Variant Perception

Where We Disagree With the Market

The market is pricing Supreme as a vape distributor with a regulatory ceiling. Our evidence says it is a 30%-plus ROIC FMCG distribution platform that has bought four UK consumer brands at sub-3x post-synergy EBITDA in fourteen months — but the same evidence forces a second, less comfortable disagreement: consensus is also extrapolating FY25 cash quality, and FY25 acquisition-adjusted free cash flow was actually negative £3.65m. Both views can be right: the platform is misclassified, and the cash conversion is being financed by working-capital draws and debt headroom rather than recurring cash. The single resolving signal is the FY26 print in late June 2026 — non-vape gross profit at or above £40m and a return to positive acquisition-adjusted FCF would settle the rerating debate; either failing settles it the other way. If we are right on platform classification but wrong on cash quality, the stock still rerates; if we are wrong on platform but right on cash quality, the bear case prints. The interesting trade is where both legs resolve in the same direction.

Variant Perception Scorecard

The 68 variant strength reflects two materially divergent disagreements that move valuation by more than 30% in either direction, both anchored to dated resolving events inside six months. Consensus clarity is high because the gap between the 7.79x P/E print and the 12.18x five-year mean, the 229.5p Fintel consensus target versus the 160p tape, and the FY25 EPS-versus-consensus history are all observable. Evidence strength is the weakest leg — the cleanest contradicting evidence (acquisition-adjusted FCF, segment gross-profit run-rates, acquisition multiples) is well-documented in the upstream tabs but partly model-implied rather than directly disclosed. The six-month resolution window is the FY26 full-year print plus the autumn 2026 vape excise levy.

Consensus Map

The consensus map shows where the market signal is strongest (vape overhang, founder alignment) versus where it is genuinely mixed (cash conversion, Drinks & Wellness optionality). High-clarity disagreements are higher-leverage; low-clarity ones are sizing inputs.

The Disagreement Ledger

Disagreement 1 — Mis-categorised platform. Consensus says Supreme is a UK vape distributor whose terminal value is capped by HMRC and whose earnings deserve a regulated-distributor multiple in the low single digits of EBITDA — that is the only framing that justifies trading at half the EV/EBITDA of AG Barr and Nichols on margins that are now in line. Our evidence says the engine of value creation is the platform itself: ROCE held at 39.7% in FY25, the gross-profit-to-overhead ratio has not deteriorated as revenue tripled, and four bolt-ons have closed at 2-4x post-synergy EBITDA in fourteen months — that is a Constellation-style compounding pattern, not a single-category distributor. If we are right, the market has to concede that 50%-plus of forward revenue is in branded FMCG categories with margin profiles that read across to Vimto/Irn-Bru, not to JT International, and the multiple gap to those peers (currently approximately 50%) closes inside 12-18 months. The cleanest disconfirming signal is the FY26 print: if non-vape gross profit comes in below £30m, the Drinks & Wellness platform thesis is stalling and the vape-pure-play framing was right.

Disagreement 2 — Cash quality is worse than headlines. Consensus reads the 12.3% FCF yield, the 1.04x FCF/Net Income, and the £40m HSBC facility headroom as evidence of a self-funding compounder; the 229.5p Fintel target only works if the dividend is paid out of recurring cash and M&A is not being financed by working-capital extraction. Our evidence says FY25 CFO £25.1m and acquisition spend £25.6m mean acquisition-adjusted FCF was negative £3.65m, year-end cash fell from £11.6m to £3.2m, and FY26 added two more deals (SlimFast £20.1m, 1001) requiring the first HSBC ABL drawdown the company has ever needed. If we are right, the market has to concede that the flywheel is debt-financed — which does not break the thesis, but does compress the multiple a serial acquirer with negative organic FCF can support. The cleanest disconfirming signal is FY26 acquisition-adjusted FCF: a return to at least +£10m says the FY25 print was the M&A-bunching artifact management claims; a second consecutive negative year says it is structural.

