Repo: The Abundant Reserves Regime Is Broken

Summary

• The Bank of International Settlements is out with their quarterly report, and it provides some background on what is happening in the overnight funding markets.
• They focus on the 4 largest banks, presumably JPMorgan, Bank of America, Citibank and Wells Fargo, and mostly just the original September Repo Revolt.
• The private sector repo market has become almost entirely dependent on the Big Four for liquidity. They are near their limits. When they are tight, the whole system is.
• The increasing size and terms of Fed repo deals, QE-level open market purchases, and half a trillion in year-end repo tell us something is seriously wrong with the abundant reserves regime.
• This is all rooted in the accelerated inflows of Federal debt, and the inability or unwillingness at current rates, for private nonfinancial sectors to absorb it. Full QE is almost inevitable.

Daddy’s Daddy

The Bank of International Settlements is sponsored by about 60 central banks, ranging from the Fed to the Bank of Algeria and everything in between. Mostly what they do is produce white papers, but they also provide crucial international counterparty settlement functions, as the name suggests. It’s located in Switzerland, which makes it a sweet gig for macroeconomists who like to write white papers and attend symposia in European capitals.

Last week, the BIS’s Quarterly Review came hot off the presses, and unusually it caused a bit of a stir from a 3-page section on the September Repo Revolt. BIS has more granular data than us, down to the firm-level, and they focused on a couple of things in addition to the oft-mentioned perfect storm of new Treasuries, corporate tax payments, and payrolls all coming at once.

• The US Treasury was coming off the debt ceiling extension in August, and was restocking their checking account - more perfect storm upper atmosphere turbulence. This pulled out $120 billion in reserves in the preceding month, half of that in the preceding week.
• The four largest US banks, presumably JPMorgan (JPM), Bank of America (BAC), Citibank (C), and Wells Fargo (WFC), have taken on a ton of Treasuries. Over 40% of their combined liquidity is in Treasuries, up sharply since 2017.
• They also suggest through indirect evidence that much of the stress is coming from the secondary FICC-cleared repo market, where hedge funds do their contracts. Let the speculation begin.

Overall, the story they tell calls into question some of what we are hearing from the Fed and the big banks, and suggests that more structural issues are at work, outside the Fed's control.

If you need a refresher on the September Repo Revolt and how the overnight markets work, you can find that here.

The Big Banks Have Too Many Bonds

Let's back up a second and review the landscape a bit. I've been focusing a lot on this chart:

Federal Reserve. Through the end of September.

Concurrent with each other, the Fed and international buyers reduced their holdings of Treasuries as a percentage of all Treasuries. Once the Fed stopped QE3, their portion of the debt inched down with new Treasury issuance, and then accelerated during QT. Overall, these two holders had 48% of Treasury debt at the end of Q3, down from 62% in 2014. Let's see where those bonds are going, and why it is pulling out reserves to the tune of $1.2 trillion over this period.

Of that 14 pp no longer owned by the Fed or the rest of the world, some has been absorbed by households/nonprofits:

Federal Reserve. Through the end of September.

Some by funds (money-market, ETFs, mutual funds, etc.)

Federal Reserve. Through the end of September.

Some by pensions (all public and private):

Federal Reserve. Through the end of September.

But most importantly for our purposes, banks and dealer/brokers have seen their holdings jump since last year:

Federal Reserve. Through the end of September.

That jump you see there is an additional $300 billion in bonds the banks have on their balance sheets since last year. With their anonymized firm-level data, BIS has added some detail here. The big buildup has primarily been with the Big Four.

In the first place, the entire banking sector, and the four biggest in particular, have become net lenders of reserves, when previously they were net providers of collateral. This has put a lot of stress on their reserves.

BIS. Through the end of September.

The whole sector has seen Treasuries as a proportion of their liquid assets rise, but the Big Four much more so:

BIS. Through the end of September.

Over 40% of liquidity in the Big Four is Treasuries. Finally, they give us this chart for individual banks:

BIS. Through the end of September.

I have some guesses on which banks are which, but I'm not sure enough to put it in here. But suffice to say Banks 3, 7, 9 and 17 have very Treasury-heavy liquidity structures.

So adding it up:

• Foreign buyers and the Fed have reduced their portion of the debt considerably since 2014.
• Other buyers like households, pensions and funds have stepped in, but it has not been enough with $2.2 trillion in new debt coming in from 2018 Q1 to 2019 Q3.
• The banks have had to pick up the slack. Now that the Big Four are providing almost all of liquidity every overnight, whenever they are tight, the whole system is.
• Banks may also be front-running another round of QE, which is starting to look inevitable.

The Part That Gets Everyone Hot

I'll let the BIS take this one:

Shifts in repo borrowing and lending by non-bank participants may have also played a role in the repo rate spike. Market commentary suggests that, in preceding quarters, leveraged players (eg hedge funds) were increasing their demand for Treasury repos to fund arbitrage trades between cash bonds and derivatives.

