Author: Xinhui Technology

Mr. Fu Peng, Chief Economist of Xinhui Group, was invited to participate in the Wiki Finance EXPO Hong Kong 2026 and delivered a keynote speech. Starting from the global liquidity framework, Mr. Fu shared his core views on the current global major asset classes and market trends, and made a systematic judgment on the underlying logic of the crypto market.

Mr. Fu Peng, Chief Economist of Xinhui Group, delivers a speech at Wiki Finance EXPO Hong Kong 2026

Today I would like to share with you some perspectives on the global major markets from several dimensions. First, I will discuss liquidity. Regardless of the type of asset, as of today, including mainstream crypto assets, they are essentially all linked to the global core liquidity.

After the 2008 financial crisis, global liquidity reached a peak from 2008 to 2021. Liquidity cannot only be viewed through news of interest rate hikes or cuts; it can be broken down into three dimensions, and I urge everyone to remember this.

Interest rate hikes and cuts are merely changes at one end of the interest rate curve, and do not represent the complete liquidity environment. Observing liquidity can be broken down into three components: the volume of water in the pool, the temperature of the water, and the distribution of funding pressure within the pool. The underlying logic can be simply understood as P/Q×G. From a professional perspective, liquidity can be tracked from the interest rate side, the interest rate curve, the Federal Reserve's balance sheet, open market operations, etc. But there is a simpler way to observe: in the current traditional trading, people generally regard mainstream crypto assets like Bitcoin as leading indicators of liquidity strength or weakness.

01. From "Crazy Speculation on Poor Quality Assets" to "Contraction": Two Faces of the Liquidity Cycle

After the pandemic in 2020, the world entered a phase of low interest rates and extreme easing with central banks expanding their balance sheets. During this period of easing, global financial assets exhibited typical characteristics: rampant speculation on poor quality assets.

For example, the GameStop short squeeze in the U.S. stock market and the explosive rise of many scam coins in the crypto market are fundamentally results of excessive liquidity—when there is plenty of money, anything can be speculated on. However, when liquidity begins to contract overall, the market will experience a "contraction" trend: funds actively distinguish between good and bad assets, and poor quality targets are abandoned by capital. This process of squeezing bubbles started in the second half of 2021.

From the second half of 2021 to 2022, a typical case is Bitcoin in the crypto market falling from over $70,000 to around $20,000, while NVIDIA's stock in the U.S. fell about 64% to 65% throughout 2022. This process is like squeezing water out of a sponge, continuously expelling market bubbles.

This year, the real core turning point occurred last November. The year 2021 was the peak of the central bank's balance sheet expansion, and the end of last year was a critical juncture where liquidity faced the dual tightening of balance sheet reduction. Simply put: water draining from the pool does not mean that funds become tense the moment balance sheet reduction starts; it is only when the balance sheet reduction progresses to a certain degree that the market will genuinely feel the pressure of funds.

Last November and December, Bitcoin was around $110,000, and I even made a bet with Li Lin that within a year, the crypto assets would likely be halved. If this comes true, it would once again confirm that the underlying logic of crypto assets is entirely bound to global liquidity.

Last November, the core indicator to observe was the Federal Reserve's SRF open market operations. This indicator represents that after the balance sheet reduction reaches a critical threshold, structural funding pressure has appeared in the market. It’s essential to understand that when banks tighten credit and liquidity contracts, it does not mean that everyone is short of money; funding pressure is transmitted in layers: high-leverage, weak-quality entities are the first to face funding breaks, while high-quality core entities still have ample funding. Layered transmission of liquidity will lead the capital market to experience "contraction": funds first sell off weak assets that are sensitive to liquidity and continue to concentrate on the most core, highest certainty assets.

