Econintersect: Charles I. Plosser, President and Chief Executive Officer – Federal Reserve Bank of Philadelphia – has been one of the mavericks in the Federal Reserve System. Because of the rotation of regional Presidents in the FOMC, his dissenting voice in the FOMC meetings is not heard. However, a speech to the Cato Institute’s 30th Annual Monetary Conference today – his dissent was literally the entire speech.
It is very hard to quantify the risks associated with our unconventional policies. But they are real. With our recent extraordinary policies, we have sailed into uncharted territory; we need to acknowledge that, proceed with caution, and continually assess the potential costs and benefits of our policy actions.
Good Intentions in the Short Term with Risky Consequences for the Long Term
Presented by Charles I. Plosser, President and Chief Executive Officer, Federal Reserve Bank of Philadelphia
Cato Institute 30th Annual Monetary Conference — Money, Markets, and Government: The Next 30 Years
November 15, 2012, Washington, D.C.
Introduction
The last five years have been an extraordinary time for the global economy and monetary policymakers. The financial crisis and the severe global recession that followed have tested our resolve, our patience, and our economic theories. To restore the health of ailing financial markets and economies, central banks have driven short-term interest rates to essentially zero, expanded their balance sheets to unprecedented levels, and engaged in market interventions that have blurred the lines between monetary policy and fiscal policy.
These extraordinary efforts were well intentioned. Although it will be some time before we fully understand the effectiveness of various actions, some have credited them with preserving financial markets and saving the global economy from an even deeper recession. Yet, these actions also carry long-term risks for our economies and for central banks.
In my remarks today, I will focus on U.S. monetary policy and discuss some of the longer-term risks arising from our policy responses to the financial crisis and slow recovery. I will then share my thoughts on an approach to monetary policy that I believe would prove beneficial in the post-crisis economy.
Before continuing let me note that my views are my own and not necessarily those of the Federal Reserve Board or my colleagues on the Federal Open Market Committee.
The Risks of Extraordinary Accommodation and Nontraditional Policies
Let me begin by reviewing the extraordinary actions taken by the Federal Reserve as it attempted to maintain liquidity and the basic functioning of our financial markets and subsequently to support an economic recovery.
During the height of the crisis, the Federal Reserve instituted various liquidity facilities for particular segments of the financial system that were under stress. These programs supported primary dealers, the commercial paper market, and money market fund investors. The Fed also gave support to specific individual institutions, including Bear Stearns and AIG, to avert what could have been disorderly failures.
After reducing its policy rate, the fed funds rate, to essentially zero, the Fed instituted several large-scale asset-purchase programs. The first of these programs of quantitative easing, commonly referred to as QE1, ultimately involved purchasing $175 billion of housing agency debt and $1.25 trillion of agency mortgage-backed securities, or MBS; the Fed also purchased $300 billion in longer-term Treasuries in 2009. The unprecedented purchases of significant quantities of MBS were intended to support housing — the specific sector of the economy in which the financial crisis was centered.
As market functioning returned to normal, large-scale asset purchases continued. But the purpose shifted to providing monetary stimulus or defending against the potential for deflation while our policy instrument was constrained by the zero lower bound rather than supporting market functioning or lender-of-last-resort activities. As the economy struggled to recover and deflation became a concern, the Fed implemented QE2, a program to purchase $600 billion in longer-term Treasury securities. Most recently, the Fed instituted QE3, an open-ended program to purchase agency MBS at a pace of $40 billion per month.
In addition to the asset-purchase programs, the Fed has been engaged in a maturity extension program, popularly referred to as operation twist — an intervention that was last attempted, with little success, in the 1960s. The objective of this program is to flatten the yield curve by removing duration from the market.1 Finally, the Fed has attempted to alter public expectations of the future path of monetary policy and the economy by issuing forward guidance about how long it expects to keep the fed funds rate exceptionally low. As of September, that expected date was at least through mid-2015.
As a result of these policy initiatives, the Fed has now held its policy rate near zero for almost four years, its balance sheet is more than three times larger than before the crisis, and the composition of its balance sheet has shifted toward mainly longer-term housing-related and Treasury securities compared with mostly short-term Treasury securities held before the crisis. Indeed, by the end of this year, the Fed will hold almost no short-term Treasuries.
Despite these extraordinary efforts by the Fed, our economy remains lackluster — unemployment remains uncomfortably high and is declining only slowly, economic growth is mediocre, and confidence in the future remains subpar.
Looking at this state of economic affairs, one might conclude that the Fed just hasn’t done enough. Since the Fed seems to be missing on the employment part of its dual mandate, some suggest the Fed can and should continue to pursue more accommodation as long as inflation remains contained. But this is not the only conclusion one could draw from the evidence. Instead, one could conclude that the factors contributing to mediocre economic performance cannot be offset by monetary policy.
