2018
Bitcoin arouses strong passions. At one extreme, the enthusiasts hail it as a new currency, immune to the manipulation of politicians and central banks. It has even been suggested, apparently seriously, that Bitcoin could be the basis for restoring world currencies to the gold, well, Bitcoin, standard.1
On the other side, Bitcoin is seen as the digital equivalent of snake oil, tulipmania and Ponzi schemes. Detractors admit that the underlying blockchain technology is interesting, brilliant even, but not in the context of a currency.
For those who have been living in isolation for the past ten years, Bitcoin is a ‘cryptocurrency’ based on a ‘distributed ledger’. ‘Miners’ gather transactions into ‘blocks’ which may be thought of as a page in the ledger. A block is then added to the ledger through a protocol known as ‘proof of work’. All members of the network receive a copy of the ledger, and the ‘proof of work’ protocol ensures that the ledger cannot be later altered.
Peer to peer Bitcoin payments are effected without the need for a ‘trusted third party’. Clever mathematics and a public ledger provide the required trust.2
The Bank for International Settlements has now weighed in on the argument. The Annual Economic Report 2018, Chapter V is entitled ‘Cryptocurrencies: looking beyond the hype’, hereafter ‘BIS V’. Although the title refers to ‘cryptocurrencies’, the analysis and criticism is almost entirely focused on the Bitcoin structure which uses blockchains and ‘proof of work’.
Alice makes a cash purchase from Bob. She hands him notes and coins. Bob looks at them, decides that they are valid Australian notes and coins and the transaction is finished. No third party verification is required.
The early dream of ‘digital coins’ was to allow the same peer-to-peer transaction. Bob would have some method of determining if the proffered digital payment was valid and that Alice had the right to make the payment. No third party would be involved.
Samid has shown that this dream is not achievable. There must be an ‘oracle’ who will verify that the digital money is valid and that Alice has the right to transfer the sum to Bob. In our everyday use of electronic money, the oracle is the bank which holds the payer’s account.3
In Bitcoin, the oracle is the distributed ledger built on blockchain technology. The intellectual appeal of the blockchain is that it provides a trustworthy system in an environment where no particular party is trustworthy. It is robust in the sense that it continues to function when one or more of the participants drops out or when one or more tries to game the system.
The central argument of BIS V is that there may be situations where blockchains provide the answer to a lack of trust, but the payment system is not one of them. The argument is made in two stages:
BIS V begins with an historical review of money, noting that history is a graveyard of failed currencies. From this review, they derive the proposition that ‘money’, to be viable, must be:
After a brief and not entirely neutral analysis, BIS V concludes, probably correctly, that trust is central to the success of a currency, that without trust in ‘the system’ a currency will soon cease to fulfil one or more of the essential characteristics of money.
BIS V then makes a small leap in logic (at 93): ‘The tried, trusted and resilient way to provide confidence in money in modern times is the independent central bank.’
And, to be fair, the leap is justified in terms of today’s payment systems. They are generally fast, amazingly cheap and remarkably reliable. Most of us would probably also admit that central banks do a relatively good job of stabilising the value of currencies.
Most of us, but not all. The very intervention of central banks is seen by some as ‘interference’ with a free market and, even worse, it is an interference which has resulted in inflation and a host of other ills that plague modern society.4
BIS V enumerates three shortcomings of Bitcoin which, in the opinion of BIS, prevents Bitcoin from becoming a serious currency:
Scalability is certainly a serious issue for Bitcoin and other cryptocurrencies. Bitcoin can currently process somewhere around 7 transactions per second. By contrast, the Visa system processes around 24,000 per second! There are research proposals to increases the Bitcoin rate, but none have yet been proven effective in real world transactions: see the discussion below.
There is another aspect of Bitcoin that causes scalability problems. The Bitcoin blockchain is big, currently about 170 Gb and growing at about 50 Gb each year, and it must be communicated to all the ‘miners’ in the Bitcoin system. BIS V notes that if Bitcoin attempted to process national payments that the blockchain would soon swell beyond the storage capacity of most, if not all, computers. It would, says BIS V, bring the internet to a halt, a claim that made many headlines.
Value stability is, of course, Bitcoins most obvious Achilles Heel. The first commercial sale which accepted Bitcoin was for two pizzas worth about $25. The purchaser paid 10,000 Bitcoins. In December 2017 the price was near US$20,000 and at the time of writing is just over US$6,000. Samid argues that a ‘self-anchored’ currency such as Bitcoin is inherently unstable.5
Bitcoin adds blocks of transactions to the ledger on average once every ten minutes. A payee cannot be certain of payment at least until the particular payment is incorporated in the ledger. As BIS V notes, there have been times when payments have queued so that finality cannot be determined until much later.
