Author: Crypto Wei Tuo Source: X, @thecryptoskanda

"Humanity may have no future, but it will never stop plundering."

Fat currency is the most efficient, covert, and iterative plunder tool in human history. It can be created out of thin air by the regime - the regime only needs to continue to prove that this currency can unconditionally exchange for the core materials that maintain the current operation of human society. The problem is that the real survival threshold of people is only food, clothing and reproduction, which is too low to continuously prove that legal currency and political power are needed. Therefore, the legal currency system must continue to raise the cost of people's survival around a "material" to justify plunder. This process is what we call "development" - constantly creating mismatched rigid needs, Rationalize the mismatches from the previous round with new mismatches.

This is the Ponzi system of fiat currencies including the US dollar. This material was once gold, and then oil. After the 1974 Saudi Arabia Agreement, petrodollars supported the US dollar system for a full 50 years. But this mechanism began to structurally fail after 2024 - Saudi Arabia accepted RMB settlement, OPEC dollars no longer automatically returned to US debt, and US dollars printed money faster year by year. It needs a new material to carry the next 50 years. And it must be done before the petrodollar completely expires. This time, it chose AI. I call it "USD-AI Ponzi". Such "AI Ponzi theory" will be scorned by the "orthodox players" of traditional US stocks. But for native players in the currency circle who are already familiar with the three-plate theory and have honed their Ponzi sense and opponent market thinking in countless zero-sum transactions - there is no need to belittle themselves. There is no need to give up their own expertise. This series aims to help these players quickly establish their understanding of narrative-driven stocks without conflicting with the underlying cognition, enter the stock trenches with their most advantageous position, and maximize their interests in this "stock market currency circle" that has not yet fully unfolded. The series is divided into three parts, the upper, middle and lower parts, this part is the first.

The final source of water for the upstream of AI - systematic cognition of U.S. dollar Pond's

The dollar itself is the best Ponzi textbook

“The mismatch between funding needs and expected returns creates a ‘gap’ that can only be filled by further mismatches”

Funding requirements: The U.S. fiscal deficit is $2T+ per year, and the government must continue to spend money; the Federal Reserve must continue to print money to meet the liquidity of the financial system. Expected returns: People holding US dollar assets (either debt or cash) - the actual purchasing power does not shrink and can outrun inflation. Gap: Government spending + interest compound growth (deficit is $2T+ per year, accumulated national debt is $36T, and interest expenses are already $1T+ per year) It is destined to be irreconcilable in the long term with the amount of debt that the economy can truly absorb. In the short term, new debt can be issued to fill it - but someone must buy the new debt. Domestic capacity is limited, and overseas structural holdings are reduced (China's debt holdings are reduced from $1.3T to $0.7T). The only option left is for the Fed to print money for monetization. And whether it is issuing more debt or the Fed printing money, it will ultimately dilute the purchasing power of each dollar - It naturally conflicts with the holder's expectation that "actual returns outperform inflation."

Therefore, the U.S. dollar always needs a "material" that can sustain the speed of the Fed's money printing and the rapid growth of demand. If there is no material absorption: the U.S. dollar pushes up prices → the purchasing power of U.S. dollar asset holders is diluted → sells U.S. dollar assets → the scale of the sell-off exceeds the ability of any buyer → the Fed can only monetize the acquisition → print more money → make inflation more out of control → more sell-offs. The U.S. dollar system keeps creating bubbles and looking for new materials to absorb U.S. dollar inflation. It is the basic need for the survival of the U.S. dollar itself. Look at the U.S. dollar through the combination of three stocks: U.S. debt is a dividend tray. Central banks, monetary funds (MMF), U.S. banks, and life insurance sink money into U.S. bonds and earn interest. The Treasury Department continues to issue new bonds, and the interest comes from taxes + new debts - borrowing new ones to repay old ones, and a relay-type deposit deposit. The geometry is exactly the same as the dividend tray, except that this dividend tray is endorsed by the U.S. military + global settlement demand to support it. In 1980, the closed loop of the Federal Reserve-Treasury-Bank was a mutual aid market. The Fed printed money to the banks, the banks bought U.S. debt, the Treasury used the money to make expenditures, the expenditures were returned to the banking system as deposits, and the banks bought more U.S. debt. A→B→C→A complete accounting cycle. The core of this mutual aid disk was the Fed's own balance sheet - it doubled, and the cycle doubled. From 2008 to 2014, it expanded 5 times, It will double again in 2020. The US dollar split is to offload the demand for US dollars to a global "material". The printed US dollars must have a place, otherwise the return will be inflation. This undertaking is the material mentioned at the beginning - historically it was gold, then oil, and now it is AI's turn.

