Alternative Lens on SMIC: Is the A-H Share Price Gap a Matter of Faith?

Recently, after the liquidity flood in the Hong Kong stock market, many stocks listed in both Mainland China and Hong Kong have seen their valuation discounts in Hong Kong narrow, and a small number of companies have even experienced valuation premiums. However, SMIC in Hong Kong has shown a completely opposite trend compared to its A-share counterpart—after a sharp rise following DeepSeek, its pricing relative to A-shares has started to decline recently.

In fact, when Dolphin Research first looked at SMIC, a group member asked why there was such a large price gap between SMIC's A-shares and H-shares, given that it is a core asset of national importance. At one point, the Hong Kong shares were only 25% of the A-shares, and even during the highlight moment of DeepSeek, it only recovered to 52%. What are the different pricing perspectives between domestic and foreign investors?

This article by Dolphin Research takes this question as a starting point, adopting a more neutral perspective, and examines the issue based on the input-output ratio:

1. As a business requiring investment returns rather than a strategic asset, what is its fundamental nature?

2. How to understand the significant structural pricing differences between Hong Kong and A-shares?

3. When investment returns "encounter" strategic assets, how should one rationally view investing in SMIC?

In addressing these two questions, we naturally also address the issue of pricing differences in Hong Kong stocks from different funding backgrounds in Hong Kong and A-shares.

Below is Dolphin Research's detailed analysis of$SMIC(00981.HK) :

From the perspective of domestic funds, SMIC, in the domestic wafer foundry and even the entire semiconductor industry, almost bears the "hope" of the entire village. In mature processes, it has the most extensive capacity layout, and in the upward upgrade of processes, it is also the absolute leader among domestic peers. It is the absolute leader in the domestic substitution of wafer foundry.

However, in this analysis, we do not focus on this "long-term and beautiful" fantasy story but treat it as a normal business looking at investment returns for rational analysis. After all, stock investment is about what returns and benefits it can bring to shareholders in the future, more precisely — future cash flow. Dolphin Research will carefully sort it out.

1. Black Hole Investment

As the domestic foundry leader, the company has doubled its annual revenue by 2.5 times in 6 years, reaching $8 billion in 2024, with a compound annual growth rate of 16%, surpassing cyclical fluctuations, and indeed showing the characteristics of a growth business from the revenue side.

But equally important is the more aggressive capital expenditure, which has quadrupled in 6 years, reaching $7.3 billion in 2024;

Key point: $8 billion in revenue vs. $7.3 billion in capital expenditure, meaning the company converted over 90% of its revenue into capital investment that year. Note that $8 billion is revenue, not profit. Calculating with cash profit is even more exaggerated. The cash profit that year was $3.8 billion (after adding back amortization, depreciation, equity, etc.), the company reinvested twice the money it currently earned into production.

The problem is, it's not just like this in 2024; it's been like this almost every year. Clearly, under this input-output trend, self-generated cash is far from enough.

As for why such high capital expenditure is necessary, it is actually very understandable: the wafer foundry industry that SMIC is engaged in is a heavy asset, high-process, high-barrier industry.

From lithography → deposition → etching/cleaning → CMP → measurement/inspection → other processes, every step requires heavy investment in equipment to accomplish.

Moreover, the chip process, if not upgraded annually, is upgraded every two to three years. Opening up new processes and investing in new products is a necessary rigid investment.

Here, Dolphin Research introduces TSMC under a conventional business model for comparison:

As a fellow wafer foundry, TSMC has maintained a gross margin of over 50% in the past five years, with its latest annual post-tax operating profit growing to $36.6 billion.

TSMC's capital expenditure is clearly driven by ROI—investing more when performance is good and tightening when performance is not. For example:

TSMC maintains an annual capital expenditure of $20-30 billion, but with nearly $40 billion in operating profit, after adding back non-cash expenses such as equity incentives and amortization, the actual free cash flow attributable to shareholders is significantly positive.

In other words, shareholders of TSMC can expect the company to use positive cash flow to reward shareholders.

