Random Heresies on Bitcoin and Fractional Reserve
Or, equity, leverage, liquidity, and duration as a guide to understanding money
Ever the irrationally exuberant optimist, I think that an under-appreciated leading indicator for Bitcoin’s success is the seriousness with which serious people seem to take it. While Bitcoin Twitter can mostly appear to be chaotic, memetastic noise, from time to time one can detect some encouraging signal.
In the space of a few days in mid-February, Lawrence White edited (but not wrote, to be clear) a Coindesk article on crypto-dollarisation, and his long-time colleague at UGA and respected monetary economics and free-banking scholar George Selgin implored Bitcoin Twitter to take fractional reserve banking more seriously. Having long found Selgin to be practically the perfect gateway drug to Bitcoin, I was particularly excited to dive to his defence, but was also quickly frustrated that Twitter proved a less than convenient medium. So, I have decided to turn to the Bitcoiner’s favourite medium: Medium.
Everybody will hate this post.
Bitcoiners will hate it because I throw shade at the Rothbardian wing of the Austrian school. People who actually understand how banking works will hate it because I speak approvingly of Bitcoin. People who only sort-of understand how banking works will hate it because I exclusively use weird accounting concepts out of context. People who love a good Twitter fight will hate it because I reiterate Selgin’s plea to actually go read his books rather than assume every good idea can be reduced to a tweet or two. My trolls will hate it because, once again, this isn’t really an essay; it’s just a bunch of random thoughts that I wanted to put all in one place so I could copy and paste a link from time to time rather than composing impromptu 17-tweet-long threads. I believe there is a word for that, and I believe that word is “blog”. This is not the New Yorker. Get over yourself. Go Color.
The random thoughts are on the interplay of Bitcoin and fractional reserve banking (FRB), and I apologise in advance for any perceived heresy. And just to round it off, Selgin already hates it because he thinks it’s all semantic shenanigans obscuring a lack of substance.
At least half my argument is about semantics. Selgin was kind enough to look at an early draft, and this is the major edit from the version he saw. My argument is not a legal or historical one. It is more of a moral and an economic one, which follows from my preferred framing of the issue of FRB in esoteric moral and economic terms. I am setting out to give FRB a very unusual framing. I do not intend to prove or disprove anything by this, and so this framing ought not to be read as an argument for or against anything. As such, I do not disagree with Selgin, nor am I really sure I disagree with anybody or any thesis in particular. I just insist that my semantics are better.
I am going to try to talk about banking in general, and FRB in particular, solely in terms of capital, equity, leverage, liquidity, and duration. It will seem very strange at first, I warn you, and may be hard to follow.
My sole desire and intention is to frame the issue such that my thinking on how Bitcoin fits into all this is readily understood by the conclusion. I feel this is somewhat merited because, as in almost every area it touches, Bitcoin is so new and so interdisciplinary that it breaks established framings. My framing is weird because Bitcoin is weird. That doesn’t mean I succeeded, but it at least explains why I tried.
To repeat, my argument is about semantics. Explicitly, deliberately, and with a goal in mind for which the meanings of words and concepts will be absolutely critical. Perhaps most importantly, and yet most boringly, “history” and “legality” is completely irrelevant to Bitcoin. Bitcoin is new and has no history, and Bitcoin is a honey badger that doesn’t care about your meatspace “laws”. It is a law unto itself.
Selgin on FRB and Fraud
It all started when Selgin set out to dismiss the historical and legal nonsense that FRB is literal fraud. His debunking of this hyper-Austrian trope at Alt-M is well worth the read.
The claim of “fraud” is, very roughly, the following:
FRB is selling the same thing to two different people, and hoping neither notice by trying to use it at the same time, or something to this effect.
To pad this out for readers unfamiliar with the debate within economics and economic history, there is often an additional gripe that, in the current system, regular people have little choice but to partake in this fraud in order to lead normal, functioning lives. That they simply have no choice but to use and abuse “fake money”, and that this in turn somewhat justifies there existing a “lender of last resort” to meet the false expectations of depositors in times of crisis. To be clear, both Selgin and I disagree with all of this, and the claim of fraud above. This entire post is my attempt to provide an alternative framing …
As a kind of warm up for the main act, we ought to contemplate how weird banks actually are as businesses. Are you a customer of a bank? You are?