Many retail traders in crypto have a misconception: that when there is no market in crypto, funds all rush to speculate on U.S. stocks. This is very retail-oriented thinking. Objectively speaking, in a tightening liquidity cycle, funds will first clear out all high-elasticity, high-speculation assets from their portfolios; cryptocurrencies and small-cap thematic stocks will be prioritized for reduction. When there is a lot of money and liquidity is loose, funds are willing to speculate on all kinds of junk assets; when money is scarce and liquidity tightens, funds will focus only on genuinely valuable core assets, which is the essence of the "contraction" market.

02. Key Turning Point for the AI Industry: Free Cash Flow to Zero, Capital Expenditure Narrative Completely Ineffective

Since last November, global capital has continued to gather around the main line of long-term productivity upgrade, namely the artificial intelligence (AI) sector. The logic of the AI sector can be likened to significant investments in fixed assets, which is easier to understand through examples of domestic infrastructure.

In 2001, the core topic in the market was large-scale infrastructure development within the country, as the old saying goes: "To get rich, first build the roads." At that time, Lin Yifu and Xie Guozhong were discussing the role of highway and railway infrastructure in driving the economy. By 2002, the two sessions confirmed the development direction of infrastructure, and in 2003, central finance and land finance matching funds were all landed, entering a long-term capital expenditure cycle. In 2004, the core targets of institutional allocation included upstream equipment and raw materials enterprises such as Sany Heavy Industry and Anhui Conch Cement.

The analogous logic fits perfectly with the current AI sector. It will not change the underlying rules of the industry just because it is labeled "AI." The first half of AI was driven by applications like ChatGPT, which spurred corporate capital expenditure willingness; starting in 2023, global tech companies focused on digital infrastructure, which includes computing power and data center construction. Large-scale digital infrastructure development will benefit upstream hardware, storage, optical modules, HBM, etc. Samsung Electronics, SK Hynix, and TSMC correspond to the concrete, cement, and construction machinery of the infrastructure era. However, the second quarter of this year marks a critical turning point for the entire industry chain, compounded by the dual variable resonance of liquidity contraction.

After the release of Google's earnings report yesterday, mature investors could noticeably detect that the market's central logic over the past two to three years has become ineffective. For 2023, 2024, and 2025, the market rules are simple: if big internet companies increase spending on AI infrastructure and expand capital expenditure, the market gives high valuations. However, after the earnings disclosures of major companies in the second quarter of this year, even if capital expenditures continue to grow at a high rate, stock prices have fallen instead.

The core reason is that investors have captured a crucial piece of data: all leading companies heavily investing in AI infrastructure have seen their free cash flows drop to zero. Last night, the most critical indicator from Google's earnings report was free cash flow. Many investors are still clinging to the old logic, thinking that as long as capital expenditure continues to grow, stock prices will rise; that era has ended.

The market pricing logic has completely switched: previously, it was about the scale of capital investment; now capital will question whether infrastructure investment can bring sustained traffic and revenue realization to break even. Free cash flow dropping to zero is a symbolic signal of the AI industry's transition from the first stage to the second stage. If companies plan to continue adding capital expenditure, they can only rely on external financing through issuing stocks or bonds, and external funds come with costs, leading investors' review standards to become extremely strict.

Here’s a refined tracking indicator: the ratio of capital expenditure (CapEx) to cloud business revenue growth rate. Currently, Google has this ratio at about 1.9, which means that for every 1.9 units invested in infrastructure, it can only create 1 unit of cloud business revenue; this is the core reason why the capital market is unwilling to continue providing high valuations.

With the overall funding environment tightening, global capital continues to seek out a few high-certainty assets, compounded with the switching of the industrial cycle, the "contraction" market will inevitably experience significant risk fluctuations.

I will take NVIDIA as an example to thoroughly outline the industrial cycle: 2022 marked the starting point of NVIDIA's industrial cycle confirmation, as that year its market value fell from one trillion to around 100 billion; after the explosion of ChatGPT, NVIDIA officially entered the value growth stage. In 2023 and 2024, NVIDIA's logic will be completely closed-loop: continuous growth in performance, global AI capital expenditure driving order expansion, market value successively breaking the one trillion, two trillion, three trillion thresholds, with extremely low stock price volatility and virtually no risk of deep correction.