AI is the largest split in the US dollar right now

Is it different this time? This time, people outside the currency circle are most likely to retort: "AI is really useful, ChatGPT really improves productivity, OpenAI has a revenue of $25B a year, how can it be a Ponzi?" Brothers, at this moment, it is exactly the same as that time: "L2 really has a sorter fee, ZK can really generate income, how can it be a Ponzi?" L2 and ZK are both ETH splits. The fees of L2 are real. But it must rely on the continuous issuance of new assets to absorb external liquidity - without the influx of new tokens, new protocols, and new applications, the revenue immediately dries up, and TVL is siphoned off. Almost all failed L2 died from this. As early as the end of 23, I predicted the biggest disaster for ETH - the main stage of asset issuance was snatched away by Solana/BSC, the ETH gas + destruction mechanism immediately fell into structural stagnation, and the price reached new lows relative to BTC. Not ETH The technological regression is due to the collapse of its split. At this time, some people may retort: "But the downstream semiconductors of AI are different - NVDA / TSMC / Marvell are real things and have real cash flow, it is always the fundamentals, right?" The real things are right, and the cash flow is also right. But the current forward profit expectations of semiconductor companies are completely based on the premise of "the dollar system continues to inject liquidity + Yun Dachang continues to spend $300B+/year on AI capital expenditures". AI The expansion rate of this chain of $1T+ per year far exceeds the real income that the terminal AI companies can create themselves - and most of this real income is essentially replacing human labor with AI, resulting in irreversible job losses. It must rely on continuous external capital injections to support it. Once the Ponzi input of the US dollar system stops → AI narrative stalls and valuations retreat → Cloud Dachang’s capital expenditures stagnate → NVDA’s forward revenue declines → Forward PE explodes → The overall valuation is repriced → Semiconductors have returned to the valuation of cyclical stocks. This is similar to L2/ZK, the only difference is that the source of external liquidity relay is different. The current "fundamentals" of semiconductors can be compared to L2's TVL, number of addresses, trading volume, and mainstream adoption in 2023/24 - they are all narratives that promote the split.

After all, Trump said at that time that he would use cryptocurrency to convert debt. As AI is used to undertake US dollar inflation, how can it not convert debt? The core is to make this liquidity relay more reasonable.

Three Thousand Worlds - Dismantling of the Ponzi structure within the AI industry chain

AI upstream itself is a mutual aid disk. Having said this, the logic that AI is a dollar split disk is clear. But within AI, the upstream (terminal applications + cloud manufacturers + chip design) itself constitutes a mutual aid disk - mutual revenue, objectively stabilizing the total market value of each company (that is, the "liquidable assets" in the mutual aid disk):

• Microsoft invests $13B in OpenAI + Azure cloud service quota → OpenAI spends $12.4B in computing power on Azure

• NVDA lends money to CoreWeave to buy NVDA’s own GPU

• Google invests in Anthropic, and Anthropic spends time on Google Cloud

• Oracle - OpenAI five-year contract bundle

The money has never left this circle, but it has created a win-win situation for all three parties: OpenAI annual revenue of $25B + Azure AI revenue 50%+ year-on-year growth + Mag-7 stock price at a new high. It is exactly the same as the Federal Reserve-Treasury-Bank mutual aid package, but it is smaller in magnitude. If you can’t see clearly that AI is Ponzi, watch Microsoft give OpenAI money to allow OpenAI to operate on Azure Just spend it up. This upstream mutual investment is the "initiation place" for AI CAPEX capital expenditures - it determines the direction and magnitude of AI capital expenditures, and also determines how many orders the entire downstream supply chain can get. Upstream mutual investment stabilizes the book liquidable assets (Mag-7 + AI company's total market value) at a high level. When any individual shareholder wants to exit, he can get back cash close to the book price. It is this expectation of "withdrawal at any time" that makes retail investors / ETF / 401k Continuous buying is the core mechanism by which the AI split market can maintain $1T+/year expansion. The dual paths within the AI split market return to the split market definition:

"A split is a Ponzi system in which the total capital remains constant at a certain point in time, but the number of equity or assets corresponding to each unit of capital is multiplied, while the price of the newly generated equity or assets is reduced proportionally to attract subsequent capital inflows."

As a split of the US dollar, AI needs to attract capital inflows. This capital inflow is divided into two paths:

• Primary market: Through capital expenditure, in the form of revenue and primary financing, it is transmitted to downstream manufacturers according to the level of the industry chain

• Secondary market: It is not based on the industrial chain level, but all AI downstreams compete for the same pool of liquidity

After talking about the upstream, the downstream is the overall split of AI:

• Semiconductor supporting: Marvell, HBM factory (SK Hynix / Samsung / Micron), optical module (Lumentum / Coherent), power supply (Vertiv / Eaton), data center REIT (Equinix / DLR)

• OEM + Equipment: TSMC, Samsung Foundry (OEM); ASML, Lam, KLA, Applied Materials (Equipment)

