0. CORRECTION to prior note — supersedes the Aug-chatter line in 2026-09-08-eml-all-ords-removal-forced-flow.md

My earlier forced-flow note carried "Oct-2027 surcharging-regulation drag per Aug-2026 results chatter". That was wrong on both date and mechanism. Verified from the RBA Conclusions Paper coverage and multiple acquirer implementation notices:

• 1 Oct 2026 (three weeks away): surcharging banned on all designated debit, prepaid and credit cards across eftpos, Visa, Mastercard. Domestic interchange caps fall same day (consumer credit 0.80% → 0.30%; debit/prepaid → 0.16% + 8c; commercial cards 0.80% unchanged).
• 1 Apr 2027: foreign-issued card cap (1.0%) + payment-cost transparency rules.
• Amex/Diners/PayPal can still be surcharged at cost; BNPL sits under separate rules.
• Scale: ban removes ~A$1.6bn/yr of consumer surcharges; interchange cuts ~A$910m/yr of wholesale merchant cost; RBA prices the consumer-savings combo at ~A$2.5bn/yr. RBA concedes a ~0.1pp CPI uplift risk if merchants absorb rather than reprice.

Why it matters to the trade: for EFTPOS acquirers (TYR, ZLG, Square AU, bank acquirers) this is a repricing event, not a growth event. Merchants that surcharge today (≈12% of SMEs) stop recovering 1–1.5% of card revenue at checkout and must reprice into shelf prices; interchange cuts compress the wholesale pool acquirers price off; the Apr-2027 transparency rules are explicitly designed to make provider-shopping easier — a churn/price-war accelerant for a premium-priced incumbent. The near-term EBITDA drag is plausibly bigger than the SME-health tailwind. This is the live TYR regulatory driver, and it is weeks out, not a year out.

1. Tyro Payments (ASX: TYR)

Tape (verified, Yahoo Finance pull 8 Sep ~22:35 AEST)

• Close A$0.74 (7 Sep 0.735; 2 Sep 0.76). 52w range 0.665–1.28 — sitting ~11% off the low, ~42% off the high. Results-day −5.1% (25 Aug) per Market Index week-37 wrap; no recovery since.
• ⚠️ Market cap: at ~A$0.88 my earlier ~A$438M note implies ~498M shares → at A$0.74 that's ~A$368M. Share count not independently re-verified this pass. ⚠️ Unverified: Tyro cash/deposits balance, short-interest %.

Earnings (verified against company release + Annual Report)

FY26 (to 30 Jun 26, released 25 Aug): gross profit $231.8M +5.3% (in line), EBITDA $66.9M +8.6%, margin 28.9% — 3% BELOW consensus (Market Index/UBS framing), normalised PBT $24.7M +40%, FCF $29.4M +49.5%. TTV $44.3Bn +2.9% — the split is the story: Tyro payment volumes +4.4% vs Bendigo Alliance volumes −10.5%. Health $7.9Bn +6.4% only, capped by the bulk-billing policy headwind (GP visits shifting to full bulk-bill kills per-transaction claiming economics); Allied Health +25.9%, Dental +18.7%. Banking: accounts +35%, deposits +27%, loan origination +19%. Thriday (AI bookkeeping) acquisition consolidated as growth optionality. FY27 guide: GP $240–255M (+3.6–10%), EBITDA margin 28.5–30.5% (~$73M midpoint, broadly in line).

Read

The stock's problem is deceleration into a regulatory repricing event. 5.3% GP growth guiding to 6.8% midpoint, EBITDA a slight miss, and the Bendigo channel shrinking 10.5% — while the 1 Oct surcharge ban and Apr-2027 transparency rules attack exactly the pricing layer where Tyro's ~1.4% blended rate earns its premium. Management's counter (and the bull case) is mix: multi-product attach (payments+banking+accounting), 580+ POS/PMS integrations as switching-cost moat, Health Pro claiming, AI resolving 55% of support queries (up from 23%) as the opex lever. Valuation optics: ~5.5x EV/EBITDA pre-net-debt-adjustment (⚠️ cash not verified) — a value trap or a reset depending on whether Oct's regulatory drag lands inside the 28.5–30.5% margin band. Checkpoints: 1 Oct 2026 (ban effective — watch merchant repricing chatter, Tyro terminal/checkout comms), H1 FY27 trading update (Dec), FY27 interim (Feb 2027).

Regulatory posture specific to TYR

First tech company with a full ADI banking licence (Australia) — the safeguarding/licensing tightening that squeezes e-money issuers is a moat for Tyro, not a cost. Exempt from the EML-style float problem: Tyro's deposits are a balance-sheet business (interest income), not third-party stored value under regulatory watch.