This alternative hypothesis should not come entirely as a surprise or as a radical point of view. The ability of monetary policy to influence employment has long been recognized as tenuous at best. Indeed, the current workhorse models in macroeconomics rely on some form of wage or price stickiness to generate real effects of monetary policy. As wages and prices adjust, the effects of monetary policy on the real economy dissipate; in other words, the effects are transitory. In addition, the experience of the 1970s clearly demonstrated that attempts to use monetary policy to pursue an employment or unemployment target can lead to extremely poor economic outcomes, jeopardizing both employment and inflation.
Of course, one might argue that even if there was only a small chance that additional accommodation could put people back to work more quickly, it would be worth undertaking. Yet, that would only be true if the potential benefits of such a policy outweigh the potential risks that such accommodation creates. I join many economists who are skeptical that further asset purchases will have much effect on longer-term interest rates. Even if they do, the declines in long rates are likely to have fairly negligible effects on employment or growth at best. On the other hand, I believe the extraordinary policies the Fed has pursued pose substantive longer-term risks: These include moral hazard, future inflation, and loss of institutional credibility.
Moral Hazard. Let’s first discuss moral hazard. In taking unconventional and unusual steps in recent years, policymakers run the risk of altering the public’s expectations of how policy will be conducted in the future. This is most frequently discussed in the context of the too-big-to-fail (TBTF) problem. In trying to stabilize the financial system, policies led creditors of large financial institutions to expect that the government will protect them from losses. This creates moral hazard and undermines the market discipline that creditors exert on a firm’s risk-taking. Without a clear set of rules or guidelines that tell market participants how such lender-of-last-resort policies will be conducted in the future, the actions run the risk of sowing the seeds of a future credit crisis. Moreover, it is unlikely that regulatory reform as embodied in Dodd-Frank has substantially addressed the TBTF problem. Indeed, some have argued that it has expanded the government safety net, thereby aggravating moral hazard.
Yet, moral hazard risks are not confined to the TBTF problem. By engaging in targeted purchases of housing-related securities, the Fed has affected expectations about what monetary policy will do in the future should the housing market take a sharp downturn. Will market participants price housing-related assets with the expectation that the Fed will protect the market from significant losses? Will investors in other markets expect similar treatment and therefore be encouraged to take excessive risk? Similarly, will holding rates at essentially zero for a long time also spur increased and undesirable risk-taking in the search for higher yields? Certainly, the Fed does not intend to create such moral hazard. Yet, the lack of clear guidelines about future monetary policy can introduce its own form of instability and uncertainty.
Future Inflation Risks. Another potential risk is future inflation. So far, the asset-purchase programs have expanded the Fed’s balance-sheet assets from about $900 billion before the crisis to nearly $3 trillion. Banks are holding $1.5 trillion in excess reserves in their accounts at the Fed. These reserves are not inflationary in the current environment. Indeed, inflation and inflation expectations remain near our goal, despite high unemployment rates or large measures of output gaps.
Yet, history tells us that central banks tend to find it easier to lower interest rates than to raise them. Moreover, it is difficult to identify the appropriate moment to begin tightening policy, even in the best of times. The tremendous expansion in the size of the Fed’s balance sheet complicates the challenges the Fed will face when it comes time to begin exiting from this period of extraordinary accommodation.
Once the recovery strengthens — and it surely will — long rates will begin to rise and banks will begin lending out their excess reserves. Loan growth could be quite rapid, and there is a real possibility that the Fed will have to withdraw accommodation very aggressively in order to restrain money growth and inflation. While economic conditions might evolve very gradually, financial markets are not always patient. As soon as the markets perceive that the Fed might begin to remove accommodation, we could see long-term rates move up quite rapidly. In such an environment, policymakers might need to tighten policy quickly to contain inflationary pressures. Will this tightening require rapid sales of housing-related assets that could potentially disrupt a recovering housing market? The bigger our balance sheet, the more difficult it will be to exit in a way that meets our inflation objective without creating instability in the real economy, thereby undermining our credibility and reputation.
Institutional Risk. The Fed’s recent policy choices also impose other institutional risks. The purchase of large quantities of housing-related securities is viewed by some commentators and policymakers as a type of credit allocation to one sector of the economy in preference to other sectors. I, and others, believe such credit allocations should be in the province of the fiscal authorities, not the central bank. Blurring the boundaries between monetary and fiscal policies can pose institutional risks for the central bank and its independence.
