The Big Idea

Rushing to reduction

Stephen Stanley | March 18, 2022

This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors. This material does not constitute research.

The last six months have seen the Fed rushing to catch up after maintaining its emergency policy stance too long. The FOMC rushed through tapering in only four months so it could begin raising rates. After last Wednesday’s liftoff, the Fed is now projecting it will hike at every meeting for the rest of the year and may move into restrictive territory by late next year. In addition, the FOMC is preparing to reduce its balance sheet as soon as the next FOMC meeting in May.

My prior call

Until Wednesday, my forecast was that the Fed would opt to announce balance sheet reduction at the June FOMC meeting and would begin allowing runoff in July, though I added that the timeframe could be moved up or back by one FOMC meeting.  My thinking was that officials would want to see the rate hike campaign well underway before initiating a second front in the form of balance sheet reduction.

What We Heard Wednesday

The Fed was deliberately coy about its balance sheet intentions at the March FOMC meeting.  The statement noted that the committee expected to begin reducing its holdings “at a coming meeting.”  While that sounds neutral, Chair Powell’s press conference comments conveyed a sense of urgency.  In response to a question, he said that “we know the economy no longer needs or wants (this) very highly accommodative stance which — you know, so it’s time to move to a more normal setting of financial conditions, and we do that by moving monetary policy itself to more normal levels.”

Then, when asked specifically about the balance sheet, he pulled out his prepared answer: “Thank you for asking. So, at our meeting today and yesterday, we made excellent progress toward agreeing on the parameters of a plan to shrink the balance sheet. And I’d say we’re now in a position to finalize and implement that plan so that we’re actually beginning runoff at a coming meeting. And that could come as soon as our next meeting in May. That’s not a decision that we’ve made. But I would say that that’s how well our discussions went in the last two days.”

In my view, the combination of these two quotes is a strong signal that the FOMC’s attitude at this point is to move forward at the earliest possible date, which would be the May 3-4 FOMC meeting.  Powell indicated that the committee has laid the groundwork for doing so, and, barring some massive financial market dislocations between now and then, I now look for the committee to follow through with a May decision to begin reducing the balance sheet.

One Vision for Balance Sheet Reduction

The FOMC has not offered much clarity yet about how it will reduce the balance sheet.  Here is what Chair Powell said at his press conference: “I don’t want to get too much into the details because we’re literally just finalizing them. But the framework is going to look very familiar to people who are familiar with the last — the last time we did this. But it’ll be faster than the last time. And, of course, it’s much sooner in a cycle than last time. But it will look — it will look familiar to you. And I would also say that there’ll be — I’m sure there’ll be a more detailed discussion of our — in the minutes to our meeting that come out in three weeks, where I expect that it will lay out, you know, the — pretty much the parameters of what we’re looking at, which I think will look quite familiar.”  By the way, those FOMC minutes will be released on April 6.

I doubt I will get all of the details exactly right, but here is one plausible scenario for how balance sheet reduction will proceed.

Timeframe.  If the FOMC votes to begin balance sheet reduction at its next meeting on May 4, then I would look for balance sheet reduction to begin in mid-May for both Treasuries and MBS.  For Treasuries, runoff occurs when Treasury securities mature, and that happens on the 15 day and the end of the month.  So the Fed can squeeze in a full month’s worth of runoff in May, even if the announcement comes after the start of the month.

For MBS, the monthly cycle for Fed purchases and redemptions has always been governed by the timing of the prepay data, which come between the fourth and sixth business day of each month.  A May 4 announcement would likely kick in with the monthly cycle commencing just after the prepay speeds report is released in the following days.  The Fed would probably release its schedule of operations twice a month, as it has throughout the QE process, or perhaps once a month once purchases slow to a crawl.

Monthly caps.   When Chair Powell says that the form of the Fed’s balance sheet reduction will be “familiar,” I believe that the most important element that he is referring to is that the Fed will employ monthly caps on the pace of runoff in an effort to prevent excessive month-to-month swings in the amount of redemptions.

As I detailed in a piece a few months back, there are effective caps on how fast the Fed can passively redeem securities.  In the case of Treasuries, in non-refunding months—that is, months aside from February, May, August, and November—the Fed’s holdings are generally in the neighborhood of $50 billion.  That makes $50 billion a month a de facto cap on the caps if the Fed seeks to keep the speed of runoff steady every month.

On the MBS side, the pace of runoff is not pre-determined.  It depends on housing turnover, refinancing and so on.  In general, the faster and higher mortgage rates rise, the slower prepayments are likely to be.  Under plausible assumptions about prepayment speeds, it looks like $25 to $30 billion a month in redemptions is about as much as the Fed can realistically expect.  As a result, $30 billion may be an effective cap on the caps for MBS.

Phasing in.  In the prior balance sheet reduction period in the late 2010s, the FOMC chose to phase in the reductions to promote a smooth transition.  The sense of urgency is clearly stronger this time around, but I still look for a short phase-in period, perhaps over three months.  My guess is that monthly caps will be $30 billion in May, $40 billion in June, and $50 billion in July and forward for Treasuries and $15 billion in May, $22.5 billion in June and $30 billion in July and forward for MBS.

Active or passive.   One question that has arisen frequently is whether the Fed would actively sell securities at some point rather than merely allowing runoff.  For Treasuries, my answer is almost certainly not for Treasuries.  A pace of $50 billion a month would be a rapid pace of reduction, and if the Fed wanted to pick up the pace, the next step would be to lift the caps, which would bring an additional $50 billion or so in redemptions four times a year, or close to $200 billion more for the year—upping the annual pace from $600 billion to $800 billion.  It just does not seem plausible that the Fed would see a need to cut its Treasury holdings faster than that.

For MBS, my answer would be maybe, but only later in the process.  If mortgage rates rise fast and far enough to shrink the pace of runoff to well below $30 billion per month, the Fed might opt to sell securities to speed things along.

Alternatively, since policymakers have expressed a preference that the Fed should eventually hold only Treasuries in the SOMA account, it may choose to accelerate the process with outright MBS selling.  However, the Fed has been quite hesitant historically to rock the MBS market.  My suspicion is that if the Fed does choose to sell MBS back to the market, it will do so quite a bit later in the balance sheet reduction process, perhaps in the second half of 2023 or beyond. Keep in mind as well that as the size of the Fed’s MBS holdings decline, the dollar amount of prepayments should also naturally shrink, all else equal.  I also would imagine that the Fed would be careful in deciding how fast to sell off its holdings.  And this begs a question: if the New York Fed Open Markets Desk suggests to the FOMC that it could only sell a relatively small amount of MBS per month—say, $10 billion—without risking excessive volatility, would the FOMC conclude that it was not worth the trouble?