At the March 23 Federal Reserve Board conference (Conference on Central Banking: Before, During and After the Crisis), three prominent central bankers – Jean-Claude Trichet (former head of the ECB), Masaaki Shirakawa (governor of the Bank of Japan) and Jaime Caruana (BIS) – argued that policy makers needed to be aware of the unintended consequences and risks associated with their easy-money and quantitative-easing policies, put in place to address the financial crisis. How to craft an exit from those policies and engineer a gradual and non-distorting shrinkage in central bank balance sheets is a prime concern as economies gradually improve, as illustrated most recently in John Taylor’s op-ed in the March 29th WSJ.
The Federal Reserve has been struggling with some of those issues. One consequence of those deliberations is the Fed’s new coordinated communications strategy. Unfortunately, the change in communications has still left financial markets with considerable uncertainty. For example, most recently the FOMC has explicitly stated that it intends to hold the target Federal Funds rate within its current 0-.25 percent range well into 2014. But at the same time, the FOMC released for the first time participants’ assumptions as to when the first rate move might take place. Interestingly, notwithstanding the 2014 date contained in the FOMC’s policy statement, 6 participants assumed that the first rate increase would not occur until 2015 or into 2016, while 6 assumed that it would occur before the aforementioned 2014 date. The wide distribution of policy views, this long into the recovery, was surprising, at least to this author. The large dispersion reflects considerable uncertainty within the FOMC about the path of the recovery.
To complicate matters, the chairman and other participants, in their speeches, rightly state that the 2014 consensus date for the next policy move is not a firm commitment but is contingent upon the actual path of the economy. Not to be outdone, market participants and media pundits keep raising the possibility of another round of quantitative easing.
With each speech and/or piece of congressional testimony, longer-term interest rates have moved either up or down, depending upon how the comments are interpreted. All of this is insignificant, however, compared with what will happen when the FOMC actually begins withdrawal of the extreme accommodation it has been pursuing. Bond investors, in particular, will not want to be subject to large capital losses on their portfolios as rates rise; and so the probability of a sharp rate shock due to market reaction to even a modest change in the policy rate is extremely large as investors run for the door, all hoping to be first out.
Right now the three key indicator variables of where policy effectiveness sits from policy makers’ perspective are unemployment, inflation, and a measure of economic slack in the economy. We know that many FOMC participants are first and foremost concerned about employment and are willing to keep rates low until the unemployment rate drops to an “acceptable” level, whether that is 7 to 7.5% or some lower number. This preference will make it especially difficult to justify a policy move to stave off inflation until it is actually observed. But because the long and variable lags in the effectiveness of policy, waiting until inflation actually occurs will be too late.
But what will happen if the economy takes off, and inflation breaks out and is fueled by a sudden increase in bank lending as the high volume of excess reserves are converted into required reserves, with the associated multiplier effect on the money supply and stimulus to aggregate demand? This is behind the concern of John Taylor in his March 29th op-ed, especially if the FOMC is behind the curve. Is there a way for the FOMC to hedge against a too-rapid increase in bank credit, thereby tempering a potential rapid expansion of aggregate demand until unemployment improves, without shocking markets and long-term investors with an aggressive increase in the policy rate?
Such a hedged policy may be possible. What if the Fed were to increase bank reserve requirements to, say, 90 percent? The appropriate number may be a bit larger or smaller than that, but for the moment follow the logic of this proposal.
While this might come as a shock to banks, that shock could be mitigated by coordination with the banking industry. The fact is that, given the current level of excess reserves, banks are voluntarily holding a much higher level than what is being proposed. Before the financial crisis, banks holdings of excess reserves were less than 4% of their total reserve holdings, whereas today they are 96%. So an increase in required reserves to even 90% would not be binding, and there would be little or no impact upon bank lending costs, nor would it affect in any way the 25 basis points the banks are receiving for holding those reserves at the Fed. At the same time, as the economy improves, there would be no effect on the increase in the opportunity cost of holding reserves instead of accommodating demands for funds.
The high level of required reserves could also function as a speed bump, limiting the pace at which lending could increase without the need for the FOMC to abruptly and rapidly increase its policy rate to prevent the economy from overheating. But most importantly, the policy would give the Fed substantially more flexibility in how it managed the shrinkage in its asset portfolio. Right now, its two main options are to either hold assets to maturity or to sell them into the market. Holding assets to maturity would be virtually impossible, if it has to rely solely on increasing its policy rate as the main tool for tempering what might otherwise be an overheating economy, since selling assets would the main tool the Fed would have to push rates up.
Selling assets to raise interest rates would also require the Fed to recognize capital losses and reduce its book-value capital, of which it has precious little. It would also avoid having to use the accounting gimmick it put in place to defer recognition of those capital losses – in other words, eliminate having to employ phony accounting to pretend it was book value solvent when it was not. Raising reserve requirements to the suggested levels effectively sterilizes the bulk of the excess reserves on the Fed’s balance sheet, and by implication the assets in its portfolio, and would permit it to hold to maturity most if not all of the assets it has acquired as a result of its liquidity policies, quantitative easing, and efforts to support the housing market.
Some policy makers follow the absorption rate of excess reserves into required reserves as an important indicator of how quickly bank credit and the money supply are increasing. Use of that indicator would not be affected, since there would still be ample excess reserves that could be employed to accommodate needed bank credit as the economy begins to pick up, while at the same time tempering the speed at which the banking system in the aggregate could expand credit and the money supply.
The real benefit of the policy being suggested, however, lies in the use of both a quantity-limiting policy tool (reserve requirements) and a price tool (the Federal Funds rate) to control the pace of the recovery and limit the acceleration in inflation. Furthermore, it would enable the Fed to explicitly articulate an exit strategy for reducing its portfolio over time and commit to financial markets that it will phase down reserve requirements at a pace that is in accordance with a rule or formula that will not unduly constrain bank lending and that is geared to both the maturing of its portfolio and the pace of the recovery. One could envision, for example, a pace of reduction that allowed for a larger amount of excess reserves initially but that gradually declined as the economy expanded, so as to use increases in market rates and bank lending rates to dampen what might otherwise be an overheating economy and an undesired increase in inflation. The FOMC could then reduce the interest rate paid on only excess reserves to lower the attractiveness of holding the residual excess reserves as a marginal stimulus for banks to begin lending without impacting the total volume of reserves. The FOMC would still also have the ability to employ changes in the target Fed Funds rate, but in this case using two policy instruments rather than one solves many problems that would otherwise be associated with the strategy that was put forth last year, which placed almost exclusive reliance upon manipulation of market expectations and a gradual increase in the funds rate.