Arthur Hayes' Full Speech at Token2049 in Singapore: The Next Bull Market Will Start in Early 2024

Speech Topic: "Money Printing, AI, and Crypto: Fueling an Epic Bull Market Mania"
On-site Recording: Logicrw, BlockBeats
Content Compilation: Kaori, czgsws, BlockBeats
On September 13, Arthur Hayes, co-founder and former CEO of BitMEX, delivered a speech titled "Money Printing, AI, and Crypto: Fueling an Epic Bull Market Mania" at Token2049 in Singapore. The on-site reporters from BlockBeats recorded, organized, and compiled the speech content, which is as follows:

In this speech, I will discuss debt and AI from a macro perspective, then narrow it down to a micro level to briefly explain why I am very fond of Filecoin. Once again, I remind everyone that this is my broad framework regarding the value of cryptocurrencies, namely the liquidity of fiat currencies and certain technologies within the ecosystem. Over the past decade, one of these two factors has always been the reason for us to enter a bull market, but we have yet to encounter a bull market where both aspects appear simultaneously.

Now I want to explain why I believe the next bull market may begin in early 2024, which will be the largest bull market in both the cryptocurrency space and the risk asset space since World War II and the Great Depression.

Fundamentally, what is GDP? I got this saying from Robin Paul, which is population growth, productivity, and debt. I want to focus on babies (population and demographic structure) and debt (money printing). A very simple statement is that rich people don't have many children. There are many reasons for this phenomenon, but if you look at the relevant chart data, you will see that we are actually entering a population growth deficit, which brings a very big problem. Ultimately, we are borrowing from the future to build today, and then hoping that more people will come in the future to fill this growth so that they can continue economic activities.

So, when there aren't enough children, who can drive this growth to repay the debt? What solutions have central banks and governments proposed to address this issue?

How do we maintain GDP growth? ------ By printing money.

Since 1970, this is the World Bank's chart on global debt versus GDP, and we have increased from 110% to 360%. The worst part of this chart is not just the terrifying 360%, but the acceleration is equally chilling. It took us 10 years to increase the growth rate by 100%, while during the COVID-19 pandemic, we only took two years to increase it from 250% to 360%.

From a global perspective, we are in the late stages of a debt disaster. Why is debt unsustainable? Each country has its specific reasons, but on a global level, we can see that we are using accelerated debt issuance to compensate for the nonexistent growth.

Even though we are in Singapore, we are under the dollar standard, and the monetary policies of most countries reflect U.S. financial policies, so I spend a lot of time discussing the Federal Reserve and financial policies.

If you are a holder of U.S. Treasury bonds, this is a very scary chart, which is the maturity profile of U.S. debt. By 2026, the U.S. must roll over nearly $8 trillion in bonds, and they must find someone willing to accept this debt at the current market yields. Now think about it, if the whole world is facing such a problem of too much debt and insufficient population growth, how will they do this? Who will buy this debt?

When the government has a pile of debt to issue, and no one is willing to buy it at an affordable yield, what does the government do? They print money.

Therefore, if I look at the liquidity issue of fiat currencies from a global perspective and from the perspective of the U.S. as the issuer of the reserve currency, this will be the largest currency issuance market we have ever seen, but no one here is willing to buy these things at these yields.

Decentralized Storage in the Age of AI

Okay, now let's shift from fiat currency to technology. What is the dominant topic today? It's AI.

Yesterday, I attended a lunch meeting with many global asset management companies, family offices, and wealth salespeople, discussing what ideas or themes we are most confident about. I think about 75% of the people mentioned something about AI. This is the trend.

The year 2000 was when the internet was going to dominate the world; you know there is always something, 1929 was about radio and railroads. In every cycle, we focus on some new technology that will fundamentally change the human experience. AI is no different, as the level of computer penetration and the speed at which human civilization has upgraded since the invention of mainframe computers and the internet is so rapid that you can understand why investors are so concerned about what AI can do.