Well, maybe not. Maybe you are a customer. And maybe I am toying with semantics here, but I put it to the reader that the majority of affirmative responders almost certainly were wrong, unless they have a liability on their personal balance sheet that must in time be paid back to a bank. As John Mulaney said, “I bought a house … well, actually, my bank bought a house and they are letting me live in it.”
Depositors do receive services in return for their deposits in a sense that makes them “customers”, colloquially. But for the purposes of this framing, I think it is more accurate to think of these as perks offered to entice their business. In addition to the interest, this is what the bank offers to secure this supply of capital.
People who take out loans are the real customers of banks. People who deposit money in a bank, and whom banks refer to as their “customers” are in fact their suppliers. Regular folks supply banks with liquid capital for a low price, banks turn around and give that liquid capital to their customers for a high price, and in doing so create a pair of illiquid assets and liabilities that sit on their balance sheets, while the price differential on the intermediation continually rolls onto their income statements. The real customers will almost certainly turn that liquid capital into illiquid — but productive — capital on which they will endeavour to earn a rate of return above the price they are paying for it. And on chugs ‘the economy’, as an entirely nonsensical noun.
This framing is weird, I admit. You thought you were a customer, when in fact you are a supplier. You thought banks dealt with ‘money’ when in fact they deal with capital of varying liquidities. In fact, it is probably even more accurate to describe the relationships not as suppler and customer, but as renter and rentier: banks rent capital from depositors and rent it out for more to borrowers. Banks are like WeWork. Reassuring, huh?
WeWork as an Anti-Bank
Actually, they are almost the exact opposite of WeWork: the problem banks have, which we will come to, is that their liabilities have a much shorter duration than their assets. It is worth noting that this is also their reason for existing, so I am not proposing “solving” this problem, merely acknowledging it creates a persistent tension and an ever-present risk that must be managed. WeWork’s assets have a much shorter duration than its liabilities. Arguably, both make money from the very fact of engineering this maturity mismatch and shouldering the risk that it goes horribly wrong.
But isn’t it interesting that the pricing differential is the opposite in the two markets? That’s the magic of liquidity, that pesky word I keep using for this capital instead of just calling it ‘money’ like a normal person.
Money is (by definition) the most liquid asset, and yet is not productive. Productivity is only possible with illiquid, production capital. But this must be purchased with liquid, financial capital. The more capital intensive your business, or the higher returns you would ideally like to make, the longer the period over which you will have to commit to such an arrangement for it to bear fruit. People want to be short money for as long a duration as possible and long capital for as long as duration possible (“duration” is finance speak for, how long, an average, until the bill comes due?). The relevant inverses are that people want to be long money for as short a duration as possible and short capital for as short a duration as possible.
When operating as a renter/rentier, therefore, if your asset is highly liquid (money, banks) then those going long this asset (i.e. having it as an asset on their personal balance sheet, so, depositors) will very likely want to do so for a shorter duration than those going short the asset (borrowers). We can’t say anything more specific without knowing more about the precise circumstances, but relatively, we can say that the intermediating institution — i.e. the bank — will likely end up with shorter duration liabilities and longer duration assets.
But if your asset is highly illiquid (commercial real estate, WeWork) then those going long this asset (landlords) will very likely want to do so for a longer duration than those going short this asset (partners on the journey to elevated consciousness). The intermediating institution — The SoftBank Vis- I mean, WeWork — will likely end up with longer duration liabilities and shorter duration assets.
And so, while it may seem that the mechanics of banks and WeWork are the opposite with respect to assets and liabilities, they are actually identical with respect to liquidity and duration.
What fun! We are now ready for the main act. Banks are in the asset management business. They do not create money, they create leverage. There’s your bumper sticker.
Now, this might seem like semantics, and in some sense it is. Bob Murphy said as much when I criticised his canonical stance on this:
And fair enough. Murphy is a well-respected economist at the Mises Institute, and Bitcoiners may be interested to know that he hosts a podcast the sole premise of which is to refute whatever Paul Krugman said that week (he and Riva should team up). I do not intend to diss him, or even to disagree, frankly. When a bank grants a loan under FRB it may very well be boosting M1 and higher monetary aggregates, depending on the bank’s own source of funding.
