Variant Perception

Where We Disagree With the Market

The market is paying 3.0× tangible book for a 27% RoTCE that the bank's own through-cycle floor — repeated in every CEO letter since 2022 — is "above 20%". That 700 bp gap is the entire variant view: it implies that an unknown share of FY25 profitability is cyclical replicating-portfolio yield and reserve-overlay timing rather than structural earnings power, and that the PTSB accretion case bundles a one-off €370–400m day-one badwill into the same number that consensus is also paying a synergy multiple on. Sell-side is unusually crowded — 14 of 16 brokers at Buy, average target €175 versus a €151.50 spot price, with Erste's €145 Hold the only below-market call — so the dispersion that would normally cushion a thesis crack is absent. We do not disagree with the operating story; we disagree with the market's confidence that the FY25 print is the through-cycle baseline rather than the cyclical peak. The three observable signals that resolve it — Q2 2026 RoTCE on a clean post-acquisition base (22 Jul 2026), the FY26 risk-cost ratio versus the disclosed 45 bp guide, and the PTSB Day-1 pro-forma CIR at close — all land inside the next nine months.

Variant Perception Scorecard

The consensus is clear (target dispersion €145–€202, with one Hold) but the disagreement is narrow rather than broad: the bank's operating engine is not in dispute. What is in dispute is whether 27% is the right RoTCE anchor or whether 22% is, and whether the PTSB accretion the multiple already discounts is operating synergy or balance-sheet badwill. Evidence is graded medium-to-high because each variant view rests on disclosed numbers — overlay-release language in the FY25 MD&A, day-one badwill quantified by Sretaw and the Irish Times, AIB's 30% Irish mortgage share — rather than on speculation. The 6-month resolution window is anchored to the H1 2026 print (22 Jul 2026), with the full window stretching to the FY26 mid-term reset around 10 Feb 2027.

Consensus Map

What the market appears to believe, and the observable evidence that this is the consensus view.

The Vienna-listing point is the cleanest piece of consensus. The market has already done the work of disentangling BAWAG from Erste/RBI; that part of the variant trade no longer pays. Where consensus is still vulnerable is in the four issues above it — and most of all in conflating cyclical FY25 outperformance with structural through-cycle earnings power.

The Disagreement Ledger

Four ranked disagreements where report evidence pushes against the consensus reading. Lead disagreement is the through-cycle RoTCE anchor, because it determines the multiple. The PTSB-badwill point is second because it is contestable on hard numbers. Risk-cost overlay timing is third because it ties to forward earnings quality. The "PTSB integration is structurally different" point is fourth because it pushes against the most-extrapolated piece of the bull case.

Disagreement #1 — Through-cycle RoTCE anchor. The consensus analyst would say "BAWAG has printed RoTCE above 25% for three consecutive years through three different rate regimes — the spread to the EU sector is structural." The evidence that complicates this is internal to BAWAG's own disclosure: the 2021 Investor Day target was RoTCE >17%, raised to >20% in 2022, and the >20% floor still appears in the FY25 letter and the Q1 2026 deck. Management is signaling 22% as the underwriting anchor; the market is paying 3.0× P/TB on 27%. If we are right, the multiple has to compress toward AIB's 1.5× as soon as a clean quarter prints below 25%. The cleanest disconfirming signal is the Q2 2026 RoTCE on the post-Knab/Barclays clean base (22 July 2026): if it lands at 25%+ with NIM holding above 3.20%, the structural read defends.

Disagreement #2 — PTSB badwill versus operating synergy. The consensus view treats the price contestation from Goodbody and Sretaw as noise — a defended seller wanting more money. The evidence cuts the other way: if Sretaw's quantification of ~€370m badwill and ~€410m PTSB surplus capital is roughly right, the real net outlay is closer to €800m than the €1.62bn headline, and the >€250m PBT-by-2028 path is being lifted disproportionately by the discount, not by operating synergy. CFO Sirucic has already hinted at routing the day-one gain into tech investment — i.e., not into reported EPS — which would be the cleanest tell that management itself does not fully believe the synergy line. The market would have to concede that some of the accretion is non-recurring rather than compounding. Disconfirming signal: PTSB Day-1 PPA disclosure plus pro-forma CIR within 12 months of close — below 42% defends, above 45% breaks the case.

Disagreement #3 — Reserve-overlay smoothing. Consensus would say "BAWAG provisions conservatively; FY24 release and FY25 rebuild are textbook through-cycle behaviour." The Forensics tab agrees that the disclosure is clean — but it also shows the optical pattern: 56 bp (COVID build) → 23 bp (release) → 86 bp (Linz) → 22 bp (maintained) → 19 bp (release; flatters FY24) → 41 bp (rebuilt) → 45 bp guided FY26. The own-disclosed conservative reconciliation is €776m FY25 versus €860m headline. The market would have to concede that consensus FY26 NP estimates of >€960m carry an embedded overlay assumption rather than a flat IFRS 9 baseline. Cleanest disconfirming signal: a quarter where overlay is released without NPL improvement — the smoothing question would then re-open.

