Unwinding Emergency Federal Reserve Liquidity Programs and the Federal Reserve’s Extraordinarily Accommodative Monetary Policy: Implications for Economic Recovery

Laurence H. Meyer*

Vice Chair

Macroeconomic Advisers

before the

Committee on Financial Services

U.S. House of Representatives

March 25, 2010

*This testimony represents my views, and not necessarily the views of Macroeconomic Advisers or my colleagues at Macroeconomic Advisers.

Chairman Frank, Ranking Member Bachus, and other members of the Committee. Thank you for giving me this opportunity to discuss questions related to the already completed exit from the Fed’s emergency liquidity facilities and plans to exit from its extraordinarily accommodative monetary policy.

When and Why Did the Federal Reserve Close its Emergency Liquidity Facilities?

As Chairman Bernanke explained to you in his earlier testimony [1], the Fed has already closed virtually all its emergency liquidity facilities.[2] How did the Fed do this? It’s important to know that these facilities basically closed on their own, as was always the plan. Requests for funding declined and then virtually stopped as markets healed. In essence, all that was left for the Fed to do was just to close the door and terminate the facilities. They are already history.

When the history is written, it will surely conclude that these programs proved to be extremely effective. They prevented an even worse financial collapse than would otherwise have occurred, and hence an even deeper recession.

At the outset of the financial panic, there was an unprecedented liquidity squeeze. Financial institutions that always fund themselves in short-term money markets and invest in illiquid longer-term assets found that they could no longer do so because of perceived insolvency risk and uncertainty about where the risks were located or concentrated. The Fed and other central banks, often in cooperation, acted, as is their role, as liquidity providers of last resort. Their swift, creative, aggressive, and sometimes coordinated actions clearly prevented a fire sale of illiquid assets. Failing to provide this liquidity would have immediately pushed the financial system to at least closer to the edge of an abyss.

The last step in the process of liquidity normalization, and one that was not well understood by the markets, was an increase in the discount rate, specifically an increase in the spread of the discount rate over the funds rate. This spread had been 100 basis points before the crisis, and was shrunk to 25 basis points as part of the Fed’s efforts to inject a massive amount of liquidity. The Fed in this case sought to further encourage banks to access the discount window, its traditional vehicle for meeting emergency needs of depository institutions, and counter the perceived stigma associated with doing so. The spread has now been increased by 25 basis points to 50 basis points. Further increases are likely, and we expect the Fed to move the spread back further in the direction of its pre-crisis level.

What is the relationship between closing the emergency liquidity programs and monetary policy exit, specifically the first hike in the funds rate?

None. Chairman Bernanke fully explained this distinction to you in his testimony.

Raising the discount rate was entirely part of the process to normalize liquidity arrangements, specifically to close emergency liquidity facilities that were no longer needed. Chairman Bernanke emphasized, as clearly as he possibly could, that this increase in the discount rate had absolutely nothing to do with the timing of exit from the near-zero funds rate, and was not therefore a signal about that timing.

This was the one of the early tests of Fed communication. This was also a test of how well the market listens to and understands that communication. In any case, Chairman Bernanke did very well, the markets did not.

What are the tools the Fed will use to unwind its extraordinarily accommodative monetary policy?

There are three actions that define the Fed’s exit from its extraordinarily accommodative policy: actively withdrawing reserves, raising the policy rate, and shrinking the Fed’s balance sheet.

The Fed will withdraw reserves by executing reverse repos involving Treasury and agency securities and by offering term deposits to depository institutions. It will raise the policy rate by increasing the interest rate paid on reserves (IOR), and it will shrink the balance sheet passively through redemptions and prepayments (runoff), and perhaps actively by selling securities, with an emphasis on MBS.

What will be the sequence of steps the Federal Reserve will take to unwind its extraordinarily accommodative policy?

The first step is preparation, and this is already well under way, but not yet complete. The preparatory steps under way include the NY Fed’s small-scale operations with reverse repos to test the plumbing, the infrastructure for its ultimate use of reverse repos to drain reserves. So far, it has conducted such operations only with Treasury and agency debt as collateral. It will soon do so for MBS. The Fed has also expanded the counterparties who can participate in reverse repos with it, and will likely expand this list further, to help the market absorb the large amount of reverse repos that it wants to be prepared to do. The Board also requested comments for a proposal to establish a term deposit facility, another way the Fed will ultimately drain reserves. It is now revising that proposal in light of the comments received. The Fed should be in a position to use these tools to begin to drain reserves by the second half of this year.[3] The NY Fed has clearly communicated to the markets that it believes that it is prudent to ensure that this infrastructure is in place so that the FOMC can exit whenever appropriate, but that this preparation has no implications for the timing of the first increase in the policy rate.