I just dispute that the concepts of ‘M1’ and ‘higher monetary aggregates’ are useful. My intention in this entire heretical ramble is to ignore these outputs of econometricians (about as seductively useless as ‘GDP’, ‘CPI’, and many more such contrivances that are similarly agnostic to ‘time’) and instead frame the entire discussion as much as I can around the concepts of equity, leverage, liquidity, and duration.
My argument is semantical only insofar as strongly advocating a consideration of the meanings of certain concepts being far more elucidating to understanding the mechanism in question than are others. What I don’t think Murphy realises is that his argument is just as semantical as mine. But I know it, and mine is also better.
Banking via Accounting, not Economics
So, banks create leverage, not money. What does that mean? First of all, I don’t want to answer “what is money?” here, as this post is on track to be heretical enough. Let’s just say its ‘the most liquid asset and hence primarily a medium of exchange’ and note that, besides, what matters to this discussion is only that, at t0, some money exists and that at t1, we must check what has happened.
Let us assume that, at t0, Alice has $100 of ‘money’ that she wishes to deposit in a bank. That can be considered an asset, obviously, but also it can be considered equity. There are no counterbalancing liabilities so far in this setup. Alice deposits $100, and the bank turns around and loans it to Bob. Now Bob has money. Is there twice as much money? Not so fast …
Let’s work backwards. Bob has $100, an asset, but also a liability reflecting the fact that this money is on loan to him. He has no equity. He is fully levered. The asset that matches this liability is with the bank. It is its “Bob’s $100 loan” on its balance sheet. So does the bank have equity? Also, no, because they have an equal liability to Alice on their balance sheet as ‘Alice’s $100 deposit’. The asset that matches this liability is with Alice. But is this money?
No. In accounting terms it is best described as ‘accounts receivable’. But it is not money; it is a claim on money that is redeemable with the bank for money. Alice still has $100 in equity (just with a different asset), the bank has $100 in assets, $100 in liabilities, and $0 in equity, and Bob has $100 in assets, $100 in liabilities, and $0 in equity. Bob is clearly levered, given the setup: he borrowed to invest. But crucially, the bank is too. Net equity is still $100, but aggregate liabilities has jumped from $0 to $200. The system is now twice levered, because two loans have been made against 100% of the available equity. Hence, we have created leverage, but not money.
Readers may have immediately objected on the grounds of, effectively, semantic shenanigans on my part. What I insist on calling ‘accounts receivable’ are, in real life, precisely what we call money, no? Aren’t they pieces of paper that (dishonestly) promise to pay the bearer on demand, and aren’t they demonstrably even more liquid than the precious metals on which they were once claims? Isn’t money money?
In making this seemingly convincing accusation, we are illicitly switching back and forth between what we are calling money in the first instance. Say the certificate is money, then note that, in this scenario, it would have been given to Bob, not to Alice. Bob is the one who needs liquid capital as a medium of exchange. Alice is the one providing it. But then we must wonder what Alice actually gave to the bank? If it wasn’t money then obviously this makes no sense. If only the certificate is money, then we are back in the situation as described initially, but we’ve just changed the labels around. It is possible that both are money (I mean, why not? we are making up the rules here anyway) and is actually the most helpful setup to grasp:
Alice gives the bank money-1, and the bank issues her a certificate for money-1, which also happens to be money: money-2. Then it lends out money-1 to Bob. Now things get interesting, and I think this is where anti-FRB-ers are likely to get the most excited. Surely now, the bank has created money? Isn’t it obvious that, because of this act, there is twice as much money as previously?
Again, not so fast. We need to add two caveats that make this realistic. The first is that money-2, by its definition, can be redeemed for money-1 at any time. This means that it is very much in the bank’s interest to hold enough money-1 to cover however many redemptions it deems likely. Second, imagine that it is actually far more convenient to hold money-2 certificates at the bank than independently, as indeed has been the case in real life, c~1700–~2020. Were this the norm, and Alice was going to give the bank both the money-1 and the engendered money-2, then it would make more sense for the bank to lend to Bob the money-2 and keep the money-1, given it would make very little sense for the bank to hold a claim against itself for something it doesn’t have.
