Long Baitao: The Nature, Risks, and New Opportunities of Cryptocurrency Exchanges from a Monetary and Financial Perspective

This article was published on August 7, 2019, on the Chain Catcher WeChat official account and organized by Gong Quanyu.

Why is it said that understanding exchanges requires first understanding banks? Where are the new opportunities for exchanges? What systemic risks do exchanges face? … Recently, Chain Catcher invited independent monetary finance researcher and Tsinghua University computer science PhD Long Baitao to give a thematic presentation on "The Essence, Risks, and New Opportunities of Exchanges from a Monetary Finance Perspective" at Catcher Academy.

Dr. Long Baitao graduated from Tsinghua University with a PhD in computer science, focusing on cryptocurrency technology and monetary finance theory, and is one of the authors of China's first book on Libra. As a serial entrepreneur, Dr. Long founded Zhixiang Technology and received 27.5 million RMB in investment, focusing on financial cloud computing, quantitative investment, and machine learning; he served as the Chief Technology Officer of China International Capital Corporation's Jiazhi Investment Fund; and held senior financial services consulting positions at Accenture and IBM, where he represented Accenture as the chief designer for the Shanghai Stock Exchange's next-generation trading system project.

Presentation by Long Baitao

Hello everyone, I am very pleased to have this opportunity to communicate and discuss with you. The topic I will share today is "The Essence, Risks, and New Opportunities of Cryptocurrency Exchanges from a Monetary Finance Perspective," which is divided into five parts:

First, we will review the theoretical foundations of monetary finance;

Second, we will look at the business model of exchanges from the perspective of monetary finance;

Third, we will analyze the systemic risks of exchanges from a financial perspective;

Fourth, we will explore new opportunities for exchanges in the future;

Fifth, we will discuss exchanges' customers, products, operations, and technology.

1. Review of Monetary Finance Theory

In the traditional financial system, the primary role of banks is as financial intermediaries, mainly absorbing deposits and issuing loans. The so-called "short-term deposits and long-term loans" means that most of the deposits absorbed are short-term demand deposits, while the loans issued have longer terms, so banks undertake the functions of term transformation and risk transformation.

From a legal perspective, once depositors deposit funds into a bank, they no longer own those funds; deposits are merely a form of debt acknowledgment from the bank to the depositor (I owe you -- IOU). All funds are aggregated by the bank to form a pool of funds for further business operations. Banks are the only institutions legally permitted to conduct pooled fund operations.

The second role of banks is to create money through loans. Generally, it is believed that banks must first absorb deposits before issuing loans. In reality, banks do not need to have deposits beforehand to issue loans; they create money at the same time they issue loans. For example, if you take out a loan of one million from a bank to buy a house, the bank creates a deposit account for you on the liability side of its balance sheet, directly writing one million in that account, allowing you to spend that amount.

In this process, the bank does not need to transfer money from anywhere else; it can be said that this one million is created out of thin air. At the same time, the bank will create a loan asset of one million on the asset side of its balance sheet, which the borrower will repay according to the agreed terms and interest rate. The bank destroys the money when the loan is repaid. Banks also create/destroy money through the purchase/sale of assets.

Thus, when banks issue loans, they simultaneously create one million in assets and liabilities on both sides of their balance sheet, keeping the balance sheet balanced.

In China, the central bank mainly constrains commercial banks' ability to create money through the reserve requirement system. Assuming a reserve requirement ratio of 10%, if a bank has ten million in reserves at the central bank, it can create up to one hundred million in deposit money, or issue up to one hundred million in loans. The central bank also requires banks to meet regulatory indicators established by Basel III, such as capital adequacy ratio, leverage ratio, and liquidity ratio, to ensure banks' ability to absorb losses and constrain their ability to create money or loans.

In fact, many countries in Europe and America no longer use the concept of reserve requirements but follow the regulatory standards of Basel III, such as the Bank of England and the Swiss National Bank.

