March 19, 2023

The Dizzying Array of Accounting Gimmicks Preventing Silicon Valley Bank’s Failure From Affecting the Debt Ceiling

The Dizzying Array of Accounting Gimmicks Preventing Silicon Valley Bank’s Failure From Affecting the Debt Ceiling

Normally I do not release pieces on Saturdays or Sundays. However, this is the third anniversary of the first piece I ever sent of this newsletter. That brief note, appropriately titled “Sign of the Times”, was more like a glorified social media post than a newsletter. As I said in that very first piece “This is the big one and I think everyone now realizes that”. I am enormously proud (and still bewildered) by how much I was able to subsequently write for those first four months. I covered nearly every complicated step of what happened.

More to the point, I'm still shocked by how overwhelming the audience response was — culminating in a globally translated viral profile in Bloomberg Businessweek. Still, even three years later, I am not used to how much my status has changed personally and how many people now take my opinion very seriously. I can’t ever express how grateful I am to all my readers. You have all changed my life forever.

It has felt appropriate then, as “financial” events have recently “sped up”, to be publishing pieces at a faster pace. This is my third piece this week, coming on this deeply special third anniversary. Before Silicon Valley Bank failed, I was putting the finishing touches on a book proposal for a book on the Federal Reserve (and I will be officially announcing sometime in the coming weeks.) I’m so excited for that next phase in my writing, and in this newsletter. To those paid subscribers who have shown me patience, I also am extraordinarily grateful.

Finally, I want to say something about the concept of crises. Getting success for covering them can sometimes be a strange feeling. I am, in a sense, “countercylical”. My peaks are everyone else’s troughs. Put simply, when global economic conditions get rough, I get more attention. As much as I’m grateful for the success my readers have brought me in so many different ways, it would be nice to make this newsletter irrelevant by moving past crises, and solving our major socioeconomic problems. The Coronavirus pandemic is not over, even if the depression has been for sometime. I hope someday I’m rendered completely obsolete as a newsletter writer, and I will be happy to move ahead to something else. In the meantime however, enjoy this piece about an especially complex set of arrangements that lie behind the current crisis response. Subscribe here if you want to support these pieces being free for the next few weeks.

I try my best to provide clear and simple explanations. I was really gratified to hear from many readers who read my overall guide to what happened to Silicon Valley Bank. They reached out to tell me that they found it to be a simple and clear explanation of quite complicated material. Sometimes however, there is no clear and simple explanation that will make what is confusing easily and fully understandable. In today’s case, that is true because what we are going to discuss simply is convoluted and confusing (even to me!) I became globally known in part for explaining the very many accounting gimmicks which lay behind the Federal Reserve’s crisis facilities. Today’s piece is a return to that same subject, but now in overdrive.

As a result, even though I am going to devote many hundreds of words to make what I am talking about understandable and easy to follow, I am still going to sound a little bit like Charlie from It’s Always Sunny in Philadelphia explaining how there is no Pepe Sylvia.

When readers think of the Depository Insurance Fund, I am sure an image pops in their mind. They are imagining some bank account somewhere filled with funds, where insurance claims are paid out of and “premiums” by banks flow in. The issue with this image is, that unless you are one of a handful of people obsessed with financial system details, you aren’t thinking about the larger institutional issues involved. To make my point clear, let's ask a simple question, where is that bank account? Is it with a… bank? Are those deposits… uninsured? If they are insured… who is insuring them? Where do they keep the funds to insure the deposits, which are meant to insure all the other deposits? Is there a depositor insurance fund for the depositor insurance fund?? Who insures that fund??

As should be obvious by now, this is something of an infinite regress problem. Of course, none of this is actually at issue because the FDIC does not have an account with any chartered bank or hold private financial assets. Instead, like social security, it has an account — really an accounting entry — with the United States Treasury. These accounts are, confusingly, called “non-marketable treasury securities''. When the FDIC needs to “draw down” the Depositor Insurance Fund, what is really happening is this “non-marketable treasury security” account gets marked down. At that point, the treasury sends a payment from its checking account at the Federal Reserve, called the Treasury General Account.

Thus, when you consolidate the Federal Government (even if you wrongly exclude the Federal Reserve), outflows to cover deposits from the FDIC are always a use of public money. Those non-marketable treasury securities are just intragovernmental accounting tools used to authorize the payments (dare I say gimmicks?) but at the consolidated level, all the non-marketable treasury securities in the DIF — or the social security trust fund for that matter — disappear. All there is is public money flowing out, and public money flowing in. That means this narrative about now “taxpayer money” (a nonsense phrase) was “not involved” in supporting uninsured depositors is nonsense.