What we see here, is that even this more complicated setup degenerates into a version of what we described at the very beginning. We could even propose that the bank then issues money-3 to Alice in receipt of her deposit of money-2, and that this is money too, but, clearly, the same logic would apply from the beginning all over. The final link to a realistic reflection of modern banking is to observe that (largely as a result of this for all n induction) we can actually safely ignore the indexing numeral and return to just talking about ‘money’.
In the real world, with money now mostly digital, ‘money’ and ‘receipts on money’ are effectively interchangeable, at least from the point of view of a regular person, as we really can no longer tell the difference. And irrespective of our potentially flawed perceptions, the actual difference is nowhere near as obvious as that between metal and paper. The analysis we must now provide is twofold, and closely related: has the bank committed fraud, and have we created more money?
I fully realise that “fraud” is a legal concept with a proper definition, but insofar as I have professed to only being interested in morality and economics, I will restrict the meaning of ‘fraud’ in this case to: make a promise that cannot be fulfilled. I realise this is highly incomplete and inaccurate in isolation, but I don’t intend to use it outside this exact context. Even the most astute and legally minded readers will notice that what I am describing would in fact be legal fraud, even if my definition is clearly incomplete.
Has the bank committed fraud? This is an interesting question. It depends on how much the bank keeps ‘in reserve’, or, in more contemporary terminology given we aren’t actually talking about ‘reserves’ any longer, how much of an equity cushion it has. I would argue that the question is not answerable in the abstract; it is only meaningful when observed in real life. If the bank always has a large enough equity cushion to meet the calling of its short duration liabilities, then it is fulfilling its promises. If it can’t meet these calls, then it is not.
This might seem like a cop-out, but the philosophical distinction is actually extremely important as it suggests a kind of category error in the alternative analysis: if we acknowledge that the act of making loans is rife with economic uncertainty, dependent ultimately on the ability of the borrower to make a higher economic return than he pays in interest, then the idea of only making promises that can be fulfilled becomes obviously silly. Banking is a business. Businesses take risks. In particular, banking is in the asset management business, and it is certainly possible to manage assets poorly.
This is why focusing on equity is so elucidating. An equity wipe-out can be interpreted from the point of view of the horrified depositor, as above, or from the point of view of the equity-holder in the bank. This represents their business going bankrupt. It means they failed and have lost everything. They will do everything they can to avoid such an outcome.
In fact, we can invoke many more entirely realistic functions of banking to evidence this claim: in the event of an equity squeeze, the bank will seek short term funding (there’s liquidity and duration popping up as pertinent). Or it may seek to liquidate more long duration assets instead (ditto), or come up with ways to incentivise depositors to leave their money (extending the duration of liabilities in the bank’s favour).
The very nature of the equity cushion is that it is a function of leverage. In the cartoon example above, the bank actually had zero equity cushion, but in reality, responsible risk management would dictate lending out less than has been deposited precisely so that such an implosion can be avoided. We can even consider the concept of ‘leverage’ more philosophically than in simple accounting terms as something like ‘induced vulnerability to volatility’ to make this point yet more obvious.
So no, I would say that FRB is not fraudulent to the exact extent that the business of banking is run successfully. I think Selgin would actually go even further than me and follow this up with an argument for competing monies as well as competing banks. Given we understand the incentives to run banks well, why not let the deposit certificates themselves compete as money in the market, on the implicit basis of their issuing banks’ competitive strengths? Obviously, this presentation is exceedingly glib, but the reader can hopefully at least imagine the direction of the argument, and is encouraged to read Selgin!
So, no fraud, but what about money creation? Equity is again our guide. Remember that the bank’s liabilities are precisely the requirement to pay back depositors on demand. So, depositors do not have that money. As in the simpler example, they have, at best, a kind of psychological accounts receivable. Of the $100 Alice gave the bank, the bank has kept its equity cushion, say $10, and lent $90 to Bob. This is a little confusing, because the ‘money’ is now spread all over, but it is easy to trace it all back to its origin: Alice has no money, but $100 in accounts receivable, the bank has $10 money, $90 accounts receivable, and $100 accounts payable, and Bob has $90 money and $90 accounts payable. There is still $100 of aggregate money, because there is still $100 of net equity.
