--- title: "Issue archive" canonical: "https://anchor.is/the-hysteresis/issues" publication: "The Hysteresis" publisher: "The Hysteresis" language: "en-US" guide: "https://anchor.is/the-hysteresis/llms.txt" issues: 14 --- # The Hysteresis > The Hysteresis exists to give a small cross-functional team — strategy, investment, and procurement — a shared weekly instrument for reading the battery supply chain earlier than consensus. Success means the team consistently identifies inflection points in cell pricing, supplier health, and policy enforcement weeks or months before they surface in earnings calls, annual surveys, or trade press. The commercial mechanism is internal-first: the publication proves itself as a decision tool inside the organization, then extends to paid subscribers and a practitioner community as upside. Revenue follows utility — if the team makes better-timed procurement calls and competitive assessments because of what they read here, the publication earned its place. [Read on the website]() ## Issue archive - [Issue 14]() — Published 2026-09-14: We traced a supply route this week that looked clean from the outside: a Samsung-Stellantis joint venture in Indiana, posting revenue, converting a line to LFP. Follow the structure and it gets complicated fast. Samsung's 51% stake is pledged as collateral on a DOE loan already cut from $7.5 billion to $4.1 billion. The anchor customer holds the other 49%. Any external sale has to navigate both constraints before a cell ships. If you're qualifying Korean supply for a US project right now, the route matters more than the logo on the datasheet. This issue follows the routes. - [Issue 13]() — Published 2026-09-07: We kept coming back to two price assessors agreeing a Chinese storage cell costs CNY 0.365 per watt-hour and disagreeing on whether that number rose or held flat. China's consumption tax took effect September 1. Neither assessment separates the tax from the underlying move. That problem showed up everywhere this week. All three Korean battery makers posted better Q2 numbers for unrelated reasons. A US defense cost threshold can shift on commodity prices without the product changing. The readings look stable until you ask what they actually measured. - [Issue 12]() — Published 2026-08-31: China's tax administration resolved the consumption tax boundary on Wednesday: clusters are taxable September 1, complete storage systems are not. You need that before September quotes land. It also means the first post-September cell price will blend pre- and post-tax transactions into one number. Anyone reading it as a clean move will be wrong. We kept hitting that everywhere this issue. Samsung's first positive quarter in seven can't be fully decomposed. Three Korean makers report utilization using three different definitions. When the instruments shift alongside the market, knowing which numbers still hold weight is the whole job. - [Issue 11]() — Published 2026-08-24: The world's largest cell maker doubled its finished goods inventory in six months while running 95% utilization. RMB 47.6 billion in unsold product, backed by RMB 372 billion in cash. We kept returning to that because it changes what the flat cell price tape means. The surplus is growing at twice last year's rate industry-wide, and the entity holding the largest share has no financial reason to move it on anyone else's timeline. The price looks stable. Underneath it is a holder choosing when to release, with the balance sheet to choose slowly. This issue is about what that tape leaves out. - [Issue 10]() — Published 2026-08-17: Three US plants say they're producing ESS cells. Korean makers all posted Q2 profit. Chinese cell prices held flat. Put those together and the supply chain looks like it's stabilizing. We spent this issue one layer down. No US plant discloses converted GWh. LGES, the most transparent Korean maker, is generating earnings while burning cash. Chinese prices hold because a supplier-credit layer absorbs costs that regulators have started forcing into the open. The stabilization is real in the headline. Whether it survives the balance sheet is what this issue is about. - [Issue 9]() — Published 2026-08-10: We set out to answer one question: how much US cell capacity can a buyer actually contract right now? Producing, qualified, compliant, not locked to someone else's program. The number is small enough to reframe the terawatt-hour of announcements sitting behind it. SK On posted 821 billion won in quarterly operating profit and we still can't isolate how much comes from recurring battery operations. China has ESS lines running a cell format the market is already moving past. The reported figure and the actionable figure have quietly diverged. This issue measures how far. - [Issue 8]() — Published 2026-08-03: We spent this week trying to verify things. All three Korean cell makers reported Q2 improvement. Only one disclosed enough for us to decompose the result to an ex-credit number. That gap between reported and provable kept showing up: in Chinese export growth heading into a narrowing corridor of tax changes and equipment restrictions, in a US cancellation count where a bankrupt facility and a converted production line register as the same event. The number always exists. Whether you can build a sourcing decision on it is what this issue works through. - [Issue 7]() — Published 2026-07-27: Samsung SDI reported KRW 80.5 billion in AMPC credits last quarter. You won't find that in the English investor deck. It sits in a KRX regulatory filing, in Korean, and it reframes the July 30 comparison: LGES looks more credit-dependent than Samsung partly because LGES actually shows the dependency. We kept running into that same dynamic all week, from Ford Kentucky's BESS pivot to lithium-indexed contracts that miss the iron phosphate cost moving against them. The risk right now lives in the layer between what gets filed and what gets presented. - [Issue 6]() — Published 2026-07-20: Lithium carbonate fell 5% this week. The China ESS cell price barely moved. The system price didn't move at all. We kept coming back to that gap. One Chinese cathode processor raised fees into a falling lithium market, which is the clearest sign the floor isn't about lithium anymore. Processing costs, a new safety standard, a shrinking export rebate: these don't soften when lithium does. If your procurement model assumes upstream relief flows downstream on a predictable lag, this issue maps where the pass-through actually breaks, and why the timeline is longer than the spot chart suggests. - [Issue 5]() — Published 2026-07-12: We kept hitting one gap: what a battery facility is called versus what a buyer can actually contract from it. Ford's Kentucky plant has a certificate of occupancy, a $3.8 billion federal loan, sole Ford ownership on file with the SEC. No disclosed output. That distance between label and proof held across every section we filed. Chinese ESS cells repriced up while most models still assumed decline. Korean makers signed real supply agreements alongside losses they can't cover without subsidies. Defense compliance starts in eighteen months and no visible supplier has a complete proof file. This issue runs the audit. - [Issue 4]() — Published 2026-07-06: We spent this week looking for denominators. Samsung SDI reported a KRW 177 billion battery segment loss last quarter without disclosing whether US production credits were inside that number. LG Energy Solution disclosed theirs and looked worse for the honesty. That kept happening. Korean utilization hasn't refreshed since March. US plant trackers count cancellations without separating them from conversions. China's cell price went flat, so the quote stopped carrying information. Every section of this issue landed on the same condition: a claim sitting on a number nobody published. - [Issue 3]() — Published 2026-06-29: BNEF's annual benchmark says storage packs fell 45% last year. That number is in every deck we've seen this quarter. InfoLink's spot data, six months fresher, shows Chinese ESS cells climbed 22%. Both are sourced, both are real, and they've stopped describing the same market. We kept pulling that thread across U.S. plant conversions, Korean earnings, defense procurement timelines. Each one came back to the same problem: the number you can cite in a presentation and the number that governs a sourcing call have quietly diverged. - [Issue 2]() — Published 2026-06-09: Lithium reversed from its two-year high. China's cell price floor didn't follow. Five forces hold it up — each on its own countdown. - [Issue 1]() — Published 2026-05-25: Korean makers all posted losses for the first time as cheap lithium inventory ran out. The next quarter's quotes carry costs no buffer absorbs. ## More - [Publication guide]() - [Issue archive]()