727 FinanceThe October pullback stemmed from TGA drainage, while the November liquidity release is the turning point; the long-term logic lies in the convergence of TradFi and Crypto driven by compliant financing and RWA (real-world asset tokenization).
Written and hosted by Mr. Z (@168MrZ)
The U.S. economy is entering a new infrastructure cycle driven by AI CapEx, while AI commercialization is only just beginning to realize the compute invested over the past year; the real risk is not a Semiconductor pullback, but oil prices re-anchoring inflation, Cloud Revenue ceasing to accelerate, and the Anthropic IPO beginning to drain market liquidity
On October 29, 2026, 168X invited Qinbafrank (@qinbafrank) for the fourth time. The conversation ranged from the FOMC, the U.S.-Iran war, oil prices and inflation all the way to Microsoft, Google, Anthropic, AI CapEx returns, Cloud, Agent, AI for Science, and then to Crypto's TGA liquidity, compliant financing and Tokenization.
Frank's core framework this time is not a simple judgment of AI bulls versus bears, but rather a decomposition of the market into two simultaneously unfolding lines: on the macro side, high oil prices and high long-end rates are still continuously raising the valuation threshold for Risk Assets; on the industry side, AI Commercialization is still accelerating, and the CapEx poured in over the past year is gradually turning into online Capacity, enterprise orders and Revenue.
So AI is not over, but it has entered the second half. The future is not all AI stocks rising together, but rather having to start answering: who actually turns CapEx into Revenue? Who owns the enterprise entry point? Who can take more profit as Tokens become cheaper and cheaper?
I. Why the U.S. economy is still so strong: AI CapEx has become America's new infrastructure
Mr. Z: Let's start with the big picture. How do you see the U.S. economy, inflation, and the Fed's rate pause this time?
Qinbafrank: I look at the U.S. economy on two levels. The first is macro GDP, and I think there is no major problem at the moment. Starting in the second half of 2025, the U.S. has essentially entered the era of AI mega-infrastructure.
Qinbafrank: CapEx by large tech companies in 2025 was roughly $400–500 billion, and this year it could approach $900 billion. That is, nearly $500 billion in new capital expenditure in a single year. That magnitude alone can provide very clear support to U.S. GDP, about 1-2%.
Qinbafrank: This is actually very much like China more than a decade ago. At that time, a large part of China's economic drive came from real estate and infrastructure; now the U.S. has swapped in new Infrastructure, with the core being AI Datacenter, Power, Chip, Networking. As long as this round of CapEx Growth continues, it is hard for U.S. GDP aggregate growth to be particularly bad.
Qinbafrank: But on the second level, looking from consumers, employment, and individual experience, it is a completely different story. Consumption among middle- and high-income groups is still strong, while middle- and low-income groups are clearly weaker, and AI's replacement of Labor is becoming increasingly obvious. Microsoft has been continuously laying off people over the past two years, and other large tech companies will continue to do similar things in the future.
Qinbafrank: So it is a very clear K-shaped Economy. AI, Semiconductor, Power, Datacenter—these industries are doing well, but people and industries unrelated to AI may feel completely different.
II. The real problem with oil prices is no longer a lack of oil: it is an extra $30–40 per barrel in shipping costs
Mr. Z: You have consistently believed that oil prices are one of the most important variables for the next two months. Where exactly is the bottleneck now?
Qinbafrank: The important reason inflation has moved back up this year has always been energy, but the logic in the first half and now is actually no longer the same.
Qinbafrank: The first half was more of a Supply Shock. After the Strait of Hormuz was blockaded, oil genuinely could not get out, so oil prices shot straight up. But after entering September, supply itself has recovered a great deal.
Qinbafrank: For example, Saudi Arabia has a pipeline from the Persian Gulf to the Red Sea, which according to the materials I have seen can transport about 6–7 million barrels per day; the UAE's pipeline bypassing Hormuz can also provide about 3 million barrels. On the other hand, Middle Eastern oil producers have also begun using more VLCCs; a large tanker can load about 2 million barrels of crude at a time, and then pass through the strait concentrated in lower-risk time windows.
Qinbafrank: So after late September, the volume of crude leaving the Persian Gulf has actually recovered to very close to pre-war levels. The problem is that volume has returned, but cost has not.
