Body
The global financial crisis of a decade ago pushed economies into deep recessions. To prevent these recessions from becoming depressions, several major central banks made large-scale purchases of long-term bonds and other financial assets to ease financial conditions and support spending.[1] These actions, known as quantitative easing (QE), were hardly without controversy. Critics warned that QE would fuel inflation, undercut creditors, and damage the credibility of central banks themselves. But as I have argued since the crisis, including in my recent Policy Brief with Brian Sack, QE was a success and if anything should have been undertaken more aggressively. The important thing now is to draw on lessons learned to fashion an approach for QE that smoothly and effectively transitions from conventional interest rate policy at the effective lower bound on interest rates.
The main reason for thinking ahead is that, with interest rates expected to remain at historically low levels, central banks are likely to encounter the lower bound again in the next recession. Having a well-articulated plan for policy at the lower bound is essential. Our Policy Brief shows that QE in the form of purchases of long-term assets in moderate steps and with possible inertia can provide economic stimulus in much the same way as cuts in the conventional policy interest rate.
How Has QE Been Used So Far?
A decade ago, central banks found it necessary to turn to QE once they had lowered their policy interest rates to what they viewed at the time as their effective lower bounds.
However, the structures of these programs differed substantially over time and across central banks. The experience of the Federal Reserve (Fed) demonstrates this dispersion. The Fed’s initial QE announcement in March 2009 involved a very large amount of securities and was viewed as an exceptional one-time action to counter a rapidly weakening economy. And yet the Fed ended up having to do a series of QE programs. By the time of its final program, in September 2012, the Fed switched its approach dramatically and announced that it would purchase a set amount of long-term assets every month, as long as needed, to achieve its objectives.
Meanwhile, in the United Kingdom, a different approach to QE was evolving. The Bank of England (BOE) revised up the total amount of QE assets to be purchased in May 2009, just two months after the initial QE announcement. It soon became apparent that the BOE was willing to revise its target total holdings in response to economic developments on a fairly routine basis, with increases announced in August 2009, November 2009, and so on. In Sweden, the Riksbank began QE purchases in 2015 and followed an approach similar to the BOE, albeit in rather smaller increments.
Japan and the euro area began QE programs in 2013 and 2015, respectively. Following the US lead, they implemented monthly purchases with no set end dates, but they were clear from the beginning that purchases would run for more than a year and possibly many years. In September 2016, Japan pioneered a new form of QE aimed at keeping the 10-year government bond yield near zero rather than promising any specific monthly rate of purchases.
A Proposed Approach: Make QE Policy Comparable to Conventional Interest Rate Policy
When a central bank lowers its policy interest rate to a level below which further declines are either not possible or not helpful, it should switch to purchasing long-term bonds, as the Fed and other central banks have done since 2009. However, the approaches the Fed used to implement QE diverged from conventional policy more than is desirable or necessary.
To achieve a monetary easing comparable to a policy rate cut of 0.25 percentage points, long-term bond holdings should be increased by about 1.5 percent of GDP, and the policy setting should be described in terms of the stock of assets held. Moreover, as with changes to the conventional policy interest rate, changes to bond holdings should have considerable inertia, so that a central bank wishing to make a large policy adjustment likely would do it in several steps. This approach allows the central bank to adjust the path of asset purchases in response to economic developments.[2]
An important benefit of this proposal is to allow financial market participants and other economic agents to better understand the central bank’s approach to QE. Indeed, the comparability between conventional interest rate policy and QE policy should make communicating policy intentions easier. Central banks should include projections of long-term bond holdings alongside their projections of the future policy rate, at least when adjustments to asset holdings are an active policy instrument.
When it comes time to reverse course and tighten policy, the central bank should start by raising the policy interest rate, in part because it is the more familiar instrument and in part because it is important to maintain the precedent that QE assets remain on the balance sheet for a long time.[3] After the policy rate is significantly above its lower bound, QE assets may be allowed to run off gradually.
Conclusion
A return to QE is highly likely once the next economic downturn arrives. Now is the time for central banks to absorb the lessons of the past 10 years and adopt a strategy for using QE when the time comes. The proposal here makes the smoothest possible transition from conventional policy to QE policy, while retaining many of the advantages of conventional monetary policy practices.
Notes
1. Central banks have long used their balance sheets, purchasing assets and making loans, to achieve other objectives such as calming financial panics and influencing exchange rates. These actions typically do have significant macroeconomic consequences, but that is not their primary goal. The Bank of Japan made large-scale purchases of government bonds in 2000 through 2006 in order to provide monetary stimulus, but it limited its purchases to bonds with short remaining maturities and thus had little economic impact. Another element of QE not discussed here is the extension of subsidized loans to the banking system conditional on banks’ expanding credit to the private sector.
2. The BOE came closest of the major central banks to adopting this proposal, especially after May 2009. However, the BOE never stated explicitly that it was aiming to use QE in a manner similar to the conventional policy interest rate nor has it given expectations of future QE asset holdings the same prominence as market expectations of the future policy rate in its quarterly Inflation Reports.
3. Expected future purchases of long-term bonds push down current long-term interest rates, and expected future sales of such bonds push interest rates up. Thus, a purchase of long-term bonds that is expected to be sold off quickly would have little effect on long-term rates.
