Introduction: Is it too late for a business company established just five years ago to chase the light now?

It's too strong—'Computing Power Leather' has gone crazy!

Xingye Technology (002674.SZ), which supplies automotive interior leather for brands like NIO and Li Auto, saw its stock price hit the daily limit for six trading days from June 18 to June 26 (excluding the three-day Dragon Boat Festival holiday). Closing at 28.83 yuan per share on June 26, up 77.2% from 16.27 yuan on June 17, it has become a 'hot stock' in the recent cross-border AI industry chain.

On the evening of June 21, Xingye Technology announced plans to acquire Qingdao Leon's indium phosphide substrate and semiconductor electronic materials business for 55 million yuan in cash, making a cross-border move into optical communications and optical module sectors. Notably, Xingye's stock price had already hit the daily limit the day before the announcement.

In the leather products sector, Xingye Technology is no small player.

Owners of models like the NIO ES6, EC6, ES8, ET5, ET7, Li Auto L series, and Zunjie S800 likely have leather trims on their car seats, door panels, dashboards, armrests, and steering wheels produced by Xingye Technology's subsidiaries. By 2025, Xingye Technology generated over 800 million yuan from selling automotive interior leather.

With the current surge in AI computing power demand driving higher requirements for data transmission efficiency, this directly fuels the expansion and upgrading of optical modules. Indium phosphide is the core substrate material for high-end optical modules at 800G, 1.6T, and above. Due to structural supply-demand mismatches, indium phosphide is seen as the 'cornerstone of light' for AI computing power.

However, since Xingye Technology lacks prior experience in R&D, production operations, and market expansion in semiconductor new materials, there are doubts in the market about whether it is chasing trends or hyping concepts. With more companies from diverse industries crossing into AI—such as Andly (concentrated juice), Lotus Seasoning (monosodium glutamate), Jinzi Ham, and By-health (health supplements)—this phenomenon is drawing increasing attention.

Objectively, AI represents the most promising industrial track ( track : track here means 'track' or 'field') globally. Cross-border investments by companies from different industries are not inherently wrong, and policies also encourage capital to empower AI-related fields. However, AI-related concepts inherently have a price-boosting effect on associated stocks. Beyond risk warnings to investors, listed companies must 'hold the line' and avoid 'false moves' for short-term gains that leave investors bearing the risk of stock price declines.

01 Leather Product Revenue Nears 3 Billion, But Profits Decline

The acquisition of Qingdao Leon's indium phosphide business is not Xingye Technology's first cross-border move into a hot sector.

In September 2025, Xingye Technology announced a strategic cooperation with Suzhou NST Electronics to jointly develop flexible electronic skin, primarily for humanoid robot tactile sensing, smart car stealth touch cockpits, and smart wearable/sports shoe gait monitoring.

Flexible electronic skin has some relevance to Xingye's core business, as it is based on natural leather and biomaterials, aligning with Xingye's focus on natural cowhide R&D and production.

However, more than half a year after announcing the flexible electronic skin collaboration last year, Xingye has not disclosed significant progress or plans for the robotics market.

To date, Xingye Technology remains a 'leather company' with four core product categories: leather for footwear, bags, and belts; automotive interior leather; furniture leather; and special functional leather. Footwear, bag, and belt leather account for over 60% (63.5%) of 2025 revenue, while automotive interior leather accounts for nearly 28%.

Xingye's automotive interior leather business is operated by its controlled subsidiary ( controlled subsidiary : holding subsidiary) Hongxing Automotive Leather, covering natural leather for car seats, door panels, dashboards, armrests, and steering wheels.

In recent years, Xingye's core business has shown signs of decline.

In 2024, Xingye reported revenue of 2.951 billion yuan, up 9.32% YoY, but net profit attributable to shareholders fell 23.97% YoY to 142 million yuan. In 2025, revenue grew just 1.19% YoY to 2.986 billion yuan, with net profit declining another 28.41% YoY to 101 million yuan.

Product-wise, the decline is clearer: In 2024, footwear, bag, and belt leather revenue exceeded 2 billion yuan, up 7.59% YoY. But in 2025, this segment fell 5.54% YoY to 1.897 billion yuan.

Meanwhile, automotive interior leather revenue rose from 720 million yuan to 834 million yuan, increasing its revenue share by 12 percentage points to 27.94%, but this could not fully offset the decline in traditional footwear and bag leather.

Alpha learned that Xingye's footwear and bag leather clients are mainly traditional leather shoe brands like Senda, Red Dragonfly, Kangnai, and Aokang. Automotive interior clients include Li Auto, NIO, Chery, JAC, and Geely.

The underlying cause of the footwear and bag leather decline is shrinking demand for traditional business formal leather shoes, mismatched with new-generation consumer preferences, coupled with high inventory and weak demand in the footwear and apparel industry, putting pressure on upstream leather orders.

Additionally, leather product inventory has age requirements, necessitating asset write-downs.

