Friday, January 20, 2012
Bitcoin is a new payment application available on the internet since January 2009. In a way, by virtue of its open source publication, it is similar to the World Wide Web, the hugely successful internet application of the internet that now enables so many others. Much like the WWW has redefined the way mankind produces and shares knowledge, bitcoin transforms the social code underlying money supply to bring about a new degree of economic freedom. Can it be seen as a new monetary reform vehicle?
The Bitcoin Database of Transactions
The bitcoin protocol specifies how to build and maintain a distributed database of transactions on the internet. Transactions are published and signed electronically (using asymmetric cryptography and key pairs). The protocol enforces confirmation of every transaction by the network nodes.
Because all transactions in the database are linked together cryptographically, a property that explains why the bitcoin database is usually referred to as the bitcoin “chain”, any change in a past transaction would require computing an entirely new database from that point on.
Signatures by private keys ensure the title of property to any given amount of bitcoins: knowledge of the private key associated to the receiving address is required to send a part or the whole of any transaction output. The bitcoin transaction database holds all the necessary information for an address owner to receive and spend any amount of bitcoin: the database and the communication protocol together bring to life a new electronic currency.
Bitcoin Money Supply
Moreover, the bitcoin protocol specifies a money supply mechanism. Bitcoins are generated gradually by the network until a maximum quantity of 21 million bitcoins is reached. Bitcoins are created ex nihilo, pretty much in the same way as dollars are created when a commercial bank lends money to someone who is buying a house. The house existed before the loan. In fact, the seller is walking away with the dollars, de facto transferring the ownership of the house to the bank. The bank essentially created the money in its ledger out of thin air: it recorded the amount of the loan as a bank’s liability upon crediting with it the account of the borrower and that same amount was recorded as a bank’s asset, i.e. as a loan that must be paid back to the bank. Even top bankers and economists, including Nobel Prize winner Maurice Allais, call it magic.
Bitcoin builds on the recognition of the fact that the monopolistic aspect of the money supply mechanism in today’s banking system deserves at least to be revisited, if only because it has lasted for several hundred years. If only because of the continued effects of the financial crises, Bitcoin allows us to experiment with a new concept for money.
Bitcoin, as a new universal currency, is a true innovation, building on numerous prior attempts to create a sustainable currency and doing so independently of any state or centralized organization.
Towards the growth of a Bitcoin Economy?
Some bankers I spoke with shrug off Bitcoin on the somewhat circular logic that it cannot be backed by any assets in the real world. Well, thanks to merchants accepting bitcoins and to online exchanges trading bitcoins just like any other currency, a Bitcoin economy is emerging. If the economy of the euro or the dollar collapses, the expectation of a backing by the central banks is based on the assumption that a government has unlimited taxation power over said economy. The current financial crisis in Greece or, to a lesser extend in the US, demonstrates that actual use of this theoretical power is a stretch of economic reality. The truth of the matter is that 90% of the money today is created in the books of the banks and that price stability is the determining factor towards continuing public support of the central banks view.
There are over 140 state sponsored currencies in use across the world today. It is hard to prove that a new complementary currency will jeopardize the world’s economic outlook only because this new currency, unlike all the others, stems from a distributed monetary system without a central authority. The innovative feature of bitcoin that truly sets it apart from anything that existed before is simple: it is cash online without the necessity for a central organization or third party. As such, it is making other forms of cash transaction comparatively less convenient. Removing the necessity does not affect the usefulness of a third party in some cases. Transaction processing between merchants and their customers will always entail some kind of dispute resolution system, regardless of the currency that is being used for payments.
E-gold, as a tentatively universal currency backed by gold, was created in 1996 and failed mainly because it relied on a central organization to manage an inventory of gold. The quantity of gold needed was supposed to grow with the e-gold economy, which was impractical, dangerous and fundamentally useless in a digital age. (Another shortcoming of e-gold was the lack of a specific protocol: relying solely on a web browser to conduct irreversible transactions makes it difficult to secure them.)
Bitcoin money supply mechanism simulates the extraction of a rare metal with a mathematical model, using a clever recipe proven in electronic signature schemes and hashing algorithms that can be found today as ingredients in most banking systems.
One can think of bitcoin as a currency backed not by gold but by a metaphoric substitute of gold, since the quantity of bitcoins is limited by design: 21 million bitcoins.
