Ralph Musgrave made a blog post “Lawrence White Tries to Argue for Fractional Reserve Banking” where he criticises some of the arguments made by Lawrence White in testimony on his fractional-reserve banking. I tend to agree with many of the things Ralph wrote, but here I’ll concentrate one one aspect where I disagree.

Ralph writes:

“In fact there is no reason to suppose that “payment services” provided by a fractional reserve bank are any cheaper or “more economic” than those provided by a full reserve bank. That is, the ACTUAL COSTS (clearing cheques, issuing bank statements, etc) are the same in both cases. However, fractional reserve offers depositors interest on their deposits. And if you deduct that interest from the costs of payment services, then of course the cost of those services could be said to be reduced. But the reality is that this is just cross-subsidisation of one bank activity by another.” [emphasis original]

Ralph missed several things here.

Storage costs

The first one (which I cannot find explicitly in White’s testimony), is that holding less than 100% reserves decreases storage costs. Even if the “excess” reserves were used in a different manner than loaned out, FRB would still have lower operating costs. Storage costs are a subset of what I call “maintenance costs of money substitutes”. One might argue that if a withdrawal request arises in excess of reserves, the transaction costs of facilitating redemption (e.g. sale of assets) would be higher in an FRB. But the operational issues of redemption are not a feature specific to FRB. Even in a full reserve system, as long as branch operations are allowed, it is still possible that someone wants to withdraw more than the available reserves in that specific branch. The withdrawal would still have to be postponed and either the reserves transported from another branch, or some assets sold (the latter might be still chosen even in a full reserve banking if it has lower transaction costs, which is entirely possible).

Bank notes

The second thing is mentioned by White:

The other bank payment instrument, redeemable banknotes circulating in round denominations, simply cannot exist without fractional reserves. Banknotes are feasible for a fractional-reserve bank because the bank doesn’t need to assess storage fees to cover its costs. It can let the notes can circulate anonymously and at face value, unencumbered by fees, and cover its costs by interest income. An issuer of circulating 100% reserve notes would need to assess storage fees on someone, but would be unable to assess them on unknown note-holders. There are no known historical examples of circulating 100% reserve notes unemcumbered by storage fees.

Now, I have a minor addition here, it is hypothetically possible to create a bearer instrument with demurrage (e.g. stamps) even on a 100% reserve banking. Whether there is a practical merit in that I will leave open, but I’ll ignore it for the time being, in order to explain the argument of White. So, let me reformulate White’s argument: In a metallic monetary system, as long as people prefer bank notes to coins, FRB will emerge. This is a straightforward logical necessity. Irrespective of what people think about legitimacy of FRB, in this particular case it is an unavoidable consequence of consumer demand. Why might people prefer bank notes to coins? I’ll address that right away.

Why are substitutes substitutes?

Here it gets a bit tricky, because anti-FRB-ists, when criticising FRB, tend to “objectivise the boundaries of goods”. They argue that an instrument issued by the bank is treated by the users of that good as a substitute because they think it’s the same good (i.e. it is a claim, an ownership title). But this is a non-sequitur. The best refutation of this assumption are so called “complementary currencies”, in particular those of type “mutual credit”. Mutual credit are a form of circulating medium of exchange which is derived from a “normal” money (e.g. the USD or EUR), but they are not based on deposit banking. Some of the more popular examples are WIR and TEM.

In other words, a subsitute medium of exchange (money substitute) does not need to be a claim. Mutual credit is not even convertible into the base money. This leaves the question open: why are then some goods accepted as a substitute medium of exchange? The Austrians know this, but magically, when talking about FRB, they forget about it. They are accepted as substitutes because they decrease transaction costs. Practically all of the anti-FRB Austrians realise this, but only when not talking about FRB: Rothbard, Hoppe, Salerno, de Soto.