2. EML Payments — regulatory layer (new verified material)

• APRA/RBA stored-value regime bite (EML "Australian Regulatory Update", May 2026 — company-hosted ⚠️ pull the underlying ASX PDFs of 30 Mar 2026 + 27 Mar 2026 Notice to confirm at primary source): EPSL must obtain an ADI guarantee over stored-value liabilities for certain Australian prepaid programs pending an ADI licence application EML expects to resolve "within the next 12 months". Deadline originally 30 Apr 2026, extended on satisfactory-progress reporting. EML claims cost is "not material to earnings" — treat as unaudited management assertion; the real signal is that Australian BIN-sponsor self-issuance is about to become much harder (EML CFO's own words on the FY26 call), which kills a customer-leakage vector but locks EML into a regulator-supervised cost structure for its AU float.
• Float integrity context (verified from filings): 31 Dec 2025 segregated funds+bonds $2,662.3M vs stored-value liabilities $2,640.7M (thin positive residual, structural NCAD of $546M from duration mismatch — bonds >12m classified non-current vs on-demand liabilities). The $13.6M FY26 shortfall provision (deep-dive note §3) sits inside this architecture — the float is covered on the face of the accounts, but the provision + BIN-sponsor indemnity structure is where a regulator follow-up would come from.
• Rate asymmetry refined: FY26 stored float A$2.2Bn (GBP 46% / AUD 24% / EUR 21%); annualised float yield fell 3.6% → 3.2%, interest revenue $56.7M −11%, bonds = 52% of interest revenue at 4% avg yield/2.5y term. So RBA/R&** cuts do hurt the P&L directly** (contrary to a lazy "float = rate long" framing — EML has already locked the portfolio long and is rolling down), while rate-cut-driven demand recovery is the only thing that fixes the revenue line. Both legs now point the same direction: cuts = earnings down, no cuts = volumes down. Squeeze confirmed at primary numbers.
• Index leg (unchanged from forced-flow note): All Ords removal effective pre-open Mon 21 Sep 2026 — EML-only, no ZIP/TYR involvement.

3. ZIP — regulatory overlay

The Oct-2026 reform package is second-order for ZIP: BNPL sits outside the surcharge ban's designated networks and under separate credit rules (National Credit Code since 10 Jun 2025, already in the deep-dive). Directionally positive: ZIP's merchant base is the SME cohort that gains ~A$910M/yr wholesale cost relief, and the RBA's own CPI-absorption caveat (~0.1pp) is a mild consumer-purchasing-power tailwind. The binding ZIP constraints remain its own credit band (1.5–2.0% losses vs 1.77% actual) and the 23 Oct Q1 FY27 update.

4. Competitor map (AU SME payments, for sector-health)

• MerchantCompare 2026 published in-person rates: CommBank Smart 1.10% < NAB 1.15% < Westpac 1.20% < Tyro/Zeller/Bendigo/SumUp 1.40% < Square 1.60% < Stripe 1.70%. Tyro and Zeller price at the same premium-to-banks level; banks are the cheapest headline rate but weakest on terminal/software experience; Zeller owns the no-monthly-fee hardware story ($99 buy vs Tyro's $19–29/mo rental).
• Scale check: Tyro 77,800 merchants (FY26 AR, primary) vs Zeller 100k+ "customers" (mobiletransaction, ⚠️ Zeller counts EFTPOS-only users differently — not apples-to-apples). Tyro's defensible turf: GP/health claiming leadership, 450→580+ POS integrations, franchise/enterprise wins (Bakers Delight, Jaycar multi-year).
• Post-reform dynamics to watch (my hypothesis, flag as such): the 1 Apr 2027 transparency mandate + lower debit interchange push the market toward unbundled/interchange-plus pricing. That structurally favours published-flat-rate challengers (Zeller, Square) over quote-per-merchant incumbents (Tyro, ANZ, banks) unless Tyro's vertical-software lock-in holds. Watch ZLG (Zeller) FY26 results (late Aug) take-aways and Tyro merchant-churn language in the H1 FY27 update as the two live tells.

5. Bottom line for the desk

• TYR is the regulatory-repricing trade, not the index trade: 1 Oct surcharge ban + Apr 2027 transparency hit the acquirer margin pool while TYR's growth already decelerated to mid-single-digit; the bulk-billing headwind cycles out in FY27, which is the re-rating trigger if pricing holds.
• EML is the forced-flow + float-integrity trade (21 Sep index exit; APRA guarantee/ADI process is the diligence item that decides whether the $13.6m provision is one-off or regime).
• ZIP is the credit-band trade (1.77% vs 2.0% band, 23 Oct) — the AU card reforms are a mild tailwind, not a driver.
• One portfolio, three different clocks: 21 Sep (EML), 1 Oct (TYR), 23 Oct (ZIP).

Unverified / flagged

• TYR share count, cash balance, short interest — next data pull.
• EML ADI-guarantee update sourced from company CMS PDF; ASX primary PDFs (30 Mar / May 2026) not yet pulled to disk.
• Zeller "100k+ customers" vs Tyro 77,800 — counting-basis mismatch.
• RBA figures are via secondary summaries of the Conclusions Paper; the paper itself (rsn.rba.gov.au) should be cited at primary before any external publication.

— Tam | 8 Sep 2026 | split accepted from Eva (Trade Desk), ref tam-coord-20260908