Like gold, bitcoins can be seen as bonds that never mature but unlike gold, bitcoins bear virtually infinite divisibility and liquidity with no vaulting costs. According to 2010 year-end estimates by GMFS, total above ground stocks of gold are 166,600 tons worth US$6,500 Billion by 2010 average gold price, of which around US$2,400 Billion are held privately or by official reserves, in the form of coins and bars. The total stocks minus approximately 30,000 tons held in official reserves worldwide as of august 2011, gives us an estimated “market” size of US$1,230 Billion for gold as a private store of value. If we were to compute a fictitious exchange rate with the dollar based on these numbers, we would project an exchange rate of around US$600 for one BTC, if bitcoins were to capture only a 1% fraction of the private market for gold as a hedging instrument.
By the same token, if the bitcoin economy were to grow up to 5% of the US GDP, i.e. approximately US$750 Billion USD, and assuming the velocity of bitcoins to be the same as that of the dollars (around 50), then one bitcoin would represent the equivalent of US$700.
The numbers translate to a $15 Billion future valuation estimate for the bitcoin network. They are consistent with the market capitalization of Visa, Inc. ($55 Billion as of august 2011) or MasterCard ($39 Billion). Buying bitcoins today is like buying stocks of a new global electronic transaction network. At $10 as of august 2011, Bitcoins can be considered undervalued even if it is likely that other universal currencies might try to step in at some stage.
One would think at first that such a valuation would provide an investor with the rationale for purchasing the most powerful computer on earth (“K” in Japan as of June 2011) to take over the bitcoin network. However, doing so would drive away today’s “mining” participants and, at least temporarily would reduce the value of the bitcoins to zero. The supercomputer would stay idle, acting as a deterrent to any seller of bitcoin, waiting for the intruder to back away. In other words, to mitigate this risk, the new operator himself or herself would have had to acquire a large amount of bitcoins beforehand to keep the bitcoin economy running after the takeover, in the hope that more sellers would return after the more or less chaotic transition.
Additionally, by the time the attack is ready, it is unclear whether said supercomputer would be able to match the majority of the hashing power of the current nodes while this cumulative power is going up day after day. The uncertainty surrounding the outcome of this kind of hostile takeover makes it more likely that a rational investor would simply buy bitcoins much in the same way he or she would buy stocks in a start-up venture.
Another, more naïve attack would consist in buying “pools” of miners that are thriving these days on the bitcoin network. By joining a mining pool, a miner aims at gaining a share of the steady flow of bitcoin expected with a large amount of aggregate hashing power. The share is prorated according to the hashing power contributed by the miner. Conversely a solo miner can go for a long period before he/she earns the 50-bitcoin reward associated with the computation of a new block of transactions. Statistically, the expected rewards are the same though, only the income flow is steadier in a pool. Hence a miner would have no incentive to stay with a pool under the control of a hostile investor. He or she would simply switch to another pool or start mining solo.
This analysis remains true even after all the bitcoins have been minted. In fact, the incentive to “mine” transaction blocks will sustain the decline in the rewards by virtue of the increasing transaction fees combined with the increase in value of the bitcoins.
Electronic Money and Trust Management
Bitcoin is backed not only by this kind of projection but also by the trust of the “network nodes” that is materialized in the aggregate computing power of this group of people. Anyone with a personal computer and a graphic card (GPU) can join in and start participating in this giant transaction processing pool that defines the bitcoin network. It requires only downloading a piece of free software known as “mining” software in reference to the bitcoin money supply rules. Today, there are several tens of thousands of such “miners” around the world.
In contrast, Ripple, launched in 2004, attempted to create a universal currency by knitting together a web of Local Exchange Transaction Systems (LETS): IOUs issued in any LETS could be forwarded to any participant in the network, across a Web of Trust mechanism. However, a Web of Trust requires a central authority to manage trust certificates and to prevent fraud, raising lots of practical issues: the thorny problem of registering people to link them to a secure digital identity can lead a technological innovation to its demise even before it reaches any sizable roll out stage.
Bitcoin proof of work protocol avoids the need for a secure web of trust, relying instead on the assumption that a majority of the computing power is in the hands of honest participants: “honest” here simply means that they will cooperate to make the network confirm legitimate transactions. Legitimate transactions on the bitcoin network are those that can be linked cryptographically to the “genesis” transaction via the “longest” chain: the chain “length” refers to the depth of its proof-of-work, not to the number of transaction blocks. The “genesis” transaction is one that took place in January 2009 and got things started. This concept of cloud computing applied to the confirmation of transactions is very consequential: only two years after the launch of bitcoin, it takes already as much computing power as the “K” computer to take over the network. Taking over the network today with such a powerful machine would not prevent the network from operating but would disturb transaction confirmations until participants find a way to overcome the attack and regain control of the operations. Recovery strategies could be applied to resume operations normally thereafter. In other words, even assuming that a government or a large organization would be able to harness so much computing power to engage in such heavy-weight counter measure against the bitcoin network, its expected outcome is uncertain at best. This property makes bitcoin as resilient as a transaction network can be.