This is inferred by BIS. Hedge fund repo deals clear in the secondary FICC market and two things have happened there:

1. High spikes in activity in 2019.
2. 20 bps spreads between the FICC repo rates and the regular repo rates in 2019.

They infer two things here:

• Demand for repo deals from hedge funds is growing to cover leverage overnight.
• "During September, however, quantities dropped and rates rose, suggesting a reluctance, also on the part of MMFs [money market funds, who provide the liquidity here], to lend into these markets."

What gets everyone hot is the implication if this is true. Hedge funds need liquidity, but can't get it in the primary market, and now they are having trouble getting enough from money market funds. What is happening in effect now is that the Fed is providing liquidity to the banks so they can provide it to the hedge funds, so they don't have to sell leveraged positions and drive stocks down.

If one were of a suspicious nature, it is easy to imagine that the Fed not only sees that this is happening, but that it is the reason they've stepped in to begin with - to prop up stocks. My own take is that this is not their goal, but is a side effect that they're not inclined to do anything about.

The Dimon Defense

This gets back to this exchange during the last JPMorgan earnings call.

Glenn Schorr (Evercore)

Curious your take on everything that went on in the repo markets during the quarter, and I would love it if you could put it in the context of maybe the fourth quarter of last year. If I remember correctly, you stepped in, in the fourth quarter. So higher rates, threw money at it, made some more money, and it calmed the markets down. I'm curious what's different this quarter that, that did not happen. And curious if you think we need changes in the structure of the market to function better on a go-forward basis.

James Dimon

So if I remember correctly, you got to look at the concept of - we have a checking account at the Fed with a certain amount of cash in it. Last year, we had more cash than we needed for regulatory requirements. So repo rates went up, we went with the checking account which paid IOER into repo. Obviously makes sense, you make more money. But now the cash in the account, which is still huge. It's $120 billion in the morning, and it goes down to $60 billion during the course of the day and back to $120 billion at the end of the day. That cash, we believe, is required under resolution and recovery and liquidity stress testing. And therefore, we could not redeploy it into repo market, which we would've been happy to do. And I think it's up to the regulators to decide they want to recalibrate the kind of liquidity they expect us to keep in that account.

And again, I look at this as technical. A lot of reasons why those balances dropped to where they were. I think a lot of banks are in the same position, by the way. But I think the real issue when you think about it, is does that mean that we have bad markets because that's kind of hitting a red line in that checking account. You're also going to hit a red line in LCR, like HQLA, which cannot be redeployed either. So to me, that will be the issue when the time comes.

And it's not about JP Morgan. JP Morgan declined - in any event, it's about how the regulators want to manage the system and who they want to intermediate when the time comes. [emphasis added]

In Dimon’s telling, this is all about regulators requiring too much reserves for the very large banks, so they could not take their $120 billion off the sidelines and get the 10% overnight rate that was available briefly. Under the Basel III structure, the large banks are required to have 30 days of liquidity under worst-case circumstances, though Treasuries count towards that. But clearly, he would like that to change.

Dimon is talking his own book here, so there is reason to be skeptical, but even taking him at his word, it’s only a partial explanation. Schorr wisely asked the question in the context of the earlier year-end repo spike, where JPMorgan did step in and get those high overnight rates. The difference, says Dimon, is that JP Morgan's reserves were lower than in December, and the bank had to keep them there because of regulations.

But he skipped out on explaining why reserves were lower in the first place. Always pay attention to what they don’t say. Now we have a better understanding how their reserves had declined so much. They have too many bonds.

In fact, when we add it all up, it's not crazy to suggest that the big banks are frontrunning a new round of QE by buying longer-term notes and bonds in anticipation of the Fed driving rates down with open market purchases.

The Fed Response Was Robust, and Getting Stronger

The Fed stepped in, and began providing a daily facility for overnight repo deals, as well as longer-term 2-week deals. As I explained a couple of weeks ago, September and October were a little rough, but things settled down in November when the Fed upped their offerings.

Federal Reserve. Through the end of September.

An extra $16 billion a day will do that. When looking over the action, two patterns emerge:

• More demand around the 1st and 15th of the month, when cash is moving around for tax payments, payroll and the like.
• More demand for the longer terms that bridged the 1st and 15th of each month.

There was a lot of worry about the year-end overnight. Last year is when these problems first crept in during the 2018-2019 overnight, and there were many predictions of repo apocalypse for the 2019-2020 overnight. To quell fears, back at the end of November the Fed started their longer-term contracts that would extend into January. Through last Monday, there had been three of these:

Federal Reserve. Treasuries only. On fully subscribed offerings, the spread is typically 0 or 1 bp.

Even though the overnight and 2-week operations were going smoothly with the extra liquidity added in November, these three operations were well oversubscribed, and the spreads spiked. No one wants a part of that year-end.

Then last Thursday, the Fed decided to stop messing around and brought out the big guns.

Damn, son. Keep in mind those offering figures include mortgage and agency debt as well as Treasury debt, which is what we are focusing on.

When you add up:

• the previous three longer-term offerings,
• the five just announced,
• and the additional $225 billion the Fed will have on offer on December 30 and 31.