For commercial banks, the main risks at the business level can be divided into two types: one is liquidity risk, and the other is insolvency risk, also known as repayment risk. Liquidity risk means that the liquid assets a bank holds are insufficient to meet the demand for cash withdrawals from depositors, which primarily manifests as bank runs. The banking system's methods for resisting risks mainly include regulation (macro-prudential and micro-prudential), deposit insurance schemes, central bank lender of last resort obligations (providing emergency liquidity to banks), and the central bank using government funds to rescue banks.

Although banks do not need to have money on hand to issue loans, they still primarily rely on attracting deposits to reduce compliance costs. Banks create money through loans and earn profits mainly from the interest rate spread between deposits and loans. The liability side of banks mainly consists of deposits, while the asset side includes loans and proprietary trading. Banks prefer mortgage lending, especially loans secured by real estate and financial assets, because the risk assessment models for such secured loans are standardized.

Banks are less inclined to offer commercial loans to startups due to high risks, lack of collateral, and insufficient risk assessment capabilities. Large banks increasingly prefer proprietary speculation, especially in Europe and America, as they can create and use money at a very low cost.

From a monetary perspective, the business model of exchanges must be understood as fundamentally banking; in other words, the primary profit-making ability of exchanges comes from the power to create money out of thin air, just like banks.

2. Business Model of Exchanges

One of the most important businesses of exchanges is attracting investors to trade; investors depositing funds essentially serves as zero-interest deposit-taking for the exchange. Commercial banks also like to attract deposits because it is the lowest-cost way to obtain funds, but there is still a cost associated with attracting deposits for commercial banks, while exchanges incur no costs for deposit-taking.

Investors' deposits form the liability side for the exchange, which is the foundation of the exchange's asset pool. Exchanges do not set up independent custody accounts for each user like traditional exchanges; instead, they create virtual accounts for them, where the account merely represents a number, indicating the debt the exchange owes to the investors.

Once the exchange has a liability side, it can develop asset-side businesses, such as providing financing (stablecoins) and securities lending (other digital assets) services for investors, corresponding to the financing and securities lending businesses of traditional brokerages or exchanges. The daily interest rate for mainstream exchanges' financing is about 1%, which can no longer be described as usury, and most exchanges can achieve up to five times leverage.

The financing services provided by traditional brokerages are typical financing businesses, where the funds lent come from real funds held in bank custody accounts. However, cryptocurrency exchanges do not have stablecoins and digital assets pre-existing for financing and securities lending; they create these funds out of thin air when needed to lend to users at no cost.

On the surface, this business seems to pose little risk to the exchange because the borrower has collateralized their assets in the exchange. However, it should be noted that the numbers in the exchange accounts no longer represent actual assets; they are merely IOUs from the exchange to the investors. Therefore, while it appears that investors are using their account assets as collateral to borrow funds from the exchange, in reality, they are pledging the exchange's IOUs back to the exchange and borrowing funds that have been created out of thin air.

This scheme may seem familiar; indeed, this is how the US dollar operates. For example, when the US economy needs more dollars but the Federal Reserve cannot conveniently issue more currency directly, the US Treasury issues ten billion dollars in government bonds, and global investors exchange ten billion dollars in currency for ten billion dollars in bonds. The Federal Reserve then purchases these ten billion dollars in bonds from the market, thereby increasing the money supply by ten billion dollars.

As a result, the entire economy has increased its circulating money supply by ten billion dollars, but in reality, nothing has changed in the world; no actual products or services have been produced, and no new value has been created. The dollar system can operate this way because the United States, as the world's leading economy, has sovereign credibility to back it up, and in dire situations, it can rely on its powerful military to reinforce this guarantee. But what gives cryptocurrency exchanges the right to operate in this manner?

Other asset-side businesses of exchanges include proprietary trading, where they take these coins from the fund pool and engage in market manipulation and speculation on other exchanges or their own. Other asset-side businesses also include project investments made by cryptocurrency exchanges. In recent years, almost all digital asset exchanges have established large investment ecosystems, many of which include numerous vaporware projects that have gone to zero, resulting in significant losses from investments in vaporware. Who bears these losses? The fund pool.