Qinbafrank: Previously, shipping a barrel of crude from the Persian Gulf to Asia might have cost only $3 or $4; now it may become $35–40. When you see Brent above $110 today, a large part of that is not the price of Crude itself at all, but War Risk + Shipping Cost.
Qinbafrank: Because tankers have to pass through high-risk waters, shipowners bear the risk of being attacked, and insurance, escort, lightering, and transshipment all become more expensive. So this has already shifted from Supply Shortage to Logistics Inflation.
III. Beyond Crude, you also have to look at Refinery: refined product recovery is slower than crude supply
Qinbafrank: And now there is a second layer, which is Refinery.
Qinbafrank: Many people only look at crude, but in fact what U.S. consumers ultimately use is Gasoline, Diesel, and what industry really uses is various Refined Products. Some refineries in the Middle East have been damaged, and Russia and Ukraine have continuously attacked each other's refining facilities in August and September, so global Refining Capacity has also been weakened.
Qinbafrank: This means that even if the crude transportation problem is gradually resolved, refined products may not necessarily come back immediately.
Qinbafrank: So why did the U.S. later discuss Diesel Supply with Russia, including hoping other countries would increase refined product exports? Essentially because solving Crude alone is not enough.
Qinbafrank: Oil Inflation now has two layers: the first layer is Shipping Cost, the second layer is Refining Bottleneck. If these two problems are not solved, Energy Inflation will be very difficult to truly disappear.
IV. The Fed's rate pause is not a dovish turn: it is essentially "wait and see" within the hawkish camp
Mr. Z: The Fed did not raise rates this time. Should the market understand this as a dovish turn?
Qinbafrank: I don't think so. I would rather understand it as Patience within the hawkish camp.
Qinbafrank: The Fed also knows that part of current inflation is caused by war and energy. If U.S.-Iran relations ease and oil prices fall on their own, then Inflation in November and December may improve accordingly.
Qinbafrank: So what it is doing now is: I am in no hurry to keep raising, I will look at a few more data points, and see whether this Oil Shock is Temporary or will slowly spread into broader, more persistent inflation.
Qinbafrank: If the Middle East eases and oil prices fall clearly, then in December it could even continue to stand pat and further observe the pass-through of energy prices into CPI and PCE.
Qinbafrank: But conversely, if oil prices remain high, Energy Cost begins to spread from oil to chemicals and industrial goods, and then to more commodities, then it is no longer a One-off Shock but begins to become persistent.
Qinbafrank: In that case, the worst scenario is not "one more hike in December," but possibly continuing to hike next year.
Qinbafrank: The market can actually digest 1 or 2 additional hikes, because you can calculate it. But if the original expectation was 2 and it ultimately becomes 3, 4, or even longer, then the entire valuation system will be repriced.
V. The real window for the U.S. and Iran is after the midterm elections: Trump is now instead boxed in by himselfMr. Z: So when is a real turning point between the US and Iran most likely to occur?
Qinbafrank: I think the period after the midterm elections will be a relatively critical window.
Qinbafrank: Right now Trump has somewhat boxed himself in. Iran's system means it can endure economic pain for longer than the US political system can. In the US there are elections, public opinion, and inflation pressure; the suffering of ordinary Iranians does not necessarily translate immediately into regime change.
Qinbafrank: So previously the US hoped that through stronger economic sanctions and blockade, Iran would be the first to crack, but the reality is Iran has proven more resilient than expected.
Qinbafrank: That's exactly the problem. Before the midterms, if Trump directly backs down, his political image takes a hit, because it would amount to launching a conflict without winning and then retreating first. But after the midterms, regardless of the outcome, his political constraints will lessen.
Qinbafrank: At that point there may be two paths: either actually carry out another, more intense military action; or realize that continuing to fight won't change anything, and begin negotiations.
Qinbafrank: I personally think the latter is not without a chance, because you're already seeing more and more third-party mediation. Countries like Pakistan, Saudi Arabia, and Turkey are all getting involved, and China also has incentives to push for de-escalation, because no one can afford another round of soaring oil prices.
Qinbafrank: And in fact, signs of more direct communication between the US and Iran are beginning to appear.
