The World Is Desperate For Increasingly Scarce Dollars

What at the beginning of the year many thought was impossible, this week it happened. The Federal Reserve cut rates for the first time since December 2008, even though you could have counted most if not all FOMC officials in that other group. Trying to reconcile this drastic change with a backdrop of the unemployment rate hovering near a 50-year low, it just doesn’t appear to make any sense.

At most, in Chairman Powell’s own view, the US economy has faced just “cross currents” this year. These will reduce the rate of expansion but aren’t expected to be much more trouble than that, according to the Fed’s models.

So, why the rate cut?

The better question is, why did the curve(s) invert in the first place? After all, the Treasury curve had flipped upside down while Powell was still raising rates. The eurodollar futures curve had inverted all the way back in June 2018. Both were predicting exactly what the Fed just did – when at the time everyone at the Fed thought it was laughably stupid and crazy.

And the unemployment rate is lower now than back then.

A good place to start looking is the repo market. More than a backbone, it is the very center of the global dollar system. Where the central bank should be, and where almost everyone still believes it is, the repo market functions as the ultimate liquidity backstop. It is the lender-of-last-resort.

The Federal Reserve, by contrast, just wants you to think that the central bank is. So long as you believe in the Fed, by its theories and models that’s enough.

Which is why the modern central bank places so much importance on symbolism. The smoke and mirrors, the frantic hand waving and jawboning. Getting down to the nitty gritty in the trenches, the Fed really doesn’t go there.

The repo market has been an agitated and disturbed hotbed of turmoil ever since around May 29, 2018. It is interesting that eurodollar curve inversion followed just a few weeks after that particular date. I wrote in early July last year:

“In other words, the market just isn’t buying the economic story. Instead, it is buying that Fed officials buy it, but that over some unknown length of time reality will be imposed in a way that, on balance of probability, will force the FOMC to reverse course.”

That same day, the Federal Reserve had published the minutes of its June 2018 meeting. It caused a bit of a ruckus because of one word: very. The highly sanitized minutes which had described the economy as “strong” before would now describe it as “very strong.” It was an upgrade, one publicized for maximum effect (symbolism).

In other words, at the very same time monetary officials were becoming more confident (or wanted you to believe they were) in their forecasts, the bond market was becoming significantly less so. It is entirely possible the latter created the need for the former.

I’ve recounted May 29 many times since it happened. We don’t know exactly what took place, but given the primary symptoms (sharp drop in bond yields) and what has followed afterward it’s not hard to piece together what went wrong in general terms. In a single word: collateral.  And that means repo.

Given how things have only gone downhill since then, it’s amazing to watch as so much reverse engineering is employed in order to avoid the simplest explanation. The reason it is necessary is simple faith in the Fed; there’s an abundance of reserves, therefore there cannot be a liquidity problem at root of all of these things. Abundant reserves don’t overtake a very strong economy turning into a very weak one in need of some kind of assistance.

With that as the fixed point for mainstream convention, it takes something which would’ve made Rube Goldberg proud for that convention not to so thoroughly unravel.

Beginning again in repo, one very prominent theory going around has it that primary dealers are getting stuck with an overwhelming volume of US Treasuries. The federal government is selling a lot more debt of late, and dealers are statutorily obligated to buy it. They intend to sell them to the public, but what if the public’s appetite doesn’t match the supply?

In that situation the dealers will have to hold what they don’t really want. With foreigners selling their holdings, it can get messy. Because they are stuck holding the bag, the dealers have to fund it somehow, some way. Repo is pretty much the only place they can go since these firms (the subs, anyway) aren’t actually banks.

Clearing through Bank of New York Mellon (which is a bank), what unfolds is a complex dance of intraday flows. It pulls in one’s reserve account with the Fed for other clearing accounts with the custodian, the end result is pressure upon the repo rate which is visible for anyone to see.

And that’s all really a distraction since the theory largely rests upon one factor: why are the dealers stuck holding so many Treasuries it is causing funding market disruptions?

Nonbank buyers, supposedly, staged a buyer’s strike because the curve inverted. Why would a money market fund, for example, buy a 10-year UST from a dealer at price which yields not much more (even much less) return than what it can get on money market rates – even the Fed’s reverse repo?

I guess we aren’t supposed to notice that the rest of the world has been buying UST’s at a furious pace. How else did the yield curve invert in the first place? Sure, the money market fund and other nonbanks like it may have preferred commercial paper or even unsecured LIBOR to a very low UST yield, but someone is buying to push every UST yield lower to begin with.

Not just someone, but a whole ton of someones. To believe in this convoluted theory, you are asked to believe that primary dealers whose entire job is to find buyers for Treasuries are disrupting repo markets and other funding mechanisms because they can’t find enough buyers for Treasuries during a time when there is otherwise so much demand for Treasuries it has upended the natural order of curves everywhere.

The law only says dealers have to buy what the government auctions; there is no law that states they can only sell what they are forced to buy to the local nonbank money fund. Far-fetched is giving it too much credit. But because it adheres to conventional doctrine, the idea is taken seriously even though it is absurd.

The data is at least unequivocal. Primary dealers are holding an immense amount of UST inventory. Is it because they are having trouble offloading that inventory into a marketplace starved for those very things? Or might it be because dealers are choosing to hold on to, and even acquire more of, the choicest, most pristine form of repo collateral?

