Fed communication has moved significantly towards a much more dovish tone in the past two months. The change in communication has been twofold, both on rates (the Fed became "patient" and "flexible" on the rate outlook) and on prospects for the so-called quantitative tightening (no longer any "autopilot" in balance-sheet runoff). In this piece we focus on the second tool of Fed policy, analyzing rationales and targets behind balance sheet normalization, which have been detailed and widely expressed in recent Fed communication released by Chairman Powell, other Fed governors and the minutes of the January FOMC meeting. An earlier end to "quantitative tightening" (QT) has become likelier, working in combination with a more dovish stance on rates.
The minutes from January FOMC meeting released on Wednesday confirmed the dovish stance of the US central bank. This was particularly the case for the so-called “quantitative tightening” (QT), as the minutes revealed that previous Powell statements about central bank balance sheet normalization were almost unanimously shared by Fed governors: "almost all participants thought that it would be desirable to announce before too long a plan to stop reducing the Federal Reserve’s asset holdings later this year." In a nutshell, the Fed looks posed to end balance sheet normalization sooner than later, probably in late 2019 and with a final size looking much higher than pre-GFC levels. Fed members’ view on rates looks more split between "several" policy makers stating that "if the economy evolved as they expected, they would view it as appropriate to raise the target range for the federal funds rate later this year," and others underlying that "rate increases might prove necessary only if inflation outcomes were higher than in their baseline outlook."
Powell addressed the QT issue at January FOMC
As widely known, Fed communication has moved significantly towards a much more dovish tone in the past two months. The change in communication has been twofold, both on rates (the Fed became "patient" and "flexible" on the rate outlook) and on the prospects for the so-called quantitative tightening (no longer any "autopilot" in the balance-sheet runoff). In short, the message on Fed strategy was quite clear: a "pause" on rates is closer, and the Fed has revised the QT outlook, as well. However, as we see it, in January the markets focused more on the change in the rate outlook, and details were lacking on the second point. In this respect, clearer indications on the process of balance sheet normalization came later, at the first FOMC meeting of 2019, by the end of the month. At that time, in our view, Fed chair outlined quite clearly Fed guidelines for managing the end of QT.
Powell sent three major messages on QT during the press conference following the January 30 FOMC meeting:
The ultimate size of the Fed's balance sheet will be calibrated on financial institutions' demand for reserves, plus a buffer. In Powell's own words: "Settling this central question clears the way for the FOMC to address a number of further questions regarding the remaining stages of balance sheet normalization. The decision to retain our current operating procedure means that, after allowing for currency in circulation, the ultimate size of our balance sheet will be driven principally by financial institutions’ demand for reserves, plus a buffer so that fluctuations in reserve demand do not require us to make frequent sizable market interventions."
The estimate of this reserve level is much higher size than the pre-GFC world and has increased considerably in the past year: "Estimates of the level of reserve demand are quite uncertain, but we know that this demand in the post-crisis environment is far larger than before. Higher reserve holdings are an important part of the stronger liquidity position that financial institutions must now hold. Moreover, based on surveys and market intelligence, current estimates of reserve demand are considerably higher than estimates of a year or so ago. The implication is that the normalization of the size of the portfolio will be completed sooner, and with a larger balance sheet, than in previous estimates."
At coming meetings, the issue will be addressed, and a decision could be taken and communicated. 2019, therefore, may mark not only the end of rate normalization, but also the year of the last leg of balance sheet normalization. "In light of these estimates and the substantial progress we have made in reducing reserves, the Committee is now evaluating the appropriate timing for the end of balance sheet runoff. This decision will likely be part of a plan for gradually reaching our ultimate balance sheet goals while minimizing risks to achieving our dual mandate objectives and avoiding unnecessary market disruption. We will be finalizing these plans at coming meetings."
The big question mark is obviously the likely target of the Fed: at the time of the latest FOMC meeting, the market consensus was already broadly pointing to a level of excess reserves in the USD 1tn area. This would be considered consistent with a balance sheet size close to USD 3.5/3.6 tn (down by around USD 400/500 bn from the current level of close to USD 4 tn).
In the Q&A following the press conference, Powell, understandably, did not ratify any number but seemed comfortable with consensus indications. This was the direct question posed to the chairman on this issue: "You (…) keep mentioning the market average or the market outlook for the size of the balance sheet. Are you endorsing the market average, which is 3-1/2 trillion? And if you’re not endorsing it, why do you keep mentioning it?” Here is Powell’s answer: “I’m not going to give our estimate or ratify anybody else’s estimate of what the equilibrium balance sheet is here today. There are estimates out there but I’m not at a point today where I’m going to be giving out numbers on that. But there are estimates and I think they’re consistent with what I said, broadly speaking."
