Fed to Banks: Use the Discount Window

The Days Ahead:

• Fed meeting and more company reports.

This Week:

• Transfer payments and income.
• They’re falling.
• But they have nothing to do with the labor market.
• Americans are delaying working to educate and train themselves.
• The Fed would very much like banks to use the discount window.
• Only half have even applied.
• GDP is too low and inflation too high.
• But this is not “stagflation.”

---

Unemployed in America

Unemployment in America has run below 4% for nearly three years and is about the same as it was from 2017 to the eve of Covid-19. By any metric, it’s rarely been lower. Record levels of people are working, so we were a bit surprised to see this from a noted economist.

Are all those transfer payments making people lazy and not want to work? We find this thought provoking. So, we dug around a little, to see what’s going on.

Let’s start with transfer payments. This is basically money moving from one group of the population to another. For example, last year, the government took in around $1,193 billion in social security taxes from taxpayers and paid out, or transferred, $1,353 billion in benefits. The $160 billion difference was paid from the  trust fund which is specifically set up to pay benefits. Similarly, some $49 billion was collected in unemployment taxes and $26 billion paid out in benefits, with the rest going to the trust fund. Medicare, Medicaid, unemployment and Veteran’s benefits are all transfer payments and all have trust funds.

Transfer payments as a percentage of personal income looks like this:

Yes, they’ve gone up from 13% to 18% since 2000. But that’s not 21%. And they haven’t changed since 2008.

It’s true that labor force participation dropped from a 67% peak in 2001 to 63% now. But are labor participation and transfer payment related? After all the labor force participation rate was 63% back in 1977. A few things happened, all of them good.

1. Younger workers went to college and training.

In the last 24 years, participation among those aged in 16 to 19 years old fell from 49% to 37% and in the 20-24 age group it went from 78% to 71%. But at the same time, college enrollment grew from 15 million to 19 million and the participation in trade school (everything from construction, culinary, auto repair and electricians) has grown 20% in the last few years. From 2000 to 2024, the percentage of the workforce without a high-school diploma fell from 10% of the workforce to 5% and those without a college degree fell from 24% to 20%. That seems like a very good trade. More educated workers and young people training for higher paid occupations rather than entering the work force unprepared. Transfer payments have nothing to do with this.

2. Women in the workforce.

Female participation rate fell from a peak of 60% in 2000 to 57%. But this too is fine. Female participation rates for 16- to 19-year-olds fell from 54% to 37% and for 16- to 24-year-olds, from 64% to 56%. This has nothing to do with women not going to work because they don’t need or want to. There were 1,218,000 women awarded degrees in 2000 and 1,760,000 last year. They’re delaying entering the workforce to better educate themselves. The participation rate for women aged 25 to 54 rose from 75% to 78%. Women are simply participating more in the higher wage and age groups. Again, transfer payments have nothing to do with this.

Transfer payments have risen as a percentage of personal income, but they don’t have anything to do with labor force participation. People are not paid to educate themselves. We’ve traded high participation amongst younger, less educated people for increased participation of older, well-educated people. That makes eminent sense in the modern knowledge economy.

Are unemployment benefits causing people to stay out of the workforce? Doubtful, given the unemployment safety is thin and temporary. The maximum amount of time someone can receive benefits is 26 weeks and 13 states provide less, with Florida the shortest at 12 weeks. The average weekly benefit ranges from $196 (Mississippi) to $426 (New Jersey), which are 16% and 35% of average national weekly earnings. It doesn’t seem that staying home is lucrative if it pays 30% to 62% of the minimum wage.

If transfer payments really are “paying people too long and too much” it would show up in the total amounts paid and in long-term unemployment.

Here’s the national unemployment benefits data:

The blue line shows the current rate of unemployment benefits paid out at $22 billion, unchanged in the last two years. In dollar terms, it’s the lowest it’s been since 2000, and around the same as in 1984. In inflation-adjusted terms it’s 20% below where it was 10 years ago. The green line shows unemployment benefits at 0.09% of personal income, the lowest on record.

Part of this is because only 1.8 million of the 6.4 million people unemployed receive benefits. Only 30% of the unemployed are eligible for benefits, down from 52% in 2000. Many people don’t have sufficient credits, work history or are independent contractors. The insured unemployment rate is 1.2% and has never been lower.   

Finally, if transfer payments were generous, the long-term unemployment rate would hike. But it hasn’t. The number of people unemployed for more than 27 weeks is 1.2 million, unchanged in two years. In the last 20 years, it’s ranged from 0.7 million to 6.6 million. As a percentage of the workforce, it’s at 0.7%, its lowest level since December 20001¹.

In our view the increase in transfer payments has had nothing to do with lower participation in the workforce. The lower number for some age cohorts is because people are taking longer to train and prepare themselves for the workforce. No one is being paid to stay out of the workforce. The dynamism of the U.S. labor force remains. Transfer payments are a popular but incorrect reason for any of its strengths or shortcomings.

The Fed Would Like Banks to Sign Up for the Discount Window, Please

The Fed’s discount window (DW) is the pressure valve for the money markets. If a bank can’t meet withdrawals, it has two choices. Shut the doors and call in the Federal Deposit Insurance Corporation (FDIC) or borrow money at the Fed’s DW.