We can stress test this. If Alice withdraws less than $10 then not much interesting happens, but say she tries to withdraw $20. Either she does or she doesn’t (duh). Say she does: then the bank liquidates $10 of Bob’s loan (securitises it and sells a slice to another bank, or some such), at which point it has the following: $20 money, $80 accounts receivable, $100 accounts payable. Then it gives Alice her $20 and the accounts payable goes to $80 without much ado.
Or, it fails. It tries really hard to transform its illiquid assets into enough purely liquid capital to satisfy Alice, but Goldman Sachs screws it over. Now it is bankrupt and goes into full liquidation. Given it only owes anything to Alice, she inherits the assets, which are $10 money and $90 of Bob’s loan. Maybe she then liquidates the loan because Goldman has nothing against her. Maybe not. It doesn’t really matter. What matters is that, in the static scenario, without profit over time, and without raising equity finance, you simply cannot change net equity without assuming that either assets change, but not liabilities, or vice versa. This might mean something like, writing off Bob’s loan because his business failed, or writing off Alice’s account because she died with no will. Notably, neither scenario (nor any that would actually explain this) has anything to do with the bank.
And this is because … drumroll … the bank creates leverage. Everything the bank itself does has an equal effect on assets and non-equity liabilities (go back and check all the different examples above) and leverage is simply the resultant ratio of assets to equity. Assets and liabilities can go up, down, and all around, meaning leverage can go anywhere, but net equity doesn’t change. Which means aggregate money doesn’t change one bit, either. The bank does not create money.
It is worth noting in passing, and with reference again to all the examples, that this does not mean the bank doesn’t affect money. I was very careful just above to say ‘net’ equity and ‘aggregate’ money don’t change, because it is still entirely possible that the way people use money changes over short periods of time as a direct result of the bank’s actions, and hence it is perfectly fair to say that money changes.
For example, in the case where the bank goes under and Alice inherits the illiquid assets, a lot less consumer spending will happen over this period of time, until Alice is able to flog Bob’s loan to Goldman and get back to life as normal. This is essentially because of the lack of clarity as to who owns what assets, how liquid they are, how much leverage needs to be unwound, how safely that can be done, and in sum how long it will take to get everything back to the way everybody thought it already was. This is when we need to get lawyers involved because the situation has become pertinently legal and historical (yuk).
This ambiguity and resulting temporary change in behaviour could easily be described as a ‘contraction of the money supply’ or something similar, and be deemed the cause of ‘poor economic growth’. You could even call it ‘animal spirits’ if you wanted to, although I don’t think this helps much as an explanation.
While I have deliberately chosen a very different terminology, I think this is basically fair enough: even though the ‘money’ is still there in aggregate, there is temporary confusion over the true liquidity of the aggregate of assets, and so the net equity is not being treated as equity, which has knock-on effects. Interesting as this is, it doesn’t change that net equity — net money — is unaffected in the long run. The bank created leverage, and unwinding the leverage created confusion, but at no point was equity either created or destroyed.
Some Real Life Examples
Once you are willing to drop the econometrics and try to see things through this weird accounting lens, in which causation is a real thing and time has a direction, all sorts of puzzling monetary scenarios immediately become much clearer. Were Hong Kong’s coronavirus-battling-fiscal-handouts “helicopter money”? No, says Selgin at Alt-M once more.
They were just moving money around, not creating any, and the difference is important. What about the Fed’s recent open market operations? Is that QE? Is that ‘injecting liquidity’? No, says Joe Weisenthal:
This one is a bit more complicated. What the Fed has actually done is swap assets that are roughly as liquid (government bonds out, Fed balances in) but one of which has an interest rate set by the market that the Fed wants to manipulate for the greater good.
By increasing the demand for government bonds with these purchases, the price goes up, and the interest rate comes down (preventing a short-term rate spike was precisely the point) without the private sector’s liquidity, leverage, or duration profile changing much, if at all. Is this creating money? Actually, yes! The Fed balances didn’t used to exist. They were invented. Equity from nothing! Tada! But as Weisenthal went to pains to point out, whether you approve of this or not, understanding the mechanism is quite important and just screaming “injecting liquidity!” is, as Weisenthal says, illiterate nonsense. Or, God forbid, something like, “we owe more money than there even is!!!” The appropriately sarcastic response to this, best delivered in retweet-and-comment form, is: “Leverage. You’ve just invented leverage.”