So what are money substitutes? Money substitutes are copies of the monetary base. They are persistently causally related to the original (e.g. by a peg, by using the same name, etc.), and they act as substitutes from economic point of view. And, at the latest since Kinsella’s Against Intellectual Property, the Austrians increasingly come to the realisation that copying is not per se a violation of property rights. Analogously, creating an instrument which is subsequently then accepted as a substitute by the market is not, per se, a violation of property rights either.


Now that we have clarified what are money substitutes, we can look at cross-subsidising. The instruments issued by the banks are copies. They have two important features that distinguish them from the originals:
  1. They have lower transaction costs
  2. They allow credit expansion
Here the error of anti-FRBists becomes apparent. They do not realise that the instruments are goods separate from the original. They draw the boundaries of goods based on their own desires, not based on how these goods are treated by the market. The result of banking is a new good, that unifies lower transaction costs with credit expansion. Even if they don’t like it, this is the economic fundament of the banks’ activities.
Since this new good, this copy, satisfies both the demand for lower transaction costs and credit expansion, these two activities economically manifest themselves as one. The merging of these two activities by the bank is simply a response to this unified demand. As long as this new good, copy, satisfies both demands, the activities of the bank are economically inseparable. Therefore, there is no cross-subsidising, similarly as the manufacturing of tires is not cross-subsidising the manufacturing of chassis as long as people demand the whole car, even if some people have a fetish for the tires and want to ban car assembly.

How to fix this?

Unlike the freebanking branch of the Austrian school, I actually agree with the gold standard branch that credit expansion [EDIT: and I mean any credit expansion] leads to distortions, such as the business cycle. So I’m at odds with both of them, as I argue that credit expansion [EDIT: I originally wrote FRB but that is inaccurate, it should be credit expansion] is economically detrimental, yet not a violation of property rights, but a consequence of market forces. I could just end here, leaving everyone baffled and annoyed. But this blog is called “Economics of Bitcoin”. And here it comes in.
To explain why, I’ll first start with a quote by none other than professor White, in Competitive Payments Systems and the Unit of Account:

“Coinage reduces transaction costs compared to simple exchange, because
of authentication and weighing. Bank liabilities also reduce transaction
costs. But these are empirical factors, and not something inherent in
all possible monetary systems
.” [emphasis added]

Now, putting it all together, if the transaction costs of monetary base are sufficiently low, money substitutes do not emerge, and thus there is no credit expansion. Bitcoin shows that such a system is empirically possible. It “fixes” credit expansion without “fixing” FRB. Is that a hack? No. Money substitutes are a hack, and credit expansion is a result of that. Bitcoin shows that a proper solution is possible.

8 thought on “Market forces and Fractional Reserve Banking”
  1. "Unlike the freebanking branch of the Austrian school, I actually agree with the gold standard branch that credit expansion leads to distortions, such as the business cycle."

    Good post.

    Don't the free banking Austrians admit that credit expansion is detrimental, insofar as credit is being lent at a rate below the so-called natural rate?

  2. Hello JP,

    thanks for the reaction. Indeed you are right, the gold branch and free banking branch disagree on the interpretation of Mises.

    The gold standard branch argues that according to Mises, any issue of fiduciary media is credit is disequilibrating, and if it is sufficiently extensive, results in a business cycle (I would put in particular Rothbard, Hoppe, de Soto and Block into this category). Some of them argue that in a fully free banking, the market forces will result in a nearly full reserve (Rothbard seems to take this approach). Some argue that freebanking will reach some sort of equilibrium state where the reserve ratio will be stable, and in this state the money supply will be less elastic and thus there won't be a cyclical behaviour (one example that comes to mind is Detlev Schlichter in his "Paper Money Collapse"). Some argue that in a free market, money and credit will be separated (I got this impression from Robert Murphy, for example).

    The freebankers (represented in particular Selgin and White) argue that the issue of fiduciary media on a free market with commodity monetary base is not disequilibrating, as it is a reaction to bank customer demand, and thus the money supply adjustments are warranted.