Qinbafrank: So I think the biggest disagreement between the two sides right now may not even be the final terms, but who takes the first step. Does Iran first loosen the Strait of Hormuz, or does the US first ease sanctions and unfreeze assets? The biggest problem for both sides is that trust has been destroyed; each fears that if it steps back first, the other won't honor its commitments.
VI. Warsh Is Remaking the Fed: Going Forward, the Dot Plot and Forward Guidance May Matter Less and Less
Mr. Z: What do you make of Kevin Warsh's new Fed communication and decision-making framework?
Qinbafrank: I think he genuinely wants to change things, not just put on a show.
Qinbafrank: The Fed had 8 policy meetings over the past year; his direction now is to reduce the number of meetings that actually discuss monetary policy, with adjustments starting next year and possibly further reductions after that.
Qinbafrank: More importantly, he clearly dislikes excessive Communication and dislikes giving too much Forward Guidance. The importance of things like the dot plot may also decline in the future.
Qinbafrank: His thinking is more like: each time, genuinely judge based on the data at that moment, rather than telling you 3 months in advance roughly what I'm going to do next.
Qinbafrank: In addition, after taking office he set up several working groups, and I think what's really worth watching by year-end are these reports. I'm mainly looking at 3 questions.
Qinbafrank: First, will the Inflation Framework change? Why must it be 2%? During the Powell era it was changed before, from a Point Target to an Average Inflation Target over a period, so whether there will be a new definition in the future is, I think, very important.
Qinbafrank: Second, how does AI Productivity enter monetary policy? AI itself is a disinflationary force in the long run, because the same resources can produce more goods and services; but in the short run, massive AI infrastructure buildout creates inflationary pressure from capital expenditure, materials, and electricity. How exactly does the Fed balance short-term Inflationary CapEx and long-term Productivity Deflation?
Qinbafrank: Third, the Balance Sheet and Reserve Regime. Since 2008 the US has developed a very large balance sheet and an ample reserves regime; will Warsh in the future shrink the balance sheet again and release more liquidity from the Fed system back to banks and markets?
Qinbafrank: If these things really change, the impact is far greater than whether or not the FOMC hikes 25bp at a single meeting.
VII. Stop Fooling Yourself with ARR: What's Really Valuable About Anthropic Is the Revenue That Has Already Happened
Mr. Z: Let's move to AI. You've been saying recently that we should stop looking at ARR and should look at real revenue instead. Why?
Qinbafrank: Because ARR and revenue that has already occurred are fundamentally different in quality.
Qinbafrank: Anthropic's full-year 2025 revenue was about $4.5 billion, and in the first half of 2026 about $16 billion-plus. In Q3, according to figures previously disclosed by the media, the single quarter was again about $16.8 billion.
Qinbafrank: That means its actual Revenue for the first three quarters of this year is already about $33 billion. If Q4 continues to grow versus Q3, then my earlier estimate of $50–60 billion in real Revenue for the full year is open to discussion.
Qinbafrank: This is completely different from "$100 billion ARR." ARR annualizes the Run-rate at a point in time; $50–60 billion is sales that actually happened this year.
Qinbafrank: So I say that after AI has developed to this stage, we should gradually push the ARR metric to the back and start looking at Quarterly Revenue, Annual Revenue, and real Cash Flow.
Qinbafrank: And at this pace, if Anthropic's Revenue exceeds $100 billion next year, I don't think that's exaggerated. It previously said it would reach $200 billion by 2028; now at least that's no longer a completely unimaginable number.
VIII. What Really Beat Expectations at Microsoft Wasn't Azure's 48%: It's That Enterprise Customers Are Starting to Take Over from Model Companies
Mr. Z: Microsoft this quarter almost hit the upper end of your forward-looking framework. What do you value most?
Qinbafrank: I think Microsoft's earnings really were good. Azure year-over-year growth reached 48%, Backlog is about $670 billion, of which about 30% is expected to be recognized over the next 12 months. Free Cash Flow is still $18.6 billion, which stands out sharply among the big cloud players.
Qinbafrank: But I think what's really important isn't the 48%, but who is buying Azure.
Qinbafrank: Microsoft previously gave a data point: a large portion of new orders are no longer from Frontier Model Companies like OpenAI and Anthropic, but from ordinary Enterprise Customers.
Qinbafrank: This time we again see Foundry's business growing rapidly, with a large number of paying customers; Copilot enterprise customers on pay-as-you-go are also continuing to grow rapidly.
