Note: This article is a dialogue between Binance founder Zhao Changpeng and the Vice President of the University of Hong Kong at the 'Hong Kong Crypto Finance Forum' organized by MetaEra on August 27.
10 key quotes at a glance:
1. The current US government is very smart and deeply understands the strategic value of Tether to the global position of the US dollar. Americans do not actually need stablecoins. Almost all USDT users are outside the US, which effectively expands the global influence of the US dollar.
2. Stablecoins are essentially tools to help underlying currencies achieve globalization. Currently, the several countries I have interacted with hope their fiat currencies can go on-chain. I believe every country should at least have a few stablecoin products.
3. As the world's largest stock market, the US attracts global investors to purchase US stocks through blockchain technology, which is extremely beneficial for its economic development. If US stocks can also be successfully put on-chain, it will further consolidate the US's dominant position in the global financial market.
4. If other countries do not participate in the RWA transformation, they will face the risk of being marginalized and eliminated. Just as the absence of Alibaba in China could lead to the e-commerce market being completely dominated by Amazon, the absence in the fintech field will also have far-reaching economic impacts.
5. Exchanges should not impose restrictions on tradable assets; all assets should be able to circulate freely on the same platform. Once assets are on-chain, they are just a token, with no substantial difference. The real difference lies in compliance: which regulatory authority needs to be applied for a license, and whether approval can be obtained.
6. Regulatory authorities require all operations to be conducted locally, which is not very meaningful in the digital currency industry. If only local residents are allowed to trade, it is impossible to generate sufficient trading volume. Without liquidity, price fluctuations will be very severe, which is actually harmful to users. Real user protection comes from a sufficiently deep order pool.
7. From what I understand, most licensed exchanges in Hong Kong are currently operating at a loss, and although they can maintain this in the short term, such a loss-making model is difficult to sustain long-term.
8. In 10 to 20 years, the scale of decentralized exchanges will definitely exceed that of centralized exchanges; this is the future trend.
9. The core logic of DAT is to package digital currencies in a stock-like manner, allowing traditional investors to gain exposure to digital currencies. This group is actually a very large market, much larger than the cryptocurrency circle.
10. The integration of AI and Web 3.0 is by no means just a conceptual hype, but rather a trend that will inevitably lead to breakthrough developments in the future. What currency will AI use? The answer is clearly not the US dollar or traditional payment systems. The currency system of AI will inevitably be based on digital currencies and blockchain. This means that the transaction volume of blockchain will experience exponential growth. In the future, everyone may have hundreds or thousands of AI agents.
The full text is as follows:
1. Discussing stablecoins: From the birth of USDT to the US stablecoin bill
In fact, I am not an expert in the stablecoin field, but the Binance platform accounts for about 70% of the global stablecoin trading volume, making us a key distribution channel for stablecoins in the industry.
Let me briefly introduce the development history of stablecoins. The earliest stablecoin technology prototype was Colored Coins, which was the first asset on-chain solution explored by the Bitcoin community. In 2014, USDT was initiated by Brock Pierce. The project initially developed slowly, and then Pierce gradually withdrew, making way for the current USDT team led by Craig Sellars, and by 2017 there was still not much progress.
When Binance was founded in 2017, we focused on cryptocurrency trading, supporting trading pairs like Bitcoin against Ethereum and BNB, but lacked fiat trading functionality. This created a user experience problem: whenever the price of Bitcoin fell, users could only withdraw Bitcoin to other fiat exchanges to convert it to fiat, and there was significant uncertainty about whether those funds would flow back to our platform.
At the same time, this is also extremely unfriendly to user experience. To improve user experience, we decided to support USDT as a 'safe haven' during market downturns. At that time, we understood stablecoins as a short-term storage tool, so the decision to support USDT was relatively simple — there was no signing of complicated cooperation agreements, nor was it a strategic partnership, it was simply the integration of the product.
