Hua'an Xu Hantian: The core contradiction of the AI wave is "the business model is yet to be resolved," with a focus on two main lines in chips

Recently, Xu Hantian, a former Texas Instruments (TI) employee and current manager at Huashan Fund, reflected on his transition from the industrial front line to financial investment in a deep dialogue with Wall Street Insights. He fully revealed his core investment philosophy: a rigorous framework driven by macro "industry prosperity" and micro "entrepreneurial spirit." He also provided profound insights on the current market's most pressing issues and analyzed the essence of the "bubble" in the AI field.

Regarding high-growth areas such as AI chips, he believes that static valuation is a thing of the past; the key lies in the imagination of "ultimate market value" and the grasp of growth "certainty."

He judges that the current AI boom is not a true bubble because its underlying technology is still rapidly iterating, and the core contradiction is "business model yet to be solved," rather than "lack of core value."

The investment class representative summarized the key points as follows:

1. The investment framework is extremely clear: first, look for high-prosperity tracks in an upward cycle from the top down ("where to play"); then, find entrepreneurs with strong resilience, objective awareness, and exceptional execution within the track ("who to bet on to win"). The correctness of direction (prosperity) determines the lower limit of returns, while the entrepreneur's vision and capability determine the upper limit of returns.

2. Technology itself is not an unbreakable barrier; it only earns you a valuable time window. The real moat is the ecosystem, data flywheel, user stickiness, and brand recognition built rapidly during this window period using technological advantages. Therefore, prioritize companies with clear strategies that are committed to building sustainable moats during the window period.

3. For growth stocks, discussing the current static PE is of limited significance; the key is to estimate their potential market value at the industry's endgame and assess the certainty of their growth. As long as the future space is large enough and the path is clear, temporary high valuations can be tolerated.

4. The core of a bubble is that growth cannot fulfill its valuation. The current core driving force of the AI industry—the technological iteration of large models—is still accelerating, and product capabilities continue to break through. Unlike the internet bubble of the past, which was only conceptual, AI now has solid products and a large user base; its core contradiction is "business model yet to be proven," rather than "product value does not exist."

5. Firmly optimistic about the long-term prospects of domestic production, external restrictions are a "growing pain" in the short term but a "huge benefit" in the long term. Any substantial breakthrough in the domestic manufacturing chain in the next year or two will be an extremely significant investment opportunity.

6. Leave professional matters to professionals. Instead of spending energy on timing with low win rates, it is better to focus research on selecting a fund manager with a clear investment philosophy, stable framework, and long-term trustworthiness. Choosing an excellent "race car driver" is far wiser than trying to drive on a bumpy track yourself From the front line of Texas Instruments to the investment decision-making room of Huazhang Fund, Huazhang Fund manager Xu Hantian has completed the identity transition from "chip maker" to "chip selector." His investment framework is deeply rooted in the soil of industrial logic and seeks Alpha in the wave of AI and domestic substitution with a high degree of certainty.

He imagines himself as an "industry CEO." Once he confirms a significant turning point in the industry and the core competitiveness of a company, he will invest heavily with a high degree of certainty.

The following are the key insights organized by the investment workbook representative (WeChat ID: touzizuoyeben), shared with everyone:

Three Profound Understandings from Transitioning from Industry to Financial Investment

Q1: Wang Xu, Vice President of Wall Street Insights (hereinafter referred to as "Wang Xu")

Hello, Mr. Xu. You have a strong semiconductor industry background but decisively transitioned to financial investment during your career ascent. Can you help us understand the deeper reasons that drove you to make this critical decision?

A: Xu Hantian, Fund Manager of Huazhang Fund (hereinafter referred to as "Xu Hantian")

This choice stems from three profound understandings I gained on the front line of the industry:

First, in the product marketing department, I established a "demand-driven" mindset.

I was responsible for collecting the demand for next-generation chips from leading smartphone manufacturers. I found that all successful chips are not developed in isolation but are based on precise insights into downstream demand and proactive planning. This concept remains my primary criterion for judging whether a technology company can succeed.

Second, in large customer sales, I realized the extreme importance of "foresight."

In 2018, I was responsible for BYD's business, and at that time, TI's BMS (Battery Management System) chips were not the strongest in the market. However, we keenly captured the customer's forward-looking demand for "functional safety" and ultimately won key projects through deep cooperation.

