To answer the question of whether fractional-reserve banking is compatible with Bitcoin it may be helpful to first inquire as to whether fractional-reserve banking is compatible with the free market.
Now, I should mention I’m more than aware of the controversial nature of this topic. My intent here isn’t to trigger another round of bitter infighting, but rather to simply present my opinions and hopefully engage in constructive dialogue.
Here I’m not going to concern myself with the macroeconomic impact of fractional-reserve banking or 100% reserve banking (although I do have a good amount to say about it), but rather I’d simply like to discuss which of the two systems is likely to emerge from a genuine free market. Before I get labeled as being part of a “moronic cult” or as “clownish”, let me say that I’ve changed my mind on this issue probably a half dozen times before arriving at my current views. I’m not wedded to either position. As compelling new evidence is presented, I change my views. In this context, I think Bitcoin can be viewed a grand economic experiment that can help us better understand money in a free market.
The Fraud Question
The issue of fraud in fractional-reserve banking can be broken down into two parts. I think it’s the failure to distinguish between these two that is a source of much disagreement. First is the question of whether fractional-reserve banking was historically conducted in a fraudulent manner. The second is whether fractional-reserve banking could be conducted in a non-fraudulent manner going forward.
The answer to the first question is empirical. I’ll admit I don’t really have an answer. I think it’s likely that fraud was committed in some instances and not in others. Admittedly, I do think there may have been some fraudulent activity in the case of the London goldsmiths. Wealthy merchants stored their gold with the King at the Royal Mint. During the English revolution Charles I raided the Royal Mint causing a loss of confidence in the safety of deposits. The merchants started looking around for a place to put their gold and ultimately settled on the goldsmiths. Now we may ask, why goldsmiths? I think the answer is pretty obvious — they had experience storing gold. If the merchants had been truly desirous of investing their gold rather than storing it, wouldn’t it have made sense to bring it to any one of the financial intermediaries active at the time rather than to a goldsmith? The fact that the goldsmiths started lending out the gold deposited with them seems, to me at least, to be prima facie evidence that fraud may have taken place. George Selgin disputes this claim in his article Those Dishonest Goldsmiths and provides some circumstantial evidence of contractual fractional-reserve banking. In my opinion the evidence is inconclusive. Again, the most likely scenario is that fraud took place in some instances and not in others.
History aside, what we really want to know is could fractional-reserve banking be practiced legitimately going forward. This is where I part with the 100% reserve crowd. The answer is obviously yes. In the common law a contract takes place where there is a “meeting of the minds”. If a depositor agrees to lend his money to a bank on the condition that the bank keep a percentage in reserve to satisfy withdrawal requests made on demand, then certainly this would be a legitimate contract. Not only do I see this as legitimate, but I’m positive this form of investment banking would be widespread given that it allows depositors to keep funds in highly liquid accounts while simultaneously earning interest. Neither am I pessimistic of the ability of a bank to effectively manage its reserves and avoid bank runs. Bank runs are typically a byproduct of government intervention instilling moral hazard in the banking system, causing the business cycle etc. In a stable economic environment, I would expect bank runs to be few and far between. To the extent that they do occur, a simple suspension clause in the contract could help the bank avoid total collapse.
So while I see the benefits of fractional-reserve investment banking, I do not believe it would translate into a fractional-reserve monetary system in a genuine free market. Why do I say this? It may be helpful to remind ourselves on how money originates in the marketplace. Here’s Carl Menger in The Origin of Money:
What do you do if you are unable to obtain what you want through a direct barter spot transaction? It makes sense for you to obtain goods with a high degree of saleableness, and then exchange these in the wider community at present or in the future. By this process, the most saleable good (or goods) becomes the medium of exchange (Menger 1892: 249).
Essentially what he is saying is that whichever good (or goods) is the most liquid will establish itself as the medium of exchange in the free market. Now, let’s remember what fractional-reserve banknotes are after all — risk-bearing financial instruments, historically denominated in gold or silver. Now ask yourself this, how can financial instruments, bearing a credit risk and a liquidity risk, become more liquid than the very commodity they’re denominated in? That’s almost like suggesting that Treasury Bills (a very liquid asset) could become more liquid than dollars. This seems logically false on the surface.
That’s not the only problem, however. There’s no reason for us to suspect that fractional-reserve banknotes issued by different banks would be homogenous, typically a requirement for money. Each bank is located in a different geographical region, is staffed by individuals of varying talents, and has differing risk exposures. How could financial instruments issued by such institutions trade on par with each other? Again, it seems very unlikely.