But after I returned from research in Singapore in June 2024, I warned various financial institutions about the risks: NVIDIA's business operations, industry supply-demand, and industrial fundamentals are all fine, but the risk all comes from off-market financial leverage.

Now, a significant cognitive misunderstanding exists among the new generation of 00s investors, who believe that stock price movements must completely align with fundamentals. This viewpoint is entirely erroneous! The capital market prices and trades based on market expectations, which will significantly lead actual corporate fundamentals.

To give an example: the current state of the industry is that there is a shortage of HBM production capacity and supply, the fundamental fact is indisputable, but it cannot be inferred that stock prices will continue to rise. This is a typical cognitive bias: financial reports and production capacity reflect the current reality, while stock prices trade future expectations, making fundamental data severely lag behind market pricing. This is my practical experience over more than twenty years in the industry.

In July 2024, NVIDIA’s stock plummeted 20% within just a few trading days, while the Japanese stock market fell 10% in one day. At that time, many researchers published reports attributing this decline to the Bank of Japan's interest rate hikes and the unwinding of yen carry trades, which was just the surface explanation. The underlying truth is: global capital is concentrating into a few certainties assets, extreme certainty breeds extreme greed, directly manifesting as investors madly increasing leverage.

For a more relatable trading analogy: let's say we are playing cards, you have a 6 and I have a 5, and you clearly know your hand is better. Regular retail investors would enter heavily, but a qualified trader would leverage fully and bet everything. Core conclusion to remember: certainty breeds greed, everything has two sides; the operation corresponding to greed is increasing leverage. The market uniformly anticipates that NVIDIA has sufficient long-term orders, traders will continuously add leverage to amplify their returns; this is instinctive for traders. Once leverage accumulates to a critical point, it will inevitably trigger severe volatility and rapid declines.

The current market is replicating similar scenarios: some memory chip stocks have no industry negatives, stable business operations, full orders, and steadily growing performance, yet their stock prices frequently plummet. Many young traders in the Korean market have seen significant profits one day, only to experience massive losses the next day. The root of the problem lies not with Samsung or SK Hynix, nor with the supply and demand of the HBM industry, but in the excessive accumulation of market leverage.

The underlying logic is completely consistent with NVIDIA's flash crash in July 2024: high certainty assets create leverage bubbles; once leverage touches the critical point, it will inevitably collapse. There is no such thing as a permanently sustainable leverage market. Here’s a straightforward risk assessment standard: when young individuals who have just graduated, with no practical experience, pour all their funds into leveraged bets on Samsung or SK Hynix, it signifies that risks are approaching. An originally niche professional field is flooded with speculative retail investors; the bubble burst is merely a matter of time. For the past few years, the market has been in a liquidity contraction cycle, and everyone is clear on a few core certain assets, but the risk lies not in the industry fundamentals, but hides in liquidity and leverage layers; this must be a point of caution.

The current market has reached a core critical point in the first stage of AI; the narrative logic relying solely on capital expenditure expansion has run its course. The market will face substantial volatility and valuation adjustments. Regarding the overall judgment on U.S. stocks this year: the index maintaining sideways oscillation is already an optimistic expectation. Some might counter: after the significant drop in March, U.S. stocks rebounded again in May and June. However, it must be understood that the upswing in May and June belongs to extreme structural trends, with only a very few individual stocks driving the index higher while the vast majority continue to decline.

The A-share market has exhibited a completely identical structure over the past year: 55% of individual stocks are priced below the corresponding levels of 3000 points, wholly reliant on a few leading stocks in the AI sector to support the index.

