LG Energy Solution Booked a Profit; a US Subsidy Booked It for Them

On the morning of July 6, 2026, LG Energy Solution (LGES), the South Korean company that is the world's second-largest battery maker, told investors it had returned to profit. Revenue for the second quarter reached 7.56 trillion won, up 24.8 percent from a year earlier and its fastest growth in more than a year. Operating profitOperating profitWhat a company earns from its core business after paying the costs of running it, but before interest and taxes; a basic gauge of whether the business itself makes money., the money a company earns from its core business before interest and taxes, came in at 113.3 billion won (about 74 million dollars). After a loss the previous quarter, that looked like a turning point, and management pointed to the reason: a surge in energy storage systems (ESS), the big banks of batteries that feed power grids and, increasingly, AI data centers, plus the first full-scale production of cheaper lithium iron phosphate (LFP) cells, which began at the company's Michigan plant.

There is a catch one line down in the disclosure. Sitting inside that 113.3 billion won profit is 241 billion won from a single item called the Advanced Manufacturing Production Credit, or AMPC: a US government subsidy, created by the 2022 Inflation Reduction Act, that pays battery makers a fixed amount for every cell they build on American soil. It behaves almost like cash, and LGES recognizes it inside operating profit, as an offset to the cost of goods sold. That is precisely why you can subtract it from reported operating profit and get a clean read on the underlying business. Do that, and the credit turns out to be more than double the profit it sits inside. The quarter is not a profit at all. LGES itself said that without the AMPC it would have posted a 127.7 billion won operating loss.

The company also fell well short of what analysts expected. The LSEG SmartEstimateLSEG SmartEstimateA forecast published by data provider LSEG that weights an analyst consensus toward the forecasters who have been most consistently accurate, so it is treated as a sharper version of the plain average. consensus, a forecast weighted toward the analysts who are most consistently accurate, stood at about 294 billion won; the reported 113 billion won is less than half of that. So the headline is not just subsidy-flattered; even with the subsidy counted, it missed.

LGES is in the middle of a widely praised strategic turn. As electric-vehicle demand cooled, it began rerouting factories away from EV cells and toward LFP storage batteries for the grid and for the power-hungry data centers behind the AI boom. The bull case writes itself: a growth market, a higher-value product, a company reading the room. So here is the narrow question this piece tries to answer with evidence rather than narrative:

Is the pivot from EV cells to data-center storage actually earning money on its own, or is the whole company being held above breakeven by one subsidy whose value shrinks and rebuilds with volume, not with margin?

To answer it, we did something simple and repeatable. For each of the last six quarters we took LGES's reported operating profit and subtracted the AMPC credit it disclosed for that quarter. Because the company books the credit as an offset to cost of goods sold, inside operating profit, the remainder is the underlying operating result: what the batteries earned before Washington's cheque. If the pivot is working, that underlying line should climb toward zero and beyond. If the subsidy is doing the work, the reported profit should track the size of the credit while the underlying line stays under water.

The picture is stark. In four of the last six quarters, LGES's operating result without the subsidy is negative. A fifth quarter, the second of 2025, was essentially breakeven at plus 1 billion won. The only clearly positive quarter, the third of 2025, was lifted by roughly 280 billion won of one-time make-good payments tied to customers taking less volume than they had contracted for, money that vanished the next quarter, when the ex-subsidy result cratered to minus 455 billion won.

Read the two lines as a gap. The distance between the reported line and the ex-AMPC line is the subsidy. Quarter after quarter, the reported result rises and falls roughly in step with that gap, not with any independent improvement in the batteries themselves. The clearest way to see it is the latest quarter on its own, drawn as a step-down.

This is the crux. The second quarter of 2026 is the first clean look at the pivot after the new LFP storage lines reached full production, and the underlying result is still a 128 billion won loss, deeper than the roughly breakeven the business managed a year earlier, when it was still mostly building EV cells. Revenue is up nearly a quarter and the celebrated new product is shipping, yet the batteries lost more money on their own than before the reroute began.

