Fed Balance Sheet Runoff: A ‘Roadmap’ To The End Of The Road

The debate about the impact of Fed balance sheet runoff is interesting for a number of reasons, not the least of which is that some market participants don’t seem to be convinced that the committee fully understands (or maybe “appreciates” is a better word) the effect it has on the system’s “plumbing”, so to speak.

Indeed, some of the cat calls emanating from the peanut gallery following the January Fed minutes centered around the notion that the apparently imminent announcement of a halt to runoff seemed to stem more from a knee-jerk reaction to anecdotal assessments of the psychological impact on risk assets than from a careful consideration of how balance sheet normalization affects the economy and money markets.

“In case you suffered from the delusion that the Fed is carefully calibrating what the normalized size of the balance sheet should be” the January minutes indicate the Fed is employing only the “roughest of approximations”, Jefferies’ Ward McCarthy and Thomas Simons wrote last month, adding that “it seems that the goal with balance sheet normalization was to reduce the size only until there was a noticeable impact on financial markets.” Long story short, Jefferies isn’t convinced the Fed has provided a “solid” enough rationale.

Read more

What Wall Street Thinks Of The January Fed Minutes

While those are the conversations we probably should be having, past a certain point they become at least temporarily irrelevant. When Donald Trump raised awareness of the balance sheet issue by amplifying an already high profile WSJ Op-Ed in his infamous “50 Bs” tweet, market participants came to believe that runoff was partially responsible for the selloff. Importantly, even if the majority of savvy folks didn’t actually believe that, they believed that other people did, and if you believe other people are panicking, you’re likely to try and get out ahead of them, even if you think their worries are unfounded. The end result was that the balance sheet issue added to market participants’ sense of acute consternation and turbocharged the self-referential, liquidity-volatility-flows feedback loop that undermined markets in December.

By the flip of the calendar, it was clear the committee needed to communicate a willingness to adjust the pace of runoff. At that point, markets were no longer interested in the mechanics of QT and just about the last thing anybody wanted to hear from Fedspeakers was some kind of hopelessly opaque argument about “carefully calibrating what the normalized size of the balance sheet should be”.

That’s not to say there aren’t good arguments in support of the notion that runoff really is “like watching paint dry”. There is no accepted “standard” (as it were) for quantifying the mechanical effect of QT on risk assets, but there are ways of quantifying the rate-hike-equivalent of balance sheet runoff. Suffice to say it is by no means clear that runoff has had an outsized mechanical effect and given that, it isn’t entirely clear that the Fed needs to slam on the brakes.

Read more on the effects of Fed balance sheet runoff

Into The Shadows: How Many Rate Hikes Is QT Worth?

As WSJ Tips ‘Earlier-Than-Expected’ End To QT, One Bank Assesses Impact On VIX, Rates Vol.

Ok, Sure! I’ll Take A Look At That Fed MBS Runoff Vs. S&P Returns Note…

But, again, rationality doesn’t matter when people are irrational – especially not in markets, where irrationality has the potential to start manifesting itself in real economic outcomes. Consider, for instance, the following passage from a Saturday post:

Sure, the Powell Fed will couch everything in terms on apolitical policymaking guided solely by their mandate and the incoming data, but there is little question that Trump’s incessant badgering contributed to the FOMC’s epochal dovish pivot in January. Even if you don’t believe that Trump’s intention to fire Powell or his dinner dates with the Fed chair (and his deputy) directly influenced policy, the president might well have indirectly influenced policy by drawing the public’s attention to balance sheet runoff, thereby heightening market concerns and exacerbating the Q4 selloff. That selloff tightened financial conditions, which in turn informed the Fed’s decision to take a pause.

See what I mean? This whole thing is inextricably bound up with itself.

Complicating this further is the fact that the Fed has, to a certain extent anyway, relied on the market not focusing too much on the “double” tightening dynamic, wherein if you add the current effective Fed funds rate to the 300bp hike in the shadow rate (from the lows in the middle of 2014, which reflects the winding down of QE), you end up with something like 5.5% worth of tightening, an amount which, as SocGen has been pretty keen to drive home, “is more elevated than recent cycles.”

Read more

Into The Shadows: How Many Rate Hikes Is QT Worth?

Once market participants decided to start obsessing over that in Q4, it became a problem. How do you explain tightening via two avenues when global growth is slowing and risk assets are in free fall?

Worse, the prospect of taking a pause and suggesting that the next move on rates could be a hike or a cut, forces markets to ponder an absurd scenario where rates are being cut while the Fed continues to tighten via balance sheet runoff. Obviously that’s untenable, if for no other reason than markets would key on that contradiction as a reason to contend that officials are pursuing conflicting policies.

So, here we are, awaiting an official announcement of the end date for runoff. Analysts have written voluminously about how things will proceed from here, and we’ve spent a ton of time documenting those efforts.