As I mentioned, the Fed’s balance sheet is not only quite large, but it now contains mostly long-term securities. As interest rates rise, if the Fed finds it must sell assets at a rapid pace to restrain inflation, it would very likely incur substantial losses. If so, the Fed may not be able to make any remittances to the Treasury for some years. While this is of little macroeconomic significance, it will not go unnoticed, particularly in an era when the government will be struggling to reduce deficits. This could place considerable short-term pressure on the Fed to prevent those losses by tightening policy more slowly than might otherwise be appropriate. If, instead of asset sales, the Fed tries to restrain credit growth by increasing the interest rate paid on excess reserves, this too would reduce our remittances to the Treasury as more of the Fed’s income would be paid out to the banks holding the reserves. Again, this is of little significance to the macroeconomy, but it is a risk to perceptions about the institution, which eventually may put the Fed’s independence at risk.
It is very hard to quantify the risks associated with our unconventional policies. But they are real. With our recent extraordinary policies, we have sailed into uncharted territory; we need to acknowledge that, proceed with caution, and continually assess the potential costs and benefits of our policy actions.
Monetary Policy Principles
I would like to spend my remaining time outlining my preferred course for setting policy in these uncharted waters. Even in such an environment, or perhaps even more so in such an environment, I believe sound and effective central banking should focus on four guiding principles.
The first principle is to be clear and explicit about the goals and objectives of policy. And in so doing, policymakers must acknowledge what policy can and cannot achieve.
The second principle is for policymakers to make a credible commitment to their goals by describing how they will conduct policy in a way that is consistent with those goals. One way to do this is for the central bank to articulate a reaction function or rule that will guide policy decisions.
The third principle is to be clear and transparent in communicating to the public the policy actions that are taken.
The fourth principle is to strive to ensure central bank independence.
During the last few years, the Federal Reserve has made some important strides in advancing these principles. Most important in my view was the statement of the FOMC’s longer-run goals and policy strategy released in January. This statement made explicit for the first time the FOMC’s goal of a 2 percent inflation target. It also explained why it was inappropriate for the FOMC to establish an explicit numerical objective for the employment portion of our mandate. Indeed, in this respect, I believe the statement helped to explain certain limitations of what monetary policy can achieve.
The Federal Reserve has also made great strides in enhancing transparency. It has expanded the quarterly Summary of Economic Projections, or SEPs, and has begun to include information on each participant’s view of the appropriate monetary policy that underlies these projections. The Chairman now holds press conferences following each of the meetings at which the SEPs are compiled.
A Systematic Approach to Monetary Policy
I believe we could take further steps to improve our communications and reduce uncertainty over the path of monetary policy and reduce moral hazard. One enhancement would be to articulate a more rule-like approach to our decision-making process. This means making policy decisions based on available information in a consistent and predictable manner. One cannot know what the future holds or what future policy decisions will be. Policy will be data dependent, but the data should feed into a decision-making process in a mostly systematic or rule-like way.
Research has shown that more systematic policy can generate better economic outcomes and that there are simple monetary policy rules that perform well in a variety of models.2 Of course, the Taylor rule is the most well known of these simple rules, but a number of variations have been studied, including growth-rate rules and first-difference rules that avoid some of the measurement issues of rules based on gaps or levels. Because we simply do not know the true model of the economy, I prefer to focus on a set of robust rules designed to give good results across a variety of models.
These robust rules tend to have some features in common. They suggest that policy should respond aggressively to deviations of inflation from target but more modestly to measures of economic slack, such as output gaps or unemployment. These robust rules also tend to exhibit inertia that prevents the policy rate from significant swings or volatility.
I believe that by using these robust rules and being explicit that such rules are important guides for policy, we could make policymaking more transparent and make policy actions more predictable. Indeed, articulating rules as guides provides the best kind of forward guidance, which would be helpful in stabilizing the economy and the path of inflation. In this approach, the FOMC would describe its policy decisions in terms of how the arguments in such rules change. For example, we would indicate that we chose to tighten policy because inflation or inflation expectations rose or some measure of resource utilization, such as capacity utilization, unemployment, employment growth, or an output gap, improved. Conversely, we would explain that we took actions to increase accommodation because inflation or inflation expectations fell or some measure of resource utilization weakened. There would be an added benefit of accountability from this approach because it would require policymakers to explain why they might choose to deviate from the guidelines offered by the rules at certain times.
