The Financial Times ran an opinion piece today that posed the question “who gets to create a dollar?” which is a common thought of mine, or at least one way to frame my favourite question ever, which is “what is money.” The topic has been covered several times on this site, but the world at large remains blissfully unaware, really, of the existential nature of money, and as a result, we find new ways to panic about it or, alternatively, to build incredibly silly castles in the air around it.
The latest panic, of course, comes from the failure of a bank on the West Coast, Silicon Valley Bank, which blew itself up in what was frankly one of the most avoidable ways to fail as a bank: they bought a lot of really long duration fixed income securities using deposit money that was highly uncertain to stick around and had only come around to them in the last couple of years. Lots of other banks have bought really long duration fixed income securities, but they did so with very stable retail deposits that had proven their “stickiness” as customer relationships. Silicon Valley Bank’s deposits came from venture capital firms and the startups they financed. The startups themselves should never have been viewed as “sticky money”: 70% of all startups fail in their first two years, and when they do, they famously “burn” through their funding. Thus most of SVB’s startup deposits should have been expected to go away quickly – certainly before their portfolio of low coupon mortgage backed pass-through securities would ever pay off. And a large portion of the rest of their deposits came from rich VC types and their fund vehicles who parked their cash at SVB while they decided what to do with it. With new funding on hold in a rising rate environment, and with VC people famously as interested in buying real estate and Maseratis as they are in keeping robust deposit balances at their friendly VC-oriented bank, again, any kid with a year out of college and the opportunity to take four hours of online banking training would have known that the deposit book was flighty – or “hot”, in the industry parlance.
So SVB’s depositors – predictably – drew down their balances starting late last year – nothing panicky, mind you, just startups burning through cash as they do and VC folks making poor luxury good purchasing decisions as they do. But SVB quickly ran out of its cash-cash, and had to start selling its really long dated fixed income securities. The trouble was, they had been bought at yields around 1.75%, and yields were inching up towards 5.00%. With an average duration of around 8 years or so, that meant these securities were trading at around 80 cents on the dollar – not because there was a risk that they wouldn’t pay, mind you, it’s just that they’d pay you back at a very low interest rate, so people buying them today aren’t willing to pay par for them in the way SVB paid par for them in 2021 and early 2022.
SVB then had to tell the market “hey, we just had $21 billion in additional deposit run off, no big deal right, sort of normal, but we took over $2 billion in losses when we sold our Treasuries and agency bonds to cover it, wow, sorry, and oh yeah, if we have more deposit runoff the losses on the rest of the book are even bigger in percentage terms, like, a lot more than all of the capital we have, sorry.” Unsurprisingly, other depositors thought “wait, if their capital is less than the current losses on their securities, they might not be able to pay off our deposits – after all, we’ve got a lot more than the $250,000 insured by the FDIC – so let’s get our money out now” and last Thursday, $42 billion of deposits were withdrawn, and sure enough, SVB ran out of value: they could not place enough collateral at the Fed to enable the Fed to release the withdrawn deposits, and the State of California – who regulated SVB – announced that the bank was closed.
As a banking professional, I thought “wow, dumb treasury guys over at SVB – or at least, dumb treasury strategy” and literally thought nothing of it. The good folks at the Fed and FDIC, though, thought “huh, a few other banks are in a similar position” and moved quickly to tell large depositors (not bond holders, and importantly, not all depositors – just ones with deposits that can immediately be called) not to worry, their money was safe, don’t go withdrawing it to put it in money market funds or with JPMorgan Chase, it’s all okay. And they told the folks at SVB that, no matter how much they had on deposit, they’d get their money back – effectively backstopping those uninsured deposits, although in practice, the FDIC will charge for any losses incurred after it figures out what to do with SVB’s bond and loan portfolios, which could easily involve a massive exchange of long dated securities for short dated securities with the Treasury Department in what’s called a debt management operation, normally kind of tricky but since the FDIC is just an instrument of the Treasury Department actually kind of just a paper exercise in this case.
As I mentioned, this happened because SVB did something terribly foolish – namely, pretend that their really “hot” deposit book that was always at risk of fast runoff was somehow stable enough to justify investing in very long dated fixed income securities. What they arguably should have done was invested in short dated securities – say, two years or less – and then if the depositors needed their money back, hey, in any quarter, a decent chunk of securities would be maturing anyway, and no need to sell and produce losses against capital. Had SVB done that, though, those dollars would have stuck around, chasing some new thing to buy. Oddly, by investing in long dated securities, SVB did the economy a huge favour: it destroyed value.
