Can bitcoin really replace the existing fiat money system and take on the monetary responsibility? Lets first look at two simple perceptual data, and we can see that this expectation is unrealistic: the average waiting time for bitcoin transaction verification is 10 minutes; so far, bitcoin experienced the most serious transaction congestion in history in December 2017, and the average waiting time for a transaction in that month is two days and two nights!
From the point of view of system efficiency, bitcoin is different from the existing electronic payment (such as Alipay), which is different from a coach to a rocket. Readers may wonder if the bitcoin is the coach or the Alipay is the carriage. The author can say responsibly: bitcoin is a carriage! In fact, as a payment tool, bitcoin is one hundred times bigger than Alipay, compared to the speed difference between the carriage and the rocket. Please note that this discussion is only from the perspective of payment tools, and does not deny the innovation and application value of blockchain technology brought by bitcoin. (refer to the authors article virtual currency: failed monetary experiments and successful technological innovation, August 30, 2020, surging Business School).
In 2009, the birth of bitcoin marked that virtual currency, as a phenomenal event, formally stepped onto the historical stage. We generally call bitcoin, a private digital currency, as virtual currency. As of March 5, 2020, there are 245500 virtual currencies developed using erc20 on Ethereum. Bitcoin is deeply influenced by the idea of private money. It not only has the grand ideal of replacing the existing fiat money, but also has the grand ideal of changing the existing financial structure. Nakamoto cong, the designer and founder of bitcoin, pointed out in the bitcoin white paper that bitcoin is a pure point-to-point e-cash, which can realize direct online payment between counterparties without financial institutions. Since the birth of bitcoin, it has attracted extensive attention from all walks of life. People once had great expectations for virtual currency. Many people regard it as a high-tech weapon to replace the existing payment system. They even boldly predict that the point-to-point payment method created by bitcoin will make the financial system completely free from dependence on financial institutions.
Remember our middle school textbooks said that the development of new things is tortuous. This sentence is also appropriate for bitcoin and many virtual currencies. At least, bitcoin has become a key means of payment for the public. However, it is not possible to use bitcoin as an indicator of large-scale payment.
Why is there such a huge contrast between reality and expectation? This also starts with the underlying technology of bitcoin.
1u3001 Transaction data structure of bitcoin: transaction, block and chain
Like the traditional electronic payment, bitcoin has abandoned the entities such as paper money and gold. But in order to realize reliable payment in the peer-to-peer environment, bitcoin creatively uses transaction records as the carrier of currency. Although there is the word coin in the name of bitcoin, there is no glittering coin in the bitcoin system. The model constructed by bitcoin is quite different from the balance system of traditional banking system, and certainly different from the ancient gold and silver coins and paper notes (on the contrary, it is similar to gold, silver and paper money in circulation mode and privacy protection). Bitcoin, designed by Nakamoto, appears in the form of transaction records. You dont own a bitcoin, but you have a record of someone elses bitcoin transfer to you. This record is called unspent transaction output (utxo), which can be understood as an electronic certificate of transferable rights.
Many works or reports tell us that blocks form chains, so we call this technology blockchain. In fact, the transactions in the bitcoin block are also linked into chains in the form of utxo records, forming a continuous transaction chain, that is, transactions into chains. Lets have a specific understanding of the characteristics of block and the formation process of chain.
Block is a data body formed by packaging multiple transaction data together. The block can be understood as a ledger in which multiple transactions are recorded. In the simplest case, the transaction will be larger than 250 bytes. The size of each block is limited by Nakamoto to 1 Megabyte (1024 KB). Then we can calculate that a block can hold up to 4096 (1024000 / 250) transactions. In the analogy of account books, Nakamoto Limited a Book of accounts to only 1024 pages at most. And each of the most basic transactions takes up a quarter of a page, so the maximum number of transactions in a ledger is 4096.
While blocks become chains, transactions also form chains. After the bookkeeper records the transactions on the account book (block) according to the rules, the time seal is pasted on the front cover, and the number of the previous account book is recorded on the front cover of the account book, thus forming an account book chain connecting the end to end. There is a feature of the book chain: the older the book, the more difficult it is to tamper with the transactions. If you want to change a transaction in block 363270 in 2015, you need to change all the books since 2015, and then make a new account.
How to prevent tampering with the books? Bitcoin creatively uses the proof of work (POW) mechanism. The miner obtains the accounting right through hash operation competition, so as to prevent block tampering. Bitcoin does not have a central server. Instead, the mining machines of all nodes run spontaneously to provide computing power for the system. Through competition, the miner obtains the right to account, and the system will reward the miner with the bookkeeping right with bitcoin. This is what we often call mining. The pow operation of the miner is to perform two hashing operations on the block header data (the sha256 hash operation is used for bitcoin), and a 256 bit hash value is obtained.
