Variant Perception

Where We Disagree With the Market

The market is buying ConvaTec at 205p as a chronic-care compounder where the only debate is whether the consensus 302p target is "right now" or "12 months out" — and the report's evidence says that frame is wrong on four specific points. Sell-side ratings are clustered Buy (Barclays, Berenberg, Jefferies, RBC, JP Morgan) and the loudest narrative — "FCF-to-equity converted at 101%, mid-20s margin by 2027 was upgraded to 24-26%, InnovaMatrix is the bottom on CMS" — is the precise narrative the working files contradict. Our disagreement is not directional (we are not bear, and we are not bull beyond consensus); it is about what the August 2026 H1 print will reveal about the cash machine, the margin floor, the regulatory cliff, and the FDA cost line. If we are right, the gap to the 302p consensus target compresses through earnings cuts rather than through a re-rate. The cleanest single resolving signal is the H1 2026 working-capital reversal — it lands inside 90 days and decides three of the four disagreements simultaneously.

Variant Perception Scorecard

Consensus Map

The consensus is not muddled. It is well-articulated, recently reaffirmed at 320-330p targets, and built on four assumptions that the report's evidence stress-tests directly. We agree with the market on cyclicality (chronic-care demand is non-cyclical) and on Insulet (a 5-7 year fade, not a 2026 cliff). We disagree on cash, margin floor, biologic-AWC perimeter, and the right way to read the FDA letter.

The Disagreement Ledger

Disagreement #1 — The cash machine is half its headline. Consensus would say: the FY25 cash conversion ratio is the strongest in the company's listed history (101% on FCF-to-equity), the buyback was funded by FCF, and the comparative was restated for transparency. Our evidence disagrees: under the prior definition the conversion was 61%, the CFO confirmed this on the call, $111M of payables release (DPO at an eight-year high) and a 25.1% receivables build mechanically unwind in the next two halves, and the metric that drives 15% of variable pay was redefined in the new CFO's first reporting cycle. The market would have to concede that FCF yield is ~3.6% rather than ~6.2% — and that the $326M buyback was a leveraged buyback, not an FCF buyback, with cash drained to $68M. The cleanest disconfirming signal is H1 2026 working capital: if DSO normalises below 56 days and DPO holds 150-165d without a payables reversal, the cash story re-validates and we move to consensus. If DSO holds above 60 with AP unwinding, our variant is right and consensus has to cut.

Disagreement #2 — The margin glidepath was walked back, dressed as an upgrade. Consensus would say: Mason raised the medium-term floor by setting an explicit 24-26% range; the upgrade was specifically rewarded by Berenberg, RBC, and Jefferies with reaffirmed buys at 320-330p targets. Our evidence disagrees: the original "mid-20s by 2026 or 2027" was set at the FY22 CMD and repeated through FY24 — implying a roughly 25% mid-point — while Mason's "closer to 24% by 2027" floor is meaningfully below that mid-point. H2-25 IFRS op margin compressed to 10.9% from H1's 15.2% on tariffs and Unomedical spend, and the company's own arithmetic is $24M per 100bps on $2.44B revenue: a 200bps lower floor is ~$48M of FY28 adj op profit, or roughly 5% of forward earnings power. If we are right, the market has to accept that the multiple gap to Coloplast (16.6x EV/EBITDA vs CTEC 14.5x) deserves to be wider, not narrower — anchoring the bear's 165p versus the bull's 310p. The disconfirming signal: H1 2026 prints ≥23.5% with the full tariff load AND management reiterates the 24-26% range without softening to "by 2028."

Disagreement #3 — Biologic AWC is regulated-margin, not moated, and the next reset is in 2027. Consensus would say: InnovaMatrix has been impaired and sized to ~$20M of FY26 revenue, the LCDs were withdrawn in December 2025, and Coloplast Kerecis remains a credible growth driver. Our evidence disagrees: the $127.28/cm² Medicare rate was not withdrawn even though the LCDs were; the AATB publicly calls the rate "unsustainable"; and the moat tab describes biologic AWC as a regulated-margin business across the peer set, not a CTEC-specific cliff. If a second CMS revision lands in CY2027, Coloplast Kerecis, SNN GRAFIX, and Integra Skin face the same drop CTEC has already absorbed. The market would have to concede that the InnovaMatrix story is the prelude to a peer-set repricing, not the bottom — which would compress relative-value targets that anchor CTEC's bull-case multiple. The disconfirming signal: the CY2027 proposed rule (typically July 2026) keeps the rate or removes the LCD risk for 2027, in which case our variant fades and InnovaMatrix becomes the contained event consensus describes.

