Conventional wisdom continues to believe that soon enough, as has been paraded by the various Fed presidents, the Fed will commence various tightening steps, commencing with the termination of reinvestments of various maturing securities holdings, a process that would lead to gyrations in the IOER (and thus the IOER-GC spread which as has been discussed recently has gone negative due to the FDIC assessment fee). Following the reinvestment decision, the Fed would next proceed to drain excess reserves using various operations such as reverse repos, term deposits, and SFBs (a process which many doubt would success when the total amount of excess reserves is set to hit $1.6 trillion shortly). The last step in the Fed's balance sheet renormalization would be to proceed with outright asset sales of its $2.6 trillion in Treasury and agency holdings (as for those billions in Other Assets, nobody knows). Barclays' Joseph Abate does a great summary of the pitfalls attendant each and every step in the process: "In asserting the supremacy of the Fed funds rate as the primary policy tool, the minutes outline the central bank’s longer term objective. The Fed hopes to eventually establish a corridor system – where the FF target is set between a lower bound of IOER and an upper bound of the discount rate. This would require the Fed to drain enough and to shrink its balance sheet sufficiently to push the effective funds rate over IOER and not merely eliminate the current -16bp spread. This might take a few years to accomplish. And in the process, the Fed would probably need to restore bank confidence in the discount window, which was shaken after the central bank was forced to disclose who had borrowed from the facility during the financial crisis." Abate concludes: "Taking all these into consideration, the April FOMC minutes indicate the Fed faces a pretty complicated task." Luckily, the Fed is most certainly aware of this complexity awaiting it as things return to normal. Of course, this whole discussion will be moot if and when the Fed, instead of tightening, proceeds with another monetary loosening episode, which as Jim Grant explained will go from QE 3 to QE n, in which case none of the below is even remotely relevant.
From Barclay's Joseph Abate
Since last August, the Fed has reinvested its MBS prepayments into Treasuries. And it has always fully reinvested its maturing Treasury holdings. The April FOMC minutes indicate that the first step in the normalization process will be a decision about re-investments – initially with respect to MBS. The minutes also suggest that Treasury reinvestment could be adjusted or terminated simultaneously with the MBS decision or it might occur later. We expect the reinvestment decision will be announced at a future FOMC meeting – a “starting gun” or flare that the tightening process has started. Consequently, term front-end rates should move sharply higher once the reinvestment decision is announced.
Clearly, given the level of mortgage rates and the average coupon of the MBS holdings in the Fed’s portfolio, reinvestment or prepayments is probably more of signal than a practical tightening in the FF-IOER spread. The Fed has roughly $280bn in maturing securities holdings in 2012. But as we have noted in the past, the relationship between the effective Fed funds rate and IOER is almost non-existent when bank reserve balances exceed $900bn. (As of Wednesday, balances were just more than $1.5trn). Consequently, full rolloffs of the Fed’s Treasury and MBS holdings is probably insufficient to “establish a tight link” between the two rates. And, as we note, the full roll of the Fed’s Treasury portfolio is impractical from a securities lending perspective as the lack of new OTRs would reduce the effectiveness of the Fed’s securities lending program.
As an aside, the April FOMC minutes also put to rest a largely theoretical discussion about what the optimal target should be – the Fed funds rate or IOER. Because IOER is an administered rate, the Fed could set it at whatever level it chose and the size of the spread to the FF rate and its balance sheet would matter far less. That said, the Fed plans to achieve its funds rate target by allowing assets to roll off without replacement and draining bank reserves via temporary operations like reverse repos, term deposits and SFBs.
At some point after firing its starting pistol with its reinvestment decision, the Fed plans to begin draining bank reserves via temporary operations including reverse repos with money funds, term deposits, and SFBs. The capacity of each of these programs to absorb surplus bank reserves is highly uncertain – probably even inside the Federal Reserve. Our sense is that the capacity of the Fed’s term reverse repos might approach $350bn ($250bn with money funds and $100bn with its traditional dealer counterparts). However, the willingness of the money funds to do term repo depends heavily on their need to maintain thick (10%) overnight liquidity buffers and the fact that, despite a 7-day put, repos with the Fed will not be very liquid. We believe that these operations instead will crowd out money fund’s willingness to purchase term unsecured bank CP and deposits.
Likewise the ability of the term deposit program to drain bank reserves may depend heavily on the willingness of non-US institutions to lock up their overnight balances at the Fed in non-lending and significantly less liquid deposits. These institutions – without significant clearing or settlement operations – hold almost $800bn in overnight deposits at the Federal Reserve. We estimate the Fed might be able to convert $250bn of overnight reserve deposits into term, although a larger amount is possible.
Asset sales – the final frontier
The FOMC spent a considerable time discussing the prospect for asset sales. With respect to its agency holdings, most participants felt that sales should occur after the first hike in interest rates. Moreover, the sales would seek to return the Fed’s portfolio to all Treasuries in a gradual process – that is, about 5 years.3 And these sales would be predetermined and pre-announced to increase transparency and reduce any distortion in the market. The FOMC further emphasized that these sales would most likely occur in addition to the rolloffs and temporary draining operations. Their timing would be determined by the speed with which the Fed’s other operations managed to establish a link between FF and IOER.
Given our estimates for term reverses and deposits – even assuming a full $200bn return in SFBs – it might be difficult to eliminate the existing spread between the two policy rates without asset sales fairly early in the policy normalization process.
In asserting the supremacy of the Fed funds rate as the primary policy tool, the minutes outline the central bank’s longer term objective. The Fed hopes to eventually establish a corridor system – where the FF target is set between a lower bound of IOER and an upper bound of the discount rate. This would require the Fed to drain enough and to shrink its balance sheet sufficiently to push the effective funds rate over IOER and not merely eliminate the current -16bp spread. This might take a few years to accomplish. And in the process, the Fed would probably need to restore bank confidence in the discount window, which was shaken after the central bank was forced to disclose who had borrowed from the facility during the financial crisis. Without this, bank willingness to use the window could be low, with banks willing to pay extremely high rates in the overnight Fed funds market to avoid “going to the window.”
Taking all these into consideration, the April FOMC minutes indicate the Fed faces a pretty complicated task.