Lloyds Banking Group plc (NYSE: LYG) — Britain’s Biggest Bank, Priced as Britain’s Best

An independent fundamental research note. All figures GBP and per ORDINARY share unless flagged; Lloyds reports in sterling under IFRS, fiscal year = calendar year. LYG is a US ADR: 1 ADS = 4 Lloyds ordinary shares (LSE: LLOY.L). As-of date 2026-07-24; ADS $6.02, ordinary 113.45p, GBP/USD 1.3315.

⚡ Claude’s Take

This block is the author’s own subjective opinion. It is general information, not investment advice, and not a recommendation to buy or sell any security. The analysis that follows (sections 1–15) is deliberately position-free and carries no price target — the single exception is this block.

Verdict: AVOID-here / HOLD for existing holders. A genuinely good, genuinely de-risking bank that has been repriced from 0.83x to ~1.96x tangible book in four years on an earnings uplift that is substantially a treasury reinvestment effect, not a franchise improvement. Fresh capital belongs in the ~1.30–1.55x tangible-book zone — roughly 75–90p ordinary / ~$4.00–4.80 per ADS — not at 113.45p / $6.02 (≈1.96x last-reported TNAV, ≈14x normalised earnings), where the multiple already embeds a perpetual RoTE the franchise has never once earned.

Two facts govern this call. First, more than 100% of Lloyds’ guided 2026 net-interest-income growth is the structural hedge. The £246bn sterling hedge — a rolling ladder of receive-fixed swaps with a ~3.75-year weighted-average life — is rolling 2021–22 vintages struck near 0.5–1.5% onto reinvestment rates “just shy of 4%.” Hedge income goes £3.4bn (2023) → £5.5bn (2025) → c.£7.0bn (2026) → c.£8.0bn (2027), while underlying NII actually fell £130m across 2023–25 and the residual ex-hedge NII is down ~25% from its 2022 peak. That is worth roughly 8–9 points of RoTE, it is ~90–95% contracted for 2026, and it mechanically exhausts as the ladder fully reprices by around 2029. It is real cash and management deserves credit for building the ladder — but it is a finite, industry-wide, treasury effect that every UK bank is enjoying simultaneously, and capitalising it at a perpetuity multiple is the error. Second, Lloyds is the biggest UK bank, not the best one. It has the largest deposit base (£496bn), the largest mortgage book (£325bn, 67% of loans), the largest app estate (21.5m users) and the most customers in Britain — and the worst cost:income ratio (58.6% vs NatWest’s 48.6%) and the lowest group RoTE (12.9% vs NatWest’s 19.2%) of the domestic majors. It nonetheless trades at a ~30% higher price-to-tangible-book than NatWest (1.96x vs 1.48x) while earning a lower return. In a scale business, the largest player having the worst cost ratio is the empirical refutation of a differentiated scale advantage: the moat here is the industry’s regulatory-and-inertia oligopoly, shared with four rivals and monetised less efficiently than the best of them.

Strip the noise and the honest number is a clean ~13.5–14% RoTE, not the 12.9%/14.8%/17.0% ladder the headlines offer: the 17bp asset-quality ratio is ~8bp flattered by model and debt-sale benefits (management itself says “closer to 25 basis points”), Q1-2026’s 17.0% carries ~1.2pp of tax benefit at a 23.2% effective rate against ~27% guidance, and 38% of the 2025 pre-tax profit increase came from a swing in “volatility and other items” driven by disposal gains. At an 11% cost of equity and 3% growth, 1.96x tangible book solves to a sustained ~18.7% RoTE — above the >16% guide, far above clean delivery, and above anything Lloyds has printed in the modern era. Meanwhile TNAV per share is +0.7% over four and a quarter years (57.5p at Dec-2021 → 57.9p at Mar-2026) despite £9.45bn of buybacks retiring 17% of the count: the compounding has been distribution, not book growth. And the buyback has now crossed the accretion line — at ~2x book it is TNAV-dilutive, with the current £1.75bn programme cutting TNAV/share by roughly 1.6p (−2.8%). Framing: a mid-to-late-innings re-rated income cyclical, not a compounder and not a falling knife. The factor data agrees — Momentum is zeroed in all four models while DividendYield (+0.41) and Country:UK (+0.68) dominate, and the nearest factor neighbours are international value and high-dividend baskets. But the lifetime record is the reminder: −94.8% maximum drawdown, a negative 20-year Sharpe, and a price still 58% below its 2007 dividend-adjusted peak after a 190% three-year run. Tag: “You’ve stopped buying a discount to book and started buying a swap ladder at a perpetuity multiple.”

Three ownership facts sharpen the point rather than proving it. Harris Associates (Oakmark) — the register’s only large, active, value-orientated shareholder, at 4.99% — sold out entirely during 2025, disclosed in prose in the annual report because no TR-1 was ever filed. Over the same thirty-one months there were three open-market insider purchases totalling £521k against £15.4m of discretionary executive selling — a 1:30 ratio by value — with the selling accelerating in both size and realised price (2024: £1.5m at ~56p; 2026 YTD: £8.6m at ~106p). And nobody at Lloyds has ever bought a share above 84p: the Chair bought roughly every 12–24 months between 28p and 47p, and has bought nothing since February 2024, through a doubling. None of that is a timing signal, and management’s absolute alignment is genuine — but the people closest to the asset have been net sellers into the entire re-rating.