This made me understand that whether in business or investment, one must think half a step ahead of the market to consider the industry direction for the next year or two.

Third, in charge of channel management, I discerned the significant weight of "opportunities of the times."

In managing hundreds of small and medium-sized customers, I witnessed many excellent companies and individuals whose success or failure often depended not on their own efforts but on whether their leaders seized the tide of the times.

I deeply realized that individual struggle has limited influence in the face of grand industrial trends; choice is more important than effort.

These three experiences have given me a deeper understanding of the essence of business. I found that I inherently enjoy researching the logic of "things" rather than dealing with complex "interpersonal" relationships. At the same time, I yearn to stand from a more macro perspective to discover and empower those enterprises that can adapt to the times and possess foresight.

Investment is precisely the best path to realize this ideal. It satisfies my curiosity about diverse industries and allows me to return to my original intention of serving the industry in the form of capital.

Investment Holy Grail: The Dual Drive of "Economic Prosperity" + "Entrepreneurial Spirit"

Q2: Wang Xu

How have these valuable industry experiences specifically internalized into your unique advantages in investment research and stock selection decisions? A: Xu Hantian

These experiences have completely shaped my current investment framework, which is built on three main pillars:

1. Economic prosperity as the guideline, individual stocks as the target: My primary job is to conduct top-down searches for high-prosperity sectors. I believe that the correctness of direction far outweighs the perfection of details.

2. Entrepreneurial spirit as the core: I place great importance on the entrepreneur's vision, ambition, and execution ability, which are decisive factors in how far a company can go.

3. Industrial logic as the measure: I always measure a company's value and potential through the lens and logic of the industry.

This framework has also evolved continuously through practice. When I first entered the industry, I missed opportunities due to my "industrial purism," as I couldn't understand why the capital market gave high valuations to some companies with mediocre fundamentals. This "tuition fee" taught me that the capital market trades on expectations and hopes.

My framework has been validated in several key decisions over the past few years:

In 2022, I discovered a leading stock in communication chips: When the AI concept was not yet widespread, I proactively identified a core upstream chip company based on the judgment that "enterprise-level demand is superior to consumer-level."

In 2023, I heavily invested in semiconductor equipment: In the pessimistic atmosphere of declining industry capital expenditures, I saw the significant certainty brought by the intensifying Sino-US competition leading to domestic substitution "from 0 to 1," and I decisively recommended the entire industry chain.

In 2024, I bet on a domestic AI chip dark horse: When the market generally questioned a company's capabilities, I found through in-depth research that its "customer-oriented" approach completely aligned with the successful patterns of the industry, ultimately leading this company to grow from a market value of hundreds of billions to a trillion-dollar leader.

The commonality of these decisions is that I imagined myself as the "CEO of the industry." Once I confirmed the significant turning point of the industry and the company's core competitiveness, I would invest heavily with high certainty.

Industry Perspective: Technological leadership is just a "window period," and the real barriers are built during this time

Q3: Wang Xu

You mentioned that you would layout the entire industry chain, so how do you accurately select the final "winner" in a large sector? Additionally, how do you convert the "entrepreneurial spirit" and "execution ability," which you value highly, from subjective feelings into assessable objective indicators?

A: Xu Hantian

First of all, my layout of the industry chain is not "sprinkling pepper," but rather constructing a "leading company portfolio in the industry chain." I strive to find the company with the greatest potential to be a leader at every key link in the industry chain (such as upstream equipment, midstream design, and downstream applications). Because I believe that in any mature industry, there will ultimately be only one absolute king.

How to find this king? The core is to deeply understand what the "barriers" of this sector really are.

I believe that technological leadership itself is not a barrier; it merely earns you a valuable time window. The real barrier is the ecosystem, data, user stickiness, and brand recognition built during this window period using technological advantages.

Therefore, my job is to identify among numerous competitors who is most likely to establish these real, sustainable barriers during the window period through their strategy and execution As for how to judge entrepreneurial spirit and execution ability, I mainly look at three points:

1. Resilience: I prefer entrepreneurs with a strong fighting spirit and an unwillingness to give up. Anyone can succeed in favorable conditions; the response in adversity is the true test.