This is no trivial interpretation. In fact, everything is riding upon which option you think makes the most sense. To the former: the Fed is still the Fed, bank reserves mean something, interest rates would be rising as everyone has been predicting (and many counting on), and the economy is otherwise truly awesome. If not for a lack of buyers hungry for UST’s, all that would be plainly obvious. And it will be again, soon enough.

To the latter: there’s something really wrong in repo and collateral, liquidity could be a big problem because repo matters in the absence of the Fed, bank reserves don’t count for much at all, and the economy is under serious threat because of all these things. So much of a threat, that when the first hint of repo spilled out into the open at the end of last May, markets began to price an increasing probability its effects would eventually be serious enough the Fed would have to turn all the way around.

As it just did.

No, no, no, says Jay Powell. Not a complete 180-degree turn. The Chairman claimed Wednesday this was nothing more than a “mid-cycle adjustment” implying there is no series of rate cuts to follow. One maybe two at most, just enough insurance to make sure the strong economy comes through these cross currents with flying colors.

Yet, while he displays his symbolic confidence in public again, the rate cut wasn’t the only “accommodation” the FOMC approved this week. The central bank policy committee also voted to end quantitative tightening (QT) two months earlier than previously planned.

QT is nothing more than the other side of QE. Quantitative easing was the brand name for large-scale asset purchases (LSAP), the buying of financial assets whose primary byproduct is a sharp increase in the level of bank reserves (double accounting; increase in assets produces the same increase in liabilities unless offset). Most people focused on the first part (what was bought) and paid little attention to the second (bank reserves).

It was largely assumed, and the Fed was perfectly happy if you did assume, that bank reserves were an increase in money supply. After all, they go into the monetary base (M0). How can they be anything else?

QT is the process of unwinding QE. As assets which were bought under those old LSAP’s mature, the asset disappears from the one side of the balance sheet and then consequently the level of bank reserves falls by the same amount (though there are several steps to how this happens). Thus, if QE was money printing, QT must be the opposite.

But even if you believe that is the case, it’s still not the whole story. In a very general sense, QE was supposed to fill in a gap left by the Global Financial Crisis in money dealing capacity. According to the theory, the Fed used its own balance sheet in place of a private financial system reeling from the debacle of subprime mortgages.

When the world finally healed enough from that, the Fed would step back and turn everything back over to the private system again. This was expected, and in this situation nobody would even have noticed QT. Whatever ultimate amount might be subtracted from bank reserves during normalization, as the Fed’s balance sheet was unwound it should’ve been made up for and then some by private dealer capacities picking back up the slack.

But the dealers are stuck with so many US Treasury securities. Unless they are holding them on purpose.

For its part, the Federal Reserve seems to be hedging. Before all this stuff first erupted in 2018, convention had it that QT could take the level of reserves pretty close to zero again – which is what they were before 2008 (another realization that deserves serious consideration in the context of what role bank reserves really play). There would have to be maybe several hundred billion left for structural reasons, including, ironically, the foreign repo pool which has exploded in the past few years, but otherwise the vast majority of them could disappear without anyone noticing.

Confronted by all these markets going haywire, the FOMC changed course somewhat by proposing an earlier end to QT and therefore a terminal level of reserves far higher than anyone had expected. On Wednesday, hedging again, officials decided that QT would end immediately. As of Thursday, it was done.

How did the markets react to the news about reserves (as well as the rate cut)?

Yields plunged. It seems there is still an overwhelming desire coming from somewhere (everywhere?) to buy up UST of all kinds, up and down the curve. If dealers truly are having trouble selling treasuries they should probably be put out of business for being so inept at the task.

Obviously, that’s not really what’s going on here. Being freed from the ideology of central banks being central and bank reserves being some form of money, we can appreciate the market action as well as the Fed’s reaction(s).

Beginning last May: market: something’s wrong in repo; Fed: no way, strong economy is actually very strong, stop complaining. Last December: market: something’s really wrong in repo; Fed: well, economy is still strong, but we’ll stop hiking rates. Happy? Earlier 2019: market: this is getting really serious; Fed: we’ll end QT probably toward the end of the year and make sure there are more reserves. June 2019: market: you aren’t listening. Fed: not only will we end QT, we’ll consider rate cuts, too.

This week: market: it’s actually getting worse (swap spreads and FX, now, too) across all indications; Fed: OK, here’s one rate cut and QT ends tomorrow; UST buyers who aren’t nonbank money funds: I better get on the phone and see if I can get that primary dealer to sell if I offer an even higher price.  

Globally synchronized growth in 2017 was supposed to hand off to 2018 all the things which would indicate the world becoming normal; interest rates had nowhere to go but up; bond vigilantes would care about deficits again; the dollar would have to fall; and the economy would be roaring and inflationary, rate hikes that would continue even accelerate into 2019.

None of those things has happened. But rather than admit they were all wrong to begin with, because mainstream convention about the monetary system was all wrong, the only explanation being put forward is primary dealers who are in a desperate struggle to sell the one asset the rest of the world is more desperate to buy. That statement says as much about the state of the world as any other.

The reality of the situation actually and unintentionally may be best explained by the abrupt end to QT. The FOMC won’t say it publicly, but it’s not what they say that matters, pay attention to what they do. They’ve decided the system really might need a whole lot more reserves than they ever had planned for. In Jay Powell’s thinking, maybe there is something wrong that more reserves might fix.

Another way of saying that?

The world really does seem to be short of dollars.