Main point: On consensus logic, we can easily show that the debate around the balance sheet is fake. The purpose of this debate was to land Warsh in the Fed Chairman’s job. That purpose has now been achieved, so let’s move on.
With Kevin Warsh ascending to Fed Chair, analyst and media attention has begun to focus more sharply on his presumed desire to shrink the balance sheet. This morning, for example, I saw on Bloomberg some very in depth and expert analysis on how difficult it may be to shrink the balance sheet without running into constraints imposed by money market liquidity considerations. Apparently, we will need to put in place regulatory reform to allow the balance sheet to contract. Otherwise, contracting the balance sheet might push us into a scarce reserves regime, which is presumably undesirable.
If this all sounds like pointless macro babble to you, then don’t worry, it is. Warsh’s stated claim is that shrinking the balance sheet will generate a form of tightening that will allow him to cut interest rates without stoking inflation pressures, presumably by easing financial conditions. What is a wash from a stimulus perspective, though, is nevertheless good policy, he says, because short-term interest rate cuts help the little guy while QE just provided a windfall to the Wall Street elite. How that part works is a matter of magic, but it is central to his argument.
To see the main problem with this claim, I do not need to drag out my longstanding view that the effects of QE were radically overstated by consensus at the time and that QT would therefore not matter much. Rather, I can just invoke that the consensus holds that QT does not matter much. The reason for this consensus is a bit odd, in fairness. The preferred argument is that QT announcements don’t have much effect within an event study framework because they inform us only on the specific timing and scale, and don’t contain much information about the path of the balance sheet over time. After all, it has always been known that QT would arrive at some point. In contrast, QE announcements contained much more news.
That argument is true, so far as it goes. But it also highlights a difficulty with announcement effects. Announcement effects are relevant only to the (apparently limited extent) that inform us about enduring effects.
Announcement effects might be a proxy of enduring effects, and therefore rationalize event studies, in an efficient market setting. But in an efficient market setting, QE is presumably irrelevant on logical grounds. So, to use event studies to scale the enduring effect of QE (or QT) seems a logical contradiction.
But that aside, there is a much simpler issue here. There must have been some point, presumably before the QT announcements came into view, when market participants shifted their view from extending QE to upcoming QT. But we never see any assessment of the effect of that transition. Instead, we have the somewhat silly claim that QE was potent and that QT is asymmetrical because of an expectations dynamic that is never fully explained.
The much simpler story here is that QE did not matter much and so its going away does not matter much. This would have been consensus, even among the great and good on Wall Street, had QE – along with its rationalizations – never come along. This used to be something that everybody knew!
Anyhow, the consensus is that a plausible range of outcomes for QT, incorporating Warsh’s presumed preference at one end and speculations of those who worry about money market liquidity at the end other, might be worth about 20 bps on the 10-year term premium. My own view is that the consensus overestimates this, although I concede that much depends on the stock-bond correlation. If we move further into a “fire” regime, then then the effects might be larger, although fiscal policy clearly massively dominates all. Anyhow, 20 bps (or less) on the 10-year Treasury yield is not something that requires a systematic offset from the funds rate. So, the consensus take on how QT works, leaving aside my own, is that this debate is entirely fake.
Meanwhile, the concerns over money market liquidity are also largely fake. It is probably true that shrinking the balance sheet will include money market liquidity to tighten – all else equal. But all else will not be equal because the Fed has many ways of supporting money market liquidity that are invariant to the Fed’s long-maturity securities holdings. The most obvious approach would be for the Fed to accumulate bills as the longer maturity assets roll off. The Bloomberg story mentioned above gets into that, quite reasonably.
I am not sure if Warsh would agree that swapping notes and bonds for bills would count as balance sheet contraction, although he should, given the official story of how QE/QT works, i.e., by presumably affecting the term premium. But if Warsh (assuming he is in charge) were to insist on smaller overall securities holdings, irrespective of maturity, then the Fed could use other means, involving repo facilities, to detach balance sheet policy from money market liquidity policy. Again, this is not me bloviating about my priors (which are correct by the way). I am just reminding you of what consensus is.
And if you are not convinced by even this argument, I can take a further step back. Does anybody think that the Fed would intentionally undermine money market liquidity? Leaving entirely aside any understanding of the plumbing at all, is that a sensible central case?
So, why is Warsh on about this? Even his harshest critics will tell you he is not stupid. The simplest explanation here is that Warsh wants a story that allows him to promise rate cuts (to Trump) on the one hand and preserve his reputation within markets as a responsible inflation fighter on the other hand. The balance sheet “debate” serves that purpose. It is its only role. The rest is entirely fake.
The thing is, we are making a transition from Warsh trying to get the job to him now having to manage having gotten the job. From here on in, reality will matter more than fairy tales. And I assume his behavior will reflect this. If it does not, then he will lose the support of the committee, which even I recognize, he will have to some extent as the starts, as a matter of institutional tradition.