Qinbafrank: This signal is especially important, because in the past everyone kept worrying: could all this Cloud Demand just be OpenAI borrowing a pile of money, then buying Microsoft GPUs, Microsoft then investing in OpenAI, ultimately forming a Circular Trade that just cycles back on itself?
Qinbafrank: Now more and more orders are coming from ordinary enterprises, showing that demand is truly starting to spread from model companies to Enterprise. That is real AI Commercialization.
IX. Cloud Will Become the AI Operating System: Enterprises Don't Care About Claude or DeepSeek, Only Results and Cost
Qinbafrank: There's another point from this Microsoft earnings call that I think is especially important.
Qinbafrank: Microsoft now has its own Models, while also connecting to OpenAI and deploying various Open Source Models. Analysts asked: how exactly do users choose?Qinbafrank: Satya Nadella's answer is essentially dynamic selection by workload.
Qinbafrank: Enterprises don't care whether the underlying model today is Claude, OpenAI, DeepSeek, or Microsoft's own model. Enterprises care about only two things: first, whether the task can be completed stably and with high quality; second, how much it costs to complete the task.
Qinbafrank: Complex tasks may invoke a frontier model; high-frequency, low-value, relatively mature work can be switched directly to cheaper open source or in-house models.
Qinbafrank: This is actually what I said before: Cloud will ultimately become an AI operating system.
Qinbafrank: It handles model routing, data connection, permission management, security, and workflow. The underlying model can be swapped, but the enterprise's entire system runs on your platform.
Qinbafrank: In particular, a large number of traditional companies are not technology companies, and they do not have the ability to build an entire infrastructure themselves. So after models become commoditized, that does not mean Cloud gets bypassed; instead, it may make Cloud more valuable.
X. Why Google both blew up and was not that bad: Cloud 103%, the problem is actually Search Monetization
Mr. Z: Google's results this quarter got a very poor market reaction. What do you think?
Qinbafrank: The headline was very bad, but after breaking it down it was not that bad.
Qinbafrank: Google Cloud grew about 103% year over year in the third quarter, already more than $30 billion in revenue for the quarter, and operating profit is also very high, with margin still improving compared with the second quarter.
Qinbafrank: So Cloud itself is very strong; its problem is mainly the original core business. AI Search now has very high usage, but monetization has not fully caught up.
Qinbafrank: In the past, when users searched for something, they would see many ads; now users talk directly to AI. If AI gives the answer and there is no equally strong ad monetization afterward, then you are cannibalizing your original Search revenue.
Qinbafrank: Also, the market saw that its free cash flow was very poor, but there were some one-off factors in there, such as the increase in TPU inventory, as well as Europe-related settlement expenses. Add those back, and actual FCF is not as exaggerated as the headline.
Qinbafrank: But the market is just that picky right now. No matter how strong your Cloud is, as long as the core business or FCF has flaws, I will cut your valuation first.
XI. The "time lag" in AI CapEx is starting to pay off: money spent in 2025 is only now turning into revenue
Mr. Z: So is it finally entering the time lag you have been talking about all along, where it starts to pay off?
Qinbafrank: Right. This is very critical. AI CapEx is not spent today and turned into revenue tomorrow. Buying land, building data centers, building power facilities, ordering GPUs, installing servers, and then actually going live takes a long time.
Qinbafrank: Large tech companies really began to scale up CapEx substantially around the third quarter of 2025, so much of the revenue growth Microsoft and Google Cloud are showing today actually corresponds to money spent a year ago or even earlier.
Qinbafrank: In other words, the capacity that began to be accelerated in the second half of last year is only now entering the production period, and over the past several quarters all cloud vendors have been saying the same thing: demand > supply, compute is insufficient.
Qinbafrank: If demand is still growing, and whatever new capacity you bring online is immediately absorbed, then revenue will naturally continue to accelerate in the coming quarters.
Qinbafrank: So I think commercialization is not peaking now; it is just beginning to pay off at scale.
XII. The real inflection point: when Cloud Revenue Growth starts to exceed CapEx Growth
Qinbafrank: But you cannot always look only at revenue here; the second stage is about ROIC.
Qinbafrank: 2026 is a year of CapEx explosion. Last year it was hundreds of billions of dollars, and this year it is close to doubling. That kind of CapEx Growth cannot be 100% every year.