Really, funds flowing out of the Treasury’s checking account are not the only form of public money involved. There is also Federal Reserve discount window lending to the new banks the FDIC has set up. As I discussed in my piece on Tuesday, the “discount window” is just a term for the part of the Federal Reserve where banks can get loans from the Federal Reserve. I will write more about the full scope of the Federal Reserve’s discount window lending this past week another time. For now it suffices to briefly say the scope of this discount window lending is absolutely massive. These loans are altogether 295.6 billion dollars, a truly unprecedented amount. For context, that is substantially more discount window borrowing than happened during the 2008 financial crisis. The new Bank Term Lending Program had a little under 12 billion dollars of lending. There is also apparently more than 10 billion dollars outstanding from the Payroll Protection Program Liquidity Facility (a silly program NotC discussed way back in April 2020). Thus, all together the Federal Reserve has 318 billion dollars of loans to chartered banks outstanding. Roughly 300 billion of those dollars went out this past week.

This brings us back to the Depository Insurance Fund. 142 billion dollars of those discount window loans went to the new banks the FDIC has created to take on Silicon Valley Bank and Signature Bank’s assets. There is a specific subsection of the “discount window” part of the Federal Reserve Act which governs lending to FDIC owned institutions entitled 10(b)(3). This is important for a few reasons. One, it clarifies that this is discount window lending (a point which a number of news outlets have incorrectly reported over the past four days.) Second, it makes clear that the Bank Term Funding Facility was not created to support these institutions. Instead, they were created to support other chartered banks, which may be having liquidity issues. This is an interesting little fact in and of itself.

Looking at such a large amount —  142 billion dollars! —  the question arises what role do those non-marketable treasuries in the DIF even play at all? Don’t these discount window loans “cover” those deposits? In some sense, that is true. In a few others, it is not. For one thing, the timing between a discount window loan and the funds depositors setting up accounts at other banks need right away might not match up exactly. I’ll return to this issue in a second. More fundamentally, it must be remembered that at the end of the day, the FDIC is dealing with primarily a capital issue, not a liquidity issue. After all, if deposit insurance were simply a matter of liquidity then the Federal Reserve could handle it all! In other words, the issue is not making payments. Instead it is making up losses to increase the net worth of this new bridge bank.

Recall for a minute how things worked before deposit insurance, and also how things worked for uninsured depositors, under the least-cost rule. The bank would enter receivership, and payments would instantly stop flowing out the door. The reason is payments tend to be, well, 100% of what’s owed. (At least in one specified monthly payment or a depositor outflow). But there are not enough assets to liquidate to cover every creditor at 100% (or at least, that is not clear yet). What that means in practical terms is that every creditor who receives 100% of what they are owed (either in full or in one payment) means every other creditor gets a smaller percentage. Instead, different types of creditors have different priorities in bankruptcy.

In simple terms: different types of creditors are put in either the front, back or middle of the line.

What deposit insurance is doing is injecting assets into the bank in order to increase its net worth and cover, at the very least, all the insured depositors. As we’re seeing this week they are covering the uninsured depositors as well. It might end up being the case that there are enough assets to cover the insured and uninsured depositors. Then it would just be the bondholders and stockholders who were wiped out. There might even be enough to cover some of the bondholders. As I pointed out on Thursday, ironically the response to the collapse of Silicon Valley Bank could possibly change interest rate policy enough to greatly improve the value of their securities. That would make far more creditors whole. In that case, there would be no ultimate impact on the Depository Insurance Fund’s accounting entry with the treasury.

Which brings us all the way back to where we started. Roughly 40 billion dollars was reported to have both “flowed in” and “flowed out” of the FDIC and thus the Treasury’s checking account. How could that be, if the FDIC’s banks got 142 billion dollars of loans from the Federal Reserve? The difference, I think, is that initially 40 billion dollars flowed out of the treasury’s checking account to cover depositor outflow before the discount window loans were arranged. Then, once those loans were made, 40 billion dollars flowed back. Thus the Federal Reserve filled up the Treasury’s checking account as a result of a direct loan to a corporation another agency of the federal government owns. In simpler terms, the Federal Reserve made a loan to the rest of the Federal Government. This is a great reminder that there are only restrictions on the Federal Reserve directly lending to the Treasury.

This still leaves us with the gap between the size of the discount window loans and how much was put into the treasury’s bank account. It’s hard to know exactly what’s going on without exact data on the balance sheets of the individual FDIC-owned banks, but I do think I have a rough idea of what happened. Recall that these FDIC-owned banks have their own checking account with the Federal Reserve, just like all other chartered banks which are members of the Federal Reserve. Also recall what these discount window loans were for in the first place — depositor outflow. These payment outflows needed to be covered —  and they were covered with discount window loans.

Thus it seems to me that the difference between these two amounts —  what was returned to the treasury’s checking account and the size of the discount window loans —  is a maximum estimate of how much deposit outflow there was. That would mean 100 billion dollars.

Now, these banks could simply have more settlement balances than they started out with before the discount window loans. However, given that before the FDIC took over Silicon Valley Bank it had 42 billion dollars in (attempted) depositor outflow in a single day, I am inclined to think that 100 billion dollars is a good estimate for how much payment outflow Signature Bridge Bank and Silicon Valley Bridge Bank experienced this week. (Naturally that provisional guesstimate is subject to revision, once we get actual data.)