This is when USDT experienced its rapid development phase:
First, after 2017, cryptocurrency exchanges entered a rapid development phase, and many platforms, including Binance, began to support USDT, driving rapid growth of USDT.
Subsequently, USDT welcomed its second wave of growth: many Asian users have a demand for US dollars, but directly opening US dollar accounts is challenging; USDT provides them with an alternative. Tether's profitability has always been outstanding, and due to regulatory pressure from the US and difficulties in bank cooperation, they have remained relatively low-profile.
In 2019, the US compliance agency Paxos proactively contacted us and proposed a partnership to issue stablecoins, leading to the later BUSD. From 2019 to 2023, BUSD's market value has grown to 23 billion US dollars, during which we invested relatively few resources, primarily engaging in some brand support and promotional activities, such as the 'fee-free withdrawal' campaign.
In 2023, the US government phased out the BUSD project. If BUSD had continued, it would have achieved a good scale of development, as BUSD's growth rate at that time surpassed that of USDT and USDC. It is worth emphasizing that when the BUSD project was shut down, all users' funds were fully refunded, which fully demonstrates BUSD's characteristics as a compliant, transparent, and secure project.
Stablecoins and exchanges have become one of the core profit sectors in the field of crypto finance. The business model of stablecoins is highly simplified: after obtaining a compliance license, users deposit funds, and the platform can issue tokens; when users redeem tokens, the platform provides cash exchange. This model features low thresholds, high liquidity, and huge market potential, with significant long-term profitability.
From a national strategic level, the US government's attitude towards stablecoins has undergone significant changes in recent years. The current US government is very smart, leveraging its commercial background to deeply understand the strategic value of Tether to the global position of the US dollar. Currently, about 100 billion USDT funds are purchasing US Treasury bonds, and Tether is widely used globally. The key is that Americans do not actually need stablecoins — they can directly use the bank ACH system for US dollar transactions. Almost all USDT users are outside the US, which effectively expands the global influence of the US dollar.
This aligns closely with China's desire to expand the influence of the renminbi internationally. Stablecoins are essentially tools to help underlying currencies achieve globalization, which should be immensely attractive to countries. Of course, as freely circulating blockchain assets, stablecoins indeed pose challenges to foreign exchange controls, but these issues can also be resolved. Currently, the several countries I have interacted with have shown strong interest in developing local stablecoins, hoping their fiat currencies can go on-chain.
In July, when the US passed the (GENIUS Act), it proposed a policy direction to limit the development of central bank digital currencies (CBDCs), reflecting a profound strategic layout regarding the global dominance of the US dollar. The popularity of stablecoins is precisely due to their high liquidity and good user experience, while some government-led digital currencies may be subject to stricter regulation and monitoring, which may negatively impact market acceptance. In fact, since 2014, more than 20 countries have attempted to issue CBDCs, but none have truly achieved success on the market level.
Blockchain technology is essentially a ledger technology, with its first application scenario being finance, making stablecoins a natural application of blockchain technology. Currently, we only see the US dollar stablecoin developing relatively maturely, while stablecoins for other national currencies have yet to rise, indicating that there is enormous growth potential in this track. Now, every country wants to develop stablecoin businesses. I believe every country should at least have a few stablecoin products.
2. Discussing RWA: Challenges and far-reaching implications
Despite the broad market prospects of the RWA track, its implementation difficulty is much higher than market expectations. The specific challenges can be summarized in the following three aspects:
1. Liquidity dilemma
From a practical perspective, products with strong financial attributes are relatively easier to tokenize, mainly because traditional financial products themselves have high trading attributes and a relatively mature digital expression. However, the tokenization of non-financial assets faces fundamental obstacles. Although theoretically everything can be tokenized — all cities, buildings, and individuals can issue coins — in practice, there are numerous problems.