The Fed will have about half a trillion dollars in repo deals that span the new year.

The first one was Monday morning, and the Treasuries were only oversubscribed by 4% with a 4 bps high-low spread, so it looks like half a trillion may be enough. Interestingly, the mortgage-backed securities went off a little rougher, with a 15% oversubscription and an 8 bps spread.

So not only are the daily offerings getting larger, but the last two weeks of the year is going to resemble a fireworks display grand finale.

And There's Also This Thing

While all this is going on, the Fed is also back to open market purchases.

As you can see, current rates of balance sheet additions are consistent with QE. Are we in QE 4?

Not quite. The three rounds of QE added over $2 trillion in Treasuries to the Fed's balance sheet. Since August, the Fed has added $199 billion, with $153 billion coming just since mid-October, when the Fed realized its repo facility was insufficient to stem the tide of oncoming debt.

So, in the first place, though we are at a QE pace, the overall scale is still an order of magnitude smaller than QE.

But the bigger difference is that while QE focused on maturities in the 2-10 year range, in this round the Fed has been purchasing mostly short-term bills, which will come off the balance sheet soon enough:

Federal Reserve

In total, about 82% of this round of purchases has been bills since October 17. Until August, bills were 0% of Fed holdings and purchases.

So, this is not QE, but it is still juicing stock prices, by my calculations, a 74 point tailwind for the S&P 500 since October 17, over half its gain through last Wednesday, the day we get the Fed’s weekly balance sheet.

Data by YCharts

It's Not Too Early To Say It

I've written pretty extensively about all these repo problems, and I usually end with an admonition that "it's too early to say that the abundant reserves regime is a failure..." But add up:

• super-sized daily repo deals,
• the need for large weekly open-market purchases,
• the increase in longer-term repo deals,
• and the December 2019 grand finale.

This tells us that it is a failure. It only works if nonfinancial private sectors keep up with the pace of new debt, and they are not.

Pre-crisis, every overnight was like a circus balancing act with a bear and hundreds of cats. The NY Fed monitored overnight reserve trading, and bought, sold, and did repo deals to keep Fed Funds where it wanted it. The first side effect was a Fed balance sheet that grew very slowly over time. The second side effect was that the incentive for the banks was to keep their reserve levels just above the penalty-free minimum. They did not get interest on excess reserves, so aggregate excess reserves remained relatively low. As such, small Fed purchases or sales could have a large effect on the overnight rate.

When the crisis hit, part of the problem was a freezing up of these overnight markets. There was already a plan in place to beef up reserves, and keep the Fed from having to mess around every overnight, and Congress moved up implementation. The Fed began paying a low, but non-zero rate on excess overnight reserves to encourage more availability for repo. Effective Fed Funds and repo should be higher, so it’s basically free money, almost 100% risk-free for the banks. But that spread has narrowed from a median of a free 4.5 bps to only 1 bp since the Fed last eased.

We are back to the nightly circus act, and because the reserves are much larger than they were pre-crisis, and the amount of bonds has exploded, the Fed's daily interventions are an order of magnitude larger than pre-crisis. These Fed interventions will have to keep getting larger, and with longer terms. The Fed will have to replace every dollar of expiring T-Bill on their balance sheet, and then some when they come due. The alternative is a spike in short-term rates. Eventually, it will be of the magnitude of balance sheet expansion as QE, and they will have to start buying the longer terms.

The driver of all this is right here:

Federal Reserve

Since the beginning of 2018 when the tax bill went into effect, the Federal government has added $2.2 trillion in new debt, an average of $318 billion per quarter, up from $196 billion per quarter in the previous eight quarters. With the Fed no longer a purchaser, and also selling bonds in QT, and foreign purchasers not keeping up either, domestic nonfinancial private sectors could not pick up all that slack. The Big Four did, but they are now stretched to their limits.

So with the pace of new Treasury debt, the private sector cannot absorb it, and the Fed is forced to start monetizing the debt again. Not QE is destined to become QE, whether the Fed wants it or not, because the alternative is a spike in rates across the yield curve. This opens up a whole can of worms that's an entirely different article.

But this does not solve the overnight issue. The private system has become almost entirely dependent on the Big Four to provide liquidity, and they are very thin themselves now. Any day they are tight, the whole system is.

It comes back to a fundamental disconnect in the way Treasuries are treated during the day, and in the overnight. During the day they are cash equivalent, considered almost as liquid as cash, but overnight they are definitely not, and in the event of a bank run they are most certainly not.

I am not suggesting changing this, but that is why banks are finding themselves upside down overnight. The regulatory changes proposed by Jamie Dimon would solve the problem until they don't. This is an underlying structural issue beyond anything the banks or the Fed have talked about.

In the interest of keeping it simple, there are too many bonds:

• The pace of new debt has soared.
• The nonfinancial private sector can't/won't absorb it all.
• The banks are at their limits.
• The Fed has to step in, or interest rates will soar.

The abundant reserves regime only works when private nonfinancial sectors are keeping up with the pace of new debt. Otherwise, it is a failure.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.