The emergence of Bitcoin (BITCOIN) as a viable currency reminds me of the early works of novelist Neal Stephenson, Snow Crash, The Diamond Age, and Cryptonomicon, in which Stephenson foresees a near future in which national governments have withered away. The cause of this was none other than the inability of governments to tax and regulate encrypted electronic commerce. Yet Federal Reserve Chairmen Ben Bernanke has said that Bitcoin "may hold long-term promise." What that promise might be, is the subject of this article.
Although Stephenson's somewhat anarchistic visions of the future make for enjoyable reading, I always thought he underestimated the resourcefulness with which national governments can generate tax revenue. His view also lacked historical perspective. Throughout our history, private forms of money have coexisted with government-backed currency in the U.S.
A good review of the history of private money in the U.S. is available in an article by Bruce Champ, a senior research economist with the Federal Reserve. The article is delightfully brief and free of technical jargon. As Champ recounts, in the 1800s private banks often issued bank notes, paper currency backed by the bank. Other private companies, such as railroads and mines paid workers in "scrip" redeemable in goods or services provided by the company. Private forms of money arose to fill a void left by the U.S. Government. When the government stepped in to fill that void, private currencies withered away.
Writing in 2007, a year before the birth of Bitcoin, Champ's observations regarding the potential for a digital currency seem particularly prescient:
With advances in information and communication technology-not the least of which is the ability to embed a wafer-thin computer chip into the equivalent of a credit card-it seemed certain that a new form of private money, "electronic money," would arise as an alternative to paper money and coins in everyday transactions. . . But so far, we haven't seen electronic forms of private money emerge as anticipated. . . Undoubtedly, too, there will be new voids in the future that will require new forms of money. Perhaps these voids will be filled by innovations in money provided by the Federal Reserve. But the private sector might also jump in and fill them, too.
Filling the Void
Stephenson's novels left mostly undefined how encrypted e-commerce would actually be implemented in order to render it impervious to governmental regulation and taxation. Looking at Bitcoin, it's as if Satoshi Nakamoto, the pseudonym for the person or group that developed Bitcoin, had read Stephenson's books and said, "Okay, how do we make this happen." All the features of Bitcoin are designed to meet the requirements of providing a viable digital currency without relying on a central bank or governmental authority:
1) Prevention of digital counterfeiting. Nakamoto's solution for this is the "block chain", a secure record of all the transactions of all Bitcoins in existence. The block chain would be maintained and updated by a decentralized network of computers operated by individuals and small groups. Each "full node" computer on the network maintains a copy of the complete block chain.
2) Controlled currency generation and transaction processing. Here the solution proved to be Bitcoin mining, which is really about Bitcoin transaction processing. Processing Bitcoin transactions into a block to be added to the block chain is rewarded with 25 Bitcoins per block. Block processing is made very difficult due to encryption requirements levied on the process in order to keep the block processing at an approximately constant rate of 1 block every 10 minutes. As Bitcoin value has increased, more and more prospective "miners" have piled on. The system has responded by increasing the level of difficulty to the point that only very specialized hardware can process a block in a reasonable time.
3) Anonymity. Finally, Nakamoto solved the problem of anonymity through the use of encryption key pairs, a standard technique where a user is issued a private and public key. The key pairs are available for the asking, and no personal information is needed. A user may be issued as many keys as desired, and typically a key pair is used for only one transaction.
Because the private key is held by only the user and the network (ideally, unless keys are compromised), the network assumes that anyone with the private key is authorized to transfer the bitcoins held under the private key. To "pay" for something, it's simply a matter of transferring the public key to the Bitcoin recipient, while authorizing the transaction with the network using the private key.
Bitcoins are fundamentally a scheme for secure, decentralized, unregulated, anonymous electronic transaction processing. This is their intrinsic worth, the void that conventional currencies or currency based credit card transactions couldn't fill. But to achieve the above properties, Nakamoto made a trade-off, sacrificing efficiency compared to conventional payment methods. This inefficiency manifests itself both in terms of the monetary and energy costs of processing a Bitcoin transaction.
According to blockchain.info, each Bitcoin transaction costs $66.04 to process, as of this writing. This is just the mining fee of 25 Bitcoins divided by the average number of transactions per block of 377, expressed in dollars, using an exchange rate of $991.92/BTC.
Because of the every escalating block processing requirements, the estimated energy cost per transaction is enormous: 1.677 kWh. Think of of two high powered servers running for an hour, just to process one transaction. And this assumes that the most advanced ASIC-based hardware is used in the block processing. In fact, the mix of hardware used in block processing, often done by mining "pools" is probably less than optimal and the energy cost greater still.
How does this compare to conventional credit card processing. Let's use MasterCard Inc. (NYSE:MA) as an example. In Q3 2013 MA processed 51 million transactions per day, and had revenue of $2.22 billion for the quarter, according to their 10Q. That works out to an average revenue per transaction of $0.48.
Estimating energy consumption by MasterCard is a little more difficult. Surely, MA doesn't devote the equivalent of 2 server-hours to processing each transaction, but we should probably include the energy overhead of MasterCard's facilities. As a rough order of magnitude estimate, let's assume a certain power consumption for each of MA's 5000 employees of 10 kW, including heating, lighting, air conditioning and computing. This works out to about 42.5 transactions/kWh. Even if the energy cost per employee were ten times greater, the 4.25 transactions/kWh would still be more energy efficient than Bitcoin.
Bitcoin's transactional inefficiency is currently masked by the fact that transaction costs are paid with newly minted Bitcoins. As the number of Bitcoins reaches the maximum limit of 21 million, block processing is supposed to be paid exclusively out of transaction fees. Presumably, the level of difficulty of block processing will be reduced (since there's no longer any currency generation) to make block processing economical. But zero Bitcoin payment per block is not projected to occur until 2140.
The Governmental Alternative
Is it possible that Bitcoin, or one of the host of Bitcoin clones, such as Litecoin, might find a technical solution that improves transaction processing efficiency. I wouldn't discount the possibility at all. The Bitcoin wiki even acknowledges the possibility that competing digital currencies might offer a better solution, thus reducing the perceived value of Bitcoin.
But the obvious solution is to relieve the condition of decentralization which currently drives the transactional inefficiency of Bitcoin. If the Federal Government were to issue a digital currency, it could avoid using the mechanism of processing efficiency altogether to limit currency size and simply set limits based on monetary policy, as it does now. The Government could even offer anonymous key pair generation as Bitcoin does, thus allowing people to use the government-backed digital currency as they would cash without fear of repercussions.
Of course, there would always be those who wouldn't trust the government to honor the user's privacy and would prefer the private alternative, which will still exist, but the vast majority of U.S. consumers won't care whether the Feds have digital currency transactions under surveillance. Recent revelations about NSA surveillance haven't met with a storm of public indignation. Far from it, primarily because so much of the surveillance is impersonal and designed to ward off the threat of terrorism.
So just as happened in the 1800s, the Federal Government will step in to provide a form of currency also being provided by private parties, and thereby render that private currency largely unnecessary. Eventually, physical currency will be phased out altogether. The dollar will live on, but in digital form.