Q&A about Fed portfolio normalization

The best way to predict how the FOMC will react to evolving economic conditions, we at the CFE have been arguing, is to listen almost exclusively to communication on behalf of the consensus, and then to, as best as possible, take that communication at face value. As we’ve documented, this approach has had an excellent track record over the last few years. [1] So what does this suggest about portfolio normalization, that is, shrinking the large portfolio of longer-term securities the Fed accumulated in the QE programs?

The FOMC has been explaining its intentions about portfolio normalization since 2011. These discussions—thoroughly documented on the Fed’s normalization webpage—have evolved some, but have been essentially unchanged since Dec. 2015. What follows is my attempt to answer some questions about normalization by taking these communications at face value.

Q: Is the FOMC going to surprise the public with initiation of portfolio normalization?

A: No. This Fed has been scrupulous, perhaps to a fault, in announcing plans well in advance when circumstances allow. At present, there is no pressing need to hurry.

Q: Will the FOMC likely to send signals about the timing of portfolio normalization soon? And if so, why now as opposed to earlier or later?

A: Given the Fed’s communications to date, it would be quite surprising if the Fed were not firming up portfolio normalization plans right now with an intention to share them over coming months. This would set the FOMC up, if the economy cooperates, to take steps around the turn of the year, give or take a bit.

The communication on behalf of the consensus has consistently stated that normalization would begin once two conditions are met: i) the ‘early stages’ of federal funds rate target normalization are complete so that interest rate normalization is ‘well-underway’ and ii) the economy appears to be on a solid footing, allowing reasonable ‘confidence in the economy’s trajectory.’[2] The first condition affords the Fed room to deliver several rate cuts if the economy falters, before having to contemplate a return to asset purchases. The second is common sense and avoids concerns about folks misinterpreting the signal in a weak economy.

These two criteria are subjective, but some things seem pretty clear. Even late last year, we had not yet seen a second target rate increase. With conditions only having warranted one rate increase per year over the previous 2 years, it was questionable when the interest rate normalization might be well underway. But now the economy has taken three rate increases in stride and, if this performance continues, we’ll see a few more rate increases in coming quarters. By year end, then, the hawks and centrists on the FOMC will be able to make a very strong case that the initial stages of interest rate normalization are complete and, so long as the economy remains at least on a solid footing, a start to portfolio normalization should begin as planned.

The FOMCs standard MO is to report that it is discussing plans, later announce plans, and then give the public time to digest the plans before commencing action. Under this MO, the FOMC should begin this sequence soon. Of course, at the March press conference, Yellen indicated that portfolio normalization plans were discussed, and so we’ll probably hear more about this in the minutes to be released today.

Q: What will the announcement mean for the intended stance of monetary policy?

A: Nothing. The FOMC has been uniformly consistent in its communications on this point: the announcement will confirm confidence that the economy is broadly on track, but will not signal any change in the desired stance of monetary policy.

The FOMC has said it would prefer, if conditions allow, to make portfolio normalization a sideshow from the standpoint of the stance of monetary policy. The plan is to set portfolio normalization on a path designed to minimize uncertainty and market disruption, and not to adjust that path in pursuit of standard monetary policy goals—recognizing, of course, that a major break in the economy in either direction could change this.

Q: Are you saying the Fed thinks that portfolio normalization has no implications for policy?

A: No. Quite to the contrary, if the asset purchases had significant effects—which remains a topic of debate—then undoing them must as well.[3] Thus, portfolio normalization will have implications for policy, but the path of normalization still need not be used as a tool of policy.

The Fed’s holdings of longer-term securities, in the standard view, are currently putting downward pressure on longer-term interest rates. The pace of normalization will determine the pace at which that downward pressure diminishes. So long as the Fed sets a mechanical path of normalization and so long as that path is sufficiently gradual to be nondisruptive, then any tightening financial conditions from portfolio normalization can be offset by pursuing somewhat more gradual increases in the federal funds rate. And any adjustments to policy in response to news about macro conditions can be accomplished by altering the path of the federal funds rate target.