In fact, traditional banks' asset-side businesses (mainly lending) also encounter bad debts. Basel III has made very clear arrangements requiring banks to meet multi-tiered capital adequacy requirements to enhance their ability to absorb these losses and their self-rescue capabilities in the face of insolvency risks. However, the cryptocurrency exchange industry clearly lacks similar arrangements.

What is even more exaggerated in cryptocurrency exchanges is that they directly write the amount of assets in their accounts to participate in trading; for example, many exchanges' so-called "forced listing" operates this way. Without evaluating the morality of such behavior, from the essence of banking operations, this is equivalent to the exchange (as a bank) lending to itself.

As mentioned earlier, banks create money out of thin air through lending. Therefore, understanding exchanges as banks means that this bank can lend to itself, giving it the magical ability to create money at will. Traditional banking regulation has long understood this risk, which is why global banking regulators strictly prohibit such self-lending behavior.

However, there have been exceptions, such as during the 2008 global financial crisis, when Barclays Bank in the UK and UBS in Switzerland created billions of dollars through such self-lending behavior to lend to their so-called investors—a Middle Eastern consortium from Qatar, which injected the borrowed funds as capital into Barclays and UBS, pulling them out of the financial crisis. The details of these two investments have not been disclosed due to their non-compliance and opaque operations.

Although the trading services and OTC services of exchanges remain the main sources of income, they are essentially just a packaging of a banking core business model. If exchanges cannot create a large amount of assets out of thin air, the income from these two businesses cannot reach its current scale.

Although there is a lack of actual operational data from exchanges, various channels indicate that the direct income from most exchanges' fund pool operations has already exceeded half of their total income. Considering that the substantial income from trading fees and OTC service fees of exchanges essentially comes from their magical ability to create assets out of thin air, it is reasonable to view the essence of exchanges as banks.

3. Systemic Risks of Exchanges

Since the magical ability to create money/assets out of thin air is so useful, where do the risks for exchanges come from? They stem from user withdrawals; if a large-scale user withdrawal occurs, it constitutes a bank run for the exchange, as the exchange has leveraged the digital assets in the fund pool to generate a large amount of assets. Whether through lending, proprietary speculation, or investment, exchanges generally move most of the coins from the asset pool, creating significant leverage. Platform tokens are similar; the vast majority of platform tokens or stablecoins issued by exchanges also repurpose the fund pool, but the asset side still corresponds to the original limited fund pool.

According to traditional banking data, exchanges have a very low core capital ratio but a very high leverage ratio. Before the financial crisis, traditional large banks in Europe and America generally had leverage ratios exceeding 50 times, even under constraints. For exchanges that currently operate without constraints and transparency, who regulates the leverage ratio and core capital ratio of exchanges? This is the most important indicator for managing systemic risks in exchanges. In the absence of regulation and industry self-discipline, the only thing that can constrain the leverage ratio of exchanges is the internal fear of the exchange founders.

The above discussion assumes that exchanges are not cheating and are conducting normal banking operations. If exchanges cheat, maliciously transfer, or hollow out assets, that is purely a scam. Such incidents have historically occurred in traditional banking, where commercial banks absorbed large deposits through high-interest rates and then absconded. Similar events have already occurred in the cryptocurrency finance sector and will continue to happen.

In traditional banking, commercial banks have the backing of central banks, deposit protection plans, and the central bank as a lender of last resort to provide liquidity support. When commercial banks face insolvency, the central bank can even use government funds for rescue. However, in the world of cryptocurrency finance, when an exchange is facing a bank run, it can only pray to God.

At this point, the exchange either goes bankrupt or absconds, which is more common among smaller exchanges, though large exchanges have not been immune either.

Recently, Binance was hacked, losing seven thousand bitcoins. Whether Binance was truly hacked is debatable; it is also possible that the market turned bullish recently, with bitcoin prices rising sharply, prompting many large holders to withdraw their coins. Seeing a significant demand for withdrawals, but not having enough coins on hand, the best excuse for the exchange might be to claim it was hacked. Ironically, cryptocurrency exchanges, regardless of size, frequently experience various security incidents each year, yet they confidently claim to be financial-grade exchanges without a hint of embarrassment.