A tempting reading of the last year is that the AMPC and the pivot are trading off against each other, that as EV lines went quiet the subsidy dried up and the company had to find a new product to replace it. The data says something more specific, and more awkward. The credit is a fixed amount per kilowatt-hour of cells made in America, so it does not care whether a cell ends up in a car or on a data-center pad. When EV cell lines idled in late 2025 and early 2026, the credit fell simply because fewer cells came off the lines. As LFP storage cells ramped at the Michigan plant, the credit began to climb back.

In other words, the pivot regenerates the same subsidy rather than replacing it with margin the company earns itself. The AMPC peaked at 491 billion won in the second quarter of 2025, fell to 190 billion by the first quarter of 2026 as volume collapsed, and rebuilt to 241 billion as storage output filled the lines back up. Storage cells qualify for the identical per-kilowatt-hour credit that EV cells did. So the machine that produced the profit headline is not the new business model; it is the old subsidy, riding on whatever LGES happens to be building.

There is a serious counter-argument, and it deserves a fair hearing. The bullish read is that this is a trough LGES chose to sit in. The company cut its 2026 capital spending by about 40 percent, largely by converting existing EV lines to storage instead of building new plants, which depresses reported margins now in exchange for cheaper capacity later. It is sitting on a North American ESS order backlog of roughly 140 gigawatt-hours built up by the end of 2025, and it is targeting about 90 gigawatt-hours of fresh storage orders in 2026, anchored by contracts such as a multi-gigawatt-hour supply deal with US utility DTE Energy. On this view, the underlying loss is the cost of retooling, and margins turn genuinely positive once the converted lines run full and the backlog converts to shipments.

That case is testable, which is the point. If the pivot is real, the ex-AMPC line, the underlying operating result, should climb durably above zero and toward the roughly 200-billion-won-a-quarter level that would signal storage is earning a real margin, not just collecting the same subsidy at higher volume. It has not done that yet. The cleanest quarter to date, with the storage lines finally running, produced a deeper loss than a year earlier. The trough thesis is plausible; it is simply not yet visible in the numbers.

Step back and the strangeness of the setup comes into focus. The AI-data-center storage boom that LGES is chasing is being underwritten, quarter by quarter, by a credit that Congress designed to reshore electric-vehicle manufacturing. Because 45X pays the same amount for any qualifying cell, the money designed to move car-battery production to America is now doing much of the work of making grid-storage production look profitable. That is not fraud or even a loophole; it is exactly how the credit is written. But it means the reported profitability of the pivot is hostage to the politics of a different industry.

And that politics is moving. Under current law the 45X cell credit is a fixed per-kilowatt-hour amount only through 2029; it then falls to 75 percent of full value in 2030, 50 percent in 2031, 25 percent in 2032, and zero after that. The 2025 One Big Beautiful Bill Act layered on a 'material assistance' rule that disqualifies batteries relying too heavily on components and minerals from restricted foreign suppliers, chiefly China, with the allowed share tightening from 2026 through 2030. LGES, as a Korean company, is not itself a 'prohibited foreign entity,' and a separate buyer-side credit, the Section 48E investment tax credit for storage projects, can even favor its non-Chinese equipment. But the load-bearing number in last week's profit headline still sits directly on top of a subsidy that is scheduled to shrink and is being fenced with new conditions. Any tightening lands straight on the bottom line.

The preliminary numbers get their audited confirmation on July 30, and the segment detail there will matter more than the headline. The single question that settles this is whether the underlying, ex-subsidy result finally crosses zero on storage margin rather than on subsidy volume. Watch two lines together: the AMPC credit, which will keep rebuilding as the converted lines fill, and the ex-AMPC operating result, which has to climb on its own. If the second line follows the first up, the pivot is real. If only the first one moves, the profit will keep arriving with a receipt from the US Treasury attached.