SCOR H1 2026: An 8.5-Point IFRS 17 Discount Behind the 79.9% Combined Ratio

Strip SCOR's Q2 2026 property and casualty result down to its attritional loss and commission ratio and the number is 76.8%, not the 79.5% quarterly combined ratio carried in the headlines (SCOR, July 2026). The eight-and-a-half-point gap is the IFRS 17 discount effect, an accounting mechanism tied to interest rates rather than claims experience, and it is the number a reinsurer's underwriting margin actually turns on.

79.9%

H1 2026 Combined Ratio

–8.5 pts

Q2 Discount Effect

76.8%

Q2 Attritional + Commission Ratio

220%

Group Solvency Ratio

SCOR published its second-quarter and half-year 2026 results on July 30, reporting group net income of EUR 171 million for Q2 and EUR 397 million for H1, up from a combined ratio of 83.7% in H1 2025 to 79.9% this half (SCOR, July 2026). Trade coverage of the release led with the ratio improvement. What it did not do, because SCOR discloses it only in the underlying component table rather than the headline, is decompose how much of the 3.8-point year-over-year gain came from better underwriting and how much came from a discount rate that has nothing to do with how SCOR's treaties are performing.

The Discount Effect: How Interest Rates Enter a Combined Ratio

Under IFRS 17, a reinsurer does not simply add up paid and reserved claims to build its loss ratio. It measures the liability for incurred claims at the present value of expected future cash flows, discounted using either a top-down curve built from the actual asset portfolio's yield or a bottom-up curve that adds an illiquidity premium to a risk-free rate (American Academy of Actuaries, 2026). When market rates sit above where they were when a claim reserve was first booked, discounting that reserve to present value shrinks it, and a smaller incurred-claims liability flows straight through to a lower combined ratio. None of the underlying claims payments changed. Only the interest rate used to value them did.

SCOR's own Q2 print shows the mechanics cleanly. The quarter's nat cat ratio was 2.9%, the attritional loss and commission ratio was 76.8%, and the attributable expense ratio was 8.2%, summing to roughly 87.9% before any discounting is applied (Reinsurance News, July 2026). The discount effect then subtracted 8.5 points, landing the reported Q2 combined ratio at 79.5%. Run the same arithmetic on any reinsurer's disclosure and the discount line is the one that moves with the yield curve rather than with claims development, which is exactly why it deserves separate scrutiny from an actuary reading the print rather than a headline writer summarizing it.

The scale of the effect is not new to 2026, and it is not small. Munich Re's non-life reinsurance combined ratio for full-year 2022 came in at 96.2% under the old IFRS 4 framework versus 83.2% once restated under IFRS 17, a 13-point swing driven almost entirely by discounting in a rising-rate year (S&P Global Market Intelligence, June 2023). SCOR's 8.5-point Q2 2026 effect is smaller than that historical example but sits in the same family: a genuine, rules-compliant accounting number that has almost nothing to do with whether the treaties SCOR wrote in 2026 are adequately priced.

Q1's 10-Point Tailwind, Q2's 8.5: A Duration-Linked Wobble

The discount effect is not a fixed addition to SCOR's combined ratio; it moves quarter to quarter with the shape and level of the curve SCOR applies. In Q1 2026, the discount effect was larger still, at negative 10 points, which SCOR attributed to benign nat cat activity, higher locked-in interest rates, and additional prudence built into best-estimate liabilities (SCOR, May 2026). By Q2, the same mechanism contributed 1.5 points less benefit. Blend the two quarters and the H1 nat cat ratio comes to 3.5%, itself higher than Q2 alone, which is one reason the full first-half combined ratio of 79.9% sits slightly above the Q2-only 79.5% print even as both benefit from meaningfully negative discount effects.

