ANZ Group Holdings Limited

The Piotroski F-Score is built for industrial firms — gross margin, the current ratio and asset turnover are meaningless for a bank whose balance sheet is itself a book of financial assets. In its place we use a nine-point, CAMELS-inspired check spanning capital, asset quality and earnings/funding.

It deliberately captures asset quality (credit risk) — historically the dominant driver of bank failure and otherwise absent from the five scoring pillars. A 7/9 result places ANZ in the "Sound" band: capital, liquidity and funding are the strongest of the majors, and asset quality is solid; the two failures are the credit-impairment charge (lifted by a A$175m Middle East conflict overlay) and a cash ROE that sits just below the estimated cost of equity.

Measures how fast the bank's income and balance sheet are expanding — operating income, customer deposits and gross loans, year-on-year. ANZ's growth is moderate: deposits and loans grew in the mid-single digits, and loan growth sits inside the healthy band. The drag is operating-income growth, where a full year of Suncorp Bank flattered the reported figure while organic income was held back by net-interest-margin compression.

Captures how efficiently the bank turns its franchise into profit — return on equity, the cost-to-income ratio, and net interest margin. This is ANZ's weakest pillar: ROE of ≈9.9% is the lowest of the majors and sits at the cost-of-equity line, while the group net interest margin (1.56%) is structurally the thinnest of the big four because of ANZ's large, low-margin Institutional book. The one bright spot is a rapidly improving cost-to-income ratio.

The regulator-facing safety metrics — CET1 capital, the Liquidity Coverage Ratio and the Net Stable Funding Ratio. ANZ is fortress-like here and scores a perfect 100: CET1 of 12.39% is "unquestionably strong", the LCR (132%) and NSFR (115%) both clear their green thresholds, and the loan-to-deposit ratio is the lowest of the majors. This is the clearest structural strength in the profile.

The income case — cash yield plus the franking credits that matter so much to Australian residents and SMSFs. ANZ offers the highest cash yield of the majors (≈4.7%) and a conservative payout (≈66–70%), which supports the pillar. The catch: ANZ franks only about 72.5% of its dividend (its earnings are more offshore-weighted), so the grossed-up yield of ≈6.2% is no better than a fully-franked peer paying a lower cash dividend.

Whether the shares are cheap or expensive — price against book value, tangible book, earnings, deposits and a dividend-discount model. ANZ is the cheapest major on nearly every multiple (P/B 1.44×, cash P/E ≈15×), which lifts this pillar well above its peers. The offset is the Gordon-growth DDM, which — like all the majors after their re-rating — implies an intrinsic value below the market price and drags the pillar down.

A separate nine-point health check covering capital, asset quality and earnings/funding — the bank-specific replacement for the Piotroski F-Score. It is shown alongside the composite, not blended into it, and it surfaces credit-risk signals the five pillars do not. ANZ's 7/9 is the strongest soundness reading in the Sentinel bank series to date, driven by fortress funding.

Normalisation. Higher-better: score = clamp((v − red_floor)/(green_ceiling − red_floor),0,1)×100. Lower-better: score = clamp((red_ceiling − v)/(red_ceiling − green_floor),0,1)×100. Band metrics (loan growth, payout): full score inside the green band, declining to the red boundaries. DDM: clamp((upside + 0.30)/0.60,0,1)×100.

Growth (62.5). G1 Op-income ≈+3.5% → 37.5 (AMBER; reported figure flattered by a full year of Suncorp Bank versus ≈2 months in FY24 — organic income was near-flat under NIM compression, so the number is not blindly rewarded). G2 Deposits ≈+4% → 50.0. G3 Gross loans ≈+4.5%, inside the 4–9% green band → 100. Group loan growth is broadly in line with system credit growth; no over-extension flag.

Profitability (26.2). P1 ROE 9.9% → 22.5 (FY25 underlying 9.6%, improving to a 1H26 run-rate ≈10.3%; RoTE 11.6%). P2 CTI 49.4% → 56.0 (improving fast). P3 NIM 1.56% → 0 (below the 1.70% red threshold; this is structural — ANZ's Institutional/Markets-heavy mix dilutes group NIM rather than reflecting mortgage-book underperformance — and it is stable, not sharply compressing). Scored on a cash / ex-significant-items basis; FY25 statutory NPAT A$5,891m vs cash A$5,787m vs cash ex-sig A$6,896m (significant items A$1,109m: restructuring, ASIC settlement, Panin impairment).

