2507 — Cirrus Aircraft Limited

Cirrus Aircraft Limited (HKEX: 2507) — deep-value analysis

Date: August 25, 2026 · Price: … (close 08/25/2026) · Market cap: … = … at …/USD Shares: 365,988,818 · Reporting currency: USD · Trading currency: HKD · FY2025 auditor: PwC (replaced by KPMG from the AGM of 06/26/2026) Primary sources: FY2025 annual report (HKEXnews 2026042403596, filed 04/24/2026), FY2024 annual report (2025042804425), H1 2025 interim report (2025092900331), H1 2024 interim report (2024091900429), listing prospectus 06/28/2024, HKEX announcements 2025–2026. All accounting figures are in thousand USD unless stated otherwise.

Calendar warning. The board meets on August 26, 2026 — tomorrow — to approve the H1 2026 half-year results (HKEX announcement dated 08/13/2026). This analysis rests on the last complete audited set (FY2025) and H1 2025. Any figure published tomorrow could materially shift chapters 4, 5 and 8. This is the worst possible moment of the year to make a decision without waiting 24 hours.

Executive summary

Cirrus Aircraft manufactures and sells premium light aircraft: the piston-engine SR20/SR22/SR22T series (SR2X) and the Vision Jet, the world’s only certified single-engine personal jet. The company is the best industrial business I’ve seen enter this tracker in recent weeks: 24 consecutive years as the leader in single-engine piston aircraft, 8 consecutive years as the leader in general-aviation jets, gross margin rising uninterrupted for five straight years (32.8% in 2021 → 35.2% in 2025), revenue grown from … to … (CAGR …), net profit from 72.4 mil. to …, a balance sheet with net cash of … and an operating ROIC around 30%. The accident rate is three times below the US general-aviation average, and the CAPS ballistic parachute has saved over 290 lives since 1999 — that’s the real moat, not marketing.

And yet my thesis is negative at the current price, for three compounding reasons.

First: the money doesn’t reach the shareholder. The company reported … of cumulative net profit in 2023–2025 and produced … of free flow to shareholders over the same three fiscal years — a conversion of 32.5%. The gap isn’t a one-year anomaly, it’s the business’s structure: total capex of 90.3 / 94.4 / … a year against depreciation and amortization of 36.6 / 45.0 / …, plus … a year of capitalized development expense (IAS 38) that at a US issuer would have been expensed. FY2025 FCFE: …, a yield of 1.34% at the current market cap. The proposed dividend (…) is 1.55 times the year’s free flow.

Second: the engine started slowing in the second half of 2025. The full year shows … revenue and … profit. Broken down by half, H2 2025 had revenue … and net profit … versus H2 2024, with operating margin 12.1% against 14.7%. Part of this is a base effect (2024 deliveries were pushed into H2 by G7-launch production problems). The rest isn’t: the order book fell from 1,135 to 1,066 aircraft (…), Vision Jet bookings from ~240 to ~221, and the value of unsatisfied performance obligations grew only 0.8% — from 1,670.7 to … — while revenue grew 13.1%. Net orders (728) were below deliveries (797).

Third: the ownership risk became concrete on June 8, 2026. AVIC, through CAIGA Hong Kong, holds 80.18%. On the same day the Pentagon added Cirrus to the Section 1260H list of Chinese military companies operating in the US, AVIC announced the sale of Continental Aerospace Technologies — Cirrus’s piston-engine supplier and a related party — to the American fund Arcline for …. The ban on Department of Defense contracting took effect June 30, 2026, and the ban on purchasing parts and services follows on June 30, 2027. On July 16, 2026, the “General Aviation Protection Act” was introduced in the House, which would reopen the review of the 2011 acquisition. None of this yet touches civilian sales — but it changes the discount rate and, more importantly, opens up a forced-sale scenario in which the price a buyer pays for 80% of a company under political pressure isn’t the minority shareholder’s market price.

Estimated value. Five triangulated models give a band of 22–… per share, with a median of …. Monte Carlo across 20,000 scenarios (owner-earnings base …, g1 8%, r …%, gt 2.5%, net cash … mil.) gives a median intrinsic value of … and a probability of undervaluation of …%. The current price of … implies overvaluation of ~27% versus the median. The consensus of the three analysts covering the stock is … — but those same three analysts had the same target when the stock quoted 67.56, and it’s since lost 33%.

Verdict: AVOID at …. It isn’t a bad company — it’s a good company at a price that assumes the cash-conversion problem resolves itself and that the ownership risk doesn’t materialize. The level at which the business becomes interesting, on the same owner-earnings base, is below … (the model median) and clearly attractive below … (the 25th percentile). A buyer who still wants exposure should wait at least for tomorrow’s results and, ideally, proof that FY2026 produces over … of FCFE.