Disagreement 3 — Governance discount is real. Consensus treats the 56% founder stake as a 7.5/10 alignment positive — the same number people-claude assigns. Our evidence says the same tab flags four offsetting items the consensus discount does not weight: the 11-year BDO relationship is overdue for tender (FRC ethical standard for FTSE 350; AIM is exempt but the spirit applies), the 3-year LTIP was replaced with a 1-year SIP at 93.33% FY25 payout, there is no internal audit function, and Sandy Chadha placed shares twice in fourteen months at flat-to-discount prices. If we are right, AIM/QCA peers with similar profiles trade at a 1-2x EV/EBITDA governance discount — roughly 15-25% of share price — which narrows only when BDO is tendered and a multi-year LTIP is reinstated. The cleanest disconfirming signal is the AGM 2026 vote: free-float dissent above 25% on either Resolution 2 or Resolution 8 forces a board commitment; below 10% confirms minorities accept the package and the discount is unlikely to close on governance alone.

Disagreement 4 — Pod margin step-down erodes earnings, not just optics. Consensus reads H1 FY26 vape +13% as proof the substitution worked. Our evidence says the same release flagged a 200bps gross-margin step-down (33% to 31%) on a segment that is 56% of group revenue — roughly £2.6m of segment gross profit at FY25 scale, plus the FY26 EBITDA margin guide is 15% versus the FY25 17.5% print. If we are right, headline revenue durability is masking gross-profit erosion that the levy in October 2026 may then compound. The cleanest disconfirming signal is the FY26 segment gross margin disclosure: above 30% confirms the substitution thesis at the gross-profit line; below 28% says the disposables-margin era is gone in pounds, not just in mix.

Evidence That Changes the Odds

The strongest single piece of evidence is the acquisition-adjusted FCF print — it reframes the cash conversion narrative without rewriting any reported number. The weakest is the governance discount stack, where the AGM has never printed material dissent at SUP and minorities may simply tolerate the structure. The mis-categorisation argument depends on the FY26 non-vape gross profit print clearing £40m; below £30m, the platform thesis is intellectually correct but economically inert.

How This Gets Resolved

Five of the eight signals resolve at the same FY26 print in late June 2026 — the catalyst calendar concentrates the rerating debate into a single information event. The autumn 2026 levy is the higher-magnitude tail; the AGM and the placement risk are credibility tests rather than thesis-resolvers.

What Would Make Us Wrong

The mis-categorisation thesis would be wrong if the FY26 print shows non-vape gross profit stuck near £25-30m — that would mean Drinks & Wellness has not scaled despite the £25.6m of FY25 acquisition spend, despite Clearly Drinks consolidating for a full year, and despite the SlimFast/1001 contributions starting to land. In that case, the market is correctly anchoring on vape because vape is still mathematically the dominant gross-profit pool, and the platform-leverage argument is intellectually correct but economically inert. The right multiple under that scenario is closer to 5-6x EV/EBITDA — about where SUP trades today — and the disagreement collapses.

The cash quality thesis would be wrong if FY26 acquisition-adjusted FCF turns positive at +£10m or above, DSO normalises below 55 days, and the SlimFast purchase price is recovered through inventory and receivables release rather than ABL drawdown. The pattern would then look like: a single-year M&A-bunching event in FY25 (Clearly Drinks plus Typhoo plus working-capital build of £18.5m all hitting one fiscal year), followed by a return to recurring FCF that funds the dividend and bolt-on M&A. If that prints, our negative-FCF concern was real but transitional, and the bull case earns its 12x mean-reversion multiple.

The governance discount thesis would be wrong if the AGM 2026 produces no dissent on either BDO or the SIP, and the next bolt-on closes at sub-4x post-synergy EBITDA with cash funding (no further ABL drawdown). The track record then says: founder concentration is genuinely an alignment positive at SUP because the operator has demonstrated capital-allocation discipline through three transformations and two regulatory cycles. We would still flag the structural items in people-claude, but the discount we are arguing for would be intellectually defensible without being economically realised — which is the textbook case where governance theory loses to operating evidence.

The pod-margin thesis would be wrong if the FY26 vape segment gross margin recovers above 32% as the ElfBar Pod ramps and the disposable inventory headwind washes through. The H1 FY26 step-down would then look like a transition cost, not a structural floor; the FY26 EBITDA margin would beat the 15% guide rather than print to it; and the substitution narrative would be vindicated at both the revenue and the gross-profit line. That is the regime that takes the stock to the bull-case £3.20.

The first thing to watch is the FY26 non-vape gross-profit segment disclosure in late June 2026 — that single number, more than vape revenue, more than DSO, more than the levy, decides whether Supreme is a misclassified platform or a vape stock that bought some brands.