Don't Put All Your Bitcoins in One Basket

Click here to view original web page at
<p&gtSo, it's still money ... right?</p> Photographer: Yuriko Nakao/Bloomberg
Photographer: Yuriko Nakao/Bloomberg

On Friday, I did a Third Way panel on bitcoins with Jerry Brito of Coin Center, a think tank dedicated to all things bitcoin. Since I was the appointed skeptic, thanks to my earlier writings on digital currency, this was a pretty tough assignment considering that one of the world's leading experts was sitting at the other end of the table.

Nonetheless, I thought it was a productive event -- I certainly learned a lot from Jerry -- and I thought it would be worth doing a post on where my skepticism lies and the hurdles I see for the future development of bitcoins.

To explain that, though, I need to talk about what bitcoins are. The fact is that there is no singular "what" -- and therein lies both their strength and their weakness, and also, I think, a lot of the confusion about the digital currency's future.

One simple way to think about bitcoins is this: Imagine that Western Union ran the money supply -- and no one ran Western Union. That is, bitcoins are -- or are attempting to be -- both a form of money and a system of accounting for payments made in that currency. As many readers probably already know, the currency's major innovation is doing this in a distributed, peer-to-peer fashion, rather than relying on some central authority to create the money or maintain the transaction ledger.

My impression is that sophisticated bitcoin analysts and investors are mostly interested in bitcoin's features as a payment system, which they hope will significantly lower transaction costs. They are less interested in it as a form of speculative investment, or a way to pay your restaurant check -- in other words, as a form of money.

On the other hand, my impression is that a lot of the enthusiasm about bitcoins comes from people who are mostly interested in them as a way to get rich or a way to pay for things -- in other words, people who think of bitcoins primarily in terms of money.

Why does that matter? Because the things we look for in a payment system and the things we look for in money are pretty different.

What do we want from a payment system? First, we want it to be reliable -- if I use it to pay you $400, it should cost me only $400, and you should be sure of getting the money. We would also like it to be fast. And we would like it to be ubiquitous, because the more nodes there are in the network, the more useful the payment system is.

What do we want out of money? Well, we want two things out of it: We want it to be a store of value, and we want it to be a medium of exchange. De-jargonified, that means we want to be able to give it to people and get something of value back. And we want to be able to save it up so that we can give people our money later and still get lots of stuff.

These things are not absolute qualities that are either on or off like a light switch; they are a spectrum, and they are not necessarily tied to each other. For example, the great money stones of the Yap Islanders are an excellent store of value -- a giant stone is hard to destroy, or even steal. On the other hand, you'd get awfully tired trying to drag one of these things around the mall, so it has drawbacks as a medium exchange.

Or consider prison currencies, such as cans of mackerel. Cans of mackerel are preferred because, ironically, no one wants to eat the mackerel, which makes it a good store of value. And within the prison, it is also a decent medium of exchange -- it is possible to price other goods in terms of mackerel. On the other hand, as soon as you leave the prison, your can of mackerel is worth a buck fifty, and you're going to have a hard time finding someone who will trade you cigarettes and magazines for your fish.

In other words, payment networks and money have at least one thing in common: The bigger the network, the more valuable they are.

That ought to be a good starting point for growth in the bitcoin ecosystem. But I think there is a danger here: If most of the network growth is driven by people who think of bitcoins as money, then a change in that sentiment could cause the network to shrink -- at which point it undermines the strength of the payment system.

If bitcoins are ubiquitous, it's easy to piggyback a payment system on it. If people lose interest in bitcoins, however, all sorts of problems crop up. The currency will become illiquid.

Liquidity is just a way of saying "how easy is it for me to trade this for something else?" A dollar is super-liquid; you can go anywhere in these United States, and many places abroad, and people will be happy to give you any number of goods in exchange for your dollars. A multimillion-dollar Manhattan apartment, on the other hand, may be a very good store of value, but it's going to take you some time to unload. The easier it is to trade your asset for something else, the more liquid we say that market is.

Like "store of value" or "medium of exchange," "liquidity" is not some fixed quantity, etched on the rocks at the beginning of time by the hand of God. Some markets are very liquid, and then stop being so, as we all found out during the financial crisis when everyone wanted to sell and no one was buying.  

