Pur Autre Vie

I'm not wrong, I'm just an asshole

Sunday, April 07, 2013

A More Viable Bitcoin?

In my previous post, I expressed my doubts as to the long-term viability of the bitcoin.  But now I want to propose a possible alternative and consider whether it could thrive.

Here is what I have in mind.  I start a company called Gresham Inc. (a reference to Gresham's Law).  Gresham Inc. issues a product that is like a bitcoin in its technical attributes, but that is issued and traded differently.  We'll call it the "talent," which is an ancient unit of mass also used as a measure of value (by reference to a talent of precious metal).

Gresham sells talents for $10 each, initially.  It also buys them for $10 each (or perhaps for slightly less).  So to conduct an anonymous transaction, an individual could buy 15 talents for $150 and send them to the transferee.  The transferee can then exchange them for slightly less than $150.  The difference between what the buyer pays and what the seller receives is a transaction fee.

Gresham uses the proceeds from selling talents to buy U.S. treasury securities (or other dollar-denominated assets).  It then profits both from the transaction fees and from its income from the assets that it owns.  Of course, Gresham must also lay out cash to maintain its secure network for transacting in talents, so its profit is the difference between its revenues and its costs.

Gresham could increase the price of talents over time.  In this way, the holder of talents owns an appreciating asset, and so may be willing to hold it for a period of time.  However, Gresham would keep the rate of appreciation low enough that Gresham profits from the difference between the rate of interest it obtains from the assets it buys and the implied rate of interest from holding talents.  So for instance, if Gresham is receiving 2% on the dollar-denominated assets it buys, it might increase the value of a talent by 1% per year.  This would be competitive with a bank account in today's economy.

Gresham would function very much like a bank or a money market fund, except that the "securities" it issues would be useful for anonymous transactions.  Gresham would need to maintain significant liquid cash reserves, because it might have to redeem a large number of bitcoins at once—again, just like a bank or money market mutual fund.

It seems to me that talents would be much preferable to bitcoins.  In part, this is because much of the "seignorage" of bitcoins is wasted in useless computations.  Gresham would capture this surplus, share some of it with the holders of talents, and thereby profit.

Of course, Gresham would probably be taken down by the legal authorities.  But then, I expect the bitcoin also runs a pretty substantial risk of being shut down, unless it simply implodes before the authorities become too alarmed by it.

Skepticism About Bitcoins

If you don't know what bitcoins are, the Wikipedia page isn't a bad introduction.  I confess I don't entirely understand the mechanics of bitcoins, so take my description with a grain of salt.

Imagine that you can generate unique serial numbers (this is a pretty easy task, I believe people have been doing it for centuries).  And imagine that each serial number can be encrypted and stored digitally in the following manner:  if Person A wants to send a serial number to Person B, he submits his serial number to a network that modifies the encryption in some way, de-activating the old encryption and generating a new set of encrypted data, which is delivered to Person B.  In this way, Person A can surrender a serial number to Person B in a way that is, in theory, untraceable.  (I suspect there is something like an "analog hole" that actually prevents these transactions from being as anonymous as people would like, as when the U.S. monitored the SWIFT system, but I am way out of my depth in these technical areas.)

So a bitcoin is just a unique serial number that is encrypted as described above.  Bitcoins are generated digitally at a fairly slow rate, and as I am writing there are apparently about 11 million of them in existence.  They are traded online.  As the Cyprus banking system melted down, the value of the bitcoin apparently surged briefly to $147/bitcoin, implying a total value of around $1.6 billion.

What are we to make of this?  Is this the end of the liberal state?  Will the dollar, vitiated by years of recklessly loose monetary policy, give way to the technologically and morally superior bitcoin?

I think not.  In fact, I think at some point the value of the bitcoin will go to zero and stay there forever.  I think this because I think the bitcoin is at best (A) an anonymous payment system combined with (B) a useless commodity, and that on both counts it will be out-competed (in the latter case, it will be outcompeted by, among other things, useful commodities).

Let's start with the bitcoin's function as a commodity.  The idea here is that it can be a good store of value, and the implication of the post-Cyprus price spike is that it might compete with the banking systems of tax havens.  So for instance, a wealthy individual might buy bitcoins in lieu of entrusting his money to a bank.