Taking real estate as an example, even in the highly volatile Hong Kong real estate market, compared to Bitcoin, the volatility is still very small. After issuing coins for such low-volatility assets, because the fluctuations are not large, the trading nature is weak, and the order book depth is insufficient. At this time, liquidity will decrease, investors will not place many orders, thus forming a vicious cycle: the order book is too shallow, leading to lower trading volume. If investors try to move funds in the hundreds of millions, it is almost impossible to execute; even if assets are on-chain, liquidity is still insufficient, making it easier to trigger unexpected fluctuations and even be manipulated in the short term.
2. Regulatory complexity
Products with financial attributes often involve a core question — is it a security? Is it a security, a commodity, or something else?
In major countries or finance-developed nations, there are clear definitions and different regulatory bodies; in some small countries, there might be just one regulatory body managing all matters. If different regulatory bodies are involved, compliance terms can become quite complex. Companies need to apply for different licenses: futures licenses, spot licenses, digital currency licenses, banking trust licenses, etc. When too many licenses are obtained, the business model can be quite restricted, and often a single business cannot even get off the ground.
3. Product mechanism defects
In my opinion, the tokenization of US securities is not yet established at the product level. The stock tokenization products we currently see, such as xStocks, do not have their token prices linked to the actual stock prices, which is unreasonable. Theoretically, if there is a price difference between the two, investors can profit from arbitrage. But the reality is that this price difference has always existed — indicating that the product's mechanism is not functioning. In other words, in the current stock tokenization track, there is no real linkage between tokens and stocks, so the entire model does not stand at the product level. Although the US is attempting various tokenization methods, it has yet to find a truly feasible solution.
4. Not doing it will lead to elimination
Despite these challenges, there is still a truly viable RWA model — stablecoins. The underlying assets of stablecoins are mainly traditional financial instruments such as US Treasury bonds, and the success of this model has verified the feasibility of financial asset tokenization.
The US dollar has already been realized on-chain through stablecoins. In the current blockchain ecosystem, almost all assets are priced in US dollars, while the euro and renminbi are mostly absent in this field. As the world's largest stock market, the US attracts global investors to purchase US stocks through blockchain technology, which is extremely beneficial for its economic development. If US stocks can also be successfully put on-chain, it will further consolidate the US's dominant position in the global financial market.
From a rational perspective, the US should actively support this development direction; while other countries that do not participate in this transformation may also face the risk of being marginalized. For instance, the Hong Kong Stock Exchange, as an important exchange with global influence, may gradually lose its influence if it is absent in this round of transformation. Other Asian exchanges like the Shanghai Stock Exchange also face the same strategic choices.
From an economic perspective, this is something that needs to be done 100%. Not doing it will lead to elimination. Just as the absence of Alibaba in China could lead to the e-commerce market being completely dominated by Amazon, the absence in the fintech field will also have far-reaching economic impacts.
Despite regulatory challenges, this trend has extremely far-reaching implications for the economy, and all countries should seriously consider relevant layouts. With the wisdom and innovation capabilities of Asians, these issues will ultimately be resolved, and one of the keys lies in seizing the moment.
For businesses and entrepreneurs, it is essential to accurately grasp the rhythm during market windows: entering too early may face survival pressure, while entering too late may miss the opportunity.
Right now, we are in a rare golden window. The US policy shows unprecedented support for virtual currencies, which will inevitably encourage other countries hoping to develop their economies to take corresponding actions. Hong Kong, as a long-standing financial center in Asia, coupled with the supportive attitude of the Hong Kong government, presents a rare historical opportunity. Therefore, everyone should seize this strategic opportunity fully.
3. Discussing exchanges: How can Hong Kong build a world-class exchange?
1. Exchanges should not impose restrictions.
I believe exchanges should not impose restrictions on tradable assets; all assets should be able to circulate freely on the same platform.
All assets, once on-chain, are just a token, whether it is a native crypto asset or a real-world asset (RWA), there is no substantial difference from the technical perspective of exchanges. Adding a new asset category typically does not require complex development, as long as it is supported on the existing chain. Currently, most RWA projects do not require independent blockchains and are more likely to issue tokens based on public chains like Ethereum, BNB, or Solana, thus the support difficulty at the wallet and exchange levels is very low. The real difference lies in compliance: which regulatory authority you need to apply for a license from and whether you can get approval. Once the license issues are resolved, there are almost no technical barriers.
