Energy storage.
Batteries without China.

Grid batteries have never sold better, yet almost nobody selling them is happy about it. Tesla installed 41% more of them in the second quarter than a year earlier, and gross profit in its energy business still fell by about a quarter. Fluence cut its revenue forecast by a fifth in September.

On paper, grid storage (battery energy storage, formally) is a simple business. Cell makers produce big lithium-iron-phosphate batteries. Integrators like Tesla and Fluence pack them into shipping-container-sized units that charge when power is cheap and discharge when demand peaks, then sell them to utilities soaking up midday solar and to data centers smoothing their load. Demand isn't the issue. The U.S. installed more grid storage in the second quarter than in any quarter before it.

The squeeze comes from where the cells are made. China makes nearly all of them, and Washington is pushing them out. To keep the federal tax credit, projects starting construction this year need at least 55% of their manufactured content from outside China and other restricted countries, rising to 60% in 2027. The main alternative, Korea's two big battery makers, makes about 4% of the world's storage cells.

That's energy storage stocks going into 2027: record demand, shrinking margins, and a supply chain being rebuilt by law.

This list of energy storage stocks traces the margin down that chain: the cell makers first, then the integrators caught in the middle, then the companies using chemistries that don't need Chinese lithium cells at all.

Whoever makes the cells sets the terms for everyone downstream.

CATL SZSE: 300750

CATL has been the world's largest supplier of storage batteries for five years running, with about 27% of global shipments in the first half of 2026. Its storage revenue rose 87.5% in that half, and its gross margin on storage held near 24%, barely below a year earlier.

That's the contrast that matters. The integrators are watching their margins collapse, and the company selling them cells is holding steady. U.S. rules are steering American projects away from Chinese cells, but most of the world can't replace CATL's volume.

LG Energy Solution KRX: 373220

LG Energy Solution is the main answer to the question "who else?" It calls itself North America's largest maker of storage batteries, and it has started producing them at five North American sites, including plants it once built for electric cars. Its storage revenue grew 4.6 times in the first half.

But the U.S. business works because of the subsidy. LG's second-quarter operating profit of 113 billion won included 241 billion won of U.S. production tax credits. Without them, it would have lost money. That's what non-Chinese supply costs right now.

These companies buy cells and sell finished systems, so they get squeezed between rising cell costs, falling system prices and the rules on where cells come from. That squeeze is the whole story for these energy storage stocks.

Tesla NASDAQ: TSLA

Tesla deployed 13.7 gigawatt-hours of storage in the third quarter, near its record. Its energy margin fell from 30.3% to 20.4% over the past year as Megapack prices dropped and it took a warranty charge tied to a supplier's faulty cells.

Tesla's way out is to make its own. A lithium-iron-phosphate cell line in Nevada is in early ramp at 7 gigawatt-hours a year, and Tesla has a $4.3 billion deal to buy U.S.-made cells from LG starting in 2027. It also has a captive customer: SpaceX, which now owns xAI, bought $405 million of Megapacks from it in the first half.

Fluence Energy NASDAQ: FLNC

Fluence shows what it costs to move production to the U.S. in a hurry. In September it cut its fiscal 2026 revenue forecast to about $2.4 billion from a $3 billion midpoint and guided to roughly $200 million of negative adjusted EBITDA. It blamed delays ramping its Houston contract manufacturing plant and said its international supply chain was working fine.

The demand is there: a record $6.4 billion backlog, including about $850 million of data center orders this year. Fluence is now a turnaround, and its fiscal 2027 plan, due with year-end results, needs to show positive cash flow without outside capital.

Sungrow SZSE: 300274

Sungrow topped Wood Mackenzie's 2026 ranking of storage integrators, and it's the clearest view of the price war. It shipped 25 gigawatt-hours of storage in the first half, up 28%. Storage revenue fell 13%, and storage gross margin dropped about 7.5 points. Its application for a Hong Kong listing will lapse in late October unless it gets a hearing.

The cleanest way around a Chinese lithium supply chain is not to use lithium cells at all. These three are betting on that, and on batteries that last far longer than four hours.

Eos Energy NASDAQ: EOSE

Eos makes zinc-based batteries in Pennsylvania, with no lithium and no Chinese cells. Revenue rose 351% last quarter, and the Department of Energy keeps advancing money on its guaranteed loan, including $87 million in September.

The problem is cost. Eos lost 71 cents of gross profit for every dollar of revenue last quarter. That makes it a binary bet: it either gets its manufacturing cost below its selling price, or the growth just burns cash faster.

Form Energy

Form Energy makes iron-air batteries that discharge for about 100 hours, built at a plant in West Virginia. In February, Google and Xcel announced a 30 gigawatt-hour Form project in Minnesota, the largest battery by energy capacity ever announced. In August it reportedly raised $750 million and quadrupled its backlog to about 80 gigawatt-hours. That's a vote of confidence in long-duration storage at a time when public companies in the space are struggling.

Hydrostor

Hydrostor stores energy by compressing air into underground caverns and releasing it through turbines. It raised $230 million in August from investors including Goldman Sachs Alternatives and Canada's national pension fund. Construction work at Willow Rock, its California project, began in July.

The cheapest battery cell an American storage project can buy is getting more expensive, not less.

Chinese storage cells cost about 21% more than a year ago, according to InfoLink, as lithium prices recovered. On January 1, China will scrap the export tax rebate on batteries entirely. Add U.S. tariffs and the sourcing rules on top. The tailwind the whole storage boom ran on for a decade, batteries getting cheaper every year, has stalled.

So the next gains have to come from somewhere else, and the market is already showing where. The average U.S. battery installed in the second quarter could discharge for 3.5 hours, up from 2.8 a year earlier. Buyers are paying for more hours per project, because more hours can replace a gas peaker or carry a data center through the evening.

In 2027, the value in storage, and in energy storage stocks, moves from cost per kilowatt-hour to hours per project. The winners will be the ones that can deliver longer duration from compliant cells, or from no lithium cells at all.