Thus, until completed, portfolio normalization will be a background factor that will have to be taken into account in determining the proper use of the Fed’s interest rate instruments. Quicker normalization of the portfolio will likely mean a slower normalization of the funds rate, and vice versa. Since nobody can be sure what effects the purchases had on longer-term rates, we cannot be sure what effects unwinding from unwinding will have to be offset. I am not aware, however, of any plausible estimates showing large effects of a gradual normalization—so long as normalization is not misinterpreted as signaling a significant change in policy.

As an aside, I’ll note that the background factor of portfolio normalization may already be making the dot-plot in the Survey of Economic Projections an even uglier 17-headed monster than before. In short, normalization becomes one more unreported factor that may differ across the 17 projections, further muddling the interpretation. For example, the median dots now show three rate increases this year. Might that be 4 rate increases worth of tightening, but with one of the increases effectively being accomplished by starting portfolio normalization? Or is normalization driving some dots, but not others? Who knows? As usual: ignore the dots.

Q: What will the FOMC say about the exact timing and pace?

A: The FOMC has said a few important things on this point, but few specifics. The FOMC has said that it doesn’t intend to sell agency securities and instead will let them run off. It has said that the first steps are likely to involve adjusting the reinvestment. And it has said that all moves will be gradual, so long as conditions permit. Beyond these characterizations the Fed has not filled in the details.

From the standpoint of the Fed’s dual mandate, details about the precise timing of the start of normalization and exact pace don’t matter much. As I’ve said before, commenting on—or betting on—issues that don’t matter much to the Fed’s dual mandate is a sucker’s game. I don’t think anybody can reliably predict much more than the broad timing and gradual pace I’ve outlined.

Q: How will Yellen’s legacy figure in these decisions?

A: Near the end of each Chair’s term, it is customary for some commentators to begin fevered arguments about legacy as a major driver of policy. This is, I believe, complete nonsense.[4] In the present case, Yellen can’t possibly be silly enough to imagine that, say, trimming the reinvestment on the portfolio in the last months of her term would have any appreciable effect on how history views her. Further, I’m pretty sure that at least the last two Chairs wanted an overarching theme of their legacies to be that they vigorously pursued the Fed’s mandate regardless of politics or ego. Fiddling with policy to trweak their images? Nuts.

Q: What about deference to the possibility of Chair turnover?

Any reasonable Chair would hesitate to set policy on some controversial or radical new path just before a Chair changeover. But policy is made by consensus of a large Committee, and over a very long period a very strong consensus has favored normalizing the portfolio when the economy has recovered sufficiently. If anything, the interests of continuity and smooth transition probably favor getting normalization underway. Waiting could mean that the next Chair arrives immediately facing an overdue policy step and with the desired groundwork not in place. At the very least, the Chair turnover argument cuts both ways.

Bottom line

So long as the economy doesn’t break sharply in either direction, we’ll see plans for a gradual portfolio normalization emerge over the next few months and a gradual normalization commencing around the end of the year—give or take bit. At least that’s what I see as a straightforward interpretation of what the Fed has consistently said at least since Dec. 2015.

There are two natural ways this analysis could be wrong: the FOMC could choose a policy that clearly conflicts with earlier statements, or I’ve gotten the supposedly straightforward interpretation of the Fed’s words wrong. The first seems highly unlikely. The second? Beginning with the minutes this afternoon, we’ll begin to see…


1. This is an ongoing theme in our Fed analysis. [back]

2. Statements like those quoted are plentiful. For example, the characterization in the Dec. 2015 and March 2017 press conferences was very similar. More statements of this variety are found on the Fed’s normalization webpage[back]

3. Strictly speaking, standard reasoning says that it is mainly unwinding purchases at a different pace than expected that would have significant effects. [back]

4. Monetary policy is always an ongoing project not much affected by a few modest decisions. Of course, the Chair might have various other projects (say, regarding internal management or other Fed matters) on which she might push to cement progress. [back]