On September 15th, CATL's A-shares closed at 316.36 yuan, dropping 6.16% - a price level not seen since September 2025. The same day, its H-shares fell over 6%, dragging the combined market value of both listings below 1.5 trillion yuan. The decline continued the next day, with shares closing at 304.31 yuan on September 16th. From the May 7th peak, the stock has lost over 30% of its value, erasing roughly 700 billion yuan in market capitalization.

Despite this market slide, the company's half-year results were remarkably strong. First-half revenue reached 276.917 billion yuan, up 54.8% year-on-year, while net profit hit 43.284 billion yuan, a 41.98% increase. Spread across 181 days, that translates to an average daily profit of 240 million yuan. For context, the combined net profits of 15 major listed automakers on both A-shares and H-shares totaled only 21.048 billion yuan - less than half of CATL's single-company figure. A year earlier, those same 15 automakers had collectively earned 35.131 billion yuan, meaning their combined profits shrank by 14.083 billion yuan in just one year. Meanwhile, CATL's first-half profit grew by nearly 12.8 billion yuan from 30.485 billion yuan the previous year. One side gained 12.8 billion; the other lost 14.1 billion.

The individual automaker numbers tell their own stories. Great Wall Motor posted first-half revenue of 102.101 billion yuan, up 10.58%, but net profit attributable to shareholders fell 61.11% to 2.465 billion yuan. GAC Group projected a loss between 4.06 billion and 4.57 billion yuan. Seres reported a loss of 1.717 billion yuan. Changan Automobile forecast net profit of just 740 million to 970 million yuan, down over 50% year-on-year. None of these companies suffered major operational failures - cars are still selling, some quite well. Many interpret this as battery makers seizing profits from automakers. It's a satisfying narrative, but inaccurate. What's actually happening is that in China's new energy vehicle supply chain, profits are migrating toward the segment with higher concentration. The vehicle assembly end hosts hundreds of brands locked in price wars, while the top two battery suppliers control over 60% of installation volume. Which end retains the profits requires no debate.

What matters isn't who's victimizing whom, but how this position was achieved and how long it can be maintained. Answering that requires examining the speed of settlements - a critical financial metric. In Q1 2026, CATL's accounts receivable turnover stood at 54.44 days, meaning customers pay on average 54 days after receiving goods. Its accounts payable turnover was 261.44 days - it takes an average of 261 days to pay upstream suppliers. That's a 207-day gap. Different data providers calculate these figures differently. Choice statistics show accounts payable turnover at 266.18 days for end-2025 and 261.44 days for Q1 2026, with corresponding receivable turnovers of 60.35 and 54.44 days. Wind and China Securities Construction Investment calculate Q1 2026 payables at just 155 days, with Q2 receivables at 53.6 days. The two systems differ by over 100 days, and articles pick whichever suits their narrative. But regardless of methodology, the conclusion is identical: cash collection far outpaces cash payment.

Looking at the balance sheet makes it clearer. At end-2025, CATL's notes payable stood at 103.3 billion yuan, accounts payable at 160.3 billion yuan - combined 263.6 billion yuan, up 65 billion or 32.9% year-on-year. By Q1 2026, notes payable reached approximately 119.8 billion yuan and accounts payable 180.6 billion yuan, totaling over 300.4 billion yuan. Three hundred billion yuan - interest-free, unpursued, simply sitting on the books as ammunition for expansion. On the other side, customer prepayments grew substantially. Contract liabilities hit 49.2 billion yuan at end-2025, up 21.4 billion or 77% year-on-year. These are customer deposits - automakers paying upfront to secure production capacity. During the tightest battery supply period from September to November 2025, some automakers' procurement staff reportedly camped outside CATL's headquarters, queuing for capacity allocations. One side pays upfront and queues; the other receives goods and settles nearly nine months later.