In the long run, future exchanges should achieve unified trading of various global assets, whether it is a building, future IP rights of a celebrity, or even personal valuations, all could circulate in the same market. This not only maximizes liquidity but also makes the price discovery mechanism more efficient.
Of course, RWA also faces some unique challenges. For example, when you tokenize a building, if you later want to sell that building, you may only be able to sell a portion of it. Because once tokens are issued, if an investor holds just one unit of the asset and refuses to sell, you cannot fully buy back the entire building or it could incur huge costs. It can be understood as the concept of 'nail households' on-chain.
Although the realization of 'global asset on-chain' still requires time, it is not out of reach for 90% of the countries globally. Compared to some countries with extremely complex regulatory systems, many countries are more likely to directly adopt unified international standards, thus taking the lead in promoting global asset on-chain and free circulation.
2. How can Hong Kong build a world-class exchange?
When discussing how Hong Kong can build a world-class exchange, I can analyze from a logical perspective.
Many countries or regions, in the early stages of cryptocurrency industry regulation, often choose strict control to reduce risks and ensure safety. Regulatory authorities worry about making mistakes, so they typically require all operations to be conducted locally: local licenses, local offices, local employees, local compliance departments, local servers, local data storage, local matching engines, local user bases, and completely independent local wallet infrastructure, etc.
This idea is relatively easy to implement in the traditional physical world, such as controlling through safes and physical isolation. However, in the digital currency industry, this distinction is not significant. Regardless of whether servers are located in Hong Kong, Singapore, or the US, the likelihood of being hacked is the same, as everything operates online.
More importantly, if operations need to be split, just building a secure wallet infrastructure often requires an investment of around 1 billion dollars. Moreover, the issue is not just about funds, but about the shortage of talent — it is very difficult to repeatedly recruit hundreds of top security experts globally to build this infrastructure. The cost of replicating a complete system is essentially equivalent to the cost of establishing a top-tier international exchange.
From a liquidity perspective, if only local residents are allowed to trade, taking Hong Kong as an example, with a population of 8 million, or a small country with an active user base of 200,000 to 300,000, it is impossible to generate sufficient trading volume. Without liquidity, price fluctuations will be very severe, which is actually harmful to users.
Real user protection comes from a sufficiently deep order pool — when there are large orders worth hundreds of millions, prices will not be broken, and when futures prices fluctuate, there is no need for forced liquidation due to sufficient market liquidity. Buying 10 bitcoins on an exchange with low liquidity incurs significant slippage, and users bear higher costs as a result. Therefore, large global exchanges can provide the most basic user protection — lowering users' trading costs.
When countries attempt to establish independent systems, it will inevitably lead to complex management challenges, which are not feasible from a business perspective. At the same time, many countries impose restrictions on tradable assets, for example, Hong Kong currently has many restrictions on listed currencies, and the product coverage is limited. From what I understand, most licensed exchanges in Hong Kong are currently operating at a loss, and although they can maintain this in the short term, such a loss-making model is difficult to sustain long-term.
However, Hong Kong also has its strengths — the speed of improvement is very fast. We saw Hong Kong launch a new stablecoin draft in May, even ahead of the US. The government is very proactive in communicating with industry participants, including engaging in dialogue with us insiders. While Hong Kong may have been relatively conservative in previous years, this is completely understandable; with changes in the global situation, Hong Kong is now showing a very proactive stance.
I believe now is a great starting point. Past limitations do not mean the future will remain limited; on the contrary, it is an excellent opportunity for exploration. This is precisely why many Web 3.0 practitioners, including myself, choose to explore opportunities in Hong Kong.
4. Discussing DEX
I believe that in the future, decentralized exchanges will definitely be larger than centralized exchanges. Although Binance may currently be quite large, I do not think it will always maintain the largest position.