It takes only a few seconds to calculate the hash value, and the miner can complete it in an instant. Therefore, bitcoin requires hash operation, that is, you must calculate the hash value specified, for example, the first 19 digits are all zero. In this way, the difficulty increases sharply. In order to calculate the specified hash value, the miner must use trial and error method to replace the random number (non repetitive random value that is used once) to calculate the hash value less than a certain value. Because the operation result of sha256 is very random, the only way to calculate the specified hash value is to keep trying nonce. Therefore, whether the mining machine can successfully mine depends on the speed of the miners operation; on the other hand, it depends on luck. Maybe the first time you change the value of nonce, you will get a hash value that is in line with the size. Maybe you can run tens of billions of times, but you cant get it. Bitcoin has set the time for automatic adjustment of hash operation, which is set to an average of 10 minutes and dynamically adjusted. This time setting is a balance between two aspects. Too little time and insufficient computing power can not make tampering attempt to retreat from difficulties and cause instability of network synchronization. If the time is too long, transaction confirmation time will be too long and mining machine experience will be too poor. (the hash operation problem is relatively complicated. We do not give a detailed explanation of these terms due to space limitation. Interested readers can read Chapter 4 of the brief history of blockchain.)
The above is the basic explanation of the blockchain principle of bitcoin. For non-technical personnel to understand the operation efficiency of bitcoin, there are three important aspects:
First, each bitcoin transaction verification process can be understood as a block out, that is, a lot of transaction data files are packed together and sealed with data hashes and timestamps, so that they will not be changed from now on (the premise is that they win the hash contest and get enough confirmation in the consensus mechanism).
Second, the bitcoin block size is set to 1 Megabyte, and each block can theoretically only record 4096 transactions.
Thirdly, the block out time is the time for miners to verify transactions, pack blocks, and hash competitions. According to the bitcoin code, the target difficulty value of bitcoin hash competition will be dynamically adjusted, so that the average block time is about 10 minutes.
Two, from the speed of single transaction, bitcoin block out time is set to an average of ten minutes, time-consuming is 200 times that of Alipay, can not be used for public daily use.
Bitcoin sets the block out time to an average of 10 minutes on the underlying technology, which means that the verification time for completing a batch transaction is 10 minutes. The average confirmation time of each transaction is about 10 minutes. For the daily payment of the public, this speed is unacceptable. Imagine that if you check out at the supermarket, the person in front of you starts payment with bitcoin and the merchant confirms the receipt of bitcoin. It takes 10 minutes. The line should be several kilometers away. Then carefully calculate with simple figures: a supermarket cashier takes 10 minutes to complete a collection, only 6 transactions an hour, and only 48 transactions can be completed in 8 hours of business hours. It can be said that the speed of bitcoin can not meet the daily needs of the public.
You can think back to the time when we use electronic payment everyday. Life experience tells us that Alipay does not have more than 3 seconds of processing time in every transaction. It only refers to the processing time of the system, excluding the time spent on user operations such as opening App and entering passwords. According to the technical specifications of China UnionPay, the non-contact card transaction time should be less than 500 ms. Three. From the speed of concurrent transactions, the number of transactions per second that bitcoin can handle is 3.7 1/10000 of Alipay.
The decision of a system transaction processing speed, but also depends on the concurrent processing ability. For example, how many cars can be driven on a highway depends not only on the length of the road, but also on the width of the road, which determines how many cars can be passed at a time. Bitcoin processes a maximum of 4096 transactions per 10 minutes, or 6.83 transactions per second. According to the data released by Alipays double eleven in 2017, Alipays peak volume of transactions per second is 256 thousand. What is the difference between the processing speed of the two? The difference is 37000 times! The first cosmic velocity of the rocket is only 474 times that of the carriage. For example, bitcoin highway is converted to 4096 transactions per minute through ten minutes, and is converted to 6.83 transactions per second, while the Alipay expressway passes 256 thousand transactions in one second. If there are 4097 transactions in the world, the 4097 transaction is destined to be completed in the next 10 minutes after 4096 transactions are completed. Imagine that in the peak travel season, there is only one toll station on the highway, which is bound to be a long queue of cars stretching for several kilometers or dozens of kilometers. In December 2017, the bitcoin system began to congest due to the sharp increase of bitcoin transaction volume due to the rise of bitcoin price. The average waiting time for a transaction in that month is two days and two nights.
4u3001 In terms of the total amount of transactions that can be carried, bitcoin is also difficult to become a large-scale payment means
Through further simple mathematical calculation, we can find that bitcoin can deal with only 590000 transactions a day in theory and only 212million transactions a year. From the actual processing transaction, taking the third quarter data of 2019 as an example, the daily transaction of bitcoin is about 300000. From January 9, 2009 to April 12, 2019, the total number of transactions in bitcoin history is about 400million.