As the former chief designer of the Shanghai Stock Exchange trading system, I can say that the code I wrote ten years ago still supports over half of China's daily securities trading volume today, without any incidents occurring. It seems that my understanding of "financial-grade" differs significantly from that of the new cryptocurrency finance elites.

Compared to traditional banks, where bank runs incur many costs due to the high costs of printing, distributing, and safeguarding paper currency, the speed and scale of bank runs in traditional banks are actually limited. However, for cryptocurrency digital assets, withdrawing coins incurs no friction costs because everything is digital, so a bank run can happen instantaneously and on a large scale. Any exchange fears a bank run, and currently, no exchange can withstand one.

At the same time, we must also recognize that creating a bank run does not require much cost because a bank run is more about public sentiment. As long as users perceive that a particular exchange is at risk, a bank run will occur. Therefore, creating a bank run does not require hackers; hiring online trolls will suffice. Trolls are much cheaper than engaging in black market activities, right?

Currently, no one is attempting to run on exchanges because there is no benefit to them; all exchanges operate under similar models, with very low capital adequacy ratios paired with extremely high leverage, posing significant systemic risks. Any attempt to attack a competitor through a bank run would be detrimental to oneself, akin to mutually assured destruction in the US-Soviet rivalry.

However, if in the future, there are exchange players whose business models do not rely on leverage, what would be their best competitive strategy? The answer is obvious—create a bank run.

If a bank run occurs across the entire market, resulting in indiscriminate runs on all exchanges, including one's own, the outcome for an exchange without leverage would merely be a loss of customers and assets, with a shrinking balance sheet, but the core capital would not be lost, so it would not go bankrupt. For exchanges with leverage, the outcome would not only be a loss of customers and assets but likely bankruptcy or absconding, so a market-wide bank run would ultimately leave only a few exchanges that do not rely on leverage or have very low leverage ratios to survive.

4. New Opportunities for Exchanges

The future opportunities for exchanges will certainly lie in stablecoins. Many exchanges want to develop stablecoins, but the development of stablecoins requires universal infrastructure, such as decentralized asset custody and clearing and settlement, as well as collateral management frameworks.

The collateral management framework includes a directory of eligible collateral and collateral risk management. The former sets certain qualification criteria for reserves or collateral required for currency issuance, such as total market value, concentration, and risk measurement, while the latter calculates and manages the volatility risks of specific assets (including market risk and liquidity risk) and determines collateral discount rates. These business capabilities can be viewed as the universal financial infrastructure for cryptocurrencies.

The trend for future stablecoin issuance is to be fully reserved, such as Libra, which is a fully reserved bank, while Binance's stablecoin operates on a partially reserved model. In addition to supporting mainstream digital assets like Bitcoin and Ethereum, the management of reserves must also consider whether typical fiat assets can be included, such as sovereign bonds denominated in RMB or USD, and the methods of generating interest for each type of reserve may also differ, presenting new challenges.

In addition, the supply and demand regulation of stablecoins, such as issuance and redemption, stablecoin monetary policy (quantity and price control), stablecoin value stabilization mechanisms, and the generation and distribution of seigniorage are all institutional aspects related to stablecoins.

Exchanges can build part of the infrastructure and systems related to stablecoins through their capabilities.

We can understand the relationship between stablecoins and exchanges from three levels.

First, exchanges are a huge asset entry point, so they can become one of the most important issuers of stablecoins in the future. However, when acting as issuers providing reserves, exchanges face competition from wallets and asset management institutions, which can influence the scale of assets comparable to that of exchanges. The earnings of issuers naturally come from sharing seigniorage.