Decentralized exchanges currently have no KYC requirements, making them very convenient and quick for users who know how to use wallets, and they offer high transparency — although sometimes it can be overly transparent, as everyone can see the orders of others.
From a regulatory perspective, we have paid a high price due to our insufficient KYC work in centralized exchanges. However, the US currently seems to have few regulatory measures for DeFi, which may bring regulatory dividends to DeFi. However, due to historical reasons, I personally find it difficult to attempt this field again.
From the perspective of user experience, decentralized exchanges offer a decent user experience, but users need to understand how to use wallets. In fact, those who used centralized exchanges in the past are well aware that the user experience is not ideal. The interface is filled with addresses, contracts, and other numbers and 'garbled text', and the operation process often requires frequent checking of block explorers, as well as guarding against various detailed issues such as MEV attacks. I myself have encountered attacks multiple times while learning.
Therefore, for users who have just transitioned from Web 2.0 to Web 3.0, most will still choose centralized exchanges because the login method of email and password and the model with customer service support make them feel more comfortable. However, as time goes on, when some users become familiar with wallets, they may turn to decentralized exchanges. Currently, the fees for decentralized exchanges are actually higher than those for centralized exchanges, but in the long run, with technological progress, the costs of decentralized exchanges should become cheaper.
Currently, many decentralized exchanges have their own token incentive mechanisms, issuing tokens for incentives. However, this incentive will eventually disappear, as tokens cannot be issued indefinitely — unlimited issuance will lead to a decrease in the current token price.
Therefore, the current market is still in a relatively early stage, and there are still these token incentives. However, in the long run, I believe that in 5 to 10 years, decentralized exchanges will become very large. I believe that in 10 to 20 years, the scale of decentralized exchanges will definitely exceed that of centralized exchanges; this is the future trend.
Although I will not lead related projects again, from an investment perspective, we have invested in many similar projects, but only small shares, and we are now providing support behind the scenes. I believe there is still considerable development space in this field in the future.
5. Discussing Digital Asset Treasury: Traditional investors gain exposure to crypto assets
Many people often understand the concept of DAT (Digital Asset Treasury) too simply, but in fact, this track is highly segmented. Ultimately, its core logic is to package digital currencies in a way that is akin to stocks, allowing traditional investors to conveniently participate in investments.
The DAT field has various layers and forms, much like traditional companies, where various models can coexist. Crypto ETFs are primarily issued in the US, but many investors lack US stock accounts or do not wish to bear their high trading and management costs. In contrast, listed companies like Strategy can often achieve asset allocation at lower costs by directly holding digital currencies. Additionally, their financing methods are more diverse, capable of raising funds in different markets such as the US, Hong Kong, and Japan. The differences in financing channels and investor structures among listed companies in different regions also shape their unique market patterns.
In the listed company model, DAT companies mainly have the following operational models:
1. Passive single asset holding model
Taking Strategy as a representative, focusing on passive holding of a single asset, Bitcoin. This model is relatively simple, with lower management and decision-making costs, allowing for adherence to established strategies regardless of Bitcoin's price fluctuations.
2. Active single asset trading model
Although only holding one type of coin, the management strategies are completely different. These companies will try to judge rises and falls for active trading, which requires assessing the trading capabilities of the managers. Due to the subjective judgment factors involved, the outcomes of this model can be positive or negative.
3. Multi-asset portfolio management model
More complex DAT companies hold various different digital currencies. Managers need to make complex decisions: how much Bitcoin to hold, how much BNB, how much Ethereum, etc., how often to adjust this investment portfolio, and when to adjust, all of which test the managers' capabilities.
4. Ecological investment construction model
This is the most complex model, where in addition to holding coins, they will also invest 10%, 20%, or more of their funds in ecological construction. For example, a company focusing on Ethereum may wish to invest to help develop the entire Ethereum ecosystem; this model is more interesting. Projects like BNB, which support other digital asset ecosystems, also have similar practices, but this requires higher management capabilities.