The second step is to revise the policy guidance contained in the FOMC statement. That guidance is intended to give markets advance notice that the time is approaching for an increase in the policy rate. A change in the guidance will be the first definitive signal that an increase in the funds rate may come sooner rather than later. Today the key phrases in that guidance are that economic conditions will likely warrant an “exceptionally low” funds rate for an “extended period.” “Exceptionally low” means an unchanged near-zero funds rate. “Extended period” is vaguer and means different things to different members of the Committee. But it is generally interpreted to mean six months or longer, a period explicitly indicated by a few members of the Committee. Chairman Bernanke will not be pinned down so precisely. In any case, these are the two phrases that will have to be adjusted—presumably in two stages—as the FOMC gets closer to raising the funds rate.

The last three steps in the Fed’s exit, as they were explained to you in Chairman Bernanke’s testimony, are: draining reserves to move away from today’s super-abundant level, raising the policy rate to withdraw the extraordinary degree of monetary accommodation today, and shrinking the balance sheet which, during the crisis period, more than doubled in size. Chairman Bernanke also presented you with an outline of the likely exit strategy, including the tools to be used, and the sequence in which the three steps could be taken: first, start to drain reserves, next, raise the policy rate (but not much later than the beginning of the reserve draining), and still later, if at all necessary, sell MBS to shrink its balance sheet.

The FOMC generally believes that reserve management is a very important complement to the direct role of IOR in raising market rates. In principle, raising the IOR rate allows the Fed to raise short-term interest rates, including the funds rate, independent of how large reserves are at the time. This is the case because no bank should be willing to lend at a funds rate lower than the rate it can earn by depositing the funds at a Reserve Bank. The IOR rate should therefore be a tight floor for the funds rate.

However, in practice, the funds rate has consistently traded soft to the IOR rate.[4] The Fed believes it can tighten this relationship by reserve management, that is, by reducing reserves, to shrink and stabilize, and hopefully ultimately close, the gap between the funds rate and the IOR rate. This is viewed as desirable, in part because it will allow the Fed to target the funds rate, much in the same way it had done in the past.

Because it would like to tighten the relationship between these rates before it begins to raise rates, the Fed prefers to start removing reserves before it hikes the IOR rate. However, even beginning to actively withdraw reserves will be taken as a definitive signal of an impending increase in the IOR rate, no matter what guidance the Fed offers at the time. So the FOMC plans to start this draining only when it is pretty sure that an increase in the funds rate is around the corner. So the first part of the sequence is to drain reserves first, and only slightly later, to raise rates.

The final step may be to sell MBS. This is the most controversial of the steps. Several on the FOMC do not see the urgency in shrinking the balance sheet faster than would happen passively through runoff. They prefer not to risk a very adverse market response that many in the markets expect in this case. This is where my sympathies lie. However, if the FOMC does sell MBS back to the market, Chairman Bernanke told you that this would take place relatively late, only when the economy is in a sustainable recovery and once markets are back closer to normal. This would also likely take place well after the FOMC begins to raise rates, and will almost certainly be gradual and very well communicated in advance to the markets. Under these circumstances, the possibility of a very adverse market response will be significantly diminished. Still, the majority appears to believe that just allowing the balance sheet to shrink passively and slowly by runoff alone will take longer than they prefer, a minimum of five to seven years, and to get it fully done, it could take a couple of decades. On the other hand, most market participants believe that there would be absolutely no adverse response to taking such a long time. In this case, I believe that relying on runoff alone is the most sensible direction.

Can the unwinding efforts be carried out without unwanted inflationary effects and, if so, how can this be accomplished?

Yes, by exiting at the right time. I cannot and neither can the FOMC guarantee that the Fed will exit at the right time. This is its biggest challenge. The Fed is operating in unchartered waters: It has never had such a challenging exit. Don’t expect the Fed to be perfect. Nobody is perfect. However, we are fortunate to have Chairman Bernanke navigating the ship. Nobody could do a better job of getting the timing right. But there will still be a debate that will linger virtually forever about whether the Fed did in fact exit at just the right time.

The right time, of course, means not too early, so as not to the stunt the nascent recovery and when inflation is still “too low” (and expected to remain so for some time), and not too late, resulting in unwanted inflation.

I will focus on the inflation risk, as this is what you asked about. Let’s start by highlighting three different forecasts: my firm’s, the consensus, and the FOMC’s. Ours is the most optimistic about growth, but all three share one property: The unemployment rate will remain very elevated for a very long time. It is likely to be above 9% at the end of this year and perhaps 8% - 8½% at the end of next year. That is relative to a full employment level that we and the FOMC place at 5% or slightly higher. Core inflation is already below 1½% and appears to still be falling. We expect it to be 1% this year and slightly lower next year. The last FOMC forecast showed inflation just below 1½% for the same period. Let’s assume that the FOMC tightened at the end of this year. What would you be asking Chairman Bernanke at the Monetary Policy Testimony in mid-February 2011? Would you be asking why the Fed waited so long, until the unemployment rate was already close to 9% and core inflation had already stabilized near 1%, before you tightened? I doubt it. So I hope you can understand why the majority of members of the FOMC are not worried about inflation today and won’t be for some time. As a result, we don’t expect the FOMC to raise the policy rate until mid-2011, and, if the pessimists are right about growth, it could be much later.[6]

It is interesting, nevertheless, that some, including a few very well respected economists, worry that inflation will be very high over the medium term. Some even believe that we are headed to hyperinflation. There is no chance that either outcome could happen as long as the Fed remains independent. So let’s be clear where the risk of very high inflation comes from. It is not the Fed. It is you.