    I think that the freebankers' argument is invalid, and that the gold standardists are probably correct that any issue of fiduciary media is disequilibrating (while I also disagree with their prognoses mentioned above about "what would happen in a free market"). I want to write another blog post about this. I already wrote about it in comments on freebanking.org and uneasymoney.com, but neither Selgin nor Glasner seem to understand my argument so I want to put it together in a bit more elaborate manner.

  3. Hi Jesse,

    yes, I'd say it's the same thing, I just attempt to put it in a more abstract, formal way while you do it based on examples and descriptions of activities. That's not supposed to be a criticism, the descriptive method is better to get people to understand how it works (and might be more persuasive), while my more abstract method is better in making to understand why it works.

  4. Does "Transaction Costs" also include storage costs? While bitcoin transaction costs may be negligable, it's conceivable that users will entrust storage of their bitcoins to "banks" (e.g. to avoid the cost of security), thus opening the door to money substitution. Observe the popularity of online wallets.

    Perhaps the conclusion should be that it's vital for bitcoin client software to provide "low cost" storage solutions in order to prevent credit expansion taking hold.

  5. Hello Darren,

    sorry for the delay, your comment got stuck in the spam and it took me a while to figure out what happened.

    The storage, I agree, is not costless. But Bitcoin is what I call "form-invariant", it can exist in practically any imaginable form natively, and there is no necessity for a deposit-banking-style service to provide the best offering. By the way, the argument that service providers might be able to provide something extra over native Bitcoin, that will allow them to practice credit expansion is not new, the first serious version of the argument I know of was presented by Eli Gothill: http://www.webisteme.com/blog/?p=192 . Coincidentally, I'll probably meet Eli this weekend at the UNsystem Bitcoin Conference Preview, and I would like to discuss it with him.

    First of all, not all online wallets are modeled on the deposit system. There are at least two that are native Bitcoin rather than deposits: strongcoin and blockchain.info wallet. They use client-side encryption, so the server can't spend the Bitcoins on its own, it only provides infrastructure for the act of managing the wallet and creating transactions. Second, there are ways of storing it which have not even been possible with previous systems, for example brainwallets (storing Bitcoins in your brain) and multi-key signatures (e.g. "bank" has one key, you have the other one, and only together you can transfer the Bitcons).

    Some of the hypothetical features might not be present in a usable implementation at the moment, but I expect them to mature and provide a wide variety of options of storing Bitcoins, to serve the consumers in the most fitting way. While I can't completely exclude the possibility of a Bitcoin-substitute to be accepted as a medium of exchange (as it's an empirical issue), I think it is unlikely.

  6. Slightly off-topic, but since you seem to be familiar with the views of prominent Austrians on Bitcoin, do you think you could make a list? E.g.,

    Doug French: favorable
    Jeff Tucker: Tentatively favorable
    Hans Hermann Hoppe: Mostly clueless
    Guido Hulsmann: Clueless
    Bob Murphy: Skeptical, mostly mum
    Tom Woods: Skeptical, mosty mum
    Robert Wenzel: On the borderline


  7. I would prefer to leave it to the people to express themselves, I don't want to put words into their mouths. But if I have to express my personal impression, here it goes.

    In general, I agree with the list you made, just I don't recall Woods' remarks on the topic and I'm not familiar with Robert Wenzel.

    Jeff Tucker read my master's thesis and he appears to be very positive about Bitcoin. Stephan Kinsella is also positive, there's an interview with him in Bitcoin Magazine issue #5 from December 2012. Walter Block and Joseph Salerno on the other hand don't appear to know much about Bitcoin.

    George Selgin is somewhere in the middle. He gets some of its aspects very well, but he made some critical remarks. Detlev Schichter is skeptical, but he understands that there is a lot of merit to Bitcoin.

    My main impression is that the typical Austrian economist is stuck with gold and thinks that if something is to beat gold, it will take a long time, and by that time it will be obvious to anyone that the replacement took place. They don't see how you can predict the outcome of competition among media of exchange, other than to refer to the historical technological and regulatory factors. Therefore, I speculate, they don't see why they should spend time on learning about it.

Leave a Reply

Your email address will not be published. Required fields are marked *