2. Objectivity: Entrepreneurs cannot live in their own world. I trust those who have a clear understanding of their own abilities and the market landscape, and who can candidly communicate challenges.

3. Closed-loop verification: Execution ability is not about what they say, but about what they do. I will conduct long-term, continuous tracking, constantly examining whether the strategic goals they set previously are being realized and whether project progress meets expectations. A significant mistake or a successful crisis management is far more revealing of an entrepreneur's true caliber than ten roadshows.

Q4: Wang Xu

In your investment, how do you view companies and entrepreneurs that seem to have strategic swings and appear "inconsistent in words and actions," yet are highly resilient?

A: Xu Hantian

This is a very good question and touches on an important evolution in my investment philosophy.

I once held a prejudice against a company that frequently changed its track, believing it was unfocused and unreliable. However, after in-depth tracking, I found that its main business had hit a ceiling, and the entrepreneur was actively seeking a "second growth curve" for the company. He had put in his utmost effort in every new direction, but for various reasons, he had not succeeded. However, he did not become discouraged; instead, he quickly adjusted and ultimately led the company to a new level through a successful acquisition.

This case made me deeply reflect: it is essential to distinguish between "opportunistic chasing of trends" and "resilient strategic exploration." The former is merely to tell a story in the capital market, while the latter is for the survival and development of the enterprise, seeking certainty in uncertainty.

My evaluation criteria have thus been optimized; the key is not whether he has changed tracks, but rather:

1. What is his original intention? Is it for speculation or to break through?

2. In every track, did he put in his utmost effort to execute?

As long as he is genuinely fighting for the future of the enterprise and has demonstrated solid execution at every stage, then this "flexibility" and "resilience" are extremely valuable entrepreneurial traits.

Valuation Philosophy: Abandon Static Valuation Anchors, Embrace "Endgame Thinking"

Q5: Wang Xu

You initially did not understand high valuations, but later dared to give a company with a market value of hundreds of billions a forward target of 600 billion. What is the leap in valuation philosophy behind this? What is the underlying logic that supports such a bold judgment?

A: Xu Hantian

My valuation philosophy is not a complete overhaul but a deepening of dimensions. I have always adhered to fundamentals, but my understanding of "fundamentals" includes not only static financial data but also judgments about the certainty of the company's dynamic growth.

I believe that discussing static valuations for growth stocks is almost meaningless. My valuation framework is based on two core questions:

1. End-game Valuation: How large can this company ultimately grow in the vast river of its industry? What is its ceiling? I will depict a future "picture" of it through various methods such as market space estimation, competitive landscape simulation, and steady-state price-to-earnings ratios
2. Certainty of Growth: From now to the "endgame," is the company's growth path clear? How high is the certainty? This depends on its technological positioning, ecological barriers, management capabilities, and iteration speed.

The reason I dare to set a target of 600 billion is based on my deep conviction regarding these two points. Through rigorous calculations, I have confirmed that its market space is broad enough; through understanding the industry, I judge that its competitive advantages are sufficient to support it in obtaining high market share and high profits; through tracking the management team, I believe they have the capability to turn the blueprint into reality.

When the certainty of growth is high enough, it deserves to have a seemingly "expensive" present. This is not a bubble, but rather a pre-pricing of future high growth potential.

Localization: Any substantial breakthrough in domestic manufacturing chains in the next one to two years will be a significant investment opportunity

Q6: Wang Xu

Returning to the chip industry, how are the profits and discourse power distributed across different links in the industrial chain? How is our progress in chip breakthroughs?

A: Xu Hantian

The discourse power in the industrial chain is dynamically changing, with the core principle being "scarcity is valuable."

• In the early stage of technology from 0 to 1, the companies that can define product design are the most scarce and have the strongest discourse power.

• When design companies flourish, the upstream manufacturing link (such as wafer foundry) that can provide stable capacity becomes the bottleneck, and discourse power shifts.

• When capacity is no longer scarce, new designs and applications that can accurately grasp the next generation of demand will again become important.

In the current AI chip field, the core bottleneck has shifted from the "design" end to the "manufacturing" end, as well as the upstream of manufacturing, namely equipment and materials. Due to obstacles in obtaining overseas advanced processes, self-controlled advanced manufacturing capabilities have become the key to determining the height of our AI industry development. Therefore, any substantial breakthrough in domestic manufacturing chains in the next one to two years will be an extremely significant investment opportunity.