With the fed funds rate at the zero lower bound and the Fed engaging in large-scale asset purchases, now may be one of those times. For example, some economists argue that because the policy rate has been stuck at the zero lower bound for some time, policy should be set at a more accommodative stance than what would be suggested by the standard set of rules. Although some of the standard rules suggest that the current stance of policy is a bit too tight, others suggest that a policy rate of zero is too low, given economic conditions. While the zero lower bound needs to be considered in setting appropriate policy, it does not mean that the systematic approach I advocate should be abandoned or simply ignored. Instead, I would argue that we use the rules as guides and then explain why the zero lower bound might suggest deviating from the prescriptions of those rules when appropriate. This approach will be particularly helpful when all of the rules begin to point to policy tightening.
I believe the systematic approach I have advocated is also preferable to the calendar-date forward guidance that the FOMC is currently providing. I also think that it is preferable to proposals to use quantitative thresholds to convey policy guidance.
The Committee first used calendar-date forward guidance in its August 2011 statement when it indicated that it anticipated that economic conditions would likely warrant keeping the fed funds rate exceptionally low at least through mid-2013. In January 2012, the Committee lengthened that horizon to at least late 2014, and in September, it lengthened it again to at least mid-2015. Date-based forward guidance is problematic. Instead, a systematic approach provides data-based forward guidance. Policy decisions should be made and explained in terms of economic conditions, not the calendar.
I also prefer the systematic or rules-based approach I have outlined to proposals calling for the Committee to indicate that it will not contemplate raising the funds rate while unemployment is above some threshold, as long as inflation is below some threshold. Some view this as a way to move away from calendar-date forward guidance by specifying the economic conditions that might prevail when the Fed begins to reverse course. Once either of the thresholds is met, the Committee will consider whether to take action. That is, meeting a threshold may be a necessary but not sufficient condition for action. In my view, this threshold approach could cause some long-lasting confusion, especially if the thresholds are misinterpreted as the FOMC’s longer-run policy goals. But how do you decide on the right numerical values? Moreover, if numerical thresholds were provided as a way to convey forward guidance for the fed funds rate, a numerical stopping rule would also be needed to convey when QE3 asset purchases could be expected to end. This means we may have multiple thresholds associated with multiple tools. It would be difficult to describe all the various conditions necessary for this multi-faceted strategy and communicate them to the public in a comprehensible and credible fashion. I am concerned that we would create more confusion than clarity. One statement of a systematic policy would be a much clearer way to conduct policy.
Others view thresholds as a way to signal that the FOMC will wait a very long time before beginning to tighten policy. The thought is that this should improve consumer and business confidence in a sustainable recovery and that this will spur increased spending and reduced saving today.
But in order for this to work, the public needs to believe that the Fed is making a credible commitment that it will not deviate from, even if it appears to be desirable to do so at the time. According to this view of forward guidance, the Fed would be trying to manage the public’s expectations in a fully credible way. I don’t believe there is any empirical evidence that we can be successful with such a strategy or that it would be of any quantitative significance if we could. Moreover, even if the action were credible, for it to be successful, the public must understand how the Fed is planning to respond to changes in economic conditions after those thresholds are reached — that is, they need to understand the Fed’s reaction function. Will the FOMC tighten quickly? Or slowly? And on what basis will the FOMC make that decision? So to believe that this approach to forward guidance will succeed in providing stimulus at the zero bound, one must have an extraordinary amount of confidence in the Fed’s ability to manage expectations with complete credibility, even though the Fed has never articulated a systematic approach to policy before.
The systematic approach I advocate is more transparent than thresholds because it gives the public much more information on how those policy decisions will be made, based on changes in economic conditions. Ironically, had the FOMC articulated and followed this approach before the crisis, it might be easier now to credibly commit to keeping rates lower for longer than the public would typically expect.
Conclusion
In summary, I believe that some of the actions the Fed has taken to address the financial crisis and the slow economic recovery, while well intentioned, have created some long-term risks for the economy and for the Fed as an institution. Excessive focus on the short term can result in long-term problems. Avoiding these risks is dependent on the Fed executing a graceful exit from this period of extraordinary accommodation. Such an exit depends on the Fed’s ability to be systematic and transparent about its policy decisions.
Over the past several years, the Fed has taken some beneficial steps toward increased transparency, which I believe will serve the economy well now and in the future. I believe the Fed should continue on this path by more clearly articulating a systematic approach to policymaking, centered on using robust simple rules as guides to both its policy decisions and the way in which it communicates those decisions.
- I note that operation twist could easily have been conducted, or fully offset, by having the Treasury alter the maturity structure of the public debt. Like the purchases of housing-related assets, this action blurs the line between fiscal and monetary policy.
- Most of the macroeconomic models economists and policymakers rely on to assess the effects of policy actions in fact assume that policy is conducted in a well-understood systematic manner.
- The views expressed today are my own and not necessarily those of the Federal Reserve System or the FOMC.