In summary, the current market environment is characterized by tightening liquidity, extreme differentiation, and a key turning point in the AI industrial cycle. I emphasize again: the long-term development logic of the AI industry has not changed, and productivity upgrade is the definite main thread, but it is essential not to hold specific targets blindly for the long term; it is necessary to lay out in phases using a complete industrial cycle approach.

I constructed a comprehensive five-layer analytical framework: industry layer, economic layer, inflation layer, liquidity layer, market layer. At present, there is no need to invest significant effort breaking down the economic layer; the industry layer, liquidity layer, and market layer are the core for analysis while the importance of analyzing macroeconomic factors has greatly decreased. Some may ask if it is still necessary to dissect the U.S. economy in-depth? The answer is absolutely not necessary. The reason is that U.S. companies are continuing to expand capital expenditures on a large scale, and the household sector completed deleveraging way back in 2008. In other words, there is no need to scrutinize high-frequency economic data; the core feature of the U.S. economy can be summed up in two words: resilience.

03. Global Market Landscape: The Only Main Thread is AI

From a market perspective, the only main thread globally is artificial intelligence; currently, global capital only follows this core investment logic. Looking at globally allocatable assets, the future core markets are limited to Japan, Korea, Taiwan, mainland China, and the United States; all other regions have extremely low allocation value, with only ASML in Europe being of concern, as other targets hold no significance for allocation.

You might ponder two questions: does the current movement of the Korean stock market relate to the domestic real economy? It is completely unrelated. Now, what about the Japanese stock market? Is it linked to Japan's domestic economy? Also, completely unrelated. Upon closely examining the core assets of the Japanese stock market, they are all upstream equipment manufacturers within the AI industry chain. The market largely focuses on Samsung and SK Hynix, but the core production equipment sourced by both companies is all from Japanese firms, fully integrating the upstream and downstream of the entire industry chain. The core target in the Taiwan region is limited to TSMC, with no other core industry companies of allocative value.

The entire AI sector fundamentally represents an industrial investment driven by productivity, possessing fixed cyclical operating laws. I will clarify the core conclusion: the point where leading tech companies' free cash flows turned to zero in the second quarter is a major turning point for the market. Before and after this inflection point, the entire asset pricing logic in the market completely reverses; this point must be reiterated.

The AI industry chain is divided into upstream, midstream, and downstream, with each segment possessing independent industrial life cycles and clear sector rotation and allocation window periods. Do not consider AI as a blind faith for long-term holdings; merely speculating on AI concepts will certainly lead to pitfalls. Many people ask me if I am pessimistic about AI; this question itself has a logical loophole. Over the past decade, the market has reached a consensus: artificial intelligence is the core main thread of the next generation of productivity, and there is no dispute about this fact. Believing in the sector does not mean mindlessly holding onto a single target at any time. As an upstream hardware core target, NVIDIA's high-speed growth cycle has already concluded, transitioning to a mature blue-chip phase starting in 2025, thus seeing a significant narrowing of gains from last year to this year. There is no need to wait too long; Samsung and SK Hynix will also enter their mature cycles, and the overall growth rate of the upstream hardware sector will slow down, with growth progressively shifting toward downstream segments.

The complete rhythm forecast for the AI industrial cycle is: 2022 saw the upstream hardware leading the market, by 2026, the software layer will undergo valuation digestion and reshaping, and around 2030 the terminal application layer will experience valuation adjustments and re-pricing. The full-scale AI industrial cycle lasts about 20 to 25 years; we have already completed 10 years, with the first ten years centered on upstream hardware infrastructure, and the next ten years focused on terminal applications.

However, currently, there exists a cyclical disconnection, with the next 10 to 18 months belonging to an industrial switching window period. During this window period, do not go all in on speculative positions; strictly adhere to the rules of the industrial cycle for phased allocations to avoid major volatility risks. Here are two key concepts to differentiate: from the programmer's perspective, AI code tools and development auxiliary tools both belong to the industrial supporting tool layer, not the terminal application layer, and the valuation logic for these two differs significantly.