Qinbafrank: What is the truly beautiful inflection point? It is when CapEx Growth starts to naturally decline in the future, but Cloud Revenue Growth continues to accelerate.
Qinbafrank: For example, Google's Cloud Growth may previously have been only 40% or 50%, then later more than 80%, and now more than 100%. If its Revenue Growth starts to exceed CapEx Growth, then the math gets better and better.
Qinbafrank: Also, the market used to worry a lot about GPU depreciation. Everyone said GPUs become obsolete in two or three years, so hyperscalers extending depreciation lives is an accounting trick.
Qinbafrank: But the reality is that older-generation GPUs still have very strong economic value today, and some long-term contracts even extend the use of old cards further out.
Qinbafrank: If the lifespan of servers, data centers, and power infrastructure is longer than the market originally expected, and chips can be replaced while facilities continue to be used, then the true payback period for the entire CapEx will also become better.
Qinbafrank: So in the next few quarters, what I most want to see is how these things start to converge: Revenue Growth, CapEx Growth, Depreciation Life, and FCF.
XIII. Coding is not the only AI scenario: Enterprise, Agent, and AI for Science are taking over at the same time
Qinbafrank: Another worry the market had before was: besides Coding, what else does AI have?
Qinbafrank: Why did Coding explode first? Because it is highly standardized, has huge repositories like GitHub, and has a clear feedback loop, so models can improve quickly.
Qinbafrank: But the second stage is not "finding another single scenario as big as Coding"; instead, many scenarios explode together.
Qinbafrank: The first category is horizontal and vertical workflows inside enterprises. Finance, Legal, Tax, Marketing, HR, and various enterprise processes may not be as big as Coding individually, but together they are very large.
Qinbafrank: Moreover, as Coding itself becomes stronger, it is also paving the way for these scenarios, because a large number of enterprise processes ultimately need code to integrate.
Qinbafrank: The second is Agent. Starting in September, major companies have all been pushing their own Agents. Personal Agent and Enterprise Agent will both continue to increase token consumption.Qinbafrank: The third is AI for Science, AIDD (AI-Powered Drug Discovery), Drug Discovery—these high-value scenarios. OpenAI, Anthropic, Google, and ByteDance are all investing in Bio and Life Science.
Qinbafrank: So it's not that there is no second scenario after Coding; rather, the second scenario has already split into many different scenarios appearing simultaneously.
XIV. How to Trade Around Anthropic 11/9: Before the Roadshow Is the Window, After the Roadshow the Draining Begins
Mr. Z: What everyone is most concerned about right now should still be the Anthropic IPO. Why do you see around November 9 as a dividing line?
Qinbafrank: November 9 is the timing I derived based on the SEC process, not something the company has officially announced.
Qinbafrank: After Anthropic publishes its S-1, under the rules it must go through a certain number of calendar days at the earliest before it can start the Roadshow. Counting forward, that lands roughly around November 9.
Qinbafrank: Before that, I think the market may still have a relatively comfortable window.
Qinbafrank: There are several reasons. First, the market has already fallen for several weeks, so a lot of sentiment has been released; second, Microsoft's and Google's Cloud Revenue at least prove that AI Commercialization has not broken; third, if Amazon's earnings also show no obvious problems, then fundamentals will give the market some support; fourth, there is also seasonality around the midterm elections themselves.
Qinbafrank: So before the Roadshow, the market may first have a rebound. What you really need to watch out for is after institutions begin seriously preparing subscriptions.
XV. Why a Good IPO Can Also Crush the Market: $400 Billion in Subscription Funds Cannot Appear Out of Thin Air
Qinbafrank: Many people don't understand what "draining" means. Draining does not mean Anthropic is bad; precisely the opposite—it may be because it is so good that all institutions want it.
Qinbafrank: I'll use the situation with SpaceX at the time as an analogy. For a large IPO, the final subscription funds are usually several times the actual issue size.
Qinbafrank: If Anthropic ultimately is also a large Deal on the scale of hundreds of billions of dollars, the market may need to prepare subscription funds in the hundreds of billions.
Qinbafrank: The problem is that institutional accounts will not just leave hundreds of billions in Cash sitting there waiting for it to list. So where does the money come from? Selling positions.
