SES AI Corp (SES) stock intrinsic value fair value estimate & owner earnings analysis

SES AI is an early-stage battery technology company, headquartered in Woburn, Massachusetts, that has spent over a decade developing lithium-metal battery cells and, more recently, has begun assembling actual product revenue. It carries out research and development in Massachusetts and South Korea, and maintains facilities in Shanghai and Shenzhen, China that are used for supply-chain development, manufacturing-process work and, since September 2025, energy-storage-system (ESS) manufacturing following its acquisition of UZ Energy, a China-based ESS maker. These are real operations of a US-incorporated (Delaware) public company, not a foreign-controlled entity; we saw nothing indicating Chinese ownership or control of the parent. For over a decade, the company's revenue model was built around joint development agreements (JDAs) with major automakers -- Hyundai, General Motors and Honda -- under which those partners reimbursed SES's R&D and pilot-line costs. Every one of those named agreements has now concluded: the Hyundai JDA ended December 2025, the GM JDA ended September 2024, and Honda's original JDA (and its successor services agreement) ends in mid-2026. None of them has yet converted into a disclosed commercial production order. In their place, the September 2025 acquisition of UZ Energy gave SES a real, already-operating ESS product business, and product revenue jumped from essentially nothing to about $7.4 million in the roughly three and a half months SES has owned it. Total 2025 revenue was $21.0 million, up from $2.0 million in 2024, but the most recent two reported quarters show that growth has not continued in a straight line: revenue was actually lower in the second quarter of 2026 than the first, and cash burn in the first half of 2026 was essentially unchanged from a year earlier, despite the revenue increase. That combination -- the legacy development-partner revenue running off on a known schedule, and the new ESS business not yet showing a clear growth trend -- is the central uncertainty in this story, and it kept us from crediting future growth we cannot yet point to in the filings. On the other side of the ledger, SES is well funded and debt-free. As of the most recent balance sheet it held roughly $155 million of cash and short-term investments (after netting a disclosed, already-committed deferred payment tied to the UZ Energy purchase), against no interest-bearing borrowings. That cushion buys time, but the company continues to spend cash faster than it takes it in -- operating cash outflow was about $58 million in 2025 and is running at a similar pace in 2026 -- so the cushion is not permanent by itself. Because the business changed shape mid-2025 with the UZ Energy purchase, and because the prior joint-development revenue is disappearing on a known schedule, we anchored this valuation on 2025 alone rather than averaging several years that describe a different business. We then built three cases for how the next ten years play out, discounting at a flat 6% (a small premium over long government bond yields, used the same way for every company we value, with risk expressed in the spread between the cases rather than in the rate). In the bear case, the joint-development revenue disappears as scheduled with nothing replacing it, the ESS business does not resume growing beyond its recent run-rate, and the company never rationalizes its cost base enough to offset the lost revenue. On that path, the operating business is worth substantially less than nothing over time -- it consumes cash rather than creating it -- and even after adding back today's cash and investments, our estimate comes out modestly negative, around -$0.48 per share. This case would likely also require SES to raise additional capital by selling more shares, diluting existing holders further; we have not tried to size that dilution precisely, but it is a real risk in this scenario. In the base case, the joint-development revenue runs off as scheduled and the ESS business holds at roughly its recent level rather than growing (we found no signed contracts or demonstrated multi-quarter growth to justify assuming more), while the company is forced -- simply by the arithmetic of finite cash -- to shrink its research spending to something closer to what its own revenue can support over the next several years. That produces continued, shrinking losses before settling near breakeven later in the ten-year window. Combined with today's cash and investments, this comes out to roughly $0.19 per share. In the bull case, the ESS manufacturing line -- which already exists and is already producing and shipping product, unlike the untested drone-battery manufacturing plans discussed elsewhere in the filings, which we did not credit with any value since they remain unbuilt and unproven -- scales up using capacity SES already owns, at a modest pace, while costs continue the multi-year decline the company has already demonstrated. On that path the business reaches modest sustained profitability within the ten-year window, and per-share value comes out to roughly $0.86. A few accounting notes: owner earnings in every year add back stock-based compensation, depreciation and amortization, and non-cash items such as gains or losses on a contingent share liability tied to the stock price, and then subtract those non-cash compensation costs and capital spending back out, so the figures reflect real cash economics rather than accounting income. We treated essentially all of the company's capital spending as necessary to keep the business running (rather than discretionary growth spending), because the revenue growth seen to date came from an acquisition, not from capital projects funding organic expansion. We could not fully reconcile roughly $0.8 million in one year's cash-flow reconciliation to a specific line item, an immaterial amount that does not affect the conclusions. We stress-tested the case that a full year of owning the ESS business would show clear, continuing revenue growth in 2026; the two most recent quarters instead showed a sequential decline, which is why we did not build acceleration into the base case. Given how much of this outcome depends on a single, recently acquired product line proving it can grow -- and how quickly the legacy development revenue is disappearing -- the margin-of-safety price is set with a wide cushion, and a figure closer to our base-case estimate would be supported only if the company discloses a real production or supply agreement with an automaker or other large customer, or shows two consecutive quarters of ESS product revenue growth confirming the business is scaling rather than plateauing.