Blockchain and Bankers

1. Introduction

2. Introduction to Digital Currency

3. Byzantine Generals Problem

4. Introduction to blockchain technology

5. Blockchain and Bankers

VI. Prospects of blockchain applications 1. Introduction

On December 3, 2013, the People’s Bank of China and five other ministries and commissions issued the “Notice on Preventing Bitcoin Risks” ^[1] , which denied the currency attributes of Bitcoin and defined Bitcoin as a virtual commodity that ordinary people can freely trade at their own risk. Two years later, what new views does the central bank have on Bitcoin? A piece of news not long ago once again set off a fire in the circle of friends of Bitcoin enthusiasts ^[2] : On January 20, 2016, the People’s Bank of China Digital Currency Seminar was held in Beijing… People’s Bank of China Governor Zhou Xiaochuan attended the meeting, and People’s Bank of China Deputy Governor Fan Yifei presided over the meeting… The meeting required that the People’s Bank of China’s digital currency research team should actively absorb the important results and practical experience of digital currency research at home and abroad… and strive to launch the digital currency issued by the central bank as soon as possible.

In short, this news tells us that the central bank is currently exploring the feasibility of issuing its own digital currency using the underlying technology of Bitcoin, blockchain. Coincidentally, Nasdaq, on the other side of the ocean, is also conducting its own blockchain experiment. On December 30, 2015, Nasdaq announced that the startup Chain issued shares of the company to private investors through the blockchain technology trading platform Linq ^[3] . In the statement, Nasdaq said that Chain was the first company to complete a private securities transaction through the Linq platform. Nasdaq CEO Bob Greifeld said:

Through this initial application of blockchain technology, we have begun a process that has the potential to fundamentally change the existing capital market trading system.

In fact, it is not just the central bank and Nasdaq. In the above picture, banks and financial institutions such as Goldman Sachs, Citigroup, UBS, and Western Union have also participated in the investment and research of blockchain technology ^[4] . So what exactly is blockchain technology? How did it attract the attention of so many bankers? Why is it said that it has the potential to bring profound changes to the capital market? I hope this article can help you solve these doubts.

2. Introduction to Digital Currency

As an information technology, blockchain is not well known. However, when it comes to the most successful blockchain-based application, Bitcoin, everyone must have more or less known about it. So let’s start with Bitcoin.

Bitcoin is a decentralized cryptocurrency. Its founder, Satoshi Nakamoto, published a paper titled "Bitcoin: A Peer-to-Peer Electronic Cash System" on November 1, 2008 ^[5] , which detailed how to establish a decentralized electronic payment system between strangers. On January 3, 2009, Satoshi Nakamoto developed the Bitcoin client and collected the first batch of 50 Bitcoins, thus Bitcoin was born. On October 5, 2009, the earliest exchange rate appeared: 1 US dollar = 1309.03 Bitcoins ^[6] . Since then, as the price has soared, Bitcoin has attracted more and more attention.

Although Bitcoin has only been around for seven years, the idea of ​​encrypted digital currency can be traced back a long time. The asymmetric encryption technology used by Bitcoin has been around since the 1970s. In 1982, David Chaum proposed the idea of ​​an untraceable cryptographic network payment system. In 1988, the Crypto-Anarchist Manifesto ^[7] even predicted the emergence of the encrypted digital market.
Since the idea of ​​digital currency has already appeared, why has there been no successful digital currency in the past 30 years? This involves some technical discussions. First of all, we must know that the ideal encrypted digital currency must be distributed. The so-called distribution, also known as decentralization, means that the operation of the entire system does not depend on one or several central nodes. In the figure below, the left side is a typical centralized system, because the operation of the entire system is completely dependent on the normal operation of the central server, while the right side is a distributed system. Even if a server is removed, the entire system can still operate normally. Regarding the comparison of the two systems, there is a wonderful metaphor in the book "Starfish Model":

If you chop off a spider's head, it will undoubtedly die; but if you chop off a starfish's leg, it will grow a new leg, and even the severed leg will grow into a brand new starfish.