Qinbafrank: Sell part of Semi, sell Momentum, sell Software, sell other AI Position, free up Cash, and prepare to subscribe. So the real Liquidity Drain often appears after the Roadshow begins, because institutions start formally doing Portfolio Adjustment.
Qinbafrank: By IPO day, when part of the funds that failed to get allocations or were locked up is released, the situation will change again; but after listing, there will be a second layer of draining, which is that retail investors and other funds will think: since there is now the purest Frontier AI stock, why should I still buy indirectly through the Supply Chain?
Qinbafrank: So in the long term, Anthropic's listing can be very bullish for AI, but in the short term it can absolutely drain the entire AI sector first.
XVI. The AI Mainline Has Already Migrated: Memory → Optical / Software → Agent / AI for Science
Qinbafrank: There is another point I have always wanted to emphasize, which is not to form the fixed mindset that "AI = Semiconductor = Memory"; the AI mainline migrates at every stage.
Qinbafrank: In Q1 and Q2 this year, Memory was the strongest; after August and September, Optical began to strengthen, and at the same time a batch of software and SaaS names that had previously been beaten down very badly also began to emerge.
Qinbafrank: Why? Because after Model commoditization and Token Price come down, the beneficiaries are not only upstream Hardware.
Qinbafrank: The cheaper the Token, the easier the Deployment, and Cloud, Software, Enterprise Integration, and Agent will instead benefit more and more. The next stage may be AI for Science and Bio.
Qinbafrank: So what trend investing really needs to do is not stubbornly defend the original Winner, but continuously identify: where the next layer of profit pools is migrating.
Qinbafrank: Memory is not finished, and Semi is not dead either; it's just that the stage like the first half of the year when everything shot straight up may be over in the short term.
XVII. The High-Yield Era No Longer Means Broad Gains: From Now On the Market Only Rewards Real Cash Flow
Qinbafrank: There is also a big backdrop now, which is that we have already entered an era of relatively high yields.
Qinbafrank: Oil prices push up Inflation, fiscal deficits cause a lot of Treasury issuance, and AI Infrastructure itself needs to issue a large amount of Corporate Debt. All these things are competing for funds in the bond market together.
Qinbafrank: When Bond Yield is high, stocks naturally also have to offer a higher Risk Premium, which is why the market is becoming more and more selective.
Qinbafrank: If your Microsoft earnings have no obvious flaws, then the market gives you positive feedback; Google Cloud is clearly very good, but if your Search and FCF show problems, the market still sells first.
Qinbafrank: So when picking stocks going forward, you must clearly distinguish: which companies are only Narrative, and which companies' businesses have truly delivered, where Revenue is real, Cash Flow is real, and the growth space for the next 2 or 3 years can still be calculated.
Qinbafrank: If a large company gets its valuation killed because of short-term headwinds, I actually think that is a good thing, because it gives you a price to get back on board. High rates do not mean all stocks cannot rise; rather, the market will increasingly "pick the best among the best."
XVIII. What to Really Watch in 2027 Q1 Is Not Semi, but Whether AI Commercialization Has Continuously Decelerated
Mr. Z: If you look at 2027 Q1, what judgment are you most confident about?
Qinbafrank: The judgment I can make is that at least through Q1 next year, I think AI Commercialization is still accelerating and has not yet reached Saturation.
Qinbafrank: Because more and more enterprises are beginning to implement AI, everyone's Token Consumption will ultimately be reflected in Compute, and Compute is largely in the hands of Cloud Provider. So at least at the Cloud layer, I think growth is still there.
Qinbafrank: When would I really start to worry about the AI Thesis? It is not Semi falling 20%, nor some stock falling 30%. It is when you see the Growth Rate of Cloud Business at Microsoft, Google, and AWS clearly declining for two consecutive quarters.
Qinbafrank: That is when you have to ask: Have all the enterprises that should use AI already used it? Have users already spent about as much as they are willing to spend? Have the Workflow that can create economic value already begun to Saturate?
Qinbafrank: This is the origin of the entire AI Story. As long as people are still using it, enterprises are still paying, and Token Consumption is still growing, the big direction is not broken.**19. Why Crypto Fell in October: It's Not That Narratives Died, It's the TGA Draining Liquidity**
Mr. Z: Let's talk about Crypto last. It surged in September, then fell back in October—what's your take?
