Long-Term Thesis

Long-Term Thesis — 5-to-10-Year View

1. Long-Term Thesis in One Page

The long-term thesis is that Safran is a regulator-enforced annuity disguised as an industrial: between now and 2035, the CFM installed base swells from roughly 45,000 engines toward 60,000, the LEAP layer compounds from a thin-margin OE drag into the dominant shop-visit franchise on every narrowbody wing, and the partnership with GE — renewed through 2050 in 2025 — locks half the economics into Safran's books for another twenty-five years past the original CFM56 end-of-life. The 5-to-10-year case works only if (a) the CFM RISE open-rotor architecture wins the next-generation narrowbody decision around 2028–2030, (b) Propulsion margin holds in the 22–24% band as the LEAP aftermarket layer adds in, and (c) management converts the headline operating income to cash at 65–75% through a full cycle rather than a single peak year. This is a long-duration compounder if those three threads hold; it is a peak-multiple cyclical if any one of them breaks. The franchise trades at 20× EV/EBITDA against GE Aerospace at 34× on the other 50% of the same CFM cash-flow stream — the spread between Safran's structurally protected aftermarket and a blended-industrial multiple is the underwriting question over a decade.

2. The 5-to-10-Year Underwriting Map

Seven things have to be true for a buyer at €298 to compound at an institutional rate of return over the next decade. Each is observable, and each has both a validating and a refuting evidence path.

Driver #1 (CFM aftermarket compounding) does the most work in the underwriting because it is already showing up in the print — the +21.6% civil aftermarket USD growth in H1-25 is faster than any plausible flight-hour growth and is the financial statement footprint of the property right earning its keep. Driver #2 (RISE) is the binary that sets the multiple beyond 2035. Drivers #3 and #4 are the financial signals to monitor that any of this is converting to cash on the way through. The investor who tracks only the first three has the thesis; the investor who also tracks the next four owns it.

3. Compounding Path

The compounding model is revenue × margin × cash conversion × capital discipline. Revenue grows at 7–9% to 2030 as LEAP deliveries ramp toward 2,000 units/year and the aftermarket installed base layers in; recurring operating margin expands from 16.6% (FY25) toward 19–20% as Propulsion sustains 22–24% and Equipment walks toward mid-teens; cash conversion runs ~70% of operating income on a clean WC basis; and the share count drifts down 1.0–1.5% per year on the live €5B buyback. The 2028 ambition raised mid-cycle (€7.0–7.5B EBIT, ~€21B cumulative FCF) is the management framing; the 2030 case is the same compounding curve extended two years.

The chart tells the cycle story plainly: revenue and ROI moved in lockstep through 2018, broke catastrophically in 2020 (-34% revenue, -77% ROI), recovered to peak in 2024, and by 2025 ROI has accelerated to a multi-year inflection as the LEAP aftermarket layer adds in. ROI growth has outpaced revenue growth in 2024 and 2025 — the operating-leverage signature of an aftermarket-rich franchise on the upswing.

The ROIC path is the most demanding test in the compounding model. Today's consolidated 5.7% understates the underlying franchise quality because the IFRS denominator carries €13.3B of goodwill and intangibles from the 2005 merger and the 2018 Zodiac deal. The path to 10%+ by 2030 requires Propulsion margin to hold the 22–24% band, Equipment to walk to 15%, Interiors to clear 5%, and no further large goodwill-creating M&A — all observable, none guaranteed.

4. Durability and Moat Tests

Five tests that determine whether Safran's franchise survives a normal 10-year stress envelope. Each pairs a competitive or financial signal with a validation and a refutation marker. The competitive tests (#1, #2) decide whether the property right holds; the financial tests (#3, #4, #5) decide whether the franchise converts to cash.

Only one of the five tests (RISE selection) is genuinely binary and existential to the wide-moat rating beyond 2035 — every other failure is a margin or multiple compression that the franchise survives. That asymmetry is the single most important fact in the durability picture: four ways to be wrong on the multiple, one way to be wrong on the franchise.

5. Management and Capital Allocation Over a Cycle

The capital allocation pattern under CEO Olivier Andriès (Jan 2021–) is the cleanest piece of evidence that this management improves rather than impairs the long-term thesis. He inherited a franchise built by predecessors — CFM JV, LEAP launch, Snecma-Sagem merger — and has spent five years compounding it rather than reinventing it. Every CMD21 target was met or beaten (revenue CAGR ~20% vs 10% promised, ROI CAGR ~30% vs 20%, cumulative FCF >€14B vs €10B), and the 2028 ambition was raised mid-cycle from €6.0–6.5B EBIT to €7.0–7.5B and from €15–17B cumulative FCF to ~€21B. The credibility pattern from History grades this team an 8/10 — they over-deliver on profit and cash, they walk down OE delivery guides when supply binds, and they do not narrate setbacks like the €(244)M SPI capital loss in FY25.

The capital allocation track record breaks into four clean periods of evidence:

The structural drag on alignment is the French State golden share (11.7% capital, 18.4% votes via double voting rights) — the State can block transformative change. The structural support is the activist on the register (TCI 8.1% capital, 9.6% votes) — a credible counterweight pushing for capital return discipline. Executive ownership in percentage terms is rounding error (CEO Andriès owns 0.012% of the company), so alignment runs through reputation and bonus rather than personal wealth at risk. The most important single piece of evidence for an investor underwriting the next 10 years is that CEO variable pay carries explicit working-capital and receivables weight (10% inventory + 5% unpaid receivables) — meaning the forensic tab's #1 yellow flag (FY25 receivables growing 35.5% vs revenue 12.5%) is a number this CEO is paid to fix. If that pattern reverses cleanly in H1-26, the alignment evidence strengthens; if it doesn't, the discipline assertion needs revisiting.

6. Failure Modes

Five failure modes that would force a long-term thesis revision, ranked by what would actually break the underwriting rather than the print. Failure Mode #1 (RISE loss) is the only one that flips the wide-moat rating to narrow over a 15-year horizon; the other four compress the multiple without breaking the franchise.

The red-team conclusion: the long-term thesis survives every failure mode except #1 (losing RISE) — that one structurally re-rates the franchise on a 15+ year fade clock. Failure modes #2 through #6 are multiple compressions the long-duration cash flows can absorb. An investor sizing this position should anchor risk management to the RISE/next-narrowbody timeline, not the FY26 print.

7. What To Watch Over Years, Not Just Quarters

Five multi-year observable signals that update the long-term thesis. Each is in primary disclosure or trackable in public industry sources; none requires management commentary to verify.