In the interest of making things a little clearer, we thought we would excerpt some highlights from a helpful new piece by Deutsche Bank’s Matthew Luzzetti and Steven Zeng who, in a note dated Friday, lay out a “roadmap” for Fed balance sheet normalization. They start by laying out the following five “pillars” of the strategy:

1. Reserves are not currently scarce and the equilibrium level of reserves is unknown at this point.
2. However, an equilibrium level of reserves around $1tn plus a reserve buffer is “reasonable.”
3. The rundown of the Fed’s SOMA portfolio is likely to be completed by year-end.
4. The Fed will likely stabilize their SOMA holdings prior to reaching the longer-run level of reserves to slow the reserve drawdown.
5. The longer-run SOMA composition is likely to tilt towards shorter duration Treasuries and hold fewer (or no) MBS

Those are the “principles”, to quote Luzzetti, and with those as the guide, he goes on to lay out a timeline.

As far as the announcement date for the stabilization of SOMA goes, the obvious choice for the Fed would be March, but May is also possible.

“Given that reserves are likely to fall to near the level consistent with $1tn plus a buffer before year-end and in light of recent signals from a variety of Fed commentary, we expect an announcement that sets the date for the removal of the runoff caps (i.e., stabilizes SOMA) at the March or May FOMC meetings, with the earlier timing more likely”, Luzzetti writes, adding that “stabilization of SOMA would be announced as beginning between July and October 2019, providing between four and six months of guidance to the market.”

Prior to that, Deutsche says the Fed will make clear the details of the post-normalization policy. “This should occur at least by the meeting prior to the start of SOMA stabilization”, he says, noting that, for instance, “if the Fed were to announce the stabilization of SOMA to begin in July 2019, they would announce the details of their reinvestment policy at the June FOMC meeting.”

At a certain juncture, reserves will fall to the much ballyhooed “equilibrium plus a buffer” level, after which the Fed will have to start expanding SOMA again to keep up with growing non-reserves liabilities. “The timing of this step is highly uncertain, as it depends on a number of unknowns, including when the Fed will commence SOMA stabilization and the true longer-run level of reserves”, Luzzetti notes, before pegging “the start date of SOMA growth as between Q1 2020 and Q1 2021.”

Here is a helpful table that summarizes all of these points:

(Deutsche Bank)

Luzzetti and Zeng go on to provide details around forecasting reserves decline. The mechanics of that are pretty straightforward if you understand the Fed’s balance sheet. The bottom line, for our purposes here, is this, from the note:

… reserve balances are expected to decline from $1.68tn currently to $1.46tn by June and $1.3tn by September. Ending the portfolio runoff anytime during those months should provide the Fed a fairly large buffer above the equilibrium level of reserves. By December, reserve balances are expected to fall to $1.1tn

Waiting too long, Luzzetti warns, risks an overshoot of the equilibrium level which means August is probably a “good” time to end runoff. He also notes that during the stabilization period, the Fed will likely let the MBS portfolio continue to shrink, replacing those securities with USTs – that’s something everyone generally expects and it is worth noting that some studies (with a note from Morgan Stanley and work by Nomura being two examples), suggest MBS runoff has a larger impact on risk assets, but that’s another discussion.

Finally, Luzzetti and Zeng detail post-normalization balance sheet growth, rehashing some of the points they made earlier in the note. The equilibrium level of reserves is uncertain, which, by extension, means the date from which SOMA starts growing again is too. Here’s one more excerpt:

If the equilibrium level falls in the high end of the estimated range, and assuming the Fed had ended the runoff in August, SOMA may need to start growing at the beginning of 2020. If equilibrium is closer to $1tn, the Fed can begin adding assets in Q1 2021. With a lower equilibrium level of $0.8tn, the Fed can delay SOMA growth until early 2022, but we think this scenario is unlikely. At this stage, we expect the Fed to be a net purchaser of $25bn-$30bn Treasuries each month to replace the MBS runoff and offset the growth in currency and other Fed liabilities.

In theory, this is all doable and can be executed adeptly, but in practice (where that means that the effort will play out against ever-changing market conditions), it depends at least in part on all manner of things, including the fickle nature of risk assets in an uncertain world.

Circling back to some of the points made here at the outset, this process has been accelerated by factors that have little to do with the mechanical impact of runoff on the the system and on the economy more generally. Because irrationality among market participants can eventually become self-fulfilling (whether that means consumer spending is affected by a reversal of the vaunted “wealth effect”, or whether it means falling stocks and wider credit spreads cause financial conditions to tighten beyond what’s desirable to curb excessive risk taking thereby imperiling the Fed’s ability to engineer a soft landing, etc. etc.), policy decisions that should be based on math and established economic theory end up becoming hostage to the madness of crowds.

Throw in an executive who made it clear on Saturday that he is far from done assailing the central bank, and you’re left with more questions than answers.

Still, Deutsche’s efforts to map this out are a welcome addition to the discussion at a time when the future of the balance sheet is the perpetual topic du jour – assuming using the word “perpetual” with the term “du jour” makes any sense, that is.

---