Disagreement #4 — The Unomedical risk channel is partner audits, not enforcement escalation. Consensus would say: CTEC's letter does not restrict production, marketing, or distribution; the high-impact tail-risk is a consent decree or import alert; absent that, Infusion Care guide stands. Our evidence disagrees: a partner-driven margin drag does not require any enforcement escalation. Beta Bionics' 28 January 2026 follow-on letter cited Convatec sets in 33% of its complaint database, with 99.7% of failed sets not returned for analysis. The next leg of cost shows up as remediation opex/capex required by Tandem, Medtronic, and Beta Bionics' own audit programmes — exactly the kind of slow-leak margin item consensus models do not pick up before the print. If Infusion Care is the highest-incremental-margin franchise and the cleanest moat, even 200bps of remediation drag on it is ~$9M of group adjusted operating profit per year, plus capacity timing risk. The disconfirming signal: the H1 print contains a remediation-cost line, partner 10-Qs disclose audit findings or sourcing changes, or — supportively — the H1 print closes the question with a clean Infusion margin and a CAPA progress note.

Evidence That Changes the Odds

The evidence ledger is deliberately not stacked. Items 1-5 push the variant view; items 6-7 fight it. Anyone reading this should treat the BMS amortisation roll-off and the protected 40% of revenue as the strongest counter-arguments to our case — they say the IFRS-vs-adjusted gap closes whether or not management lifts a finger, and that the moat lives in two franchises that are not directly exposed to our four disagreements. The variant survives because three of the four disagreements concentrate in the headline numbers (cash conversion, margin floor, biologic AWC perimeter) that drive the sell-side multiple — and only one (Unomedical partner channel) hits the protected 40%.

How This Gets Resolved

Three of these signals (1, 2, 4) land in the same 4 August 2026 H1 release. That release is the variant referendum — by design, not by accident. Signals 3 and 5 are slower-burning (CY2027 CMS cycle; partner 10-Qs through 4Q26) and signal 6 is rolling. If the H1 prints clean on (1) and (2), the variant collapses to two slow-burn debates the market will re-price gradually rather than violently. If the H1 prints dirty on (1) or (2), all four disagreements compound and the 165p bear path comes into play before consensus can revise.

What Would Make Us Wrong

The strongest fight against our variant is the BMS-era amortisation roll-off and the two protected franchises, and any honest reader has to take both seriously. The roll-off is mechanical: $95M of $134M FY25 acquired-intangible amortisation runs off by mid-2026, narrowing the IFRS-vs-adjusted wedge whether or not management lifts a finger. That fight tells us our cash-quality disagreement (#1) does not survive a long enough horizon — the headline EPS gap closes by FY27 even if our redefinition view is right. The two protected franchises (180 Medical at 59% own-brand, Infusion Care with binding OEM contracts) tell us that 40% of revenue earns moat-grade economics that sit above any of our four disagreements; even in a worst-case H1 print, those franchises do not depend on the cash redefinition or the biologic AWC LCD perimeter.

A second honest fight is that consensus might have already adjusted without re-rating the price target. Yahoo's average target of 302p is anchored on calls written before the CMD; the post-CMD action shows reaffirmed buys at 320-330p but several brokers have not refreshed. If a wave of preview cuts already lands ahead of August — Berenberg or Jefferies trims to 260-280p in late June — the variant has already been monetised in the tape and the stock will not move much further on the H1 print. The technical position (52-week low, broken death-cross regime) is consistent with a market that has partially internalised what we are flagging.

A third honest fight is that the H1 print could just be a clean print. Tariffs phase in H1 and out in H2, the Unomedical CAPA is ConvaTec-paced not FDA-paced, and the working-capital build was Q4-loaded enough to mechanically reverse by 1H 2026. If management lands ≥23% with DSO normalising and the FCF-to-equity bridge tracking, the variant is wrong on the order of operations — we caught the noise, not the trend — and the consensus 302p is the conservative target.

The fourth honest fight is that the margin walk-back is reading too much into a CEO speaking conservatively in his first reporting cycle. Mason was the architect of FISBE alongside Bitar, the playbook he is anchoring is the one he co-authored, and the lower-bound choice (24% vs 25%) may simply reflect tariff and FDA cost realism that the prior guide did not capture. If H1 lands ≥23.5%, our variant on disagreement #2 collapses; the floor was tactical, not structural.

The first thing to watch is the H1 2026 working-capital reversal — DSO normalising below 56 days without any factoring disclosure resolves disagreement #1, validates the cash-compounder narrative, and removes the variant strength from three of our four claims in a single print.