Why does this matter, if we really just want to use bitcoins as a payment system? Because someone has to hold that liquidity risk. 

Try to imagine a system in which we're mostly just using bitcoins to do lightning-fast currency transactions -- I go to Mexico and use an ATM service that buys bitcoins with the dollars in my U.S.-based bank account, then immediately turns the dollars into pesos and spits them out to me in cash. The idea is appealing, but requires us to ask: Who is buying those bitcoins from me? Someone needs to be making a market in all these other currencies so that the system has enough liquidity for me to use it seamlessly wherever I go.

One way that could happen is if loads and loads of people are using bitcoins, so that at any given time, there are lots of people who want to buy bitcoins with pesos and to sell them for dollars. Another way to accomplish the same thing is to have an intermediary who holds all those bitcoins and sells them for dollars while buying them for pesos. Someone always has to be at the other end of every transaction.

This is what financial institutions do with lots of asset classes, including currencies, so that you can buy or sell even if there doesn't happen to be anyone on the other side who wants to do that trade at the moment. So we could certainly have institutions that do that for bitcoins.

But there are a couple of things worth pointing out. The first is that most currencies are pretty liquid, so it's not like they are permanently holding onto that cash -- they're just making sure that you can buy what you need at 10:47 this morning, even if no one comes along who wants to sell that much cash until 3:17 this afternoon. Making a market in an illiquid market is a very different, and considerably more expensive, business. Which is the second thing we should point out: The folks who make the markets generally want to get paid for their trouble.

We can imagine a situation in which institutions basically just held all the bitcoins and used them to make transactions. However, they'd first have to get the cash to buy the bitcoins -- and then they would charge you a nice fee for making them hold a big, illiquid asset. Illiquid assets aren't merely hard to sell, but they're also hard to price; those sorts of markets are vulnerable to big price swings, not because anything has changed about the underlying asset, but simply because people get forced to sell, or buy, at a moment when there's no one on the other side who's interested in trading at the "fair" price. That means they have to offer a big discount to sell, or pay a big premium to buy. And since any such institutions would also effectively be in the currency-trading business, they'd also be assuming the risks of holding all the other currencies that people want to change their bitcoins into.

If my perception is correct that most of the people who are currently holding bitcoins are doing so because they view them as a speculative financial asset -- or a substitute for a credit card, or a way to evade government surveillance of their transactions -- rather than a way to make payment accounting dramatically cheaper, then liquidity will dry up when they figure out the drawbacks of those uses. In the absence of liquidity, transaction fees will rise to compensate the market makers and the people who maintain the block chain (the public ledger of all transactions). Then it might not be cheaper or more convenient than existing payment systems.

There's at least one scenario in which a collapse in the monetary value of bitcoins could actually help: Institutions could buy up all the bitcoins, then use them as a sort of low-cost back-end system. But note that for this to happen, everyone now holding bitcoins has to lose a bunch of money as the value of their holdings plunges close to zero. And this creates some problems on the ledger side of the transaction, because maintaining the ledger requires considerable computer resources, both storage and processing. Currently, people are paid for that computer processing in two ways: First, "miners," whose efforts maintain the block chain, are periodically rewarded with new bitcoins; second, very small transaction fees are offered to process transfers. If the value of bitcoins falls, however, these rewards may prove inadequate, meaning that some new way will have to be found to finance this activity.

There are other risks that I see, too, such as the risks that bitcoins will become more embedded in criminal networks and governments will move to shut the system down; the risk that theft will become too widespread as ordinary users, less security-savvy than bitcoin enthusiasts, adopt it; and the risk that governments will regulate the hell out of bitcoin transactions, adding costs until they are not really that much cheaper than what we already have. But the primary risk I see is simply that the allure of bitcoins as money will wear off -- and that when it does, bitcoin as payment system will also stop working very well.

I could be wrong about all of this, of course. I'm not a hard-core bitcoin skeptic, and I'm sure that there will be at least some enduring uses found for its rather neat technology. What I'm skeptical of is the idea of bitcoins as a ubiquitous substitute for credit cards or dollars. I think it's far more likely that in the future, however many people are using bitcoin-derived technologies, no one outside of a few technologists, and some disappointed investors, will even know their name.

To contact the author on this story:
Megan McArdle at

To contact the editor on this story:
Brooke Sample at