Reference sources for targeted reserves on the Fed radar screen
Coming back to the chair's statements, the two most likely "Surveys and market intelligence" sources cited in his statement, which we reported in point 2) likely refer to: a) the Senior Financial Officers Surveys and b) the Primary Dealers Survey (run by the NY FED). Getting into more detailed messages from these two surveys, we report the main findings from both of them below:
a. Senior Financial Officers Survey
In the September SFOS, a specific "question on lowest comfortable level of reserve balances" was put to banks: "Senior financial officers at each bank were asked to report the approximate lowest level of reserve balances that they would feel comfortable holding before taking active steps to maintain or increase their bank’s reserve balances given the prevailing constellation of short-term interest rates relative to the IOER rate."
"In aggregate, the lowest comfortable levels of reserve balances reported by all respondents summed to about $600 billion, a little less than half of these banks’ average reserve balance holdings in August 2018. Recall that, in aggregate, the respondent banks held roughly two-thirds of total reserve balances at the time of the survey."
On the back of the latter statement, it is reasonable to infer that aggregate demand for reserves from the whole system should be close to USD 850/900 bn. Regarding the buffer Powell referred to at the latest FOMC meeting, a level of USD 100/150 bn is considered appropriate by the consensus: this would total around USD 1 tn amount for the overall reserves.
b. Primary Dealers Survey (run by the NY FED)
In a speech on MP normalization on October 26, 2018, NY Fed Simon Potter had already cited the September survey of primary dealers: "Our regular surveys give us a sense of how some market participants are viewing many of these components. In particular, the surveys ask respondents about their expectations for the size and composition of the Federal Reserve's balance sheet, on average, in 2025, under the assumption of no return to the zero lower bound. The responses serve as a good proxy for expectations about the long-run size of the balance sheet. The median expectations are for the balance sheet to be $3.6 trillion on average in 2025 and for reserve balances to be $750 billion, or about $1 trillion below their current level."
This survey confirmed market consensus expectations close to a USD 3.5 tn in the long term, while a more recent survey added some color on the expected composition of the Fed’s balance sheet, which should see a greater weighting of Treasuries vs MBS and a higher level of reserves, closer to USD 850 bn. It is interesting also to add that two years ago, former chairman Bernanke had already provided his guidance, quite in line with the current consensus. On January 26, 2017 he argued that "Taking into account growth in nominal GDP and bank liabilities, the critical level of bank reserves needed to implement monetary policy through a floor system seems likely to be well over USD 1tn today, and growing."
On the policy mix, Powell reiterated the "dominance" of rates over balance sheet:
"First, as we’ve long emphasized, the federal funds rate is our active monetary policy tool.
Second, as far as the particular details of normalization are concerned, we will not hesitate to make changes in light of economic and financial developments. This does not mean that we would use the balance sheet as an active tool, but occasional changes could be warranted.
Third, we repeat a sentence of the normalization principles we adopted in June 2017. While the federal funds rate would remain our active tool of policy in a wide range of scenarios, we recognize that the economy could again present conditions in which federal funds rate policy is not sufficient. In those cases, the FOMC would be prepared to use its full range of tools, including balance sheet policy."
This looks very much consistent with previous indications from Fed officials on dynamics between rates normalization and balance sheet normalization: this was described in detail in the aforementioned speech of NY Fed Simon Potter, which is helpful for understanding Powell's latest statement, too.
"As the portfolio shrinks and the level of reserve balances declines, how will we know if we are transitioning from an environment characterized by abundant reserves to one characterized by scarce reserves? (...) At high levels of reserves, the responsiveness of rates to changes in reserve levels is fairly low, and the demand curve is said to be 'flat.' As excess reserve levels decline, the responsiveness of rates increases. At low levels of excess reserves, the overnight rate responds sharply to small adjustments in the level of reserves, and the demand curve is considered "steep."
"When overnight rates—including rates on federal funds transactions—edge higher, what tools does the Federal Reserve have to ensure that they generally remain within the target range when reserves are abundant? The federal funds target range is an important feature of the FOMC’s public communications, and maintaining control of the federal funds rate and other money market rates is therefore taken quite seriously. Public confidence in our ability to maintain rates within the target range is important for ensuring that expectations for the FOMC’s future policy stance are properly incorporated into the term structure of interest rates, and thereby appropriately affect financial conditions and the broader economy."