Capital & Liquidity (100.0). C1 CET1 12.39% → 100 (major-bank thresholds: red 10.25 / green 11.5). C2 LCR 132% → 100. C3 NSFR 115% → 100 (at the green threshold). All three clear their green thresholds, giving a perfect pillar — the strongest capital/funding profile of the majors. Calibrated to the APRA major-bank category.

Dividends (76.2). Cash yield = A$1.66 ÷ A$35.26 = 4.71% → 85.5 (highest cash yield of the majors). Gross (franked) yield = 4.71% × (1 + 0.725 × 0.30/0.70) = 6.17% → 66.9. Franking blended ≈72.5% (FY25 final 70%, 1H26 interim 75%) — partial because a material share of ANZ's earnings are sourced offshore (NZ, Institutional), so despite the higher cash yield the grossed-up figure is only in line with a fully-franked peer. Payout ≈66–70% sits in the 60–75% green band → no sustainability cap applied. No ordinary-dividend cut within three years.

Valuation (65.8). P/B 1.44× → 79.3; P/TBV 1.62× → 88.3; P/E(cash) 14.94× → 61.2; P/D 0.145× → 100 and P/A 0.088× → 100 (V4 avg 100; P/A band custom-calibrated: green <0.10, amber 0.10–0.15, red >0.15 — ANZ's large low-margin balance sheet lowers P/A). DDM (V5 = 0): r = rf + β·ERP = 4.80% + 1.00×5.75% = 10.55%; g = ROE×(1−payout) = 9.9%×0.32 = 3.17% (below the ~4–5% GDP cap); D₁ = 1.66×(1+g) = A$1.713; V = D₁/(r−g) = 1.713/0.0738 = A$23.20; upside = (23.20−35.26)/35.26 = −34.2% → 0. Sensitivity: the value is acutely sensitive to (r−g). A ±50bps move in g shifts intrinsic value by roughly ±A$1.7. The beta assumption is the key swing: substituting the raw 5-year measured beta (≈0.57) lowers the cost of equity to ≈8.08% and lifts intrinsic value to ≈A$35 (near parity). We adopt β = 1.00 for consistency with the Sentinel major-bank series and APRA-style valuation convention; the model directionally confirms full valuation rather than precise fair value.

Composite. 0.20×(62.5 + 26.2 + 100.0 + 76.2 + 65.8) = 12.50 + 5.24 + 20.00 + 15.24 + 13.16 = 66.1 → MODERATE. No pillar excluded; no weight redistribution required.

Peer set. Australian major banks — CBA, WBC, NAB (excluding ANZ itself). P/B median ≈2.0× (CBA ≈3.7 inflates; WBC 1.67, NAB ≈2.0); P/E median ≈18× (CBA ≈27, NAB ≈17–19, WBC ≈16); ROE median ≈11.4% (CBA 13.8, NAB 11.4, WBC 10.0); CET1 median ≈12.35%; cash-yield median ≈4.0%. A generic GICS-sector median is inappropriate for banks because business mix changes benchmark values.

Cash vs statutory & significant items. ROE, EPS and CTI scored on a cash / ex-significant-items basis per Australian broker convention. FY25 cash profit of A$5,787m is depressed by A$1,109m of significant items; the underlying figure of A$6,896m (flat YoY) is used for scoring. 1H26 carried no significant items. 3σ outlier capping: no metric required capping in this report.

Banks publish capital and liquidity metrics only at half-year and full-year results, so the latest CET1/LCR/NSFR (31 Mar 2026) are legitimately the most current available and are not stale. Book-value, tangible-book and balance-sheet per-share items are drawn from the 1H26 result (Mar-26) and FY25 annual report (Sep-25) pending the FY26 result (due early Nov 2026); flagged amber for the half-yearly cadence rather than as an error. Balance-sheet scale items (customer deposits ≈A$730b, total assets ≈A$1.2tn) are indicative FY25 figures and are used only for the P/D and P/A multiples, which are secondary. Market price, bond yield and dividend data are current to 3 July 2026.