So given these considerations it seems likely to me that while fractional-reserve banking may play a critical role in finance, the monetary unit is likely to remain either the commodity itself or a 100% backed substitute. Here’s where I’ll start to tie Bitcoin into this. Surely there could be fractional-reserve Bitcoin investments such as savings accounts, but imagine what it would take to have a fractional-reserve Bitcoin monetary system. People would have to start issuing risk-bearing Bitcoin financial instruments and the general public, including merchants, would have to decide that they would rather accept these (most likely heterogenous) risk-bearing instruments in place of Bitcoin proper. Does that seem even remotely plausible? Now there are Bitcoin financial instruments out there (see Bitcoin Trading Co.), but none are getting a even a whiff of being a medium of exchange. Now one could object that Bitcoin is still in its infancy, just give it time. But again, no amount of time is going to make a risk-bearing asset supplant the very commodity it’s denominated in as a medium of exchange.
Now, I’m not the first person to make the above claim, yet despite volumes of literature devoted to this subject, very few attempts have been made to refute it. The most common objection, and one I’ve encountered in the few times I’ve mentioned this to professional economists, is something like this: “Obviously, you’re wrong. A fractional-reserve monetary system did evolve on the free market. There’s no point in even addressing your claims.”
I’m sorry, but that is not an argument. Using a similar line of reasoning one could easily say: “Obviously you’re wrong when you say that lasting housing shortages cannot occur on the free market, just look at New York City! There’s no point in even addressing your claims.”
There are one of two possibilities here:
1) A flaw in my reasoning is responsible for the discrepancy between what my theory predicts and what we actually experience.
2) Some sort of (most likely government generated) distortion in the market is responsible for the discrepancy.
Saying it must be (1) without expounding on why my reasoning is wrong, while ignoring possibility (2), is not a serious argument.
Now I want to turn to the few genuine attempts to rectify this discrepancy. The first comes from Professor Larry White in his article Accounting for Fractional-Reserve Banknotes and Deposits—or, What’s Twenty Quid to the Bloody Midland Bank? published in the Independent Review. White’s argument essentially comes down to this: while accepting that 100% reserve checking deposits can allow for monetary exchange, he suggests that there is no convenient way to charge storage fees on 100% reserve banknotes. Since the banknotes are designed to circulate, there is no way of identifying the person currently in possession of a note to bill them for storage. The bank will have no choice but to subtract the fee when the note is redeemed. This will result in each note in circulation having a different value relative to par, which would be unworkable for a medium of exchange . Thus fractional-reserve banking evolved out of convenience.
Before I address this argument, first let me say that Bitcoin renders it null and void. Since bitcoins are traded electronically, banknotes (to the extent anyone would bother to produce them) would be little more than a novelty. Still, I find this argument particularly odd. Even if we accept the premise that there is no convenient way to issue 100% reserve banknotes, it doesn’t follow that fractional-reserve banknotes should become more liquid than commodity money or 100% reserve deposits. We could have a banknote-less monetary system could we not? Additionally, White’s premise is almost certainly false as well. This kind of argument reminds me of the “public goods” arguments made by many academic economists. Because they, from their ivory towers, can’t conceive of a way to earn a profit on allegedly non-rivalrous, non-excludable goods then neither can an experienced profit-seeking entrepreneur (thank God broadcast television was invented before the public goods theory!). Similarly, just because White can’t conceive of a way to conveniently provide 100% reserve banknotes, doesn’t mean nobody else can. After reading his article I jotted down a list of possible ways to conveniently issue such notes:
1) The bank may offer “free storage” or “free banknotes” as a means of drawing people to its other financial products. Goldmoney.com, for example, charges 0.12% per annum to store gold. That’s practically negligible. It’s not out of the realm of possibility that the bank may just subsidize that 0.12% (certainly for notes) in order to sell financial products earning 5%, 7%, 10% interest or more.
In fact, for those familiar with the Bitcoin industry, this is the exact business model of Coinbase. Coinbase offers free bitcoin storage to its customers as a means means of incentivizing them to buy its other products — namely merchant services and bitcoins sold at a small markup over spot. And let’s not pretend storing large quantities of bitcoin is easy. It’s arguably more difficult to keep bitcoins safe from loss and theft than gold. Coinbase operates various computer hardware/software and pays to store the bitcoins in bank vaults, yet its customers pay nothing for this service.
2) The bank could offer “free note withdrawal” while making up for it by charging a larger storage fee on deposits. For example, if we assume a deposit-to-note ratio of 3 to 1, the bank could simply charge .16% per annum for storage on deposits instead of .12%. For someone like me who keeps an average of about $1,000 in my checking, that would amount to a whopping 3¢ more per month.
3) Who’s to say that banknotes can’t depreciate? Consider, merchants who accept credit cards payments must pay fees to the issuing bank. That means they receive varying amounts for their products depending on what type of payment the customer uses. Some fees can be as high as 4% of the transactions. Cash, of course, has zero fees. Yet, with the exception of a few gas stations here and there, nearly all merchants charge the same price no matter what type of payment you use. Why would we automatically assume that they wouldn’t accept a banknote at a discount to par, especially when the discount is likely to be far less than credit card fees? Most notes wear out too quickly to circulate much longer than a few years. If a banknote circulates for three years, it would be worth 99.64% of par. A merchant wouldn’t accept a note with that value relative to par, but he would take an American Express card at 96% of par? Again, that doesn’t seem likely.