The mechanical sensitivity matters more than the specific quarterly swing. A discount benefit built on the gap between current market yields and the yields locked in when older reserves were first booked is, by construction, a function of the level and duration of SCOR's liability portfolio relative to the prevailing curve. If yields hold near current levels, the tailwind persists roughly in its current range, continuing to flatter reported underwriting margins relative to the attritional print. If yields fall meaningfully, the arithmetic reverses: newer reserves get discounted less aggressively, older reserves that were locked in at higher rates start to look relatively expensive to hold, and the combined ratio absorbs the reversal as a real charge even though claims experience has not deteriorated. That asymmetry is the central risk in reading an 8.5-point discount benefit as durable rather than as a rate-cycle artifact.

This is not a hypothetical concern confined to SCOR. Reinsurers reporting under IFRS 17 do not prescribe a single interest rate curve; different discount rate methodologies can produce materially different combined ratios for economically similar liabilities, and comparability research has found dozens of distinct EUR yield curves in use across insurers that disclose the methodology at all (S&P Global Market Intelligence, June 2023). A reinsurer holding a longer-duration liability book, or a book concentrated in casualty lines where claims settle over many years, carries more discount-rate sensitivity in its reported ratio than a property-heavy peer whose claims pay out within a year or two. SCOR's own casualty and life reinsurance mix gives it meaningfully more of that duration exposure than a pure property-cat writer.

A EUR 64 Million Arbitration Charge Inside Life and Health

The discount effect is the mechanical story in P&C. The one-off story sits in Life and Health, where the insurance service result fell to EUR 49 million in Q2, down from an underlying EUR 113 million once a EUR 64 million negative experience variance tied to an arbitration outcome is excluded (SCOR, July 2026). The full experience variance line came to negative EUR 60 million against CSM amortization of EUR 88 million and a risk adjustment release of EUR 28 million, with onerous contract impacts subtracting a further EUR 10 million.

The detail worth underwriting attention here is source, not just size. A EUR 64 million hit to reported earnings that originates in an arbitration outcome is not a mortality or morbidity signal, and it should not be read into any inference about the adequacy of SCOR's life and health pricing assumptions. It is a legal and contractual dispute resolution landing in the same period as the ordinary experience variance line, which is precisely the kind of one-off that IFRS 17's granular disclosure is supposed to let an analyst isolate. CEO Thierry Léger characterized the quarter in broader terms on the earnings call, calling it "another strong set of results this quarter, demonstrating consistency and resilience of its earnings" (SCOR, July 2026), a framing that holds up better once the arbitration charge is separated from the underlying EUR 113 million result than it would if the reported EUR 49 million were taken as SCOR's clean read on the health of its life reinsurance book.

For anyone tracking IFRS 17 CSM mechanics more broadly, the Q2 print is a useful worked example of how the standard's building blocks interact in a single quarter: a stable amortization release, a risk adjustment unwind, and a volatile experience variance line that can swing on a single legal outcome large enough to cut a reported result by more than half.

Pulling Back on Non-Cat US Property While Holding Cat Flat

SCOR's June and July 2026 renewals give the discount-effect story a capacity-side counterpart. The reinsurer reduced participation in non-cat US property and US casualty lines while holding its property catastrophe position flat, even as overall gross premium volume grew 3.2% to EUR 6.455 billion on the back of Specialty Lines growth of 19.8% and Alternative Solutions expansion of 133% (Artemis, July 2026). P&C renewal business as a whole fell 4.8%, with gross price change across the renewed book down 4.4%, a spread between non-proportional pricing (down 9.5%) and proportional pricing (down only 0.6%) that tracks the broader mid-year softening already visible across the sector.

Léger was explicit that the pullback reflects capacity discipline rather than a retreat from the market: "During the June-July 2026 renewals, SCOR continues to grow in its diversifying lines, applying underwriting discipline in a competitive environment," and he expects the January 2027 renewal to bring "plenty of capital available" given benign loss activity that has not triggered reinsurance attachment points across much of the market (Artemis, July 2026). Holding property cat flat while cutting non-cat US property is a specific signal: SCOR is defending the segment where it judges pricing still clears its return threshold and shedding the segment, non-cat property, where competitive softening has moved the terms below that line first.