If digital currency chooses a centralized system, then the entire system will be paralyzed if the central server fails, which will cause many thorny problems. For example, how to ensure that the custodian of the central server will not tamper with the transaction information privately? How to ensure that the central server will not be hacked or closed by the government? These problems are not groundless worries, but have a basis in reality. For example, E-gold was previously sued and investigated by the US government under the Patriot Act ^[8] . It can be seen that moral risks and legal issues are fatal to digital currency, and these cannot be easily solved by computer programs. Therefore, a distributed system is the ideal choice for digital currency. However, establishing a distributed transaction system is much more difficult than a centralized one in terms of technology, because in this process, a classic problem in the field of communications needs to be overcome - the Byzantine Generals Problem.

3. Byzantine Generals Problem

The formal statement of the Byzantine Generals’ Problem (hereinafter referred to as the “consensus problem”) is: How to reach consensus on information in a distributed network that is not based on trust? This statement may sound a bit obscure, but its essence is not complicated. The following example is not completely consistent with the consensus problem, but it helps us understand it ^[9] .

Imagine that in the distant Byzantine era, there was a rich city-state with everything from gold, silver, jewelry, and silk, and its lord Doraemon enjoyed all this luxury and glory. On the outskirts of the city-state, four Byzantine generals, Nobita, Fat Tiger, Suneo, and Shizuka, all coveted Doraemon's wealth, so they decided to join forces to capture Doraemon's city-state. According to the strength comparison of both sides, more than half of the generals must launch an attack at the same time to defeat the enemy, so the winning condition is that at least three of the four can agree on the attack time. So what are the chances of the four generals winning?

The answer to this question depends on how the four people cooperate. If it is a centralized system with a leader, such as Fat Tiger (equivalent to the central server), then their victory is a foregone conclusion, because it is very simple to reach an agreement on the attack time. All Fat Tiger needs to do is call Nobita, Suneo, and Shizuka to a meeting to discuss it. Even if everyone disagrees, Fat Tiger can make the final decision. Let's go back to the assumption of the Byzantine Generals' Problem. In a distributed network that is not based on trust, what are the chances of the four generals winning?

First, due to the lack of trust among the four generals, the possibility of gathering in the small dark room for a conspiracy meeting was ruled out (what if they were kidnapped by Fat Tiger in the small dark room?); secondly, since there was no leader, the opinions of the four people would be equally valued. In this case, the four generals could only agree on the time of attack by sending messages between their respective camps through messengers. For example, if Nobita thinks that 6 o'clock in the morning is a good time to launch an attack, he will send a messenger to tell Fat Tiger, Suneo and Shizuka his opinion. At the same time, Fat Tiger may think that it is better to launch a surprise attack at 9 o'clock in the evening, Suneo prefers to attack at 3 o'clock in the afternoon, and Shizuka hopes it is 10 o'clock in the morning. The three of them will also send their messengers at the same time. In this way, after the first round of communication, each of the four generals has four attack times to choose from, and they will each inform the other three of their selected time in the next round of communication. Since the decisions of the four people are made independently, there are 256 possible final choices. Only when more than three people happen to choose the same time, a consensus is reached, and there are only 64 such results, which means that the probability of reaching a consensus is only 1/4. This is only the case of four generals. What if the number of generals is 10, 100, or 1,000? A little calculation will show that as the number of generals increases, the hope of reaching a consensus will become increasingly slim.

Replace the generals in the above example with nodes in a computer network, the messengers with communications between nodes, and the attack time with information that needs to be agreed upon, and you can understand the dilemma described by the consensus problem. The ability to reach consensus is of great importance to a payment system. If you remitted money to your family to buy a car, and when you go to the bank to verify the next day, the counter tells you "There are three versions of records in our system about how much money you remitted", you obviously wouldn't dare to deposit money in such a bank. Before the emergence of Bitcoin, the consensus problem was difficult to be perfectly solved. To ensure consensus, a centralized system is needed (unless the nodes meet specific conditions), and decentralized consensus cannot be guaranteed. So how does blockchain technology solve this problem?

