Samsung SDI Q2 2026 — Profitable Again, but the Battery-Segment Bridge Stays Blank

Samsung SDI's battery segment reported KRW 159.3B operating profit in Q2, its first positive quarter since Q3 2024. The trailing sequence — KRW -630B, -339B, -177B, then +159B — moves in one direction and the steps are getting larger. What the segment earned without US production credits cannot be determined. Samsung books its 45X credits as AMPC and discloses the figure only at consolidated level (KRW 107.7B in Q2), with no segment allocation. A tariff refund also supported the quarter and was not quantified. The gap I flagged in the Q1 decomposition is unchanged: LGES's credit dependency can be read quarter by quarter, Samsung's cannot.

Trailing Decomposition

Figures from Samsung SDI earnings releases and KRX preliminary filings. AMPC from KRX filings at consolidated level. USD at approximately KRW 1,370/USD for Q2 2026.

¹ DART H1 2026 filing: 860M units produced against 1.178B unit capacity at the Cheonan and Tianjin small-battery plants. Power tools, micromobility, consumer electronics. It carries no information about EV or ESS lines.

Three rows are blank across all four quarters and a fourth carries a single entry that measures something else. Those are the rows a buyer needs to size a supply commitment: segment profitability net of credits, product mix, and capacity actually in use.

Pass 1: 45X Isolation

AMPC rose 34% sequentially, from KRW 80.5B to KRW 107.7B, the highest reading in the series. That is consistent with Samsung's reported ramp of US ESS production at StarPlus — the Indiana joint venture with Stellantis — and with its new prismatic EV cell launch. Samsung does not confirm what drives the increase at application level.

The mechanical consolidated ex-AMPC figure of KRW 96.1B is positive for the first time in the series. On the Q2 earnings call, management said the consolidated result would have stayed "slightly positive" without the tariff refund. That helps with sign and not much else: subtract the refund from KRW 96.1B and "slightly positive" accommodates anything from a few billion won to KRW 50–60B. The refund was not quantified.

The battery-segment ex-AMPC row stays blank for the fourth consecutive quarter, because Samsung reports AMPC at consolidated level only and that consolidation includes electronic materials. No public basis exists for splitting it. Attribute the full KRW 107.7B to batteries and the segment's ex-AMPC profit is KRW 51.6B; attribute some share to electronic materials and the battery figure rises. Editorial inference: most of it almost certainly belongs to batteries, since Samsung's US manufacturing footprint is overwhelmingly cell production. "Almost certainly most" does not earn a table cell.

The AMPC series itself warrants caution. It went from KRW 19.5B in Q3 2025 to KRW 79.8B in Q4, a fourfold move Samsung has never explained. Production disruption in Q3, a recognition-timing shift, or a genuine step-change in eligible output would all produce that pattern, and the company has not distinguished between them. Sequential AMPC comparisons at Samsung carry noise the issuer does not help the reader filter.

Pass 2: Utilization Trend

There is no data series to read. The 73% figure in the DART filing covers small-battery lines at Cheonan and Tianjin, and as the Issue #11 sidebar set out, it tells you nothing about EV or ESS capacity.

No named third-party source published a Q2 or H1 2026 utilization estimate for Samsung SDI's EV or ESS lines in the material reviewed for this piece. The nearest thing is a management forecast reported by ETNews: Samsung expects Hungary utilization above 70% in H2, after it reportedly fell to roughly 40% in Q4 2025. Forward-looking, single site, not a reported actual.

SNE Research put Samsung's global ESS shipments at 6.4 GWh in H1 2026, up 20% YoY. There is no disclosed ESS capacity denominator to convert that into a utilization rate.

The question this section normally answers for a Korean maker — is the rate of utilization decline slowing, flattening, or turning — has no answer at Samsung for the lines that matter. LGES discloses a companywide battery utilization rate. SK On's can be partially inferred from SK Innovation's consolidated disclosures against Hyundai production data. Samsung discloses small-battery utilization and stops there.

Pass 3: Guidance vs. Actuals

The Q1 release guided "gradual recovery" from Q2 and put quarterly profitability in H2 2026. No numerical revenue, margin, or shipment target accompanied it.

Q2 delivered positive operating profit a quarter ahead of that timeline. Samsung attributed the result to stronger-than-expected revenue across business units, with the battery improvement coming from high-power UPS, BBU and power-tool cells plus higher European EV-battery sales.

Direction and timing both landed where guided or better. The tariff refund muddies even that. If the refund was large enough to flip the quarter, the underlying business may still be running to the original H2 schedule; management's "slightly positive" remark suggests the organic result also crossed zero, but a single adverb is doing work a number should have done.

The more useful figure is the ratio. A KRW 336B swing in operating profit, from -176.6B to +159.3B, arrived on battery-segment revenue that grew 4.9% QoQ to KRW 3.519T and remains below the Q4 2025 peak of KRW 3.622T. That is margin, not volume. Mix shift, cost reduction, AMPC growth, the tariff refund, or some combination — Samsung's disclosure does not permit decomposition, and the components have different half-lives. One-time items and a growing credit line do not compound the way structural cost reduction and sustained mix improvement do. Management's Hungary forecast implies EV volume recovery ahead, but if that recovery had begun in Q2 it did not show up in the revenue line. Timing, product mix, or pricing pressure would each explain that, and the available data does not separate them.

Sourcing Decision Frame

Q2 changed the direction of the trajectory and left the disclosure gap where it was. A counterparty earning KRW 159B a quarter is a different risk than one losing KRW 177B, and that difference is real regardless of what produced it. What a supplier qualification decision needs, though, is whether the earning power survives without credits, and the filing does not say.

The forward ESS position raises the stakes on that question. Samsung is the only non-Chinese prismatic cell supplier manufacturing in the US, which matters to a buyer whose specification calls for prismatic format and matters considerably less to the many BESS integrators that have historically been format-agnostic. StarPlus is targeting October LFP cell production with customer supply by year-end. The New Carlisle acquisition is intended for ESS conversion, supported by an L&F cathode contract starting 2027. Hana Securities estimated ESS at 20.6% of consolidated Q2 revenue rising to 35.1% by Q4 — sell-side, not issuer disclosure. New Carlisle still has no disclosed purchase price, revised capacity, confirmed chemistry, or production-start date; as of August 28 every quantified figure for that asset traces to sell-side estimates. StarPlus LFP yield, accepted GWh, and named customer acceptance remain unconfirmed.

This analysis relies on Samsung SDI's English-language IR materials and Korean-language earnings-call reporting by ZDNet Korea and ETNews. The Korean-language call transcript was not independently reviewed.