Q7: Wang Xu

NVIDIA founder Jensen Huang also admitted that its market share in China's high-end chip market has dropped to zero. Do domestic manufacturers have the capability to seize this huge market?

A: Xu Hantian

We need to look at it from two levels:

• Inference chips: For the part of the market used for AI application execution, domestic manufacturers are fully capable of taking over and are rapidly replacing.

• Training chips: For the top computing power chips used to train large models, we indeed need more time to catch up.

This will temporarily affect the training speed of our top large models, but we are also compensating through compliant overseas computing power layouts and other means. The core is that as long as we do not stop exploring large models and chip research and development, this huge market will ultimately be filled by local enterprises.

Q8: Wang Xu

How much do you rely on AI tools in your daily investment research? To what extent has it changed your working mode?

A: Xu Hantian

I am a heavy user of AI tools. I am optimistic about a track, provided that I am deeply using it myself.

AI has greatly improved my work efficiency: ·Information Acquisition: Upgrading from "Search" to "Q&A," AI can provide research information in a more structured and in-depth manner.

·Content Creation: AI becomes my "writing assistant," efficiently completing detail filling and language polishing after I establish the core logic.

·Meeting Minutes: AI's real-time transcription and key point summarization functions free me from tedious recording, allowing me to focus more on thinking and communication during meetings.

AI is reshaping the productivity of knowledge workers; not embracing it means falling behind.

AI Bubble Analysis: Continuous Technological Iteration Means It's Not a True Bubble

Q9: Wang Xu

There are concerns in the current market about an "AI bubble," especially regarding the massive infrastructure investments and high stock prices. How do you assess whether the current enthusiasm in the AI field is a healthy boom or a dangerous bubble?

A: Xu Hantian

The core metric for judging a bubble is: whether the speed of industry iteration can keep up with the speed of valuation expansion.

As long as the capabilities of large models continue to evolve at an astonishing pace, the core engine driving the industry forward is still running. This is exactly what we see at present.

What everyone is worried about is actually the "monetization" challenge. This is fundamentally different from the internet bubble of the past. The internet bubble was "story (concept) > product," while AI is "product > business model." AI already has a solid product foundation and a large user base; it just hasn't found the optimal, scalable monetization path yet.

This is more like an arms race initiated by top players to seize the entry point of the next era. They are investing heavily to pave the way for future "harvesting." As long as the trend of technological iteration does not reverse, it is only a matter of time before the business model is successfully established.

Of course, we need to track this dynamically. If the monetization path remains unclear in the coming years, we need to be wary of structural risks and focus on the segments of the industry chain with the highest certainty.

Chip Industry Chain: Focus on Two Main Lines

Q10: Wang Xu

In the process of catching up, what are the main differences and development paths between domestic AI chips and overseas giants?

A: Xu Hantian

In the field of AI chips, chips serve as "tools" for large models. Currently, the iteration paradigm of global large models is still led by overseas players, so the core strategy for domestic AI chips at this stage is "following" and "compatibility."

You must be compatible with mainstream software ecosystems (such as CUDA) to allow downstream users to migrate and use smoothly. At this stage, taking a unique approach or starting from scratch carries a high risk.

Of course, in some specific areas where we are "choked," we may instead trigger "asymmetric innovation" due to limitations. For example, in the storage chip field, we have achieved performance catch-up through innovative architectural design despite process constraints.

In summary, for general computing platforms like GPUs, our path is "first follow, then seek to surpass"; while in some specialized fields, we have already begun exploring innovation paths with Chinese characteristics Q11: Wang Xu

Looking ahead to the next few years, which sub-sectors in the entire chip industry chain do you think are most worth paying attention to in China?

A: Xu Hantian

I focus on two main lines:

1. AI-driven global technology innovation chain:

The "highway" of computing power infrastructure: As the demand for computing power and connectivity speed from large models grows exponentially, high-spec PCBs, high-speed optical modules, and CPO (Co-Packaged Optics) will continue to benefit.

The "new species" of edge AI hardware: AI-native applications will give rise to entirely new hardware forms, such as AI PCs, AI phones, and AI glasses. These "new species" will bring new opportunities in the industrial chain.