Unlike TSMC, SMIC's thinner cash profit cannot fully cover the high capital investment. Without key technological breakthroughs, it is likely to maintain this state within foreseeable limits.

In this case, in Dolphin Research's long-term expectations that smooth out cycles, its long-term cash flow expectation grows into the following scenario—continuous negative free cash flow, requiring continuous replenishment.

2. Who Will Replenish the Blood?

From the perspective of future technological competition, these investments are essential. But the key question is, when self-generated cash is insufficient, who will replenish the blood? Next, let's look at how the blood was replenished in the past.

After reviewing the historical financial reports of $SMIC(688981.SH), it can be roughly summarized into three sources: equity financing, bond financing, and "special injections." And these three channels have been roughly equal over the past six years.

a) Issuing New Shares: In 2020, the company listed on the domestic STAR Market, raising $7.5 billion through new share issuance; combined with previous convertible bond financing, a total of $8.4 billion was raised over seven years.

b) Borrowing: This includes issuing bonds, bank loans, and bill issuance, with a total net borrowing of $9.4 billion over seven years.

c) Non-controlling Shareholder Injections: Occurring annually, especially frequent before 2022. A total of $9.4 billion was raised over seven years, with $7.6 billion since 2020.

How to understand these types of financing?

① Equity Financing: The STAR Market listing raised funds, with the IPO pricing below RMB 28, rising to 90 on the day, followed by a long valuation digestion period, currently at 80. For those who participated in the IPO, it was undoubtedly profitable.

② Debt Financing: Among these, as a large central enterprise in an emerging industry, external borrowing is almost a given. By the end of 2024, the interest-bearing asset-liability ratio is only 23%, and being in a strategically supported technology industry, borrowing costs are low, naturally "welcomed" by shareholders.

Dolphin Research will focus on the third point, which is relatively special and typically reflects its "strategic leader" status:

Through sorting out related subsidiaries (SMIC South, SMIC Beijing, SMIC Shenzhen, and SMIC East) from 2020 to 2024, Dolphin Research found that SMIC raised approximately $7.6 billion during this period through subsidiary establishment and capital increase, roughly matching the amount of non-controlling shareholder injections.

The main sources of injections are almost exclusively state-owned funds such as the National Integrated Circuit Industry Investment Fund, Beijing State-owned Assets, Shenzhen State-owned Assets, and Shanghai State-owned Assets.

In other words, relying on "capital market fundraising and state-owned fund support," SMIC continues to endure the current operational "deficit" predicament.

However, it should be noted that state-owned fund "blood transfusions" have shown a significant decline in recent years, with the company receiving only $280 million in non-controlling shareholder injections in 2024, far below the $2.6 billion in 2020, and some industrial funds have even shown withdrawal demands.

Assuming that the only source of financing is net borrowing (borrowing new to repay old + additional borrowing) maintaining $1.5 billion annually, without state-owned fund injections and company equity financing, the estimate is as follows:

SMIC needs to maintain an annual capital expenditure of $7-8 billion, and besides using the annual operating cash flow earned + $1.5 billion in net borrowing for capital investment, there is a $1.2 billion gap in 2025 and an $1.1 billion gap in 2026.

Looking at the company's current cash consumption rate, with $5.2 billion in cash on hand in the first quarter of 2025, SMIC is likely to need new funding or a private placement when cash on hand is reduced to around $3 billion by the end of 2026, reaching the 2018-2019 level of $2-3 billion.

It should be added that SMIC's $5.2 billion in cash is not net cash, as there is nearly $3 billion in long-term debt due by the end of the first quarter this year.

Therefore, Dolphin Research can only assume that SMIC not only borrows to repay old debts but also borrows an additional $1.5 billion (estimated based on past historical average levels).

Of course, when truly lacking funds, as a leader in a strategic industry, injections are obligatory when needed, so there is no real cash risk. But before technological breakthroughs, it will essentially become a "policy loan + government capital + social capital" supported black hole business model.

