Wide, Wide Sea of Bank Liquidity

The Days Ahead:

• Inflation reports and retail sales.

This Week:

• The Fed announced it would buy more Treasuries…
• That helps bank liquidity.
• We look under the hood of the money flow between banks, money market funds, the Treasury and the Fed…
• And why it’s good for risk assets.
• Bank loan officers are wary…
• And small businesses’ cost of loans is at a 24-year high.
• Unemployment and elections…
• Fun to look at but not predictive.

---

Wide, Wide Sea of Bank Liquidity

We recently wrote about the Federal Reserve’s (Fed) balance sheet and its effect on liquidity. The story was that Dallas Fed President Lorie Logan was concerned that the run-off, or reduction, in the balance sheet may cause liquidity problems for banks. Her basic point was if the run-off happened too quickly, banks would have to buy more Treasuries and that would deplete the deposits and reserves of the banking system. (Deposits and reserves are similar so we’ll use “reserves”). Lower reserves could impair bank lending and the daily functioning of the money markets.

As we’ve noted before, daily transactions, borrowings and money settlement generally go off without a hitch. That’s great because every day all these flows through the financial system: around $22 billion through the Treasury account, $512 billion in non-cash domestic payments, $7,500 billion in foreign exchange, $1,800 billion in interbank clearing and $4,300 billion in Fedwire transfers. Yes, many are double counted but any error on such large numbers can cause major problems.

The basic issue today is around four buckets of money: the Treasury General Account (TGA), the reverse repurchase operations (RRO), the Fed’s Balance Sheet and money market mutual funds (MMMFs).

MMMFs are now around $6,000 billion. They’ve doubled in size since 2018 and went up 26% in the last year. A typical fund probably had a total return of less than 0.25% per year from 2009 to 2021 as rates languished near zero. They now return around 5% so have, understandably, become more popular.

MMMFs normally hold short-term safe securities, such as 3-month T-Bills. They’re safe and allow the fund to price the security at par. But MMMFs also use regular cash bank accounts because they may pay more than a short-term investment. The MMMF can use the cash to buy a Treasury security from the Fed on which it receives a higher rate than the bank account. This is a repurchase agreement.

In a normal repurchase agreement, a MMMF owner of, say, a T-bill, goes to the Fed, sells it, and agrees to repurchase it the next day. A reverse repurchase (or RRO) works the other way around because the MMMF has the money and the Fed has the security. The Fed sells the bond to the MMMF and promises to repurchase it the next day. It sounds complicated but it’s just a temporary swap of cash for a bond.

During Covid-19 and the Fed’s zero rate policy, MMMF cash holdings were yielding practically nothing and the Fed’s securities were yielding more. Every night all the counterparties would call up the Fed, ask, “How much can I get if I buy your securities overnight with my cash?”

The Fed would give a number, sell the security and buy it back the next morning. In fact, the MMMFs gave the Fed so much money that the RRO ballooned to $2,500 billion and 52% of all MMMF assets.¹ (see below)

In all this, the RRO went up and that was good for banks because they had the reserves from the MMMFs. It didn’t matter for bank liquidity that the funds held securities overnight. Banks’ reserves were high. Liquidity was plentiful.

Then this:

The green line shows the RRO facility peaking at around $2,500 billion and falling to $468 billion today. In late 2022, the MMMFs invested more in securities than bank cash because the securities earned more. They didn’t need to use the RRO as much.

Where did all that money go? Some went to buy Treasury securities, which meant bank reserves went down, and some went to the Treasury rebuild of the TGA (blue line above). The TGA is like the government’s checking account. When it goes up it means the Treasury is selling securities and taking money from investors. Bank reserves go down. When the TGA goes down, the government is spending money and liquidity reenters the financial system. Bank reserves go up.

The final part of this is the Fed. When the Fed buys a Treasury, it does so from banks. At a Treasury auction, banks buy all the Treasuries on offer. During QE, the Fed then turned to the banks and bought the Treasuries banks had just bought at auction. When the Fed buys Treasuries, bank reserves go up. When the Fed stops buying from the banks, as in 2022, banks must buy more Treasuries and their reserves go down.

You may think that with the Treasury borrowing more and building its TGA, the Fed buying fewer Treasuries and the MMMF using banks less, bank reserves must be taking a pummeling. Well, yes and no.

The blue line is the level of bank reserves which soared in 2020 to 2021 during the Fed’s peak Treasury buying. They then went down in 2022 when the Fed stopped buying all those bonds. But these reserves were more than enough to handle the increased Treasury borrowing in 2023 and the draw down of the RRO by the money market funds.

The Fed doesn’t want to see the reserves drop to levels anywhere near 2019, which caused a panic in the money and rate markets. In late 2019, interbank rates went from 2% to 10%, causing the Fed to step in with large a repurchase facility. Thankfully, nothing broke. But it all started with the low level of reserves in the system.

Last week, we saw the Fed managing the TGA, MMMF, RRO and balance sheet interactions. It reduced the Treasury runoff in its balance sheet from $60 billion a month to $25 billion. That announcement and a few adjustments to other programs means that the Fed will now buy around $270 billion more Treasuries than expected between now and September. This will help banks to build more reserves. And that’s also good for risk assets.