From 2023–2025, Xingye proactively accrued 350 million yuan in asset impairments, primarily to clear old inventory and bad debts. In 2025 alone, inventory depreciation provisions reached 148 million yuan, temporarily suppressing reported profits. Excluding impairments, 2025 operating profit was 255 million yuan, down just 7.6% YoY.

After clearing large impairments, Xingye's profit rebounded in Q1 2026, with revenue of 590 million yuan and net profit attributable to shareholders of 23 million yuan, up 19.31% YoY.

This acquisition of Qingdao Leon's indium phosphide substrate and semiconductor electronic materials business is part of Xingye's search for a second growth curve. The acquisition plan is relatively 'precise,' focusing on acquiring assets, teams, patents, trademarks, and existing business contracts—a pragmatic approach.

However, the issues are clear: Xingye lacks prior management experience in semiconductor new material R&D, production, and market operations, with no existing equipment, personnel, or customer channels related to the semiconductor industry.

02 'Indium Phosphide' Business Company Founded Just 5 Years Ago, Overseas Giants Dominate Market

How much growth can Qingdao Leon's indium phosphide substrate and semiconductor electronic materials business bring to Xingye Technology? Currently, it appears limited.

On the books, Qingdao Leon's net asset value is under 25 million yuan, with estimated business revenue of under 5 million yuan for January–May 2026, accounting for a negligible share of Xingye's 2025 total revenue. Xingye believes the acquisition will have minimal impact on current revenue and net profit.

Tianyancha data shows Qingdao Leon was founded in 2021, just 5 years ago, with only 17 insured employees.

Equity penetrate (股权 penetrate : equity penetration) reveals Qingdao Leon's majority shareholder is Qingdao Huaxin Jingdian Technology Co., Ltd. ('Huaxin Jingdian'), both controlled by Xiao Di. In 2024, Qingdao Leon was named a 'Chuying Enterprise' (promising SME) by Qingdao.

According to media reports, Huaxin Jingdian, Qingdao Leon's parent company, focuses on R&D and manufacturing of semiconductor materials, including sapphire, indium phosphide, and gallium oxide substrates, with annual production of 12 million semiconductor substrate wafers.

Currently, Qingdao Leon's specific products and performance are undisclosed, with public information only showing it has delivered indium phosphide products externally.

Acquiring a business from such a 'young' company did not require Xingye to pay much—the agreed 55 million yuan is manageable relative to the company's 691 million yuan in cash and cash equivalents at the end of Q1 2026.

But from a business perspective, indium phosphide is a high-tech barrier compound semiconductor material requiring substantial R&D investment, testing Xingye's 'endurance.'

Moreover, indium phosphide products face extremely long certification cycles among global top clients, typically over 2 years, with high entry barriers. Certification delays could prevent bulk shipments as planned.

Globally, the indium phosphide market is highly concentrated, with Japan's Sumitomo Electric, U.S.-based AXT (AXTI), and France's II-VI dominating overseas mainstream shares. Among domestic leaders, Vital Materials Group achieved mass production of 6-inch indium phosphide substrates first in China; Yunnan Germanium Industry will expand production of 300,000 high-quality 4-inch equivalent indium phosphide single crystals annually; and Bogie Technology has taken equity stakes in indium phosphide-related businesses.

Qingdao Leon's existing indium phosphide and other businesses are weak in capacity and scale, facing significant market expansion challenges.

According to Omdia, global indium phosphide substrate demand totals 2.6–3 million wafers, but effective supply is only 750,000 wafers, creating a supply-demand gap exceeding 70%.

Research reports cited by 'China Finance Online' show that the AI computing power boom is driving indium phosphide—a core material for optical communications—into a super-demand cycle. As a key substrate for optical chips, indium phosphide handles core electro-optical signal conversion, with over 80% of demand coming from AI data centers.

According to LightCounting, its market share in optical chips will reach 58% by 2025, with the market size expected to hit 6.9 billion USD by 2031.

But competition is intensifying.

Global leaders are accelerating expansion: Coherent started expanding a 6-inch indium phosphide wafer plant, securing strategic investment and long-term purchase agreements from NVIDIA; JX Metal plans to invest 120 billion yen by FY2030 to boost capacity 7–10x; AXT's subsidiary Beijing Tongmei received funding for expansion, with current orders exceeding 100 million USD; Lumentum is advancing a ~40% expansion plan, expecting AI data center demand for indium phosphide to grow at an 85% CAGR through 2030.

For the indium phosphide industry, 'bottlenecks' exist in two senses: First, indium phosphide is core to optical communications. Second, while ~70% of global indium supply comes from China, indium phosphide substrates are a highly barricaded, monopolized segment of the indium phosphide industrial chain ( industrial chain : industrial chain), with ~90% of the global market share controlled by Japan's Sumitomo, AXT, and Japan's JX Metal.

The market potential is significant, but competition is fierce. For Xingye Technology to grow in this field, time may be its biggest enemy. With Qingdao Leon generating under 5 million yuan in revenue in the first five months, how long will it take for its indium phosphide and other businesses to achieve exponential growth? Is Xingye prepared for long-term development?