The December 2018 minutes suggested that there was more debate among policymakers on both the long-run monetary policy implementation framework and nearer-term decisions about the balance sheet than Powell suggested during his press conference (“I don’t see us changing that”). The staff presented on the risks around moving to a long-run operating regime in which excess reserves are lower “than at present” and the composition of the Fed’s securities portfolio is also different than it is now. The staff presentation covered a lot of ground, but the Committee still seems to be, based on the minutes and recent speeches, well behind where it should be with respect to these issues related to the implementation framework.
First, there is no excuse that we can think of for why participants have not agreed on a long-run operating framework, and, in turn, whether they will shrink the balance sheet somewhat further or dramatically further. Markets are worried about the former, and they would be even more concerned if they thought the Committee was even considering the latter. We all know that the current framework of abundant reserves will be retained (albeit with a lower level of reserves than there is currently), though the specifics must be decided. Come on, please announce this decision!
Second, the guidance that the size of the balance sheet would be the smallest level consistent with the effective conduct of monetary policy seems to have caused some confusion, including within the Committee. That can’t mean going all the way down to the kink in the demand curve where there’s a slope. There must be a reserve buffer, and it is hard to understand why this even needs discussion. To be sure, where that kink is can’t be determined precisely except by reaching it. The Committee understands it has to feel its way.
Third, the minutes and recent speeches have begun to clarify that balance sheet normalization is not on autopilot, but too late and with too much confusion. Not everyone is on the same page and this has only recently become apparent. This is simply a problem with the original balance sheet normalization plan. While the FOMC is actively tightening policy by raising the fund’s rate, it makes sense to keep the balance sheet normalization going passively—on autopilot. But as the consensus for further tightening weakens—by mid
year, in our view—it makes no sense to continue passively tightening via the balance sheet. Stop with one tool, but keep going with the other? In any case, the balance sheet will be falling below $4 trillion, not a bad ending place, allowing for a reserve buffer. And, of course, if the FOMC lowers the fund’s rate it cannot be offsetting that by passively withdrawing reserves. So a good discussion by the staff, but participants are behind the curve.
Fourth, we all may have thought that the first easing would still be far away. But it looks like it could come sooner than thought, and the Committee appears unprepared, based on the fact that they are seeking input from outsiders in their review that includes a conference in June, with an announcement following in 2020. Wow, the FOMC may be on the brink of easing by then, if not earlier. Hopefully, participants have been discussing this and will be prepared. This is about both how the proximity to the zero bound changes appropriate rate policy and what the Committee will do at the zero bound. On balance, the Committee appears to be well behind the curve, and a critical curve at that.
The staff warned that, as reserves decline, the effective fund’s rate (EFFR) could become more volatile and rise above IOER. They thought it would be difficult to discern whether such upward pressure would be transitory or persistent. The latter would imply reserve scarcity, i.e., that the kink in the demand curve has been reached.
The question of how much to reduce reserves generated considerable debate among policymakers. One key issue was the level of tolerance policymakers had for “noticeable upward pressure” on the fed funds rate.
In theory, the Policy Normalization Principles (“to hold no more securities than necessary to implement monetary policy efficiently and effectively”) would encompass the option of reducing reserves to the leftmost part of the flat portion of the reserve demand curve. But policymakers noted the risk that being at this portion of the curve would likely come with a “significant” surge in volatility that would necessitate new measures (outlined below).
Instead of having reserves at the absolute leftmost part of the flat portion, policymakers proposed “a buffer of reserves” to make sure they are well within the flat portion. That seems clear enough. The SOMA deputy manager at the New York Fed (Lorie Logan) noted that recent surveys of market participants indicated an increase in the median expected level of longer-run reserve balances (compared with an earlier survey of financial officers).
To ensure effective implementation of monetary policy under these circumstances, the staff reviewed, and FOMC participants commented on, certain measures:
1. Further lowering the IOER rate—currently set 10 basis points below the upper bound of the fed funds target range and 15 basis points above the lower bound.
2. Using the discount window to make reserves distribution more efficient. (Both #1 and #2, which use existing tools, were commented on by “several” participants, who noted the limitations of such approaches under certain circumstances.)
3. Slowing the rate of decline in reserves in approaching its longer-run level.
a. “Some” commented on doing this using standard temporary open market operations, in this case adding reserves via repos, which would partially offset the effect of balance sheet runoff. b. Ending runoff with a “relatively high level of reserves still in the system and then either maintaining that level of reserves or allowing growth in non-reserve liabilities to very gradually reduce reserves further.”
c. Slowing portfolio redemptions. (“Several” participants expressed concern about a slowing being misinterpreted.)
4. Using new ceiling tools. (“Some” participants expressed interest.)
Notably, this is the first time the FOMC minutes mentioned slowing the decline in reserves—either by slowing portfolio redemptions or by offsetting runoff via other means—as a viable solution for operational issues. It was discussed by “participants”—the minutes didn’t specify a number, but it was likely a sizable contingent. That participant considered the possibility of “ending” redemptions with a high level of reserves still in the system, in addition to more moderate measures such as merely reducing runoff, also signaled more willingness to adjust runoff than had previously been disclosed. However, this discussion remains about how best to transition to the longer-run framework. The discussion about adjusting runoff was not about making any near-term adjustments in response to market concerns about monetary policy normalization. Participants clearly are concerned that any adjustments to balance sheet runoff might be misunderstood as signaling a change in what the FOMC sees as the appropriate stance of monetary policy.
Long-run Size of the Balance Sheet
FOMC participants discussed the Fed’s nonreserve liabilities, which together with the eventual level of reserves will determine the size of the Fed’s balance sheet. There didn’t appear to be any new information in this area. “Many” participants noted that there was not a strong connection between the Fed’s monetary policy decisions and the size of non-reserve liabilities, such as currency, the Treasury General Account, and liabilities to foreign official institutions. These liabilities are expected to generally track nominal income growth (and are exogenously determined). “Participants” cited the “social benefits” of such liabilities as well. There’s no hint at this point that policymakers are considering changes to the Fed’s nonreserve liabilities.
“Participants” directed the Fed staff to formulate new communications techniques “to provide information to banks and the public about the likely long-run level of reserves.” In our view, one option would be divulging the staff’s estimate of the likely range of the long-run level of reserves. The public could then infer more easily the likely end date of runoff.
Long-run Portfolio Composition
Fed staff also reviewed the long-run composition of the Fed’s securities portfolio, including how the maturity distribution of U.S. Treasuries would look and how to manage the remaining MBS holdings.
There appears to be no firm consensus on the appropriate configuration of the Fed’s Treasury portfolio. The apparently limited discussion of this topic suggests the FOMC is very far from reaching a consensus. ▪ “Several” noted that having a portfolio weighted more toward shorter maturities would provide greater scope for lengthening that maturity during a downturn.
▪ “A couple” noted that a portfolio with a maturity profile matching the profile of all outstanding Treasuries would provide a more neutral effect.
One area where there is a consensus is MBS. Participants don’t think the Fed should have a substantial presence in the MBS market. Participants considered what they might do with their MBS holdings, which would remain even after the size of the balance sheet is normalized through runoff. “Several” commented on reducing MBS holdings “somewhat more quickly” using “very gradual” sales “sometime after” the size of the balance sheet is normalized.