Apologies for the jargon. It boils down to:

1. The Fed is going to buy more Treasuries.
2. This means banks don’t have to buy those Treasuries and reserves go up.
3. The government will stop building the TGA.
4. If the government doesn’t have to build its TGA, it means reserves go up
5. If MMMFs slow down their use of the RRO, reserves go down

Our guess is that the Fed and Treasury are working together to keep the money markets as stable as possible. The Fed is slowing the run off in the balance sheet, Treasury is keeping the TGA high and issuing more bills for the MMMFs to buy. The Fed will also keep an eye on the RRO to make sure it doesn’t run down too quickly.

The Fed is being very prudent. Reserves aren’t low now and the Fed would like to keep it that way too. Last week we saw Fed actions that should ensure the reserves situation remains stable. It’s not the same as a rate cut but it’s proactive and a positive policy signal. That’s good!

Senior Loan Officers Say…

The Fed surveys bank loan officers four times a year to measure credit conditions. It asks who’s borrowing, what for and at what rates. That way we  know how banks feel about mortgages, commercial and industrial (C&I), commercial real estate (CRE), construction, and consumer loans. The Fed supervises 5,100 banks ranking from J.P. Morgan with assets of  $3,395 billion to Bank of Little Rock, at place 2,129 with $300 million. There are around 3,000 banks with less than $300 million in assets, many of them community and single branch banks, the smallest of which is Kentland Bank in Indiana with $2.8 million in assets.

For a while banks have been reluctant lenders. Domestic banks grew their commercial loans by only 1.8% in the last year and their consumer loans by 0.4%. Meanwhile, nominal GDP rose by nearly 9%. Loans as a percentage of GDP are 46% of GDP compared to 48% in late 2022 and 50% from 2012 to 2020. When loan growth isn’t keeping up with GDP, you know there’s a reluctance to extend credit.

The green line in the above chart shows the net percent of banks tightening lending standards. If the line is above zero, it means more banks are tightening than easing. It’s at 15% which might seem better than 50% last fall but the survey is a “rate of change” measure not an absolute measure. If 99% of banks were tightening standards and 1% of banks were easing in month 1 and in month 2, 100% of banks tightened, the green line would show up as 1%. As long as that number is above 0%, banks are tightening credit.

The blue line is the net percent of banks increasing spreads and is a rough proxy of wider profitability. That too is way above zero. But we’d worry more about the green line because “higher standards” covers things like terms, pre-payment clauses, loans to value, late payments penalties, and higher collateral.

It’s not as if loan demand is high either. The recent survey of small businesses from the NFIB showed the average interest paid on short-term loans was 9.8%, the highest in 23 years and 0.5% higher than when the Fed stopped tightening in July of 2023. It’s also the fastest rate of doubling on record. It rose from 4.9% to 9.8% in 24 months. It took 8 years to double from 4.7% in 2015 to 9.4% in 2023. No wonder loan demand is weak and lending trails growth.

This would all suggest that, yes, Chair Powell was right when he said last week:

“I do think the evidence shows pretty clearly that policy is restrictive”

And went on to repeat it 16 times. We’d agree. Loan officers don’t want to lend and borrowers don’t want to borrow at these rates.

Is There a Connection Between Unemployment and Elections?

We know there will be a heck of a lot of commentary in the coming months about the election and the economy. Every jobs, inflation, house price, and confidence survey may come with an election point. You know them. It’s often along the lines of “in the last _________ elections, an incumbent/opposing candidate/party has always/never won/lost the popular/electoral college when ________ happens.”

It’s tough because the sample size is small with only 26 presidential elections since 1900 when we started to see any meaningful economic data. There are just too many variables. Simplifying it down to a few economic measures risks making any forecast look foolish.

But, hey, we’ll start. The chart shows a win or loss by an incumbent party against the unemployment rate for the 18 elections since 1952.

What can we conclude? Not much. It’s unusual to see unemployment going rapidly up and for the incumbent party or candidate to win. It happened once in 2004. The other three combinations of unemployment down or up and an incumbent win or loss are evenly matched. In other words, it seems to be that an incumbent is only in trouble if unemployment is climbing fast. Other than that, unemployment doesn’t seem to predict much.

The Bottom Line

The market was quiet this week digesting the news from last week’s Fed meeting acknowledged that inflation was persistent but there weren’t likely to be any rate hikes. The Atlanta GDPNow shows growth estimates of 4.2% for the Q2 2024. That’s way above the Q1 rate of 1.6% and is based off population changes. We don’t think that will be close to the final number, which we’ll see in late July, but it does indicate things are going well.

The 10-year U.S. Treasury settled in to the lower end of its 4.3% to 4.7% range which we’ve seen since April. We’re in a period of “delayed” cuts, so we don’t expect the longer end of the bond market to change much.

Stocks had a good week with the S&P 500 up 2.8% and just 1.2% off its record high from early May. The S&P 500 Equal Weight Index also had a good week. That’s the index where every stock is 0.2% of the index, so the top 5 companies are 1% of the index, compared to the regular S&P 500 where the top 5 are 26% of the index.  We like it when we see the market broaden out.

---

Art of the Week: Andre Derain (1880-1954)

---

1. If you’re interested you can see the RROs in the annual reports (linked here and here) of money market funds where they’re listed as bank accounts along with the security they borrowed from the Fed that day. ↩︎