Wait a second, Freilinger – I can hear you say – why is value destruction a good thing? Well, simple: we have too many dollars in “the system” right now. The Fed and the US government created them, beginning back in 2008 during the last crisis and with a huge spike during the pandemic when all the stops got pulled because people were worried about a plague-induced general zombie economic apocalypse; now they need to be destroyed because they are, in essence, the real zombies created by the pandemic: zombie dollars that have nothing, really, to buy, and thus simply chase up the price of all goods through a monetary effect called inflation.
Really, any time someone who bought a long term fixed income security at par in 2021 or early 2022 – when the amount of excess dollars was at its height – was helping the Fed by creating an almost automatic forward destruction of value. In fact, the Fed did this by expanding its balance sheet at the time to buy such securities; it’s now sitting on a gigantic paper loss, but as the central bank, it doesn’t care about paper losses; it’s retained earnings account is utterly meaningless, so it can sit on its losses forever and the economy won’t notice (or it will, but it will be happy about it, because it will limit momentum towards price inflation). Banks who did so are generally a bit annoyed: they’ll have lower earnings for years to come, but as long as they didn’t overindulge – or invest hot deposit money that is unlikely to stick around – it’ll just be a drag on capital accumulation. But that’s also a good thing: bank capital accumulation at least somewhat results in more lending, which results in more dollars being created. Since the Fed wants to bleed all of the excess zombie dollars out of the economy, bank earnings drags are a good thing as long as it isn’t too much of a drag – and viewed across the banking system as a whole, it really isn’t, in fact its a long way from being too much of a drag.
But individual banks may have made poor decisions. Indeed, having taken a look at data from all banks at the end of December 2022, only a couple of other banks with assets above $25 billion or so were even close to the level of “whoa, what a terrible decision to buy so many long dated fixed income securities” of SVB. Signature Bank, which was seized and put into receivership last Friday as well, was one of those. The other half dozen had much, much larger insured, small retail depositor bases, or in one case, had the majority of deposits in forms of trust accounts, which aren’t really at risk for either withdrawal or loss on withdrawal as the purchases are made largely at the direction of specific trustees.
Those individual banks should fail. Perhaps the odd decision on the part of the FDIC was to backstop the uninsured deposits, but taking a look at SVB and Signature Bank’s capital position, it’s not that odd: they each had enough equity and subordinated debt capital to absorb the securities portfolio losses, meaning the actual financial risk to the FDIC was quite small. Indeed, it’s smaller today, as the markets acted predictably against their own interest but directly in line with the Fed’s zombie dollar killing strategy, selling risky assets at a loss and buying government bonds, reducing the amount of money chasing price inflation and making it more likely that banks will just burn off the excess via earnings drags for the next decade or so.
Which brings me to my ultimate point. What is money?
To begin with, money is different in aggregate than it is in quantum – which makes it a lot like physical reality. Describing quantum behaviour gets into all sorts of weirdness which, when you describe physical systems on a macro level, you don’t see, and indeed you observe certain macro impacts which can’t directly be observed or even proven to exist at the quantum level. So if you think a lot about quantum mechanics, and specifically the boundary conditions between gravity and quantum mechanics, this stuff will probably be like “oh right, I get that”. If not, continue with me.
Money in aggregate is the instantaneous sum of all expressions of value in a given closed human exchange system or what we can call Value, using upper case and bold letters the way mathematicians do when they define special sets like R, the set of all real numbers, and I, the set of all integers. By “closed system”, I mean a system of human value expression and translation – or a “market” in the Adam Smith sense – which allows for fungible instantaneous exchange. A Unit of Money is 1/nth of that Value at a given moment in time, which by convention we describe as a dollar or a pound or what have you. And Units of Money can be created or destroyed, but Value is always an integralised construct of both real things – primary goods, secondary goods – and virtual things – desires, ideas, services, threats, shames. Value is changing constantly based on things like productivity but also based on social and personal sentiment.