If you run a bank and call in the FDIC, your board, shareholders and depositors will be mad but people will eventually get their money back, up to around $250,000 per customer….although in most recent bank runs, like Silicon Valley Bank in 2023, people got all their money back. But if you’re a senior manager of the bank the FDIC will invite you to seek opportunities elsewhere.

There are bank problems that only the FDIC can solve but the DW solves many liquidity shortfalls.

It’s not complicated. If a bank needs money, it goes to the Fed’s and takes an overnight loan currently 5.5% to 6.0%. Rates are always higher than the Fed Funds because the Fed doesn’t want you borrowing from them at, say, 5.0% and depositing the money at 5.25%. There are borrowing conditions but the basic theme is that if you need money to expand, for seasonal purposes or emergency credit, the DW is open for business.

Any bank can use the DW but there are a few steps.

One, you have to apply in advance.

Two, you have to pledge assets. The Fed will take pretty much any collateral from treasury bills to commercial and personal loans.

But, three, there’s a small catch. The Fed will apply a discount (which is how the DW got its name) of anywhere from 1% for a Treasury bill to 40% for a mortgage and 80% for undeveloped land or a boat loan. Once the Fed accepts your application and your assets, you’re good to go.

When the day comes you need money, you pledge the assets on the Fedwire system and cash is available within minutes.

The Fed is only too happy to provide liquidity in an emergency but they must first have the application and assets lined up. You would expect every bank would set up a line of credit. It’s free and quick. But the number of banks that are set up remains frustratingly low for the Fed.

In 2023, there were 9,537 banks and credit unions, 597 less than in 2021. But only 5,418 or 56% have signed up to use the DW. Only 2,917, or 30%, have pledged assets. Together they’ve pledged some $2,756 billion which is 12% of the total assets of the commercial bank system.

Credit Unions seem especially reluctant to enroll with 32% signed up and only 19% with pledged assets. The total pledged assets for credit unions is $130 million or 5% of total assets.

Why the reluctance? It goes back to the 1913 start of the discount window. Discount window borrowing is anonymous but it’s often seen as sign of financial weakness, especially if the borrowers’ names are leaked or analysts work it out from market activity. Before 2003, the DW was a very ponderous activity. Banks would apply, the Fed reviews capital ratios and assets, and then forwards the loan proceeds.   

During the GFC, the Fed realized banks weren’t using the DW and so set up something called the Primary Dealer Credit Facility (PDCF), which basically lent money on any terms against any good collateral. There was no stigma involved! The Fed closed the PDCF in 2021, and transferred the same terms over to the new, improved DW.

But banks again proved slow to sign up for the DW facility, until they suddenly needed it.

Here’s DW borrowing in the last 20 years. It’s normally around zero until all heck breaks loose. The Fed advanced $115 billion in three months in 2008, $51 billion in two weeks in 2020 and $152 billion in one week in March 2023. Even so, in 2023, some banks had no DW line of credit and ended up borrowing $673 billion from the Federal Home Loan Bank (FHLB) System, which is not meant to be a lender of last resort.  

What’s to be done? The Fed wants banks to use the DW but they’re clearly reluctant and probably unprepared. The Fed is about to use the carrot and stick.

On the carrot side, the Fed will broaden collateral eligibility to mutual funds, and corporate bonds and perhaps reduce the discount across the board. On the stick side, the Office of the Comptroller of the Currency (OCC) and FDIC will propose that all banks tap the DW at least once a year. That will force them to apply, pledge capital and run a trial loan.

If that happens, the stigma factor will disappear overnight and the spikes that we see in crisis times in the graph above will even out. Banks aren’t overly pleased with the proposals but given the speed of last year’s run at Silicon Valley and First Republic, which went from solid to closed in less than a month, it seems like a good idea to us. Meanwhile the Fed will keep gentle pressure to, please, sign up for the DW.

The Bottom Line

Markets were steady for most of the week but took a turn down on Thursday when the Q1 GDP numbers came in at 1.6% compared to estimates for 2.2%. The weak point was household spending at 2.5% compared to 3.3% in the final quarter of 2023. Net exports and inventory were also weak. The core inflation number was also at its highest since June 2023. The market was fine with one or two inflation misses, but this third one is looking like a pattern. The Fed always comes back to inflation as the reason to cut or stick with rates. The latest number won’t help the “reasons to cut” argument.

The road to cuts was always going to be crooked. This latest number shows slower growth and higher inflation than anyone would like. But the calls of “hey, this is 1970s stagflation” seem premature and exaggerated.

Stocks fell 0.8% for the day and are now 4.4% off their highs. The 10-year Treasury rose from 4.60 % to as high as 4.76% on the news which is its highest level since November 2023.

Meanwhile the Yen keeps ploughing lower. It’s at the same level it was in 1987. We’d expect the Ministry of Finance to start some intervention.

---

Art of the Week: Jane Culp (b. 1955)

---

1. As for other transfer payments, it’s true that social security payments were $422 billion in 2000 and are now running at $1,424 billion. But the number of recipients has also grown from 38 million to 66 million and that’s entirely due to demographics. ↩︎