Here’s a good one: what happened in 2008? If your answer has anything to do with housing, other than tangentially with reference to leveraged MBS exposure, you are in the illiterate nonsense camp, probably with a dose of politically motivated disingenuousness for good measure. As I have to patiently explain to my mum once every few years, no, it was not Bill Clinton’s fault. But neither was it a, vast right-wing conspiracy. Politics makes people so damned stupid. Anyway …
The housing market is a complete triviality in comparison to the credit transmission system, and more importantly, the effect on this system caused by banks and insurance companies unwinding leverage (largely on each other’s balance sheets) due to poor asset management (of securitised cash flows from loans collateralised with housing, yes, but look how long it took to get there!).
What the Fed did actually is more akin to injecting liquidity, although, again, the mechanism merits further inspection: the Fed bought swathes of failing credit instruments at well above their market value, on purpose, so as to try to prevent banks from needing to unwind their leverage at all, by making their profile of leverage, liquidity, and duration as good in fact as it had until recently been in their minds and their accounts. The goal was to minimise the likely enormous disruption to the credit transmission system and also likely disastrous knock-on effects in the so-called ‘real economy’. The Fed did this with created money, naturally.
You can object to this all you want — point to prior credit expansions, subsequent asset bubbles, questionable accounting of collateral, moral hazard, the Cantillon effect, and so on — but it is best to do so from a position of understanding what actually happened in the first place. Then you get to be extra sassy.
Even if you are on board with all the above in purely analytical terms, you might strenuously object to the idea that your savings can suddenly become ‘accounts receivable’ in Bob’s business. As if that’s not horrifying enough on its own, what if Bob’s business then fails and this loan is written off (as was alluded to as a possibility)? Your savings are gone! Surely you deserve something like deposit insurance? And surely that would be ‘money creation’? Equity springing forth from the void?
Yes, all correct, and I’m even on board with the moral angle, to a point, but I think there is a contingent assumption here that has gone unexamined. You don’t need to put your money in a bank. I slipped that in quite a ways up on the grounds that it is ‘more realistic’, and this, I believe, is the crux of the issue.
In the modern world, it is more or less required that you have your money in a bank account. You will find it difficult to get paid otherwise, for almost every salaried job, and you will find it impossible to pay for anything other than physical goods, in person. If you ever want to use a credit card you will need a bank to issue you the card in the first place, and you will need a bank account to pay the balance. And if you want to pay for things like utilities, or a mortgage, you will likewise have a very hard time.
The fact that, for all intents and purposes, everybody who isn’t a drug lord or a Bond villain does need to keep their money in a bank is the seedling of the rest of the parade of horribles. It means you are forced into being a customer of an asset management business against your will. The obvious dubiousness of this is then used, in turn, to justify literal money printing as a ‘lender of last resort’. And the means created for literal money printing are then used for all kinds of far less honourable or misguided reasons.
In addition, it is almost certainly logically true in a kind of practically unprovable, counterfactual sense that the forced lending distorts the interest rate and leads to a misallocation of capital relative to real consumer and saver preferences. In other words, it is probably true that much more leverage exists than anybody really wants, making the entire system more vulnerable to unpreventable volatility, and yet more likely to need to be bailed out with printed money. So, in a sense, banking is at the root of the illicit money creation machine, but, once again, it is valuable to be clear on exactly how.
I am in agreement with anti-FBR-ers if that’s where their argument ends. But not if they want to have it both ways. You can require banks to hold 100% reserves to satisfy your moral compunctions, but then you aren’t really talking about ‘banks’ any longer. You are talking about piggy banks. You can only complain about maturity transformation in the context of the lenders being unconsenting, but not in the context of asset management. Banking is asset management. Regular people should choose to opt in, for sure, but they can’t then also claim immunity from the consequences of poor management of their assets.
Fraud as Legal, Moral, and Economic
We will shortly arrive at Bitcoin, and the conclusion of this heretical ramble. But first, I would like to address a point originally raised by George, which readers may have had at the back of their mind despite my best efforts to ward it off with my pre-primetime edits. I can frame the issue with my semantic shenanigans all I like. But at the end of the day, “fraud” is a legal concept. You can infer whatever moral or economic consequences you like, but the law is the law. Either FRB is legal or it is not.