Units of Money are also changing: in a world of gold, it changed based on discovering new gold mines and melting down prior gold that had been locked into jewellery and losing gold by having ships get sunk in hurricanes. In a world of fiat or “declared” money, it changes based on what central banks and their agents, private banks, create or destroy – and now in a world of crypto, fiat money can be destroyed by being siphoned into buying nothingness in the form of bitcoin or ethereum, which themselves can be created or destroyed (or rather, for bitcoin, lost to the world of the living due to misplaced passwords). Oh, and Units of Money are the discrete mechanism by which ValueN in a given closed exchange system can be exchanged for ValueM in a separate closed exchange system. At some point, if enough instantaneous exchange takes place between two systems, ValueN and ValueM, the systems can be considered simultaneous and undifferentiated, but that takes time to realise and because human beings are rooted in time and place, usually isn’t recognised until well after it occurs. In any event, back to our story in the Value space of US dollars.
The Fed created more units of money back in 2008 at a time when banks were rapidly destroying it; their desire was to keep the number of Units of Money constant so that psychologically, the observance of deflation – the drop in prices of items of value in the exchange marketplace – wouldn’t accelerate what was already happening in consumer and business sentiment, which itself was destroying Value. This largely worked, although the destruction of Value was severe enough that, for most of the next ten years, central banks and banks themselves found it difficult to restore the ability to grow Value such that people felt good again (remember Value is partially a function of sentiment, not entirely based on actual items). By 2019, though, that long slog was largely over, such that central banks and banks could begin to see a path towards restoring a normalized balance between how Value grows (with population, with innovation, with productivity, with the growth of humanity’s conception of what can be value-able) and the number of Units of Money in existence across the world’s various and not-entirely fungible exchange markets. Then the pandemic hit, which did more than cause a Value crisis – it fragmented the world’s exchange markets in a fundamental way that really hadn’t happened since the 1930s. Central banks sort of panicked – though in fairness, there was no playbook – and they stimulated the production of massive amounts of new Units of Money as a palliative.
We’ve learned since then that this wasn’t as necessary as was thought; Value did not plunge as a result of Covid, and if anything, humanity has used the experience to expand its language of exchange and of how to create value and meaning, even as certain real asset markets have detached from one another. But there are too many Units of Money around even for that expansion, and now prices are climbing as too many Units chase not enough Value. So the central banks need to destroy Units without causing a parallel (but false) psychological or sociological sense that Value is actually contracting, instead of the Units of Money contracting while Value remains constant.
In this exercise, ironically, the failure of banks is a good thing. It destroys Units of Money in that the failed bank will liquidate securities and loans that were created at par for a value below par; the lost Units of Value literally evaporate. It would be less disruptive if the banks just limped along for awhile – zombie banks actually consume zombie Units of Money over time – but that can take a very long time (just ask the Japanese). Arguably – and the US argues this despite much of the rest of the world disagreeing – it’s better to just allow for failure to be clean and simple, to rip the band-aid off quickly and destroy the zombie dollars in one go. We will, on a certain level, play out that thesis over the coming weeks and months, but past banking crises in the US as compared to their evolution in other regimes suggests the US has got it right.
In that sense, even taking emergency measures such as a quick “guarantee but not really” of uninsured deposits at SVB and Signature is quite clever. In a stroke, the Fed and the FDIC wiped out the substantial equity and subordinated capital of the two banks – in aggregate around $40 billion – while the failures themselves made it more unlikely that any further immediately recognised losses would be required. As a counterpoint, Credit Suisse will probably limp along forever, keeping something like $80 billion of zombie dollars trapped inside a balance sheet that someone, really, should shoot with a twelve-guage, a la Woody Harrelson in the B-movie classic Zombieland.
The more pieces I read about money, the more I’m convinced that people feel a need to address it as though it were a real thing. As I hope is clear from this essay, I don’t think it is “real”, certainly not in a classical sense. Money is a purely human construct, and exists simply to fulfil a human need. As such, we should be prepared for its “nature” to transform over time and space, as the needs we experience change and as we charge the thing we designate as “money” with different tasks. I think, though, that at its heart, money is our way of expressing our need to explore how and what we value, and how we desperately hope that we are creating Value across the entirety of our very social species. The special secret of money is that it is nothing more than our purest and very much most human attempt to see whether we’re worth anything at all.
When a bank blows up, therefore – even if it’s your bank, and even if some dogmatic deposit insurance bureaucrat limits your claim to the statutory maximum – you shouldn’t really feel anything, any more than you “feel” anything when you drop a vase and it breaks. Your sentiment and your emotional state may change, but the world in its vastness has not, and if anything, the act of destruction – of a vase or of dollars – is the world giving you, and itself, and opportunity to change the function of what meaning is and can be. Pick up the pieces, tear up the checkbook, and create anew.