BIS V is probably correct about finality being a Bitcoin problem, but perhaps it is not as serious as imagined. After all, an Australian doesn’t need too many grey hairs to remember when finality took five days! Business seemed to survive in spite of it. Still, a modern payment systems should achieve finality faster.
A final criticism of Bitcoin is the huge energy use. BIS V notes that Bitcoin uses the same electricity as a medium size country such as Switzerland. Other estimates put power usage on a par with Singapore or Ireland. The ‘proof of work’ protocol uses vast amounts of computing power and, therefore, vast amounts of electricity.6
Since bitcoin is an autonomous distributed system, it cannot respond directly to BIS V. Fortunately, there are many cryptocurrency advocates ready and willing to answer the claims of BIS V.7
Some specific references are given at the end of this note, but the responses generally fall into one of two categories. The first is, to misquote the English philosopher, ‘They would say that, wouldn’t they?’.8 The BIS is, after all, central banking personified, and all cryptocurrency proponents have a deep and abiding distrust of central banks.
The second class of criticism argues that BIS V is hopelessly out of date. There is active research going on, tests being done, new systems being built that will answer each and every criticism of BIS V.
There is, of course, no arguing with the first category of criticism. The second category requires a few observations. First of all, the claims are correct. Cryptocurrency researchers are indeed addressing the problems exposed in BIS V, a fact noted in BIS V itself at footnote 27.
The most important current research is (probably) the Lightning Network and the Casper version of the ‘proof of stake’ protocol. Both are aimed at increasing the throughput transactions, thus letting the blockchain technology scale to be a serious payment system. In addition, the Casper protocols are intended to reduce the power requirements of the existing Bitcoin network by replacing ‘proof of work’ with ‘proof of stake’.
There is so much excitement, cynics say ‘hype’, surrounding blockchain technology that it is hard for the non-expert to evaluate the likelihood of any of this research meeting the criticisms of BIS V. It is, however, easy to note that none have been implemented for anything like the time that Bitcoin has existed and so they have not been subjected to the intense scrutiny that has been focused on the Bitcoin protocol.
Security in a cryptocurrency is everything. The security breaches related to Bitcoin have never been against the Bitcoin protocol itself. The breaches have always been against currency exchanges and other ‘services’ that have grown up around Bitcoin. The breaches have succeeded, in other words, against conventional databases, not the distributed ledger.
One criticism of BIS V does not fit into the two categories: one report notes that the existing payment system requires a lot more electrical power than the Bitcoin network. If we consider power required for ATMs worldwide, the power used by banking computers and the SWIFT network, and the power required to provide security for normal financial institutions, then Bitcoin looks rather frugal.9
This is probably true, but it hardly seems fair to compare the power usage of payment systems that provide for a large proportion of the world’s population with that of Bitcoin which is insignificant as a payments provider.
Is there a future for cryptocurrencies? It may be too soon to tell, but even the BIS recognises that there are some failures of the existing payments system.
BIS V itself mentions the appalling service offered to expatriates wishing to send relatively small amounts of money to family in their home country. Exchange rates are terrible, service costs are high and often unpredictable. Every intermediate bank takes a bite out of the sum being transmitted. It is little wonder that ‘informal funds transfer systems’ have prospered.10
Not mentioned in the BIS Report, but dear to the hearts of cryptocurrency supporters, is the high handed actions in 2010 of the major credit card companies in prohibiting payments to Wikileaks. The action was widely considered proof that payment institutions could not be trusted. Since the action was clearly politically motivated, and since the political winds are subject to change, it must be admitted that the Bitcoin proponents have a point.
Disintermediation of the payment system is attractive to some and disturbing to others. All agree that it would undoubtedly have widespread financial consequences. Bitcoin and similar payment structures are outside any direct control of central banks and individual governments. BIS V notes that cryptocurrencies such as Bitcoin ‘can only be regulated indirectly’ and discusses some of the possible approaches. They also note that ‘Since cryptocurrencies are global in nature, only globally coordinated regulation has a chance to be effective.’ Good luck with that.11
BIS V is correct in its criticism of Bitcoin price stability and of scaling issues. If BIS V is correct in saying that stability may only be maintained by constant adjustments in the money supply, then it is certainly possible to achieve something like that through an automatic algorithm. It is very important to note, however, that a simple feedback loop can make things worse instead of better.12
The Lightning Network, proof of stake or some other innovation might solve the scaling problem, but there are more radical possibilities. The blockchain is a brilliant concept, but at the same time its foundation is a very primitive mathematical structure. Using a more complex mathematical model is another possible way of achieving a higher throughput while still allowing a distributed system operating with no trusted third party.