I can keep going, I’ve got another four options on my list, but I think I’ve made my point. Larry White’s argument suffers from an extreme lack of creativity.
The only other real argument I’ve heard came from Jonathan M.F. Catalán in a small debate I had with him on this topic on his blog. If I remember right, his argument was something like: “The storage costs may be so prohibitive that people may prefer the fractional-reserve financial instruments over 100% gold deposits/notes”. I’ll just list off my thoughts on this argument:
(1) If storage costs are that prohibitive, one has to wonder how gold became money in the first place. (2) Let’s not forget that banknotes are not without storage costs either (3) The storage fees I quoted from goldmoney.com are so low that this argument seems implausible. (4) If the bank offers free storage as I postulated above, this argument will not hold. And (5) Bitcoin once again renders this argument moot since storage fees are zero.
Needless to say, I don’t feel either of these arguments sufficiently overcomes the absurdity of financial instruments becoming more liquid than their underlying commodities.
I can’t conclude this post without at least giving an account of why I think there’s the discrepancy between what theory would predict and our actual experience. Let me start with a story about myself that might serve as anecdotal evidence. When I opened my first checking account as a teenager, I naturally assumed the money I deposited in that account was simply kept in a vault somewhere. It never occurred to me that the bank was loaning out my money. Why did I make that assumption? Well, the checking account didn’t pay interest. I knew that savings accounts paid interest, so the bank must be loaning out those funds. But it never occurred to me that they were doing the same with my account that didn’t receive interest. It wasn’t until a few years later when I first started to become interested in economics that I found out the truth while reading on my own time. When I first read banks loan funds from checking accounts, I just assumed the author didn’t know what he was talking about. But I checked other sources, and sure enough, it was right. I couldn’t believe it. Now what percentage of the public thinks like I did? I’m tempted to think it’s rather high. And I can virtually guarantee they aren’t taking the time to educate themselves on money and banking either.
The reason I tell this story is because there is an assumption among free bankers that “everyone knows” that the banks loan out the money that is deposited in demand accounts. Given that most of the public is financially illiterate, I highly doubt that is a valid assumption.
Take a look at the following fractional-reserve notes:
Is there anything on them to indicate that they are, in reality, financial instruments? Is there any statement regarding credit risk? Any statement regarding liquidity risk? Does it say anywhere on them they constitute a loan to a bank (an interest free(!) loan nonetheless)? Now a free banker would likely say: “They didn’t need to explicitly state it on the notes, it was common knowledge.” If it really was common knowledge, then that is correct, it wouldn’t need to explicitly stated on the notes. However, if it isn’t common knowledge, a standard interpretation of contract law would require that information be disclosed.
Suppose I’m selling you are car. It’s perfectly legal to sell a car that is broken and does not drive. However, given our customs, the “default” position is that it is assumed that the car is in normal working order unless otherwise stated. If I want to sell you a broken down car I must disclose that fact to you, else I mislead you into making a purchase you otherwise would not have made.
Before I started studying economics, if you showed me those notes and told me they were from the “gold standard” era, I would have assumed there were 100% backed. I would have never known they were really risk-bearing financial instruments. My guess is that there is a large portion of the financially illiterate public, both back then and now, who would do the same. So, if I’m right that knowledge of fractional-reserve banking isn’t common knowledge among the public, doesn’t that imply that a little disclosure is in order?
If any percentage of the public accepts those notes in trade under the impression that they are backed 100% by gold when they otherwise would not have, the notes will be more liquid than they otherwise would have been. It’s not inconceivable to me that the much of the liquidity of the fractional-reserve notes came from the general public accepting them without knowing any better.
Not to mention the banks have zero incentive to actually disclose this information to their customers without a legal requirement to do so. The more their notes resemble actual warehouse receipts, the more liquid they will likely become given the lack of financial literacy among the public. Notice how the only thing written on the notes is, “Payable to the bearer on demand”. Not necessarily untruthful, but not exactly full disclosure either.
So what happens if a significant percentage of the public is driving up the liquidity of fractional-reserve banknotes under the mistaken belief that they are fully backed? Well, Gresham’s Law would kick in. That is, the “bad” money would drive the “good” money out of the marketplace. And of course once governments got involved in the in the 18th and 19th centuries this only served to institutionalize the practices that originated in the early days of banking.
So to wrap it up, the market case for a fractional-reserve monetary system is not nearly as compelling as it is made out to be. I am virtually 100% certain that we will never see Bitcoin transform into a fractional-reserve monetary system given that it is nearly impossible to disguise bitcoin financial instruments as bitcoin itself. Had that been the case with gold and silver, the banking system would likely have evolved on a completely different path than it did.