This dynamic often gets framed as a moral issue - bullying both upstream and downstream. That framing misses the point. Payment terms aren't a personality trait; they're a measure of supply chain position. Fifteen years ago, Japanese manufacturers dictated terms in the lithium battery chain, demanding payment before delivery while Chinese battery makers queued with cash. Today, the position has simply changed hands. But when you squeeze capital efficiency from the entire chain, you also squeeze competitors' cash flow. Automakers forced to wait hundreds of days for payment will inevitably seek alternatives. In early September 2026, Li Auto announced a 2.65 billion yuan investment in Sunwoda EVB. The structure: 1.4 billion yuan for new registered capital, the remainder into capital reserves. Post-investment, Sunwoda's registered capital increases from 14.515 billion to 15.915 billion yuan, with Li Auto holding 8.79% directly, plus additional shares through its subsidiary Chongqing Chezhiyuan, totaling 11.17% - making it the second-largest shareholder. Chongqing Chezhiyuan, Li Auto's industrial investment platform, had previously invested 400 million yuan in Sunwoda's Pre-A round in 2022. From 400 million to 2.65 billion - a sixfold increase in four years.

Beyond capital, there's business. The all-new Li L8 launched in June 2026 uses Sunwoda cells exclusively, with CATL removed from that model's supply list. Battery packs are produced by a Li Auto-Sunwoda joint venture, with Li Auto defining the product and Sunwoda handling manufacturing. The newly filed 2026 Li i6 also introduces Sunwoda and CALB as suppliers. What's notable: Li Auto and CATL signed a five-year comprehensive strategic cooperation agreement in 2025 covering all-material-system batteries for Li's entire lineup. GAC signed a decade-long agreement. So this isn't a rupture. Li Auto simultaneously extends a long-term agreement with CATL while sending 2.65 billion yuan to another supplier. Automakers aren't trying to overturn the table; they're reclaiming control over battery system definition - product specifications, performance parameters, and standards. These matter far more than saving a few hundred yuan per battery. The same pattern is spreading industry-wide. Xiaomi has added CALB and Sunwoda as suppliers for its Longjia battery, expanding from two to four vendors. Harmony Intelligent Mobility models source batteries from multiple suppliers. Companies like Geely, Great Wall, GAC, and SAIC are pursuing self-development, joint ventures, or supplier investments to build their own battery systems they can control. This is 2026's hottest industrial story, with increasingly dramatic headlines. But what matters more is whether all this money and effort are actually working.

Looking at real-world data: in H1 2026, national power battery installations reached 335.6 GWh. CATL installed 154.25 GWh, capturing 46.04% market share - up 2.99 percentage points year-on-year. For passenger vehicles alone, the company reports 46.7%, up 5.6 percentage points. As for the subsidized challengers: CALB installed 20.71 GWh with 6.18% share, down 0.33 percentage points. Sunwoda installed 9.15 GWh with 2.73% share, down 0.30 points. SVOLT installed 6.40 GWh with 1.91% share, down 0.90 points. All three declined. The biggest loser on the list isn't CATL - BYD's installations of 57.42 GWh gave it 17.14% share, down 6.42 points year-on-year, falling below the 20% threshold. After two years of de-CATL rhetoric and billions spent on secondary suppliers, the besieged target's share rose while the supported challengers shrank.

The reason isn't complicated: battery manufacturing can't be bought overnight. Sunwoda's own books tell the story. Full-year 2025 revenue was approximately 20.093 billion yuan with a net loss of 3.169 billion yuan. Q1 2026 revenue was 6.125 billion yuan with net profit of just 13.0452 million yuan - barely breaking even. Thirteen million yuan versus CATL's 240 million per day. By Q1 2026, Sunwoda's total assets were about 60.062 billion yuan against total liabilities of about 50.272 billion yuan - an 83.7% debt-to-asset ratio. The profit foundation is paper-thin, necessitating round after round of fundraising. A 1.68 billion yuan Series C round was disclosed in late May 2026, followed within two months by an 805 million yuan Series C+ round with Sungrow Power contributing 655 million and Tianqi Lithium 150 million. Then Li Auto added 2.65 billion in September. This reveals the true cost: building a viable second supplier requires supporting its capacity, R&D, and cash flow - expenses that far exceed any discount extracted from CATL. Automakers aren't buying immediate cost reduction; they're purchasing insurance against future supply chain constraints.

