I recently had a fascinating chat with the economist Peter Šurda to discuss how nonpolitical cryptocurrencies like bitcoin could alter the future of fractional reserve banking

Peter is also a software developer experienced in the online payments industry and will present at the Bitcoin 2013: The Future of Payments conference in San Jose in May. His 2012 master's thesis at Vienna University of Economics and Business was entitled Economics of Bitcoin: Is Bitcoin an Alternative to Fiat Currencies and Gold? He's an abstract thinker, but the implications of his work are tantalizing: that digital money like Bitcoin opens up possibilities for banking without central planners or a lender of last resort, where interest rates and reserve requirements are driven purely by the market.

The debate between the full reserve bankers and the fractional reserve bankers is an old one and it has been explored in depth by the Austrian school of economics. More recently, the debate has been broadened to include the dynamics of introducing the bitcoin cryptocurrency, which is the functional equivalent of digital gold, since its supply is predictable and fixed. (There are currently 10.9 million bitcoins in circulation with a total fixed supply of 21 million expected to be mined before 2140, 99% of them by the year 2032.) The Austrian school economist Michael Suede and the technologist Eli Gothill have speculated that fractional reserve banking can indeed appear within a bitcoin monetary environment. This is where we join up with Peter.

JON MATONIS: I enjoyed your blog post, "Market Forces and Fractional Reserve Banking." Do you consider fractional reserve banking to be compatible with Austrian economics?

PETER ŠURDA: First of all, I would like to separate fractional reserve banking and credit expansion. On one hand, there are ways of increasing the money supply, in the broader sense, which do not require fractional reserve banking or changes in the monetary base such as a system based on the principle of mutual credit like LETS [local exchange trading systems], or a fiat currency that uses bitcoin as reserves (i.e. they are not claims in the sense that Ludwig von Mises uses them, but they act as full substitutes). From the opposite direction, fractional reserve banking does not necessarily lead to credit expansion.

I agree with the full reservists that credit expansion has the effects described by the Austrian Business Cycle Theory. However, I agree with the free bankers that fractional reserve banking is not necessarily a violation of property rights and other ways of increasing the money supply also are not necessarily a violation of property rights.

So I think that the economic and legal analysis are two separate issues and need to be addressed separately. I avoided the legal analysis in my thesis and concentrated on Austrian Business Cycle Theory and policy issues, but in an earlier draft I have several pages about legal aspects too, and I discussed the topic with [the legal theorist] Stephan Kinsella.

JON MATONIS: How does a nonpolitical cryptocurrency like bitcoin alter the landscape in the "full reserve" versus "fractional reserve" banking debate?

PETER ŠURDA: Austrians have made arguments in the past that lead to the conclusion that fractional reserve banking does not necessarily lead to credit expansion, even though they never explicitly formulated it this way and might not have realized the connection. The reason is that if credit instruments do not decrease transaction costs over the monetary base, they are unlikely to act as a part of the money supply. Bitcoin shows that this is not only a hypothetical but empirically possible to implement. With Bitcoin, it is much less likely that credit expansion will occur.

In other words, we need to separate two things. Why do people want to hold fractional reserve banking instruments, which may include the interest payments as one of the reasons, and why do people want to use fractional reserve banking instruments as a medium of exchange which, I argue, requires that the fractional reserve banking instruments decrease transaction costs. That they historically manifested themselves through a common instrument is an empirical quirk and not an economic rule. The ability to loan money is beneficial. Contrary to many Austrians, I agree that maturity transformation can be beneficial, and if the loan ends up being a liquid instrument, it also can be beneficial. But if it is so liquid that it becomes a part of the money supply, that's when it has a detrimental effect on the economy.

For full reservists, Bitcoin shows that the question of fractional reserve banking is less important than they thought. Fractional reservists, on the other hand, need to think about the nature of the mechanisms equilibrating the money supply. I tried to explain the issue to [the economists] George Selgin and David Glasner in comments on their websites, but I wasn't successful in getting my point through.

JON MATONIS: If bitcoin is digital gold, does that portend a future where a bitcoin standard (akin to the gold standard) can emerge or partial bitcoin backing for other currencies?

PETER ŠURDA: They probably can emerge, but the more important question is whether they would be preferred to bitcoin. Only something that provides a significant improvement would be preferred. I only know two potential candidates for that: Ripple and OpenTransactions.

JON MATONIS: In a bitcoin world, is fractional reserve banking only possible with offline substitutes (such as physical coins or cards, which can be traded hand-to-hand, containing the private key to a bitcoin address) or an intentional "fork" in the block chain ledger?