A universal currency like Bitcoin, which is using digital signatures and asymmetric cryptography, has the interesting additional property that it can go back and forth from digital to fiduciary status. The old boundaries between electronic transactions and cash transactions are blurred: the file containing the key pairs (public receiving address and secret key to sign the outgoing transfer of the amount received) can be printed. Let us say the public key is left apparent and the secret key is hidden underneath a cover: the cover can only be removed in a non-reversible manner. In doing so we have created a new e-note containing the bitcoins received on the public address that cannot be spent until the secret key is revealed. The e-note can be traded as long as it is not tampered with. The amount received on the public address can be printed also in a tamper-proof process: the amount received on the public address becomes the denomination of the e-note (7). When the cover of the secret key is removed, the e-note amount in bitcoins can be redeemed electronically for any payment using bitcoins. In further contrast with old notions of fiat money, e-notes without denomination are also possible: the amount received on any address can be checked by the recipient in the public bitcoin transaction database.
Decentralization and Deflation
By definition, a truly decentralized universal currency must start without a central authority regulating its money supply mechanism: this precludes establishing any form of correlation between the money supply and any set of economic parameters or measurements. Such systemic correlation would yield endless discussions between the users, hence requiring a governing body capable of moderating the discussion and enforcing the rules by some yet unknown universal democratic standards.
For the same reason, the new transaction software must be free software to escape the limitations and opacity of proprietary software. As Richard Stallman summed it up in his now famous statement: “The point of free software is either the users control the software or the software controls the users”.
To a large extend, state sponsored currencies like the euro or the dollar are created in a black box, with a lot of media attention focused on interest rates rather than money supply. A central organization promoting a currency system based on proprietary software would not be a game changer.
Therefore the money supply mechanism must be hard coded and published in the specifications from the outset, with little or no room for the currency to wiggle out of it to its possible demise. In the same logic, it is neither possible nor necessary to predict the rate of adoption and growth of the user base for the new currency: these numbers cannot be factored in with precision. The money supply model is therefore deflationary, defining a maximum quantity, unless the rules of generation are bound to the number of users by a user authentication protocol. This requirement is not compatible with an objective to design a decentralized currency since user authentication requires issuing identity certificates either in a web of trust or with a certification authority.
The Bitcoin specifications not only fulfill the requirement for a limited money supply but also make provisions for transaction fees to provide a sustainable incentive for miners to keep mining even after the rewards for the generation of new bitcoins have dwindled to zero.
Because Bitcoins are traded electronically, unlike gold, they are infinitely divisible and enjoy a high velocity, so a deflationary spiral can only reduce the scope of bitcoin to the function of a store of value, a more practical process than is used for gold. In fact, the deflationary spiral would have adverse economic consequences only if bitcoins were the exclusive currency in a given territory. That’s not the case: as a complementary currency, bitcoins are supposed to coexist with the local state sponsored currency not to replace it. Prices will most likely be expressed in local currencies. In an electronic online transaction, the price expressed in a universal currency can be easily adjusted in real time for exchange rate variations. Only for off line transactions, price stability is a strong requirement for a new universal currency.
In short, deflation will augment the attractiveness of bitcoins as a store of value and will only marginally affect its application as a trade currency.
Where do we go from here? In a global economy, the inception of one or more universal currencies is bound to happen as soon as technology permits it.
Bitcoin, as the first, is paving the way for new applications. In particular, bitcoin can greatly enhance the efficiency of money transfer where it is lacking, specifically development aid that was once famously characterized (by economist Peter Bauer) as “an excellent method for transferring money from poor people in rich countries to rich people in poor countries”. Bitcoin can leverage the generalization of mobile phones in developing countries to enable a money transfer directly to the recipient, bypassing all state bureaucracies and banking intermediaries. The institution or non-governmental organization responsible for the transfer could simply assign bitcoin addresses to recipients and their local merchants then fulfill the money transfers and payments in bitcoins.
The technology is enabling both a new kind of transaction network and a new universal currency.
By analogy, it is worth noticing that the World Wide Web does have a governing body, namely W3C, a non-profit organization made of more than 300 members among the largest companies in the high tech sector. Clearly, any leverage applied by a government to one of its constituents in W3C can be balanced out by the others if it does not fit the bill of the general interest. Because this principle holds successfully for the technology enabling new ways of producing and sharing as valuable an asset as knowledge, one is permitted to hope that a similar organization can also deal one day with the Bitcoin protocol specifications at a technical level to maintain its immunity from the hazards of macro-economic measurements.
Reprinted under Creative Commons.
For further reading:
"The future of money. What is it?", Warren Coats, Cayman Financial Review, January 18, 2012