But CATL can't rest easy either. Its moat doesn't lie in payment terms but elsewhere - and those walls have aging foundations. First, capacity: first-half battery system capacity was 525 GWh with production of 498 GWh, an 86.94% utilization rate, and 764 GWh under construction. That construction figure deserves scrutiny - a few years ago, 1,000 GWh was an industry-wide aspiration; now a single company's pipeline approaches that scale. Overseas, first-half revenue was 87.129 billion yuan, up 42.35%, representing 31.46% of total revenue. What's truly valuable is the gross margin gap: overseas business delivered 29.97% versus 21.16% domestically - nearly 9 percentage points apart. The trend is also divergent: domestic margins fell 1.78 points year-on-year while overseas rose 0.95 points. From January to May, overseas market share was 33.7%, up 3.7 points. Behind this sits four completed overseas bases. The Erfurt plant in Germany achieved profitability in 2025; Debrecen in Hungary involves a 7.34 billion euro investment for 100 GWh planned capacity, with module assembly lines starting production in June 2026; the Spain project with Stellantis is a 50-50 joint venture worth 4.1 billion euros for 50 GWh; Indonesia adds 6.9 GWh. There's also the asset-light LRS model - technology licensing and engineering support with no equity stake, generating patent royalties and service fees. Ford's Michigan plant, using this model, began production in June 2026 - the first automaker-wholly-owned LFP battery facility in the US.

The second pillar is energy storage. First-half storage system revenue reached 53.261 billion yuan, up 87.54% - nearly double the growth rate of power batteries - with a 23.96% gross margin, exceeding power batteries' 20.63%. In the sodium-ion space, July saw two orders within one week: 5 GWh from Netherlands' Alfen and 2 GWh from Bulgaria's Solarpro. Behind the products lies capital: by end-June 2026, CATL held or had filed 60,319 patents domestically and internationally, with cumulative R&D investment exceeding 90 billion yuan over the past decade. On the product front, the third-generation Shenxing super-fast-charging battery supports equivalent 15C charging, reaching 10% to 80% in under 4 minutes, alongside the third-generation Kirin, Kirin condensed-state, and second-generation Xiaoyao super-hybrid batteries. Customer relationships aren't simple purchase contracts either. CATL is the second-largest shareholder of Avatr, co-built a 25 GWh plant with Changan, and a 15 GWh facility with BAIC. Its Chocolate battery-swap stations number approximately 2,000 across 180 cities in 31 provinces, supporting over 20 swap-capable models. Switching suppliers means dismantling not just a contract but an entire asset ecosystem spanning cells to energy replenishment infrastructure.

How thick is this wall? It depends on when next-generation technology arrives. CEO Robin Zeng stated in June that on a 1-to-9 technology maturity scale, solid-state batteries currently sit at level 4. The company's plan is small-batch production in 2027, with million-unit vehicle adoption before 2030 unlikely. The flip side: within this window, existing capacity, patents, and customer embeddedness remain protective. Beyond the window, unpredictability reigns. This script is familiar in Chinese manufacturing - the most profitable link in a value chain eventually gets internalized by downstream players. From display panels and photovoltaics to semiconductors and EVs, the pattern repeats. The more upstream profits, the stronger downstream integration incentives - this isn't emotional, it's arithmetic. CATL itself isn't standing still, expanding into lithium mining and materials recycling upstream, and battery swapping, storage systems, and data center power downstream, rebranding itself as a zero-carbon technology company rather than just a battery maker. Neither side plays the villain. Automakers pursue definition rights; battery makers pursue margins - both within their legitimate interests.

One detail deserves special attention. On July 24th, CATL's board approved a buyback plan of no less than 20 billion and no more than 40 billion yuan at a maximum price of 573 yuan per share, with all repurchased shares to be cancelled. The 40 billion ceiling set an A-share record for single cancellation buybacks, surpassing Gree Electric's 15 billion in 2021. The first repurchase occurred on September 11th - 604,300 shares, representing 0.0131% of total share capital, at prices between 330.15 and 331.61 yuan, totaling approximately 200 million yuan. A company earning 240 million daily spent 200 million on its own stock. At this pace, filling the 40 billion ceiling would take roughly 200 trading days - exactly matching the buyback deadline. And the 330 yuan bid is just 57.6% of the 573 yuan ceiling. Four days later, the stock fell 6.16%. The company's own modest commitment at this price level speaks louder than any analysis. As for whether de-CATL succeeds, that's secondary. The only question worth tracking: how much longer can China's power battery high-profit window remain open?