PETER ŠURDA: Hypothetically, the reserves can be offline and the substitute can be a clearing system like Ripple, so there are other possibilities too. But if I understand your point correctly, offline "substitutes" might have a higher chance of actually becoming full substitutes because they might have more obvious advantages.

JON MATONIS: As the recent block chain fork episode demonstrates, there is a need for offline bitcoin transactions to continue. Is this demand sufficient for a money substitute to evolve, such as offline substitutes with full or partial bitcoin backing?

PETER ŠURDA: This is primarily an empirical question, so we can't be completely sure about that. I think the probability for this is significantly lower than with the currencies that we've known historically. The end result is also path-dependent; for instance, it depends on how quickly bitcoin matures and/or adapts to changes compared to the potential substitute.

Fractional reserve banking does not come into existence magically. It must follow economic rules. With gold and similar commodities, fractional reserve banking comes into existence for these reasons: On the demand side, there is a demand for money substitutes, because they provide something that money proper does not; and on the supply side, money substitutes carry maintenance costs for the issuer (e.g. storage of gold) and these need to be offset somehow. The issuer can charge on holding (e.g. demurrage of bank notes), transacting (e.g. check clearing), or, obviously, externalize the costs through fractional reserves. From the point of view of an individual user, fractional reserve banking appears to be the least costly alternative. So obviously fractional reserve banking wins.

Putting it together: If there is a general demand for money substitutes, this leads to fractional reserve banking. Unless it's illegal. Then it might not. Solution: Have money which does not lead to the creation of money substitutes. Bitcoin shows that at least hypothetically, this is possible. I might even go a bit further and make this statement: If on a free market money substitutes do not develop even though there is no legal or technical obstacle for them, it means that the choice of money is Pareto-optimal since no change in the monetary system leads to an increase in utility.

JON MATONIS: Does a demand for positive return on bitcoin balances lead to an environment of competitive bank lending with risk-adjusted interest rates? And will this lead to an environment of fractional reserve banking with depositors offered higher interest rates in exchange for the additional risk premium of running a fractional portfolio?

PETER ŠURDA: Yes, I would say it does, but until there are industry niches that primarily use bitcoin, it is probably not much different from gambling.

This might lead to negotiable credit instruments with maturity-mismatching or maturity transformation, depending on which economic school you use for terminology. However, I don't think this feature alone is sufficient for these instruments to be accepted as full substitutes whereas George Selgin appears to think it is. Now, whether to call such a situation "fractional reserve banking" even though no credit expansion occurs is unclear. I lean towards yes, but there could be other interpretations.

JON MATONIS: How do you see bitcoin changing interest rate structures and lending practices?

PETER ŠURDA: Using Bitcoin for loans only makes sense for those businesses that use bitcoin as a unit of account, unless, of course, you're just speculating on the market but don't actually sell any goods or services. I think this will only occur at much higher levels of liquidity or until we can be quite sure that it deserves the label "money." Until these higher levels of liquidity are reached, the price of bitcoin will probably be quite volatile, which reduces the likelihood that people use it as a unit of account.

However, there could be niche market segments that use bitcoin as a primary medium of exchange and [bitcoin] mining is the most obvious candidate. For these, the unit of account function would make sense even if the global market penetration is lower.

Assuming one of these thresholds is crossed and the money supply remains inelastic (i.e. no significant credit expansion), the interest rate of bitcoin should be a good reflection of the time preference of those market participants that use it as a unit of account. Bitcoin also makes it much easier for lending to occur in a decentralized manner, I think. Rather than a small number of "too big to fail" institutions, we should see smaller specialized teams that act as facilitators without owning the liabilities or being liable themselves.

JON MATONIS: Can a free market fractional reserve system (as opposed to a central banking fractional reserve system) coexist with full reserve banking? Or will one drive out the other? 

PETER ŠURDA: I think that if money substitutes emerge, fractional reserve banking will out-compete 100% reserve banking in the market. I deal with this a bit in an earlier draft of the thesis. If they don't emerge, on the other hand, we'll have a money supply equivalent to the monetary base and debt will not cause changes in the money supply. It would be viewed as merely highly liquid credit. I don't think they can coexist for a long time assuming the same underlying money in the narrower sense, of course.

Jon Matonis is an e-money researcher and crypto economist focused on expanding the circulation of nonpolitical digital currencies. His career has included senior posts at Sumitomo Bank, Visa, VeriSign, and Hushmail. Currently, he serves on the board of the Bitcoin Foundation. Follow him on Twitter @jonmatonis.