We could get to very high inflation in one of two ways: a major and dramatic but almost inconceivable policy error by the Fed, essentially a total disregard for its price stability mandate, or action by the Congress to force the Fed to monetize deficits to avoid soaring interest rates that would otherwise occur. The former simply could never happen. But the latter is entirely up to you. It is precisely to avoid such an outcome that governments around the world have delegated monetary policy decisions to independent central banks. Given this concern, Congress should be careful to avoid doing anything that even gives the appearance of infringing on the independence of the Fed, including authorizing the GAO to audit monetary policy.

Have the Fed’s extraordinary monetary policy actions worked?

The Fed’s extraordinary policies include first pushing the funds rate to a near-zero level and then keeping it there for an extended period, along with turning to nonconventional policies once it faced the zero nominal bound, specifically “credit-easing” policies to provide further stimulus.

The FOMC has had a near-zero rate policy only because it cannot lower short-term interest rates into negative territory. If it could, it would have. In the absence of that option, monetary policy should be understood as being dramatically restrictive, even at the near-zero rate. We estimate that the Fed would have to lower the funds rate to -4% today to align it with the appropriate degree of stimulus. Absent being able to do this, the only option was to implement unconventional policies, despite the great uncertainty of how well these would work. The major form of nonconventional policy is “credit-easing” policies. Credit-easing policies specifically involve intervening in and buying assets, especially those with longer maturities, in markets where the flow of credit is impaired as a result of the financial crisis, with the hope of both increasing the flow of credit and improving the terms on which credit is offered, or otherwise moving to lower longer-term rates relative to the near zero funds rate. In this episode, the Fed refers to these policies as Long Term Asset Purchases, and these include both the purchase of longer-term Treasury securities to push down long-term rates relative to short-term rates, and MBS purchases to lower the spread between mortgage rates and longer-term Treasuries. The Fed has already ended its Treasury securities purchase program and will end its MBS program next week. The decision to end programs that were providing additional stimulus mainly reflected concern that further expanding the balance sheet would make exit even more challenging and increase the risk of a policy error, and therefore unwanted inflation.

Lowering the funds rate to near zero certainly helped lean against the recession and then support recovery. The Fed did as much as it could possibly do with its conventional instruments. The effectiveness of credit easing policies is more controversial and different economists reach different conclusions. Qualitatively, we agree with the NY Fed’s recent study which estimates that the purchase of Treasury and agency securities had important effects, lowering longer-term interest rates and narrowing mortgage spreads.[7]

Nevertheless, there is considerable skepticism, including some inside the FOMC, that renewed purchases of longer term Treasuries and MBS would be successful in providing further stimulus. This assessment is relevant to how much additional stimulus the Fed could achieve by, for example, renewing its asset purchase program, if the economy were to slip into a double dip recession. The realistic answer is not much.

If monetary policy is not sufficient to spur economic growth, what options are available?

Here the answer is obvious. Once the Fed is driven to a near-zero funds rate, what we call the “zero nominal bound”, the burden of stabilization policy shifts to you. This is very challenging given the difficulty of Congress delivering action in a timely manner. This is why the conventional wisdom has been to encourage Congress not to try to engage in short-run stabilization policy, and leave that entirely to the Fed. But given the Fed is at the zero nominal bound, the only option for significant further stimulus is fiscal stimulus. Fortunately, in the current episode, Congress did pass a timely and sizeable fiscal stimulus. One can argue about the size and composition, but one cannot argue that fiscal stimulus was not needed. Most economists, as reported by the CBO, believe that the fiscal stimulus was effective. Our estimate, and this is close to the consensus view, is that it raised growth by 1½ percentage points from the third quarter of last year through (prospectively) the second quarter of this year.

However, the stimulus is almost completely temporary, as was appropriately the intention. As a result, the positive effect will reverse and the effect will turn negative, even by the end of this year. This swing is one reason why forecasters generally see as the real test of the recovery the performance of the economy after the middle of the year when fiscal stimulus first diminishes and then reverses. To be sure, while a permanent stimulus would have been more effective, it would have been irresponsible to do so, given that the budget deficit is already on an unsustainable course, as everyone knows.

It is nevertheless true that the view that the stimulus package was effective is not universally held among economists: One uses different models and gets different results. I should point out, though, that many of the contrarian views that are politically driven don’t really hold up to even a very modest degree of scrutiny.