For cedants renewing non-cat US property or US casualty programs in the second half of 2026, the SCOR data point argues for treating capacity as conditional on portfolio quality rather than assuming the record aggregate reinsurance capital pool translates into uniformly available terms. A reinsurer posting a 79.9% combined ratio and 220% solvency is not retreating from the market for capital reasons; the selective pullback is a pricing judgment specific to where SCOR sees margin eroding fastest.

Comparing Reinsurers Without Comparable Discount Disclosure

SCOR's practice of publishing an explicit discount-effect line in its combined ratio bridge is more transparent than the disclosure most of its peers provide, and that transparency is exactly what makes cross-reinsurer combined ratio comparison harder rather than easier. A reader can adjust SCOR's 79.9% down to something closer to its 76.8% attritional-and-commission core. Most peer disclosures do not offer an equivalent line item, so the same adjustment cannot be made consistently across the sector.

Reinsurer	Period	P&C Combined Ratio	Prior-Year Period	Full-Year Target
SCOR	H1 2026	79.9%	83.7%	<87% (2024–2026 plan)
Swiss Re (P&C Re)	Q1 2026	79.5%	86.0%	<85%
Munich Re (P&C Re)	Q1 2026	66.8% (80.3% normalized)	n/a	~80% normalized
Hannover Re (P&C)	Q1 2026	83.6%	93.9%	<87%

Swiss Re's P&C Re combined ratio came in at 79.5% for Q1 2026, down from 86.0% a year earlier and against a full-year target of below 85% (Swiss Re, May 2026). Munich Re's P&C reinsurance combined ratio was 66.8% in Q1 2026, a figure inflated by an unusually benign major-loss quarter; Munich Re's own normalized ratio of 80.3%, adjusted for expected loss experience rather than for discounting, is the number the company itself points to as aligned with full-year guidance (Reinsurance News, May 2026). Hannover Re posted an 83.6% P&C combined ratio in Q1 2026, improved from 93.9% a year earlier (Reinsurance News, May 2026).

Lined up as reported, the four ratios cluster in a band from the high 60s to the low 80s, and a reader could reasonably conclude SCOR and Swiss Re are running neck and neck while Munich Re is running well ahead. That comparison is not wrong, but it is incomplete in a specific way: none of these figures isolates how much of the reported number is discount benefit versus underwriting margin using the same methodology. Munich Re's normalization strips out loss-experience noise, not discount effect. Swiss Re and Hannover Re do not publish an equivalent discount-effect bridge line at all in their headline releases. SCOR is the one reinsurer in this set that hands an analyst the tool to make the adjustment, which means the fairest comparison available is SCOR's disclosed attritional-and-commission ratio of 76.8% against the others' undisclosed equivalents, not their headline combined ratios against each other.

Reading the Print Behind the Ratio

The practical takeaway for anyone using SCOR's 79.9% combined ratio in a peer benchmarking exercise, a rating agency capital model, or a reserve adequacy review is to treat the discount-effect line as a rate-cycle input rather than an underwriting result. SCOR's attritional-and-commission ratio of 76.8% in Q2, held roughly flat against a nat cat ratio that stayed low and an expense ratio near 8%, is the closer proxy for the pricing discipline SCOR's underwriters are actually exercising this half. The gap between that number and the 79.9% headline is a function of where the yield curve sits today, and it will narrow or reverse as that curve moves, independent of anything SCOR's treaty underwriters do differently.

The comparability problem this creates is not unique to SCOR, and it is not going away as more reinsurers mature their IFRS 17 disclosure practices. Until Swiss Re, Munich Re, and Hannover Re publish discount-effect bridges with the same granularity SCOR provides, cross-reinsurer combined ratio league tables built from headline numbers alone will keep mixing genuine underwriting improvement with interest-rate accounting that has nothing to do with treaty pricing. An analyst who wants to know which reinsurer's underwriting actually got better this half, rather than which one benefited most from the shape of the yield curve, has to go looking for the component table every single quarter, because the headline print will not tell that story on its own.