For example, in the company's recent financing plan: the company plans to sell 14.8% of SMIC Ningbo's shares (specific transaction price not yet announced), which can help SMIC focus more on its main business areas such as 12-inch wafers, while also supplementing the company's cash flow situation.

[Note: SMIC Ningbo is positioned in the mature process special process semiconductor wafer foundry business, mainly involving RF front-end, MEMS (Micro-Electro-Mechanical Systems), and high-voltage analog devices. Due to weak downstream demand, SMIC Ningbo has suffered losses of over 800 million yuan for two consecutive years.]

As can be seen, the sale of SMIC Ningbo's equity is the monetization of mature assets, a process involving social capital participation. Although it can supplement the company's liquidity in the short term, in the context of technological breakthrough expectations, capital injection is a long-term and continuous process.

To maintain an annual capital expenditure of $7-8 billion, it will be difficult to avoid further financing through private placements, bond issuance, and other means in the future.

3. How to Understand the Significant Structural Pricing Differences Between SMIC's Hong Kong and A-shares?

In fact, after analyzing this far, it is easy to understand why there is such a large pricing difference between SMIC's Hong Kong and A-shares, given that it is a core asset worth hundreds of billions of dollars in the domestic semiconductor industry.

For domestic funds, the breakthrough in the domestic semiconductor industry is hoped for in SMIC: it has the most extensive capacity layout; among domestic peers, it is the closest to advanced processes; and it has continuous national funding support. If there is ultimately a breakthrough, SMIC is likely to be one of the front-row C-position members on the day of success.

This investment logic,

① For domestic funds: It is truly suitable for patient capital, belief capital, and long-term money with a ten-year or twenty-year dimension, embodying the expectation of narrowing the gap in the semiconductor industry and the strategic pattern of "supporting the country." This stringent condition actually filters out the vast majority of market funds that are assessed quarterly and the scattered retail funds.

② For overseas capital: 1) SMIC is not scarce in terms of assets, with options like Samsung, TSMC, UMC, and GlobalFoundries; 2) There is no need for and likely no pattern capital and belief capital; their allocation is more based on a cross-asset allocation with a return period, and if an asset is a "capital black hole business" within a visible period of about five years, there is no need to persist; 3) Investing in such assets, like investing in other H-shares, requires bearing exchange rate risks and considering additional capital costs.

It is precisely because of the different investment logic in viewing the underlying assets that it is seen as a scarce "treasure" by domestic funds but as ordinary by foreign investors.

To some extent, just as CATL is seen as a globally leading core asset in the Hong Kong stock market, willing to give a valuation premium; but in the eyes of domestic investors, it is an overcapacity, highly competitive red ocean asset, leading but not scarce. In Dolphin Research's view, the reasoning behind the discount and premium of these two companies in the Hong Kong and A-share markets is similar.

This difference in the underlying logic of viewing assets leads to a structural valuation difference that is difficult to eliminate when the asset fundamentals do not undergo significant changes.

For this reason, even for SMIC in the Hong Kong stock market, it is actually the southbound funds that are willing to buy: SMIC's southbound funds account for nearly 30% of its total market value in the Hong Kong stock market, far higher than other companies.

Looking at the stock price changes over the past year, the company's stock price rose from a low of HKD 16 to nearly HKD 60, achieving more than double growth. At the same time, the proportion of Shanghai-Shenzhen-Hong Kong Stock Connect holdings increased from 27% to 39%. This further confirms that the continuous buying by southbound funds led to the significant rise in SMIC's stock price in this round.

4. How Should SMIC's Investment Opportunities Be Positioned?

Summarizing SMIC as an investment target, we can see:

a. A business with heavy technology, heavy assets, and barriers; regardless of the call for domestic substitution, before technological breakthroughs, its fundamentals cover mature cyclical businesses, and its performance fluctuates with the cycle.

b. But as a domestic industry leader, regardless of hope, it must shoulder the responsibility of technological breakthroughs, requiring continuous blood transfusions and rigid investments, making it essentially a "policy loan + government capital + social capital" supported black hole business model.

