In the aftermath of the demise of bitcoin bank Mt. Gox comes a few startling revelations. A newly released “Crisis Strategy Draft” confirms some unfortunate truths long suspected of its handling of clients’ bitcoin deposits. (The document might be a fake, but I’ll accept it as legitimate until proof surfaces otherwise.)
The first unfortunate truth is Mt. Gox’s admission that it is a fractional-reserve bank. Seeing this problem for what it is, the company is now looking to enter into agreements with its partners to “erase a significant portion of [its] debt.”
I previously relied on the ample literature of how fractional-reserve free banks are supposed to work in order to illustrate why it was clear Mt. Gox was a fractional-reserve bitcoin bank (here, here and here). Here we have another case where free banking theory is, unfortunately, demonstrated to be true in practice. The free banking literature says that any liquidity problem faced by a member bank can be solved by asking for lines of credit and concessions from other banks. As long as the illiquid bank is fundamentally solvent there should be some deal struck to extend it credit to get it through its time of trouble until it regains its footing. Mt. Gox is seeking the help of other “Bitcoin main players” to assist it in injecting new bitcoins onto its books.
This strategy of seeking the assistance of other banks may seem positive in that it extends the life of an otherwise troubled bank. There is ample evidence to demonstrate that it is exactly these types of agreements to support each other that ultimately develop into larger centralizing arrangements. In a similar application, Philipp Bagus and I have documented how the Federal Reserve System emerged from a series of small arrangements between private fractional-reserve banks during the Free Banking Era that culminated into an agreement to centralize reserves and coordinating policy amongst the individual banks, first at the hands of the clearinghouse system, and later more formally through the creation of the Fed (available here as a pdf).
This bodes poorly for a bitcoin community which prides itself on the fact that it does not need government oversight or aid in solving its problems. There is a more troubling revelation in the Mt. Gox crisis strategy document however.
On its current balance sheet the company lists its assets as consisting of 2,000 bitcoins (plus $32 mn. in fiat), and offsetting liabilities of 750,000 bitcoins (and another $55 mn of fiat claims against it). There is a clearly a big hole to fill. More troubling is that Mt. Gox notes that this theft of its bitcoins took place over a five-year period. Furthermore the company now confirms that the loss is due to the “transaction malleability” issue with the bitcoin protocol (which I discussed here as a reason Mt. Gox held only fractional reserves).
In other words, over a five-year period the bitcoin bank went from a (presumably) 100% reserve ratio to holding less than 3% reserves… and no one noticed!
Here is the critical fault with fractional-reserve banking that rarely gets discussed. When someone deposits a good it is not because he does not want to use it. Nor does that good represent some idle resource until it is asked for. People who support fractional-reserve banking of both the centralized and “free” varieties are both of the opinion that deposits are idle cash and no one is harmed when a bank puts them to good use.
It is true that customers who make a deposit are not making physical use of their good as long as it is deposited, but that does not mean that the good is not being used. Deposits are made when 1) a client wants the good kept safe for when he does want to use it, or 2) when the client is unsure of when in the future he wants to make use of the good. In other words, the deposit is being used even when it is supposedly “idle” in an account – it is a security hedge for the depositor for that moment when he does want to use it.
It is admittedly true that the depositor will not miss his deposit much until he asks for it. But this does not mean that the bank is free to make use of it. After all, the depositor either wanted the good kept safe or he was unsure of when he would need it back. Since the depositor doesn’t even know when he will ask for its safe return what makes the bank so sure it can?
For five years Mt. Gox seemingly functioned with an ever lower reserve ratio and no one noticed. An epiphany happened when the reserve got so low that the company was unable to make good on the withdrawal requests of its clients. It tried to hide this outcome through a series of ever more desperate policies, all used previously by other liquidity-constrained fractional-reserve banks (and which I documented here and here).
Previously I noted that there is an important lesson to take from all this. It is one thing to have a currency that is not centrally controlled, and bitcoin ably shows how the market can create a money without the “helping hand” of government. The banking system matters too though. Having a fractional-reserve banking system leaves depositors open to problems even if the underlying money deposited is sound. Those empty-handed depositors of Mt. Gox are learning this painful lesson now.
There is one more lesson we can all take from this: fractional-reserve banks function until they don’t. For many years Mt. Gox survived because no one doubted its solvency. Then in a short period of time disaster struck and the bank was exposed for what it was. This pattern of fractional-reserve bank surviving for a time while they sow the seeds of their own demise has happened many times throughout history. These episodes have always ended with tears shed over insolvent banks, or government help being called for by both bankers and depositors interested in saving their financial interests. Let’s hope the demise of Mt. Gox is taken as a warning by the bitcoin community and that this same fate doesn’t befall other bitcoin banks.
(Originally posted at Mises Canada.)