Therefore, DAT is not simply 'holding coins'; different models correspond to different management costs and requirements.
The DAT companies we currently support tend to prefer the simplest first form. We prefer to focus solely on single-asset companies, particularly BNB, as it is easier to judge and does not require excessive participation in daily management. In a bull market, listed companies generally benefit, but in a bear market, especially in the US, companies often face lawsuits. If the strategy is clear and simple enough, the risk of litigation can be relatively reduced, and the company's legal costs can also decrease — after all, lawsuits are extremely expensive.
Our goal is to minimize operating costs while promoting the philosophy of long-term holding. We do not wish for the company to use funds for extra investments, but rather hope they can participate more deeply in supporting ecological development.
The significant meaning of the DAT model is that many companies' financial departments, listed companies, and even state-owned and central enterprises cannot directly buy digital currencies, but through the DAT model, we can effectively allow these traditional investors to gain exposure to digital currencies. This group is actually a very large market, much larger than the cryptocurrency circle.
In the DAT projects we participate in, we usually only play a role as small supporters. Most of the funding for these projects comes from traditional stock markets or other channels, which greatly helps our ecological development, drawing many groups outside the crypto circle to purchase digital currencies.
We generally do not take the lead, nor do we manage these companies, but rather find suitable managers through ecology and connections. Managing listed companies is not our expertise, but there are many in the industry who have relevant experience; we prefer to collaborate with them for synergy.
6. Discussing the integration of AI and Web 3.0
Frankly speaking, the combination of AI and Web 3.0 is still not ideal at present. However, I believe this trend is by no means just conceptual hype, but rather a trend that will inevitably lead to breakthrough developments in the future. A few months ago, I raised a question: What currency will AI use? The answer is clearly not the US dollar or traditional payment systems, as AI cannot complete KYC. The currency system of AI will inevitably be based on digital currencies and blockchain, completed through API calls or broadcasting transactions.
This means that the transaction volume of blockchain will experience exponential growth. In the future, everyone may have hundreds or thousands of AI agents completing tasks such as video production, multilingual translation, content distribution, reservations, and message replies in the background. Their frequent interactions will generate massive micro-payments, and the trading volume of crypto finance will conservatively estimated to grow by thousands of times. For example, a blogger can set the first 1/3 of an article as free, charging only 0.1 yuan for each reading of the remaining 2/3. If hundreds of thousands pay, they can earn tens of thousands in income — this model cannot be realized in the traditional financial system but can be easily supported through the integration of AI and Web 3.0.
Trading will also become more global. I can simultaneously employ engineers and designers from China, India, and even around the world; AI will automatically handle settlements and payments. However, currently, most so-called 'AI agents' in the Web 3.0 field remain at the pseudo-product stage reminiscent of Memecoins: presenting novel content in the front end while calling established large model APIs like ChatGPT in the backend, lacking real usage value. What we truly need are AI tools that can accomplish actual work and create economic value, and top large model companies are also striving to explore this direction.
However, the development of AI requires extremely large amounts of funding. The competition for computing power among large models is exceptionally intense, with astonishing costs. It is reported that OpenAI currently possesses about 1-2 PB of computing power, with an annual cost of approximately 6.5 billion US dollars per PB, and its expansion plan is to increase by 10 to 100 times — the investment will be astronomical, not including chip expenses. No VC, company, or even nation can bear such a huge financial burden alone, which is why the AI industry has begun to explore new paths for financing through the lens of Web 3.0.
Fundamentally, AI should be regarded as a public product. Currently, many large models are too closed. Allowing token holders to share profits, making the model more open-source, decentralized, and accessible to all may be a more reasonable direction for development. I have discussed this matter with several top large model founders. Although everything is still in the early stages, this trend will inevitably come.
Although the combination of AI and Web 3.0 has not yet been perfected, its future development prospects are still highly anticipated.