1. Self-controllable driven domestic substitution chain:

This is a highly certain long-term opportunity, covering everything from upstream equipment and materials to midstream EDA, IP, design, and manufacturing, and down to downstream advanced packaging and testing. There is enormous room for domestic substitution throughout the entire chain.

Portfolio Construction: "T-D Portfolio" - Leaders as the Foundation, Flexibility as the Engine

Q12: Wang Xu

Can you elaborate on the "T-D Portfolio" strategy you mentioned during the roadshow?

A: Xu Hantian

The "T-D Portfolio" is a figurative expression of how I construct an investment portfolio to balance certainty and flexibility:

40% (Foundation Position): Invest in large-cap companies that have industry "leader" status or potential; they serve as the "stabilizers" of the portfolio.

40% (Flexible Position): Invest in "enablers" (commonly known as "shovel sellers") that provide key supporting solutions for the leaders. These companies are usually smaller in market capitalization but have high technological barriers and significant marginal changes, serving as the "engine" for excess returns in the portfolio.

10% (Observation/Left-side Position): Allocate to some early-stage directions that are on the verge of technological breakthroughs but have not yet fully clarified their paths. This part is the "seeds" sown for the future.

10% (Liquidity/Others): Retain some cash or allocate to other assets as a "regulator" for the portfolio.

Q13: Wang Xu

In conclusion, could you summarize your investment framework in a few sentences?

A: Xu Hantian

My investment framework can be summarized as one center and five levels:

One central belief: Firmly believe that China's technology industry is in a great, globally resonant innovation cycle.

Five levels:

1. Concept Level: Be a friend of industrial trends, embracing the two historic opportunities of AI and domestic substitution.

2. Strategy Level: Use "prosperity" as a compass to top-down search for the fastest-growing and most logical tracks.

3. Stock Selection Level: Delve into industrial logic to find companies with strong "entrepreneurial spirit" and clear "barrier-building" paths.

4. Risk Control Level: Treat drawdown control as a touchstone of research and investment capability, dynamically optimizing the portfolio through continuous verification of fundamentals and market sentiment.

Q14: Wang Xu

Tech stocks are highly volatile. How do you implement the idea that "drawdown control is a reflection of research and investment capability" into specific investment operations? A: Xu Hantian

I view drawdown control as a dynamic, research-based feedback system rather than a simple stop-loss discipline.

1. Prediction and Assessment: When building a position, I have a clear projection of a company's upside potential and driving factors. As the stock price rises, I continuously assess whether the current price has over-leveraged the future. If market sentiment far exceeds the fundamentals, I will proactively reduce my position to realize profits.

2. Reflection During Declines: When my holdings experience a significant pullback, it serves as a stress test for my research framework. I will immediately reflect: Is my core logic wrong, or is Mr. Market wrong? If my judgment logic remains solid, and the decline is merely an emotional release, it actually presents a good opportunity to increase my position.

3. Distinguishing Sources of Profit: I will clearly define what I primarily earn from investing in a company—whether it is from earnings growth (EPS) or valuation increase (PE). For the latter, when the macro environment or market risk appetite reverses, I will more decisively reduce my position.

Through this system, I strive to ensure that every adjustment to the portfolio is well-founded, thereby achieving more robust long-term returns.

Advice for Investors: Abandon Ineffective Timing and Choose a "Race Car Driver" You Trust

Q15: Wang Xu

The last question, for ordinary investors facing high-volatility industries like semiconductors, what investment advice do you have?

A: Xu Hantian

My core advice is: Leave professional matters to professionals.

First, ordinary investors need to clearly recognize their circle of competence.

If you happen to be a practitioner in this industry and have a unique understanding of the industry's rhythm, then moderate timing is feasible.

But if you are an outsider trying to predict short-term fluctuations in a complex industry, the difficulty is no less than predicting the weather.

Rather than wasting valuable energy on low-probability timing, it is better to invest it in choosing a professional fund manager you truly trust.

This is similar to how we value entrepreneurs when selecting stocks. Look for fund managers with clear investment philosophies, stable frameworks, and consistent actions, and then grant them long-term trust. This is the most efficient and secure way for ordinary investors to share in the growth dividends of the technology industry.

Source: Investment Workbook Pro Author: Wang Li

For more insights from industry leaders, please follow↓↓↓

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial conditions, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investing based on this is at your own risk