Sure, fraud is mostly a legal concept, but it is not an exclusively legal one. The law does not exist unto itself, as a model of natural language. It frames and is framed by economics and morality. I don’t think it is too much either of a stretch or a simplification to say that fraud is illegal because it is immoral, and that everybody behaving immorally all the time is really bad for business.
Forget banking; I frankly think it is silly to say that any business operation is an exclusively legal matter. The law sets the rules of the game, but one plays the game by morals and economics. As an investor, I assure you, I have never once thought about the legalities behind purchasing equity in public markets, but I think a great deal about the economics and morality of doing so. You might say that is because the legalities are so thorough, so slick, so well-understood, and so helpful, that I am able to deal only in abstractions as a result. I say, yes, absolutely, but that doesn’t mean everything I deal with is really a legal issue. “Economics” and “morality” are perfectly viable conceptual primitives.
In fact, at the extremes, economic and moral issues will simply overwhelm legal issues. You can concoct scenarios in which banks never go under, but they are purely academic. In real life, even in such a plausible scenario of moving to zero reserves, banks would need to raise equity at some point, and their investors would look at the economic issues (maybe the moral ones too if dire enough) before deciding whether to let this ‘infinite loan’ machine continue. If billions of dollars of capital evaporate when a bank goes bankrupt, the only legality anybody will ponder is whether or not they are going to get the par value of the stock certificate, or if they should mark it as zero and get on with their lives.
These might be beside the point for an academic or legal analysis, but FRBs don’t exist solely in academic papers and legal contracts. They exist in the real world and they need an economic and moral basis to continue existing, and to then be discussed in academic or legal terms at all. You can’t pass a law saying no banks shall go bankrupt and imagine yourself to have outlawed the need for morality or economics in banking.
Well, not without a central bank, at least, but that’s a whole new can of economic and moral worms. I should have said, you can’t pass a law saying blah blah blah with sound money and free banking.
Speaking of which, what about that Bitcoin all the kids are talking about?
Magic Internet Money
One of the many reasons Bitcoin is interesting as a monetary system is that its default state of possession and spending is not via a depository institution at all. The equivalent of hiding money under your mattress is entirely natural in Bitcoin, and so users do, in fact, have a choice as to whether to trust their money to an asset management business. Morally, at least, this seems certainly to be an improvement. We won’t get the bad part of FRB.
But what about the good part? Many on Twitter like to band around Lightning, and occasionally Liquid, as regimes that are somehow ‘backed by Bitcoin’. But, to once again borrow a phrase from Joe Weisenthal, this is illiterate nonsense. The proper conception of both Lightning and Liquid is that you have to sink Bitcoin as working capital to participate at all. It gets locked up in cryptographic escrow while the channel/peg is open. Very clearly there is no new equity. Perhaps less clearly, there isn’t even a new asset, so there is no new liability, and hence no new leverage, either. In other words, nothing about this has anything to do with banking. It’s just a funky workaround to patch the blockchain’s computational constraints. It’s understandable that people would come up with shitty analogies given the whole thing is a solution to a problem that has never before existed.
The real excitement for me, and probably quite some way off, is whenever we do get Bitcoin banking. An obvious reason to be excited is that credit extension is obviously socially useful, and a monetary system in which it never happens is an unsatisfying proposal. But then again, one in which it always happens is not so hot either, so perhaps Bitcoin can find a Goldilocks level of real consent and fairly priced interest?
But the more exotic reason is that open source and programmable money implies open source and programmable lending. Bitcoin has properties that no previous money has ever had and so I think it is reasonable to be excited by the potential for unlocking a long tail of previously infeasible finance, even if we have no idea what any of it will actually look like.
For example, access to depositor funds, or interest paid on them, could be programmatic, dynamic, and transparent — perhaps parameterised by some measure of the performance of the loan, or the lending institution’s overall reserve ratio. Different deposit classes could exist with arbitrarily fine-grained distinctions between all these inputs. Borrowers could pledge as collateral the income stream from other transparent loan products, or even a slot in a mining pool, or an open lightning channel, what with it routing transactions, collecting fees, and functioning as a form of productive but mildly illiquid capital. You are nuts if you don’t think this will eventually be securitised. At the risk of triggering the odd maximalist, they could utilise atomic swaps to pledge assets on other blockchains.
