PTSB is deal #15 of a 14-deal flywheel — and the first one structurally different from the prior 14.

• Track record. 14 self-funded bolt-ons since 2015, each acquired below book and integrated in roughly 12 months. The September 2021 four-year guide was beaten by +54% on PBT, +50% on EPS, +56% on DPS.
• What's different about #15. $1.9B all-cash bid for Permanent TSB expands assets ~40% in one stroke (prior deals were 5-10%); 26% premium versus the historical 0.6-0.8× book purchase discipline; the scheme commits BAWAG to preserve PTSB's 98 branches — inverting the cost-arbitrage that built the platform.
• Where the spread shows up. Irish mortgage NIMs already compressed 15-20bps in FY25 inside a 3-bank oligopoly that will defend share against a foreign acquirer. 20bps of further compression erases roughly 30% of the deal synergy.

The operating gap predates the rate cycle and held across two prior integrations.

Roughly 90% of retail and SME originations move through digital, broker, and partnership channels — the universals carry branches and capital-markets divisions the cost ratio cannot escape. ING, ABN, Commerzbank, and Erste cluster tightly at 49-58% CIR and 8-19% RoTCE; BAWAG occupies the empty quadrant. The 3× tangible book is the only multiple that reconciles with the gap — and it leaves no margin if FY26 RoTCE drifts toward 22%.

FY24-25 carried roughly $275M of non-recurring income — and the buffer that produced it is now empty.

• The overlay release. FY24 risk costs printed 16bps after a full ECL management-overlay release; FY25 stepped to 41bps; Q1 2026 ran at 46bps. Forensic note: model FY24 reported earnings 10-15% below face value at a normalized provisioning rate.
• The Day-1 gains. Knab booked $77M of bargain-purchase income through other operating income in FY24; Barclays Consumer added $22M in FY25; investment-property revaluation added $56M in FY25. Cumulatively ~9% of two-year net profit, low quality.
• Why it matters. Strip both and trailing P/E moves from 12.8× to ~15× — a market multiple, not a discount. The bull math (3.4× TBV on a rolled-forward book) only holds if 27% RoTCE is the run-rate, not the FY24 peak.

The 21 July H1 print answers two of three load-bearing questions; PTSB closing answers the third.

• 21 July 2026 — H1 risk costs and CIR. Underlying risk costs below 40bps and standalone CIR converging toward 33% on the ex-acquisition book invalidates the bear's earnings-flatter case. Above 45bps with CIR stuck above 35% confirms it.
• End-Q2 — the 250bp CET1 build. 200bp from an H1-only dividend skip plus 50bp from RWA actions and SRTs has to land as guided. Any 'alternative actions' language at H1 reopens the equity-issuance tail the multiple does not price.
• Q4 2026 / Q1 2027 — PTSB closing. Clean closing with a badwill comparable to Knab's $77M and a calendared High Court hearing closes the deal-break tail. Slippage past Q3 plus a third Day-1 gain in three deals converts a clean track record into a recurring earnings-quality flag.