that of corporate bonds, was a lot weaker.
Wait, central banks clearly are hoping for an effect that lasts for more than a few days, right? Is it there?
This is a much harder question. For a few days, we have a good grasp of the demand and supply movements in the bond market. But for a few weeks, and not to mention a few months, the number of things that can happen in the bond market becomes impossible to trace and handle: rise and fall of risk appetite, actions of central banks in other countries, and many other local and global, visible and invisible shocks. Unless one has a convincing way of taking care of all these factors (good luck convincing half of the economists you meet!), it is not feasible to find out the longer-term effects of these bond purchase programs. Without that link established, we just have no idea how any such program is moving the economy to the right, or just any direction.
Bottom line: central banks have some ability to boss the yield curves around, at least for a while.
Back to Japan. The ability of the BOJ to move the yield curve notwithstanding, it is unclear to me how this shift from targeting the quantity to the price of the bonds is going help with raising actual and expected inflation (which is now the major, if not the only, goal of the BOJ). It does stabilize the profitability of the banks, which usually borrow short and lend long, but with corporate and mortgage borrowing rates already quite low in Japan, the whole “yield curve control” idea seems a lot like a stalling move.
1. One major recent paper on this topic: Greenwood, R., & Vayanos, D. (2014). Bond supply and excess bond returns. Review of Financial Studies, 27(3), 663-713.
2. For a study on the recent Operation Twist by the Federal Reserve, see: Swanson, E. T. (2011). Let's twist again: a high-frequency event-study analysis of operation twist and its implications for QE2. Brookings Papers on Economic Activity, 2011(1), 151-188; for a similar study on the UK experience, see: Breedon, F., Chadha, J. S., & Waters, A. (2012). The financial market impact of UK quantitative easing. Oxford Review of Economic Policy, 28(4), 702-728.