4. Introduction to blockchain technology

The solution of blockchain technology is actually very simple, which is to retain an arbitration right. All information determined by the holder of the arbitration right is unanimously recognized by everyone, so that consensus can be easily reached...

Wait! Smart readers may have already called the police after reading this. Isn’t “one person has the right to arbitrate” a centralized system? It is natural to have such doubts, because in our minds, “one person has the right to arbitrate” is one thing, but in fact it is not. It is just misleading to us by stereotyped thinking. In fact, it includes two things:

There is a person who has the right to arbitrate = There is a person + has the right to arbitrate

This seemingly redundant division actually contains the key idea to solve the consensus problem. The referee has the final say on fouls on the court, and the judge has the final say on wins and losses in court. Although we are accustomed to the arrangement that the arbitration right is monopolized by a certain person or organization, have you ever thought about this question: Who stipulates that the arbitration right must be tied to a certain individual? When you realize this problem, you are very close to Satoshi Nakamoto's answer. The solution provided by blockchain technology is very simple: retain the arbitration right, but do not bind it to a certain node. Instead, randomly assign the arbitration right to a certain node in the entire network every time the account is recorded, and then other nodes follow the node to complete the accounting task. In this way, it is easy to reach a consensus on the ledger, and because no one knows who has the final say on the next account, the decentralized characteristics of the system are also retained.

The above is a very brief introduction to blockchain technology. If your curiosity is still not satisfied, then we can take a look at how blockchain technology is implemented in the Bitcoin system. (If your curiosity has been satisfied, you can jump directly to the last paragraph of this section, which will not affect the understanding later.) In the Bitcoin system, many transactions are generated every moment. Once these transactions are generated, they are broadcast to the entire network to ensure that every node can receive them. However, they are not scattered in the system in a disorderly manner, but are packaged into a block (Block) every once in a while and recorded in the Bitcoin ledger. Connecting these blocks together in chronological order becomes a blockchain (Blockchain). In other words, the term blockchain actually has two meanings. It is a technology for reaching consensus in a distributed system, and in the Bitcoin system, it specifically refers to the ledger containing all transactions. The Bitcoin code is open source, and anyone can download and run the client on their computer, making their computer a node that maintains the Bitcoin network. Each node has a ledger containing all transactions, and regularly records new transactions in its own ledger. As mentioned before, every time accounting is done, there will be a node leading everyone to complete it together, so a key question arises: how is this node selected? How to ensure that the node will not forge or tamper with the transaction content?

This goal is achieved through mining. The above is a schematic diagram of the mining process. Each horizontal line in the figure represents a transaction, and the rectangular box represents a block containing many transactions, which together form the general ledger. The mining process is implemented through a function called SHA-256. Its function is to input any file and return an irregular 64-bit hexadecimal number. The dotted part is what the miner needs to input each time mining. It consists of three parts: information about the previous block, newly generated transactions, and a random number. After the input, SHA-256 will return a result. The first node that gets a small enough 64-bit number will broadcast its random number and SHA-256 result to the entire network for others to verify. At present, the requirement for "small enough" is that the first 17 digits are all 0. The difficulty will be dynamically adjusted with the computing power of the entire network to ensure that a new mine is generated approximately every 10 minutes. The verification process is that other miners input the random number they received into SHA-256 to see if they can get the same result. If the result is different, it means that the ledger of the miner who sent the broadcast is different from everyone else, and there is suspicion of privately forging and tampering with transactions, which leads to verification failure. This ensures that only honest nodes can mine new mines. The node that passes the verification becomes the owner of the new mine. It will package the new transaction as the SHA-256 input content into a new block and record it in the general ledger, and other nodes will do the same. At the same time, the new miner will receive two rewards. One is the handling fee for all transactions in the new block, which is generally 0.0001 bitcoin per transaction, regardless of the transaction amount; the other is the newly issued bitcoin, which is currently 25 bitcoins issued for each new mine, and the amount is halved every four years. After each mining is completed, all miners will participate in the next round of mining competition together. The probability of each person mining is proportional to the computing power they contribute to the Bitcoin system. The general ledger relies on such a sophisticated and fair mining mechanism to continue.