• Upstream energy + terminal: SMR nuclear power / power PPA / perilla leaf / long tail narrative small cap

Each downstream individual company has a dividend plate - invest in fab/capacity (sunk cost) → wait for upstream orders to be fulfilled → profit dividends / buyback. SK Hynix invests in HBM fab, Lumentum invests in optical module production lines, TSMC invests in wafer foundry, all of the same disk type. Of course, it is emphasized again that the Ponzi attribute of the downstream does not come from itself, but from the upstream AI transmission. Excluding the repricing caused by the AI ​​boom, the valuations of these companies will return to the level of cyclical stocks. The Ponzi mentioned here refers to this premium, not their original "fundamentals"

First level vs second level, two structures

When we get here, we actually only see primary market liquidity transmission (revenue + capital expenditure). However, secondary market speculative funds are not limited to the supply chain, but follow narrative clusters. When the increase in the upper level is no longer α but β (for example, NVDA has finished rising, and the market thinks so), speculative funds will look for underpriced candidates under the same narrative, no matter what level they are in the supply chain. For example, Bai Mao Stock God @aleabitoreddit "AI Stuck Neck Narrative" Under cluster:

• Vertiv (Data Center Power/Cooling)

• Marvell (NVDA ASIC cooperation)

• Equinix / DLR (Data Center REIT)

• Lumentum (optical module)

• Lam / KLA (Equipment)

These are different layers in the supply chain, but they are the same cluster in the secondary market, competing with each other for the same pool of "AI stuck neck narrative liquidity".

So the specific form of the collapse condition of "split rate is too high" in the secondary market is not "the sudden emergence of more track companies", but:

The popularization of this narrative logic has led to more targets with lower market capitalization and lower levels of the industrial chain being quickly discovered and gaining attention, resulting in faster and faster capital rotation in the secondary market - a "diversion" tactic familiar to trench players.

The corresponding playing style of P can be explained as: buy at the bottom, go out and call for orders to build, and make a profit (if it doesn’t end, let’s talk about it later). His perilla leaf theory is a theory that is logically reasonable and difficult to falsify, very similar to the meme as cult theory proposed by @MustStopMurad at the time.

Mr. Bai is absolutely right, but the most important point of his trading strategy is actually "buying at the bottom" - such as Raspberry Pi and SIVE.

The rationality of "trench trading" in AI split trading in the stock market

There is no need to argue whether stock trading should be "value investing" or "long-term holding." In fact, even in the US stock market, the life cycle of narratives has been greatly shortened since the dotcom bubble in 2001:

1. Information dissemination speed: newspapers (days) → TV (hours) → Internet (minutes) → social media + algorithm recommendation (seconds)

2. Cross-border capital mobility: Especially after 2014, Asian hot money joined the battle. Chinese capital brought a completely different system of capital speculation tactics - cryptocurrency is essentially a branch of the evolution of cross-border Asian hot money flow, and this branch brings a completely different logic from the previous US stock institutions.

3. Institutional HFT: Contrary to what many people imagine, most institutions currently are not "value investments", but high-frequency algorithmic quantification - accounting for 50% of the current US stock trading volume. The algorithms of institutional trading teams continue to converge in the competition, showing a highly crowded "buildings rise together and collapse together" feature.

4. Retail investor participation has increased: The conclusion that "U.S. stocks are dominated by institutions" has long been a thing of the past. As various brokerages have promoted retail trading, the trading volume of U.S. retail investors has exceeded 25%, making GME meme stocks possible. Retail investors, especially young people, have been shortening their attention span in the past decade - this has become a physical limitation in their trading logic.

Traditional markets call this reflexivity, and in the three-disc theory this is the "audience-determining pattern".

At this point, I believe my readers already understand:

• The upstream mutual aid disk is more stable than the downstream split disk; in the downstream, the closer to the core of the AI stuck neck (TSMC / ASML / HBM), the more stable it is, and the closer to the end (Perilla Leaf Theory / Long Tail), the more fragile - the mother disk's ability to lock in capital and expectations does not decay along the supply chain, but along the center distance of the secondary narrative cluster.

• The goal of the trade is to exit before these splits signal even a short-term collapse

When playing with the market, there is only one trick: buy low and buy early.

The kid who shined Rockefeller's shoes may be a son chosen by heaven for stock trading. However, at a specific point in time in the entire market, the amount of money left to take over is a limited value, and the number of 100,000 Heavenly Soldiers is limited. Arriving too late, he could only become the "price" and become Rockefeller's exit liquidity.

How to judge whether you bought low or early, and where you are in the queue of liquidity hunters, is a lifelong practice for every marketer.

This is not substantially different from the style of play of the trench P boys: what you judge is not the fundamentals, but how many people believe in this "fundamental narrative", and then make capital first - but the judgment tools and starting points are different. So, on what basis should we judge? Grand narratives can deceive people, but liquidity expectations cannot.