Business

Figures converted from INR at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Bottom Line

Sagility is the world's only listed pure-play U.S. healthcare payer-BPM operator: an INR cost base translating regulated, non-discretionary U.S. payer back-office work into rupee earnings. The engine is wage arbitrage protected by HIPAA-grade compliance, 18-year top-client tenure, and a clinical mix that AI cannot easily strip. The risk the market debates is AI commoditisation; the actual risk worth watching is client concentration at ~60% top-3 wallet share. Valuation belongs on through-cycle FCF and incremental capital efficiency, not the headline P/E that looks expensive against generic Indian BPM and cheap against the Capgemini-WNS transaction multiple.

How This Business Actually Works

Sagility sells one core unit: a domain-trained human — offshore claims processor, U.S. nurse reviewer, coder — wrapped inside a HIPAA-compliant workflow, sold to U.S. payers under multi-year MSAs with rolling SOWs. Revenue is USD (89.7% payers, 10.3% providers FY26); cost is INR. Everything else — the technology pillar, the AI demos, the practice areas — is built around that single transaction.

Takeaway: the unit being sold is a regulated-human-hour denominated in USD and costed in INR; that single arbitrage is the entire return engine.

The economics get interesting on a ROE decomposition. Reported ROE is ~11% — pedestrian for a 24% EBITDA-margin business. The cause is a balance sheet stuffed with intangibles from EQT's 2021 carve-out of HGS healthcare: ~$997M of fixed assets (mostly goodwill and customer intangibles) sit on ~$1,030M of equity. Stripping carve-out goodwill, management-reported ROCE against working capital plus delivery infrastructure runs above 50%. That is what the business earns on incremental capital; headline ROE is the LBO legacy.

The Playing Field

Sagility is small in absolute revenue but the highest-margin focused player in the listed BPM peer set, and is now the only listed pure-play in U.S. healthcare BPM after Capgemini took WNS private in October 2025. The relevant peer cohort is not "Indian IT services"; it is U.S./global BPM vendors with named healthcare segments, plus Firstsource as the local valuation anchor.

Takeaway: Sagility is the only dot in the top-right — 100% healthcare focus and the highest operating margin in the cohort. Concentrix and Conduent prove that scale without the right mix destroys returns.

Three things this peer set proves:

First, scale without mix doesn't earn. Concentrix has 13x Sagility's revenue and runs at negative operating margins because the mix is voice-CX. Conduent's healthcare-payer segment is 35% of revenue and the consolidated business still loses money. The moat is not delivery-footprint size — it is the share of revenue tied to regulated clinical work.

Second, the right comp is WNS pre-acquisition, not "Indian IT". WNS had a similar fiscal year (March), a similar delivery model, and a 23% healthcare mix; Capgemini paid ~$3.3bn at ~2.5x revenue. That sets a transaction floor for a pure-play. Sagility trades around 2.6x EV/revenue — already at parity — so the WNS comp is no longer cheap insurance.

Third, Firstsource is the local valuation anchor. The only India-listed BPM peer, trading on ~25x P/E with 16% margins, 9% USD revenue growth, and 17.7% ROE. Sagility's ~20x P/E with higher margin and faster growth but lower headline ROE is the trade the market is making — a focus-and-margin premium offset by an LBO-bloated balance sheet.

Is This Business Cyclical?

No: this is one of the least cyclical businesses in BPM, because the work being outsourced — claims adjudication, member enrolment, provider credentialing, payment integrity, utilisation management — is non-discretionary for U.S. payers. Revenue tracks membership (stable to growing) and claims volume (anchored to enrollment), not GDP. The FY22–FY26 record confirms it: revenue compounded through post-COVID wage inflation, the FY23 Indian BPM wage cycle, the FY24 Medicaid redetermination spike, and the FY25 hospital-cost crunch.

What is cyclical is seasonality, not the cycle. Q3 and Q4 (October–March) capture the U.S. Annual Election Period and Open Enrolment surge — management disclosed seasonal revenues at 6% of FY26 vs 3% of FY25, a known and bookable working-capital swing. The medium-term cyclicality that matters is on the cost side: INR/USD FX (rupee weakness = direct margin tailwind), Indian wage inflation, SEZ tax-holiday rolloff. None are demand cycles; all are input-cost cycles.

The Metrics That Actually Matter

Generic ratios (P/E, ROE) miss the picture because of the goodwill-loaded balance sheet and FX-and-wage cost-stack mechanics. Five metrics actually move the stock.

Takeaway: the concentration concern is real but trending in the right direction — top-3 share has fallen ~600 bps over 3 years while the count of $20mn-plus clients has more than doubled.

Two often-quoted metrics to discount: headline ROE (distorted by carve-out goodwill) and INR-reported revenue growth (flattered by FX; the honest read is constant-currency).

What Is This Business Worth?

Value this as one economic engine, not sum-of-the-parts. Provider RCM is only ~10% of revenue and shares the same delivery infrastructure, cost base, and contract structure as the payer book — separating it would be precision theatre. No listed subsidiaries, no holdco structure, no orphan assets. The right lens is normalized FCF and reinvestment runway, sense-checked against the closest transaction comp (Capgemini-WNS, ~13x EV/EBITDA at deal close).

Underwriting frame: at ~20x trailing P/E and 8.4x EV/EBITDA, the stock is paying for ~13% earnings growth and 24–25% steady-state margin. The Capgemini-WNS deal anchors the strategic ceiling at ~2.5x EV/revenue (Sagility already there); the Firstsource listed-peer P/E sits in a similar zone. The asymmetry: the WNS deal removed the closest comp — anyone wanting clean exposure to listed U.S. healthcare BPM has exactly one ticker. That scarcity is not in the public-comp multiple yet.

What I'd Tell a Young Analyst

Five rules for following this stock without losing your way.

Watch concentration, not the AI debate. The dominant tail risk is a top-three payer renegotiating or insourcing, not GenAI deflating the book. Concentration is below 60% and trending down — every quarterly deck shows the direction. The AI debate is real but second-order; the regulated clinical mix gives two years of cover.

Track constant-currency organic growth, not the INR-reported line. Reported growth was 29.1% in FY26; organic CC was 15%. The gap is FX + BroadPath, not operating performance. Anyone selling the stock on "30% growth" is mismeasuring.

The right peer is not Indian IT. Anchor on Capgemini-WNS at ~13x EV/EBITDA and EXL Health at the segment level — not TCS or Infosys multiples.

Don't be impressed by reported ROE. Eleven percent ROE is the LBO-goodwill artifact; underlying operating returns are multiples of that. Management's "adjusted ROCE" (above 50%) is what the business compounds incremental capital at, and it explains why a 24% EBITDA business gets a 20x multiple rather than a 10x one.

The thesis breaks if margins slip below 22% AND organic CC growth slips below 10%. Each alone is recoverable. Both at once means the FX-tailwind story has unwound while client-side pricing has softened — the only configuration justifying a serious de-rating. Until that combination appears, the default underwrite is through-cycle FCF compounding around the WNS transaction multiple, with option value on EQT eventually selling the residual ~51% stake.