The above is how Bitcoin works. Even if you don’t understand some parts, it doesn’t matter. As long as you know that it can allow all nodes to save the same general ledger in a decentralized network without establishing trust, it will be fine. Bitcoin has been in existence for seven years. Although there have been constant voices declaring that Bitcoin has failed, according to a column called “Bitcoin Obituary”, Bitcoin has been declared dead 90 times ^[10] , but the entire system is still running well. Bitcoin and its underlying blockchain technology have withstood the test of time.

5. Blockchain and Bankers

After so much preparation, the bankers are finally here. Since Bitcoin is essentially a payment system, which is also one of the most important infrastructures of banks, it is natural that blockchain technology attracts the attention of bankers. So what does blockchain technology mean in the eyes of bankers?

Let's start with a simple inter-bank transfer. The process is usually as follows: the payer initiates a transfer request, the payer's bank accepts the request and submits it to the clearing agency for the corresponding business. The clearing agency is responsible for handling the transfer of funds between the payer's bank and the beneficiary's bank, and finally the beneficiary's bank adds the funds to the beneficiary's account. Of course, the actual operation is more complicated, this is just a simplified process to help understand.

Payer -> Payer's bank -> Clearing house -> Payee's bank -> Payee

In this process, the clearing agency in the middle only appeared later. In the early days, inter-bank transfers were achieved by opening accounts between banks. Recalling the previous comparison chart of centralized and distributed systems, the early model obviously belongs to the distributed system on the right. The biggest disadvantage of this model is that the accounting cost is very high. If each transfer between two banks needs to be "recorded in a separate account", then a payment system consisting of 100 banks will need to record 9,900 accounts. Each bank must frequently reconcile with all other banks to ensure the correctness of the accounts, especially in the era when transfers were delivered by postal mail. It is conceivable how much work is required to maintain such a system, so clearing agencies came into being.

The operation of the clearing house is very simple. Each bank opens an account with the clearing house, and then all transfers are made between the accounts of each bank by the clearing house. This greatly reduces the bookkeeping costs, because each bank only needs to reconcile with the clearing house. However, this arrangement also brings about a crucial change, that is, the payment system has changed from a distributed system to a centralized system. The clearing house has become the middleman for all transactions. When Bank A wants to transfer money to Bank B, it does not do it by itself, but relies on its trust in the clearing house to hand over the transfer to the clearing house to complete.

Why is this transformation so important? Because trust has a cost. When we swipe our cards to make purchases, we trust middlemen such as UnionPay, Visa, and MasterCard. The cost of trust is included in the various fees charged by the bank to us and by the middlemen to the merchants. When we transfer money to friends overseas, we trust middlemen such as Western Union and MoneyGram. The cost of trust is included in the cross-border remittance fees. Different payment services correspond to different middlemen, but the same thing is that you have to pay the cost for trust. So can blockchain technology reduce the cost of the entire system? The answer is yes, and different institutions use blockchain, and the effects are different.

For clearing institutions like UnionPay, the cost-saving effect is obvious. The performance of blockchain on Bitcoin has proved that it can provide a fully automatic and near-instant processing solution for complex payment systems, thereby significantly reducing the labor costs of clearing institutions and increasing the speed of account arrival.

For banks, the cost-saving effect is not limited to the above two points. Let's review the key problem that blockchain technology solves: reaching consensus on information in a distributed system that is not based on trust. In other words, with the help of blockchain, bankers can establish a reliable payment system without relying on trust in others, so what is the meaning of the existence of middlemen? So in the eyes of an ambitious banker, the ideal payment process is of course as follows:

Payer -> Payer's bank -> Payee's bank -> Payee

In addition to saving the cost of middlemen, blockchain technology can further reduce bookkeeping costs, because all banks record the same general ledger containing all transactions, which saves the trouble of reconciliation between each other.