We’ve gone on a journey so far in this piece, and I thank readers for their patience. This brings us to the “debt ceiling”, which motivated this entire piece in the first place. A number of people have expressed concern that covering uninsured depositors will impact the “x” date —  the date that the debt ceiling fully binds. At that point, the treasury has to either default, unconstitutionally raise taxes, or cut spending…Or, well, mint the coin.

I actually wrote two pieces about the debt ceiling, minting a trillion dollar coin and the Federal Reserve’s role as the government’s banker earlier this year. Sadly, I forgot to promote them over this newsletter (oops!) Nevertheless, those two pieces caused quite a stir, so I promise this isn’t the last you’ll hear about trillion dollar coins on NotC!

All that said, this commonplace concern about this FDIC action moving the “x” date forward is unfounded. But it’s unfounded for an interesting reason. The debt ceiling counts these gimmicky non-marketable treasuries towards the debt ceiling. That’s true even though they have no financial or monetary impact on the economy. This is a case where doing legal analysis requires that you not even consolidate different government agencies together. Thus, whenever the Depository Insurance Fund is marked down, the treasury has that much more room to issue more treasury securities.

Of course, it turns out that even that was not necessary because the money that flowed out of the treasury’s checking account was refilled right back because of, essentially, a Federal Reserve loan to the federal government. So one form of public money made it so that we didn’t need a different form of public money. The “x” date will only be affected if there is more than 100 billion dollars of losses, and the depository insurance fund is marked as negative. In that case, they borrow from the treasury! This, of course, is silly. Different accounting gimmicks could make it possible to continue to borrow from the Federal Reserve.

Even sillier, this is not a practical concern, given the timing up the upcoming debt ceiling fight. Recall that resolving a bridge bank and finding a private buyer takes time. The “x” date is estimated to be roughly in June, perhaps July. Even in the unlikely event losses got that massive- perhaps because of more bank failures — the FDIC could simply move slowly. Since the Federal Reserve is covering all the payment outflows with discount window loans, the depositor outflow can be fully covered without the treasury sending any payments. Even if the balance held by the Depositor Insurance Fund was dwarfed by bank losses, the FDIC could simply not finish resolving that situation and “realizing” those losses so that they weren’t forced to draw on their “line of credit”. This is the case even though it's still public money going out the door. It also looks remarkably similar to Silicon Valley Bank’s plan to not realize losses. In that case, it was so that the bank wouldn’t fail. In this case, it would be motivated by not hitting the debt ceiling.

In other words, this part of high finance relies heavily on using one accounting gimmick, so that they don’t trip over another, which would trip over another. This is quite the game of avoid the dominoes.

Let’s step back for a moment. It’s when you take the full view of all this complexity that I really begin to feel like Charlie from Always Sunny. We have an accounting gimmick called the Depository Insurance Fund to legally authorize a certain amount of payments from the treasury to FDIC owned banks, and thus depositors. That accounting gimmick, in turn, technically owns another accounting gimmick which is called a “non-marketable treasury security”. But because that accounting gimmick has “treasury security” in the name, it counts against the debt ceiling. The debt ceiling, of course, is another accounting gimmick. The fact that “non-marketable treasuries'' are called treasuries makes it so that marking down the balance of the Depositor Insurance Fund doesn’t impact the debt ceiling. But it will arbitrarily start if that balance is marked as negative. But only because there isn’t an accounting gimmick to facilitate borrowing from the Federal Reserve, rather than the treasury — if you desire to realize losses sooner rather than later.

So far so simple, right?

Meanwhile, the Federal Reserve in practice makes sure that you don’t have to realize those losses until you are good and ready! They simply lend these FDIC owned banks whatever they need, and any excess funds fill up the treasury’s checking account and mark up the Depositor Insurance Fund. All of this could be simplified with much simpler accounting gimmicks that are far more understandable to the public. Of course, more understandable accounting gimmicks might be seen as politically inconvenient.

Indeed, we know they are! Compare this dizzying complexity and Rube Goldbergesque series of accounting gimmicks and creaky set of arrangements to the relative simplicity of depositing a coin at the Federal Reserve. Depositing a coin with a big face value — say one trillion dollars — might seem silly, but it takes a sentence to describe. The Federal Reserve is the Treasury’s banker. “The Treasury deposited a coin worth a lot of money in its banking account” is something a five year old could understand — even if they would also laugh. When you really understand the accounting gimmicks that propped up the Federal Reserve’s coronavirus crisis response, but also all these other areas of public finance, it is absolutely astounding and galling that a coin which will avert government default is where secretary of the treasury Janet Yellen draws the line and dismissively declares a potential solution to a huge problem an  “accounting gimmick”.

More fundamentally, it’s completely silly that deposits — which remember are bank created money and are seen as money by their owners — are technically and “ultimately” covered by issuing treasury securities. Why? If there’s any place that direct monetary finance makes sense, it's in making sure money is money. Give these FDIC owned banks a platinum coin, and tell them to deposit it at the Federal Reserve. I think this is a perfectly reasonable compromise. Give me a call Secretary Yellen — I’m willing to negotiate.