In short, the calibration of the size of the balance sheet and the volume of excess reserves aim at keeping policy rates within the targeted range without unwanted volatility in short term rates, while the recourse to balance sheet policy would occur only if non-conventional tools were needed.
FOMC minutes and Fed speeches in February gave more clues about target and timing
Following Powell’s statements, in February we had more clues on the balance sheet size target and the timing for the end of Fed QT. The main sources of new indications came both from speeches and from the release of the January FOMC minutes.
On the question about "when" QT is likely to stop, the FOMC minutes made it quite evident that we should expect 2019 to be the year that the runoff ends: “Almost all participants thought that it would be desirable to announce before too long a plan to stop reducing the Federal Reserve’s asset holdings later this year.” Furthermore, many pointed out that the decision and the statement that communication should arrive soon as "before too long" may refer to one of the next two FOMC meetings.
If the consensus is broad about the timing, the targeted level of reserves does not look so unanimous among Fed members. In their latest speeches Brainard, Clarida, Williams and Harker referred to a range between a minimum USD 1.0 tn and a maximum of USD 1.3 tn, therefore on average close to but actually even above the USD 1 tn level we mentioned before.
If Fed is going to stop draining liquidity when excess reserves fall to USD 1.1/1.2 tn, this would mean a level of Fed assets stabilizing at around 17% of GDP, as the chart shows, still quite high with respect to pre-GFC levels, at that time close to just 5%.
On the future composition of the Fed balance sheet, another relevant topic to be addressed despite representing more a medium- to long-term issue, the aforementioned latest survey of primary dealers showed an expected switch in favor of US Treasuries from current composition of QE portfolio. US Treasuries are expected to rise to 76% from the current 55%, and Agency MBS to fall to 19% from 40%. These trends in balance sheet composition look to be in line with recent Fed messages. Cleveland Fed President Loretta Mester, for example, on composition said, "My preference would primarily be Treasuries (...) And I would skew it towards short-term Treasuries (...) "It would be a shorter duration than this current balance sheet."
Regarding this topic, the very latest speech by the Fed vice-chairman for banking supervision, Randal Quarles, confirmed the preference for Treasuries: "I favor a return to a balance sheet with all Treasuries securities, allowing our mortgage-backed securities (MBS) holdings to run to zero." Also, the message of Quarles on duration also looks very much in line with the statement by governor Mester, quoted above, adding the rationale behind this view: "In regard to duration, moving to shorten the duration of our holdings could increase the Fed’s ability to affect long-term interest rates if the need arose. However, it might be preferable to have the composition of our Treasury holdings roughly match the maturity composition of outstanding Treasury securities, minimizing any market distortions that could arise from our holdings."
On the year to date Fed communication has broadly indicated an earlier end for balance sheet run-off. In January, Chairman Powell outlined the rationale behind calibrating the size of the balance sheet and the need to consider the volume of excess reserves in order to keep policy rates within the targeted range without unwanted volatility in short term rates. Powell clearly mentioned that estimates of needed excess reserves have increased considerably in the past year. Minutes from latest FOMC meeting revealed that the view on an earlier end to QT is almost unanimous among Fed members. They also indicated that the central bank is likely to announce details of its plan soon, namely at one of the next two FOMC meetings. On the back of the range of estimates from surveys and recently indicated by some central bank members, we can infer that Fed balance sheet could stabilize later in the year at much higher levels as a percentage of GDP vs pre-GFC percentages. In a nutshell, and despite balance sheet normalization, liquidity in the system is likely to remain abundant by historical standards, while on the composition, the Fed is likely to progressively increase the weight of US Treasuries vs MBS in its current portfolio, probably with an overall lower duration. A more dovish Fed attitude on rates normalization is therefore likely to be accompanied by "prudent and patient" balance sheet management, too. In terms of market implication, the combination of both stances should help reduce volatility spikes and perceived risks linked to the QT impact.
Unless otherwise stated, all information contained in this document is from Amundi Pioneer Asset Management (“Amundi Pioneer”) and is as of March 1, 2019.
The views expressed regarding market and economic trends are those of the authors and not necessarily Amundi Pioneer, and are subject to change at any time. These views should not be relied upon as investment advice, as securities recommendations, or as an indication of trading on behalf of any Amundi Pioneer product. There is no guarantee that market forecasts discussed will be realized or that these trends will continue. These views are subject to change at any time based on market and other conditions and there can be no assurances that countries, markets or sectors will perform as expected. Investments involve certain risks, including political and currency risks. Investment return and principal value may go down as well as up and could result in the loss of all capital invested. This material does not constitute an offer to buy or a solicitation to sell any units of any investment fund or any service.