By comparing the two, we can find that although they are on the road of exploring blockchain together, the destinations of clearing institutions and bankers are not the same. Obviously, the bankers' plan can reduce costs more significantly. Does this mean that if the bankers succeed, we ordinary people can also benefit from it? The conclusion is probably not so optimistic, and the reason is not difficult to understand: if a wolf and a tiger catch a sheep at the same time, the wolf eats less and the tiger eats more, and the sheep has nothing to do with it. As long as we are unable to bypass the bank to complete the payment ourselves, the wish for the bank to pass on the cost savings to us will only be a wish forever. So in the eyes of an ambitious depositor, the ideal payment process is of course as follows:

Payer -> Payee

There is no such thing as a bank or clearing house. This is exactly the significance of Bitcoin. For the first time, it technically gives everyone the ability to bypass banks and complete payments directly between payers and payees. When migrant workers in Mexico, India, Africa and other places send money home, remittance companies take 5% to 12% of the money. Even in the United States, credit card companies charge a fee of 1% to 2.5% for each transaction ^[11] . However, the handling fee for each Bitcoin transaction is only 0.0001 Bitcoin (less than 0.3 cents), regardless of the amount. Think about what these numbers mean to bankers. From this perspective, Bitcoin benefits far more than just its users.

6. Prospects of blockchain applications

Finally, let's talk about Nasdaq, which we mentioned at the beginning. If it is natural for bankers to pay attention to blockchain because Bitcoin itself is a payment system, then where does Nasdaq's enthusiasm for blockchain come from? This requires us to recall the definition of the Byzantine Generals' Problem: How to reach consensus on information in a distributed network that is not based on trust? The entire definition does not mention the transfer of money, but uses the word "information", which means that no matter what this "information" is, as long as it is in the conditions described by the consensus problem, blockchain technology is a potential solution. In Bitcoin, the information that needs to be agreed upon is the general ledger containing all transactions, which can also be understood as "a record of each transfer of Bitcoin ownership." If this information is replaced with "a record of each transfer of securities ownership," wouldn't it become the business of the stock exchange? When we can understand the consensus problem in this way, the appeal of blockchain technology to Nasdaq is self-evident. In fact, securities trading also has a structure similar to the payment process:

Buyer -> Buyer’s broker -> Stock exchange -> Seller’s broker -> Seller

In what ways will blockchain improve the efficiency of exchanges? What is the ideal trading process in the eyes of ambitious brokers and ambitious buyers and sellers? Compared with our previous analysis, everyone should have a clear idea of ​​these questions.

Of course, the information in the consensus problem is not limited to the transfer of ownership. In theory, all digitized content can be processed by blockchain, and many interesting blockchain applications have been developed, such as decentralized microblog Twister, decentralized chat software Bitmessage, decentralized cloud storage Storj, decentralized domain name system Dot-Bit, etc. Most of them have the distinctive mark of Bitcoin: lower cost, better security, information that cannot be tampered with, and tenacious vitality like a starfish...

That will fly, flies at last.

Reward address

https://zhuanlan.zhihu.com/p/20519827

References

Notes (↵ returns to text)

1. http://www.pbc.gov.cn/goutongjiaoliu/113456/113469/999049/index.html↵

2. http://www.pbc.gov.cn/goutongjiaoliu/113456/113469/3008070/index.html↵

3. http://www.coindesk.com/chain-issues-investor-shares-nasdaq-linq/↵

4. http://fintechnews.ch/blockchain_bitcoin/introduction-bitcoin-blockchain-for-fiancial-services/1968/↵

5. https://bitcoin.org/bitcoin.pdf↵

6. http://newlibertystandard.wikifoundry.com/page/2009+Exchange+Rate↵

7. http://nakamotoinstitute.org/crypto-anarchist-manifesto/↵

8. https://en.wikipedia.org/wiki/E-gold↵

9. http://www.8btc.com/bitcoin-and-the-byzantine-generals-problem↵

10. https://99bitcoins.com/bitcoinobituaries↵

11. https://www.washingtonpost.com/news/innovations/wp/2016/01/19/rip-bitcoin-its-time-to-move-on/↵