Second, exchanges are important usage scenarios for stablecoins. The main use cases for stablecoins are payments, exchange mediums, and value storage. Exchanges primarily serve the latter two purposes. Exchanges can be seen as the largest channel for stablecoins. The distribution of seigniorage from stablecoins will ultimately have a mechanism for channel distribution, so exchanges, as the largest channel providers for stablecoins, can participate in the distribution of seigniorage.

Third, as issuers, exchanges contribute reserves to participate in the issuance of stablecoins, and the usage scenario for these stablecoins is the global network of that stablecoin. From this perspective, the entire stablecoin network becomes an extension of individual exchanges, and exchanges and the stablecoin system are actually mutually channeling.

Exchanges all want to create stablecoins, but out of hundreds of stablecoins, only one or two may ultimately survive, as the stablecoin system is very complex. Many people do not fully understand what work is required to build a complete monetary finance system.

I anticipate that the vast majority of exchanges will not create their own stablecoins. Many people mistakenly believe that simply collateralizing fiat currency at a bank to issue stablecoins constitutes building a complete monetary finance system, reflecting ignorance and arrogance regarding monetary finance theory and practice. Therefore, it is essential to enhance the understanding of financial theory in the blockchain and cryptocurrency fields, respect the objective laws of finance, and promote the development of blockchain and digital currency businesses with a scientific attitude.

5. Common Issues with Exchanges

Finally, I will briefly discuss issues related to exchanges' customers, products, operations, and technology.

From a product perspective, I believe that cryptocurrency trading will gradually disappear in the future because the emergence of cryptocurrency trading has historical reasons, and the main justification for its existence now is to contribute excess fee income to exchanges. If exchanges improve their asset risk management levels (market risk and liquidity risk—the main management goals of collateral management frameworks), then any digital asset can be exchanged in real-time for fiat-valued assets, eliminating the need for cryptocurrency trading.

Another reason for the popularity of cryptocurrency trading in the past was that people still harbored some fantasies about mainstream cryptocurrencies becoming standard payment methods, such as hoping Bitcoin could become a universal payment tool, but attempts to make Bitcoin a universal payment tool have proven to be failures. I still maintain the view that the only possible universal payment tools are fiat currencies or stablecoins; all other cryptocurrencies will ultimately only present as assets.

Additionally, there is a need to enhance the professionalism of derivative products; currently, mainstream exchanges are very lacking in the professionalism of derivative products. There is also a need to reduce fees, as they are still quite high.

From an operational perspective, exchanges must change their traffic-first mindset, as the ultimate goal of a traffic-driven approach is to exploit users. If you want to build a good exchange, you must change this traffic-first mindset in the future. The reason people still talk about traffic is that institutions, or "Old Money," have not yet entered the market on a large scale, but the trend is becoming increasingly evident, so exchanges need to be prepared.

Many people involved in exchanges have a misconception that running an exchange is technically simple. I can directly state that the security, stability, reliability, and professionalism of all exchanges do not meet the requirements of financial-grade exchanges. Fortunately, the problem is not severe because user demand is still at a relatively low level. However, if we look to the future, if you want to enhance your competitive ability, you must invest more effort in technology to truly achieve a financial-grade exchange.

Regarding the business model of exchanges, I will briefly mention two points. First, reduce leverage or eliminate it entirely. As mentioned earlier, for exchange players without leverage, the best competitive strategy is to let other exchanges be bankrupted; widespread panic in the market will destroy them, and you will survive.

Second, the business model of exchanges must truly achieve community involvement, but the premise is that the business model, governance structure, and operational methods of the exchange must be completely transparent, genuinely informing users/investors about how profits are made.

If you cannot even be transparent about how profits are made, how can you talk about achieving a community-oriented business model? Currently, all exchanges rely on directly creating money out of thin air, earning interest spreads, and manipulating markets to make big profits, yet they can only claim to earn from trading fees and listing fees, which currently account for less than half of their total income.

Many people view brokerage and similar businesses as the trend for the development of exchanges, but this is merely superficial; it is just revolving around the traffic of exchanges without changing the essence of exchanges.

Thank you all for listening. My presentation ends here, and I look forward to more exchanges in the future.