And of course, as Nic Carter has explained at length, and I suspect Selgin may be partial to as well, Bitcoin could hypothetically be an excellent base money on top of which currencies could be issued. We are back in money-1/money-2 terrain. Or, more tangibly, in the free banking era. Once again, Bitcoin has novel properties that arguably make it superior even to gold: it is low stock-to-flow, yes, but it is nearly free to verify, free and easy to self-custody, nearly free to send, and far, far more resistant to seizure. Bitcoin banks could have the never-before-seen feature of publically auditable reserves. The always quotable Elaine Ou has the canonical soundbite on this: “Financial institutions make people feel safe by hiding risk behind layers of complexity. Crypto brings risk front and center and brags about it on the internet.”
How to scale Bitcoin (without changing a thing)
Why Bitcoin banks need to prove their solvency
This seems to be enormously economically preferable. I’d add that the debate over exchange proof-of-reserves fits nicely into this framework too. Peter Todd put it brilliantly:
Which, I would bungle into the terminology developed here by saying that exchanges may well be operating fractional reserve in the entirely immoral sense that their customers think the exchanges hold their savings as zero duration deposits when in fact they only hold a receivable note that may turn out to be worthless. The exchange may not have done this at all, but without an audit we will never know. Besides, what are you even doing? Get a full node, already, will you! This also seems to be enormously morally preferable. Double whammy!
This might all sound very exciting, but all the tech bundles mentioned above are not really products so much as tools. Without meaning to cause any painful eye rolls, what is so exciting about Bitcoin being open source and programmable is that it can rightly be considered a platform (ugh, I know). A clunky, obtuse platform, sure, but a platform nonetheless (whose clunky obtuseness is a friggin brilliant thing, by the way — you don’t move fast and break things with hundreds of billions of dollars of bearer assets; you move slowly and break nothing).
Besides, its being a platform is the essence of the ‘unlocking the long tail’ described above — and wherever such things are honestly described, for that matter.
To take another tack, what is a stablecoin if not a dollar with zero default risk? An asset that is pure equity and nobody’s liability? That’s a weird framing, I admit, but luckily I just wrote a whole blog post setting it up so that you will grasp it right away.
This could easily segue in an another post entirely (I leave the door open for willing contributors) but I would argue that a far more important specifically short-term signal than Bitcoin adoption will be stablecoin adoption. Provided you can get your head around the tech (no easy ask, I admit) the financial tradeoff involved in purchasing and holding a stablecoin balance is practically risk arbitrage, particularly if you can self-custody. The stablecoin market took 3 years to grow to around $3bn and then three months to grow to around $9bn:
Now, I really don’t want to get dragged into arguments about the viability of the peg, whether Bitcoin pumping keeps it all afloat, etc., etc. — I am really only interested in the following very cool, very dangerous idea: an uncensorable, self-custodied, digital dollar. Bitcoin, but stable. A “stable-coin”, if you will. I don’t care how it works. Argue about that elsewhere.
And so, now that we are all experts on exponential growth, it shouldn’t be too hard to run the above growth numbers out a bit further and deduce that it might matter. And what does it mean? What would it mean? In a word, deleveraging. It’s a digital bank run. A digital pandemic bank run. Awesome sauce.
Further removed from the easily imagined, people will use the novel building blocks to create equally novel credit products that bring to market untapped supply and demand in ways we cannot predict. And on top of all of this, imagine if we had a financial system that didn’t collapse every time people panicked. That would be nice, right? Let’s ty to imagine that for a little longer …
What is Known and What is Not Known
We know there won’t be QE. There won’t be equity ex nihilo. But there will be leverage. Arguably, there must be leverage, and much of the leverage will be in forms never previously imagined or possible. Duration and liquidity will be given new life.
Like it or not, much of the leverage will be in the form of ‘fractional reserve banking’. The sooner Bitcoiners get on board with that, the better.
Thanks to George Selgin, Nic Carter, and Gemma Barkhuizen for edits and contributions. Just to stress again, Nic and Gemma are my knucklehead friends, whereas George did not need to do this at all. Thanks again, George.
Gemma doesn’t have Twitter but she and Sacha are working on a TikTok. Their plan seems to be getting increasingly involved without an end product in sight.