Disagreement #4 — Irish mortgage incumbency. The consensus view, supported by Fitch and the Tánaiste statement, is that PTSB integrates like Knab and Barclays — disciplined cost integration, accretive within three years. The evidence that complicates this is the structural difference: Knab was an orphan Dutch direct bank that nobody was defending; Barclays Consumer Bank Europe was an orphan continental cards book, similarly unopposed. PTSB enters as the #3 Irish mortgage lender against an AIB/BoI duopoly that controls ~50% of new mortgage lending and has the entire balance sheet to defend the franchise. AIB has already started: its FY25 CIR rose 4 pp on defensive investment. The market would have to concede that BAWAG's M&A track record is conditional on competitive structure rather than universal. Disconfirming signal: PTSB-MoCo combined mortgage flow share through the Irish Central Bank BPFI series in 2027.

Evidence That Changes the Odds

The eight pieces of evidence that move the probability of the variant view, ranked by their power to shift a PM's underwrite. Each is sourced from upstream tabs or external research with named provenance. Each carries a fragility column — the way the evidence could be misleading.

The single highest-value evidence item is the first one: management's own through-cycle floor language. It is the only piece of evidence that comes from BAWAG itself, has been repeated for four years, and directly contradicts the assumption embedded in the multiple. Items two and five are the cleanest peer/structural reads. Item three is the most market-checked piece — Sretaw, Goodbody and Irish Times independently arrive at the same badwill range. Item four is the bank's own one-sentence admission of overlay timing. The remainder are corroborating signals.

How This Gets Resolved

Six observable signals over the next twelve months, each with a verifiable threshold. None of these are vibe checks; each has a date and a public document.

What Would Make Us Wrong

Several things would force this variant view to fold, and the discipline is to name them before the market does.

The first is the simplest: BAWAG keeps printing 25%+ RoTCE on clean post-acquisition bases and management raises the through-cycle floor at the FY26 mid-term reset. If the floor moves from "above 20%" to "above 22%" — or, more aggressively, "above 25%" — the multiple is no longer paying for cyclical over-earning; it is paying for what management itself underwrites. The deposit franchise has already passed three rate-regime tests (negative rates, ECB hike, ECB cut) and the spread to the EU sector has widened, not narrowed, through each. If H2 2026 deposit beta drifts down rather than up while NIM holds above 3.20%, the replicating-portfolio rollover concern is empirically wrong. The FY25 RoTCE spread to the EU sector ROE is roughly 1,620 bp — if it stays above 1,500 bp through end-2026, the operational moat is stronger than the variant view treats it.

The second is the PTSB integration math. If the Day-1 PPA confirms ≥€350m of badwill and the pro-forma CIR lands below 42% within twelve months, the synergy-vs-badwill objection collapses — the market gets both, not one or the other. AIB's FY25 CIR rising 4 pp on defensive investment is a leading indicator, but if AIB's mortgage-share defense is competitively impotent (as it was during the 2022-23 Ulster Bank exit when PTSB itself absorbed €6.8bn of loans without a price war), the "duopoly defends" thesis evaporates. BAWAG's prior 14 deals all cleared the >20% RoTCE hurdle; the base rate is high, and we are betting against the base rate on the structural difference. If we are wrong, PTSB compounds the moat rather than testing it.

The third is the overlay-timing read. If FY26 risk costs land at the guided 45 bp with NPL stable below 1% and Stage 3 coverage holding, the conservative read becomes the right read — overlay rebuild was prudent, not optical. The forensic grade of Watch (24/100) sits exactly there: yellow flags, not red. The variant view treats the conservative reconciliation (€776m vs €860m headline) as a haircut to consensus; if FY26 prints prove that the rebuild is the real cost of the consumer-cards mix shift, then consensus is anchored to the right number and the haircut is not warranted.

The fourth is the people read. The CEO bought €1.19m at €147.50 a week before Q1 26. Five of six MB members have run the bank since the 2017 IPO. There has been no insider selling since the IPO. Glassdoor is 15 reviews with a real selection bias. ISS Audit pillar at 10/10 is a flag, but is also unlikely to manifest as accounting risk on a Deloitte-audited bank under direct ECB supervision. If the variant view is right that "operational moats are people-led", the people read pushes against us — this team has eight years of repetitions and is buying at the highs.

The first thing to watch is the Q2 2026 RoTCE print on the post-Knab/Barclays clean base, releasing 22 July 2026 — at or above 25% with NIM above 3.20% defends the multiple; below 24% on a clean base is the bear's primary trigger.