I can see at least two problems with this.  First, the price of the bitcoin fluctuates wildly.  Apparently the bitcoin traded as high as $147 and then fell to $117 as the Cyprus crisis unfolded.  Now, imagine that you decided to store $10 million of your money in bitcoins and bought at $140.  If the price falls to $120, you have just lost about 14% of your money, or $1.4 million.  And of course, there is the potential to lose much more.  By comparison, taking principal losses on bank accounts is quite rare (which is precisely why Cyprus was such a shock).

Moreover, bank accounts typically pay interest.  So they provide income, while (usually) posing minimal risk of principal loss.  Bitcoins may provide "income" if they appreciate rapidly enough, but for bitcoins to compete with bank accounts two things would be required:  the price of bitcoins would have to increase indefinitely, and the rate of increase would have to be high enough to compensate for the risk of losing a double- or triple-digit percentage of principal.

Suffice it to say I don't think bitcoins will meet this requirement, unless people are willing to issue bitcoin-denominated securities or pay interest on bitcoin-denominated loans.  If that happens, then one could earn a return on bitcoins in the same way that one earns a return on dollars.  But it seems very unlikely to me that a bitcoin-denominated capital market will ever spring up.  Note that there are already plenty of commodities that act as stores of value, including gold.  And you can even buy gold-denominated financial assets.  Gold also has industrial uses, meaning that its value can't go to zero, unlike the value of the bitcoin.

So what about bitcoins as a technology to facilitate anonymous transactions?  Well, here the problem is that, as far as I can tell, one can easily produce an equivalent technology that doesn't involve the price fluctuations inherent in the bitcoin market.  A system could take many forms—a similar "currency," an encrypted portal, or simply an offshore banking system with robust privacy protections.

Now I have heard people say that a competitor cannot emerge because the bitcoin is a natural monopoly.  The cable companies, who watched their natural monopoly in television evaporate with the emergence of satellite TV, will chuckle at this conceit.  The thing to recognize is that the fact that a market would, in a perfect world, take the form of a monopoly does not mean that it will actually be a monopoly.  A monopoly will generally only persist if there are barriers to entry.

So are there barriers to entry in the "bitcoin market"?  The biggest barrier to entry is that any competitor would, at first, be small and relatively illiquid.  But this is easily solved.  Imagine a firm seeking to compete with the bitcoin.  It could issue a similar currency, but announce its willingness to buy or sell the currency at, say, $10.  (This is equivalent to "pegging" the currency to the dollar.)  Immediately, the new entrant's product would be vastly more liquid and more stable than the bitcoin. The peg could be removed gradually (for instance, by letting the currency "float" within a wider and wider range).  At worst, the currency would become as illiquid and unstable as the bitcoin already is.

Why would the firm be willing to peg its currency, even temporarily?  The answer is that the seignorage on the bitcoin is potentially tremendous.  Seignorage is the profit that flows from issuing a currency.  So for instance, when the central bank of the United States prints money, it uses the money to buy assets (traditionally treasury securities, more recently various private-sector financial assets).  Those assets generate income, which is remitted to the U.S. treasury.

So imagine that instead of the current scheme for generating bitcoins, the issuer simply sold them periodically.  At current prices, you could generate about $360,000/day by issuing bitcoins.  Now, that is your gross revenue—you would have to maintain a payment network for people to be willing to use your currency, and that would cost money.  But I imagine you could make a nice profit on revenues of $360,000/day even if you had to maintain a fairly expensive network.  You would be earning about $130 million per year, and I can't imagine a network would cost more than that to establish and maintain.  The anticipation of profits could motivate someone to "peg" a currency for a period of time, allowing it to be more stable and liquid than the bitcoin.

Do I expect a competitor to the bitcoin to emerge?  I don't know.  I think that depends on how well the bitcoin fares over the next few years.  But in any case, I expect the bitcoin to fail as a currency.  In fact, I expect it to be dominated by existing currencies to such a degree that predictions of its triumph will come to be seen as delusional.