China Netcom platform, which specializes in online third-party payment and clearing, processes 1.18 billion transactions a day, nearly three times the total amount of bitcoin transactions in history. It should be noted that this figure is only part of the transaction volume of online third-party payment in China. Some transactions of online third-party payment in China are transferred to clearing through online payment, some transactions are transferred to clearing through UnionPay, and a considerable part of them are self circulating in payment giants. In theory, bitcoin system can only carry 212 million transactions per year, equivalent to about one fifth of the daily clearing volume of Netcom. The cumulative historical trading volume of bitcoin in nearly a decade is only one third of the daily clearing volume of online clearing. 5u3001 From the choice of digital currency architecture of central banks, this paper proves the deficiency of current blockchain technology
The blockchain technology brought by bitcoin originated from the folk cryptography community, but it soon attracted the interest of government departments, especially central banks. Many central banks, led by China, the United Kingdom, Canada, Japan and Singapore, have launched blockchain technology experiments. Most central banks do use blockchain technology in digital currency experiments, but only in wholesale business (i.e. transaction and clearing and settlement business between financial institutions), rather than retail business (i.e. daily payment and transfer business of ordinary citizens).
There is no doubt that the pressure on the wholesale business is far less than that on the retail business. Although this choice and positioning is mainly for financial business considerations, it also reflects the attitude towards blockchain technology to a certain extent. The United Kingdom, Canada and other countries that have carried out central bank digital currency experiments all claim that their interest in blockchain is mainly settlement efficiency. In other words, the settlement between financial institutions on the blockchain can realize transaction is settlement, which no longer needs the intermediary processing of the central bank. Note that the efficiency here only refers to settlement efficiency, not the technical efficiency (transaction speed, carrying capacity, etc.) we discussed earlier.
According to public data, Chinas central bank digital currency DCEP is the worlds first central bank digital currency to enter the experimental stage in the retail field. However, DCEP does not completely copy the blockchain technology, only introduces the distributed ledger technology to realize transaction traceability, which is similar to the transaction chain mentioned above. In issuing, clearing and settlement, payment, recovery and destruction, DCEP adheres to the central behavior center, and also takes the centralized system as the core in technology.
From the technology selection of central banks in the experiment of digital currency, it shows that the current blockchain technology is not the first choice of mainstream official in the field of mainstream currency and payment system.
As a kind of technological innovation, bitcoin has revolutionary significance. However, bitcoins single transaction confirmation speed is too slow and its concurrent processing capacity is too low. These technical characteristics determine that bitcoin can not become a payment method that can be used by the public on a large scale, and it is difficult to replace the existing legal currency system and take on the monetary responsibility. The achievements of bitcoin similar technologies in the field of payment and settlement can only be applied to wholesale businesses with low timeliness requirements and limited transaction volume.
Of course, the follow-up virtual currency has been trying to improve and improve compared with the special currency, but it still can not be used as a payment method for large-scale public. For example, the peak transaction efficiency of eth is only about 15 transactions per second. It has been reported that EOS has reached 3996 transactions per second, but it can also be used as a payment tool for small countries, regions or industries at most.
Since 2019, the lightning network often mentioned hopes to realize bitcoin transaction without delay and low cost through offline technology. However, the current payment capacity of lightning network is extremely limited and there are a series of technical deficiencies. In 2016, Poon and dryja systematically put forward the idea of lightning network in a paper. The principle is to put a large amount of small transactions on the nodes of the lightning network, instead of submitting them to the bitcoin system. The bitcoin system is only used as the background mortgage and settlement system of the lightning network. In January 2018, the lightning network system was launched in the bitcoin main network. As of March 6, 2020, there were 11624 lightning network nodes, 36289 payment channels and 895 bitcoin payment capacity. The research on Lightning network shows that the structure of lightning network tends to be centralized, and the network operation efficiency depends on the payment capacity of large central nodes, rather than the scale expansion of point-to-point network.
Bitcoin and many other virtual currencies cant become a payment method that can be used by the public on a large scale. The development of technology in the future also allows blockchain to become a large-scale payment system. However, the inherent value, transaction cost, compliance and other problems of virtual currency still exist, which may still be difficult to be the monetary responsibility. This is the limitation of virtual currency such as bitcoin as private currency, and it is also a highly controversial topic. Lets leave it to the next time.
(the author Wu Yun and Zhu Wei co authored a brief history of blockchain. This paper is based on the paper virtual currency: a failed social experiment of private money published by Wu Yun and Zhu Wei on the 6th issue of financial supervision in 2020? u300bThe author has rewritten this edition of the surging news. Interested readers may refer to the original text.)