Therefore, Dolphin Research says that, in essence, what matches its business status is long-term money and belief capital calculated in decades. But this does not mean that ordinary funds should not participate:

SMIC's heavy asset investment + cyclical performance + growth expectation characteristics (a-b) combined, it is like a heavy asset in a growth track where others have long run far ahead, bearing the hope of the entire village, ready to start, but its bones have not truly grown, always in the "growth reserve" heavy asset.

Before it truly starts, it is essentially a cyclical business with heavy assets, and cyclical businesses are essentially PB valuations: this type of business, referring to peers like UMC, fluctuates mostly within a 1-2 times PB range based on the cycle position.

More PB comparisons among peers can be found, "TSMC 7.7 > SMIC 2 > GlobalFoundries 1.9 > UMC 1.5". TSMC's PB is the highest due to its significant assets concentrated in advanced processes, generating high economic benefits and having a clear industry-leading advantage. Among other mainstream wafer foundries, SMIC already enjoys a relatively high PB multiple.

In terms of revenue scale, SMIC, UMC, and GlobalFoundries are relatively close in size. Although SMIC is slightly ahead in process progress compared to the other two, its gross margin performance is significantly lower than UMC and GlobalFoundries.

Considering UMC's PE situation, SMIC's current PB's relatively reasonable range is between 1-2 times. When SMIC's PB value significantly exceeds 2, it indicates that the company's stock price already contains too much optimistic expectation.

In summary, before technological breakthroughs, stripping away the glamorous exterior, SMIC is actually a heavy asset target with a 1-2 PB. The difference is that, compared to UMC and TSMC's technology vs. mature stable division of labor, it has a second "growth reserve" expectation attribute:

a. Whenever PB is below 1, because there will always be stories of domestic substitution and domestic technology emerging, with the support of storytelling, it has higher PB elasticity, meaning below 1 PB, it has more certain returns and higher upward elasticity compared to peers;

b. But without substantial technological breakthroughs in fundamentals, whenever storytelling and grand narratives (and semiconductors are never short of grand narratives) raise it to 2x PB, its cyclical characteristics make it easy for those chasing high prices to be left holding the bag.

c. For most ordinary investors, when the stock price is at 2x PB, many who hold long-term beautiful narratives (i.e., using long-term logic for short-term investment decisions) become passive long-term investors, finding after three to five years of holding that there is almost no mid-term holding return.

After reading this article, at the high price of SMIC, can retail investors who are full of enthusiasm and rush in think more calmly before investing, and clearly understand how long they are prepared to hold and what kind of money they want to earn?

Dolphin Research's Historical Articles on SMIC:

Earnings Season

May 8, 2025, Conference Call "SMIC (Minutes): Mobile and Other Expectations Revised Down, Prices Will Continue to Decline"

May 8, 2025, Earnings Commentary "SMIC: Guidance Big "Crash," When Will "Domestic Chips" Break Through?"

February 10, 2025, Conference Call "SMIC (Minutes): Capital Expenditure Plan Flat with Last Year"

February 10, 2025, Earnings Commentary "SMIC: "National Subsidy" in Hand, Hard to Withstand the Cycle?"

November 8, 2024, Conference Call "SMIC: Next Year's Potential Risk is "Increment Without Price Increase" (24Q3 Conference Call)"

November 7, 2024, Earnings Commentary "SMIC: Can It Bear the Hope of the Entire Village?"

August 9, 2024, Conference Call "SMIC: 12-inch Capacity Near Full Load (24Q2 Conference Call Minutes)"

August 8, 2024, Earnings Commentary "SMIC: Soaring Against the Wind, Delivering Explosive Guidance"

May 10, 2024, Conference Call "SMIC: Full-Year Revenue Growth Will Exceed Industry Average (24Q1 Conference Call)"

May 9, 2024, Earnings Commentary "SMIC: Long Low Valley, Finally About to Emerge"