04. Karen Walsh and the Shift in Liquidity Paradigm: Central Banks No Longer Providing a Backstop, Crypto Assets Reach Maturity

Lastly, I will return to the topic of liquidity, which is also a highly relevant core variable related to crypto assets. Why is the newly appointed Federal Reserve Chair Karen Walsh a key signal? Her appointment signifies a complete rewriting of the core policy framework established by Bernanke after the 2008 financial crisis.

I wrote an analysis note in January: this personnel adjustment signifies that central bank policy is returning to pre-2008 old paths. Briefly outlining the policy background: the 2008 financial crisis exposed enormous systemic financial risks. Policymakers learned from the experiences of the Great Depression in 1929: completely laissez-faire markets cannot stabilize themselves during a crisis. Therefore, after 2008, Keynesian stimulus policies were widely implemented globally.

Bernanke and Yellen, former chairs of the Federal Reserve, all followed the same core idea: after the outbreak of a financial crisis, central banks must step in to stabilize the market. However, any policy carries duality, similar to leverage logic in investment. Leverage can quickly amplify returns; it can also lead to complete account liquidation. Central bank interventions can quickly calm market fears, preventing a repeat of Great Depression crises, but long-term unconditionally backstopping the market by central banks breeds speculative sentiment, leading to large-scale asset bubbles.

In the market, there is a professional term called "the Fed put option": as long as the market declines, funds dare to buy recklessly, and all traders bet that the central bank will intervene to rescue them. When the market forms a unified expectation that gains belong to investors while losses are underwritten by the central bank, all financial assets tend to become severely overvalued.

The core message of all of Karen Walsh's public speeches can be summarized in one sentence: the central bank only performs its legally mandated duties. The central bank has two core legal objectives: stabilizing employment and controlling inflation; it will not regularly support the stock market. Nowadays, technological progress continues, and productivity steadily improves; the central bank is now in a position to exit the long-term backstop model.

This can be compared to parental guidance: when children enter high school and demonstrate independent survival skills, parents cannot do everything for them as it will foster dependency. After Karen took office, many market participants misinterpreted it as a sign for interest rate cuts or a return to quantitative easing; the core focus is actually on balance sheet reduction, which has lower correlation with short-term interest rate movements. The core issue is how to orderly complete the balance sheet reduction while returning the central bank's function to a standard position prior to 2008.

This signifies that the largest global liquidity easing cycle in human history, which has been in place since 2008 until Karen's appointment, has thoroughly ended. Thus, do not harbor fantasies: over the next 5 to 10 years, we will not replicate the comprehensive flood of liquidity from 2008 to 2026 that saw all asset categories rise simultaneously. Funds will flow back to genuinely long-term valuable core assets; this is a crucial turning point in terms of liquidity, which will comprehensively rewrite everyone's investment strategies. Investment logic has shifted from previously diversifying allocations with simultaneous rises in various assets to now focusing on a small number of high-quality core targets.

The crypto market will also undergo similar changes. Many traders have observed: Bitcoin and Ethereum's market capitalization is gradually stabilizing, volatility is continuously declining, market liquidity is tending to stabilize, and the participants are becoming more institutionalized. The above characteristics are all typical manifestations of core assets surviving after the clearing of bubbles. The narrative logic of the past, which revolved around speculating on air crypto coins, has completely lost its effect.

Liquidity is the topmost core influencing factor for all financial assets; this year everyone must thoroughly grasp this analytical logic. Following the liquidity framework to further dissect the fundamentals of industries and companies will clarify the analysis of various assets considerably more. Today, time for sharing is limited, and I cannot go through the five-layer analytical framework in detail.

I hope to exchange underlying logic and analytical methodologies with you; once the foundational thoughts are well-organized, observing short-term micro fluctuations in the market will not lead to excessive entanglements. My sharing concludes here, and I hope it brings you inspiration. Thank you all.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。