Computers can think for themselves; what changes will this bring to our civilization? There may be trillions of dollars of value created by AI. Chat GPT has been a major highlight in the past six months, being the fastest adopted technology in human history. Why are all these startups saying we will apply LLM models in a specific area of the economy to improve the productivity of companies using our products?

So what does this mean? The biggest stock in AI right now is NVIDIA, and since the launch of its publicly consumable products, NVIDIA's stock price has nearly doubled, with a price-to-earnings ratio reaching 100 times. If you think about it carefully, this is absolutely absurd. How much does NVIDIA need to earn in the next 5 to 10 years to reach its current valuation? Some might say this is a bubble, and I agree with that viewpoint, but I think it will get worse.

Thus, we have a fiat currency liquidity bull market driven by insufficient baby numbers and massive debt, and we have a tech bull market driven by AI. I have written some papers trying to connect cryptocurrencies and AI. Fundamentally, I believe AI is a personal economic unit; if you want to anthropomorphize it, it doesn't care about human laws and regulations because governments cannot kill an AI by force. Yes, you can try to eliminate computers and so on, but that would almost destroy the social connections we have. Therefore, I believe a single government can hardly unilaterally punish an AI, so, double happiness, we will have pets and AI as reasons we should own cryptocurrencies.

So, what does AI need? AI needs computing power and data storage. When I consider my portfolio and how my family office can enter the AI field, I realize that I have missed the best opportunity to invest in some of the startups that are rising today. I won't get liquidity in 2026 or 2027, whether it's unlocked tokens or ITO windows, because a company needs five to seven years to mature enough to launch its own products. So I need to find something that can be implemented immediately.

Now, I can buy NVIDIA stock at a 100 times price-to-earnings ratio, but I can consider data storage and the angle that may relate to cryptocurrencies. So let's focus on this. Some people in this room are venture capitalists; I won't say you are fools, but the people putting money in your hands might be fools.

So, what will happen next? A bunch of venture capital firms will say, I want to create an AI fund. Then you have to pay various fees, right? Then they will turn around and see thousands of billions of dollars of investment in any pitch deck that has AI in it. When they get this funding, they will use it to buy PPUs and hot storage.

So, if I don't want to profit from the AI frenzy, I want to enter the field of capital flow. This capital is flowing into these two areas because without these things, no matter what AI solution you claim to have, it essentially won't run.

Why does AI need decentralized storage? If I believe AI won't use a lot of data storage, shouldn't I just use Amazon? You put data in a centralized company that can shut down your service at the government's command, change prices at any time because it's their data center, and can cancel contracts at any time. What will AI do? Will it sue someone in court? That won't happen. Fundamentally, AI will create an opportunity for decentralized storage.

So now I need a shitcoin; I want a listed token that I can buy, which has completely collapsed, dropping over 90% from its historical peak of $300 in 2021 to now $3, and the most aggressive sellers have all sold off, so when this token starts to rise, it has a lot of room for growth. The incoming capital doesn't need to be much to push this token up 10 or 20 times, and it also needs to be related to the AI industry chain.

The Chosen Filecoin

Filecoin is considered a worthless virtual currency, dropping from a peak of $300 to now $3, a decline of nearly 99%. However, people are still using it.

There is data existing on the Filecoin network, and there are real customers using this system, and the market still has a lot of room for development. Although AWS is the market leader, it is a centralized solution. If end-users value the benefits brought by data decentralization, then there is room for growth in the decentralized storage market.

I invested in a Filecoin storage platform called Seal Storage, which is helping to bring large data users into the Filecoin network. They have a project called Atlas, which is a data project for the European particle accelerator CERN. They are using SEAL to reduce data costs, and SEAL earns Filecoin rewards by placing this data on the network. The University of California, Berkeley is also using SEAL for similar purposes, and SEAL has collaborative projects with NASA and several other universities in the U.S.