Qinbafrank: On the Crypto side, I think it's actually fairly clear—it's Liquidity. When we talked at the end of August, my framework at the time was that late August to mid-to-late September would be relatively good, and late September to October would see a pullback, because liquidity would start tightening.
Qinbafrank: That's basically how it has played out. The TGA (Treasury General Account) went from roughly $800–900 billion in August, climbing all the way up, and is now above $1 trillion—essentially, the Treasury is withdrawing Cash from the market.
Qinbafrank: So Crypto being weak in October isn't because all Narratives suddenly died—it's that the Money Supply environment is wrong.
Qinbafrank: What to watch next is simple: whether the TGA starts releasing in November. If the Treasury starts putting Cash back into the system and market Liquidity improves, then Crypto could naturally restart.
Qinbafrank: The framework I gave earlier was that liquidity would gradually bottom out around late October, and if it gets released again in November, the crypto market has a chance to turn around with it.
**20. The Biggest Main Theme for Crypto's Next Phase: Compliant ICOs + Tokenization, TradFi and Crypto Are Converging**
Qinbafrank: On the regulatory side, I don't think there's been no change—it's just that many things take time to land.
Qinbafrank: The first line is Crypto Financing. In the past, U.S. companies raising capital either did an IPO, a Regulation D private placement, or used exemptions like Regulation A and Regulation CF. But Crypto Projects couldn't raise capital from the U.S. public in a normal way for a long time, so everyone doing Token Sales blocked Americans.
Qinbafrank: Now regulators are studying a new digital asset financing exemption—essentially giving Crypto Projects a path to legally raise capital from the public in the future.
Qinbafrank: This won't land immediately, because there's still public comment and the regulatory process, but if it's actually implemented, I think it can be understood as a new kind of Regulated ICO.
Qinbafrank: The second line is Tokenization. Even if the CLARITY Act stalls at the congressional level, that doesn't mean regulators can't do anything. The SEC and CFTC, within their own authority, can still push forward rules on Custody, Tokenized Securities, DeFi, AMM, Derivatives, and so on.
Qinbafrank: I think the Tokenization trend is already irreversible. More and more Treasury, Stock, and Security will enter the On-chain Financial Market.
Qinbafrank: So what truly benefits long term isn't just "stock tokens" themselves, but the underlying Infrastructure, Settlement, Custody, Liquidity, and DeFi Rail. The ultimate direction of Crypto and TradFi isn't to replace each other, but to gradually become the same financial market.
Mr. Z: To wrap up this episode, let me condense Qinbafrank's framework into 6 judgments.
First, the U.S. economy is not in a full-blown recession for now; AI CapEx has become America's new infrastructure investment, but K-shaped divergence will continue to widen as AI replaces labor.
Second, the most critical macro variable over the next 1–2 months is still Oil, and it has now evolved from a pure crude supply issue into a dual problem of Shipping Cost + Refinery Capacity. Whether the U.S. and Iran can enter real negotiations after the midterm elections directly determines inflation and the Fed's next step.
Third, AI Commercialization hasn't peaked—rather, the time lag from the past year's CapEx is starting to pay off. Revenue at Microsoft, Google, and Anthropic all prove that enterprise Adoption is still deepening. The key next step is when Cloud Revenue Growth truly surpasses CapEx Growth.
Fourth, AI has entered the second half of the industry. Memory and Semi won't disappear, but the main theme will continue to migrate toward Optical, Cloud, Software, Agent, Enterprise Deployment, and AI for Science. In a high-rate environment, the market will only reward real Earnings.
Fifth, before the Anthropic Roadshow, there may still be a window of sentiment recovery, but once the Roadshow starts, large-scale institutional subscriptions will require selling other positions to free up Cash—the short-term Liquidity Drain is worth watching closely.
Sixth, Crypto's October correction is essentially still liquidity contraction; whether the TGA releases in November is the most direct Trigger, while the longer-term structural main theme is compliant financing and Tokenization driving TradFi Assets to fully enter On-chain Finance.
The signal that would truly turn the AI Thesis bearish isn't Semi falling another 20%, but Cloud Commercialization clearly decelerating for several consecutive quarters. Thanks to Qinbafrank for coming to 168X for the fourth time!