Conviction: medium-high. Two live qualifiers keep it from high. (a) Publish-date risk: Lloyds reports H1-2026 results and an entirely new multi-year strategy on 30 July 2026 — six days after this report date; the 2022–26 plan culminates this year and next-phase targets are unpublished, so a credible 2029/2030 return framework could legitimately reset the analysis. (b) Motor finance is genuinely de-risking: the £1.95bn provision was not topped up in Q1-2026 — the first non-increase since 2023 — and it came after the FCA’s final rules, though the scheme is currently suspended by the Upper Tribunal pending challenge. The single fact that flips me bullish: RoTE holding ≥16% through a full Bank of England easing cycle with cost:income sustainably under 50% on a flat income line — that would prove the returns are franchise, not hedge, and justify paying up. The single fact that flips me bearish faster: a further motor-finance top-up, or the 2026 cost:income guide missed because remediation returns — the <50% target has essentially zero room for another conduct charge, and at ~2x book the downside to a low-teens-RoTE multiple (~1.35x, ~78p) is roughly a third.

📈 Stock Price Action — Five-Year Event Map

Factual price history, not a recommendation. Price moves are FACT; attributed causes are INTERPRETATION. No price target, no support/resistance, no chart patterns. USD ADS closes (unadjusted); 1 ADS = 4 ordinary shares.

The arc. Over five years the LYG ADS fell from $2.37 (20-Jul-2021) to a $1.71 trough on 12-Oct-2022 — the depths of the UK gilt crisis — then rose more than 3.5x to a $6.25 peak on 3-Feb-2026, its highest level since 2008. It closed at $6.02 on 24-Jul-2026, inside a 52-week range of $4.20 (28-Jul-2025) – $6.25 (3-Feb-2026), just −3.7% off the 52-week high and having recovered +24.6% from a sharp $4.83 low on 27-Mar-2026. Five-year total return is +210.5% (+143.7% price-only); on a dividend-adjusted basis the shares remain 58% below their October-2007 all-time high. Sterling was broadly flat against the dollar over the period, so essentially the entire ADS move is the local-currency LLOY price, not FX.

Cycle narrative. (1) The Bank of England’s first post-Covid hike turned a decade-long headwind into a tailwind and the ADS added 30% in 21 trading days — the best one-month window of the five years. (2) That reversed violently: Ukraine, a recession scare, and the Truss mini-Budget and gilt crisis took the ADS to a five-year low of $1.71 and a −40.2% peak-to-trough drawdown. (3) From that trough the stock roughly doubled over 21 months as higher Bank Rate flowed through net interest income and the hedge repriced. (4) The Court of Appeal’s 25-Oct-2024 DCA judgment is the clearest company-specific de-rating event in the five-year record, and the resulting overhang is the most plausible reason LYG has lagged every European bank ADR peer since. (5) 2025 was macro-dominated — a soft December CPI relieved the January gilt scare, FY2024 results added conviction, and the April tariff shock and pause produced the year’s two largest daily moves in opposite directions within four sessions. (6) The de-risking leg: the Supreme Court judgment was received calmly, the November Budget spared banks entirely, and the FY2025 print carried the shares to their highest level since the financial crisis. (7) The peak did not hold — a dovish 5–4 BoE hold cut the rate outlook, the Iran escalation hit European bank equity broadly, and two idiosyncratic negatives landed in March; the shares lost 22.7% in six weeks with no adverse earnings news, which quantifies how much macro and geopolitical beta is embedded in a name whose entire earnings power keys off one country’s rate path. (8) The recovery has been driven by the removal of tail risk rather than by fundamentals: final FCA rules landed inside the existing provision, Q1 beat by 33%, and the mid-June ceasefire restored the sector bid.

1. Executive Summary

Lloyds Banking Group plc is the United Kingdom’s largest retail and commercial bank — a £944bn-balance-sheet, £66.7bn-market-cap domestic lender operating through Lloyds Bank, Halifax, Bank of Scotland, Scottish Widows, Black Horse and MBNA, serving roughly 26–28 million customers with 21.5 million mobile-app users and 60,061 employees. It is best understood as a very large, very concentrated bet on the United Kingdom: 67% of the £489bn loan book is UK residential mortgages, essentially all revenue is sterling, and the earnings line keys off a single country’s rate path, housing market and fiscal politics. FY2025 statutory total income was £19,422m (+8%), pre-tax profit £6,661m (+12%), attributable profit to ordinary shareholders £4,196m, EPS 7.0p, and reported return on tangible equity 12.9% — 14.8% excluding a Q3 motor-finance charge. Q1-2026 was materially stronger: PBT £2,025m (+33%), EPS 2.4p, banking NIM 3.17%, cost:income 51.9% and RoTE 17.0%.

The investable question is what is actually driving that improvement, and the honest answer is a treasury mechanism rather than a franchise one. Lloyds converts ~£103bn of near-zero-cost personal current accounts and ~£28bn of tangible equity into income through a £246bn sterling structural hedge — a rolling ladder of receive-fixed swaps with a ~3.75-year weighted-average life. Vintages struck in 2021–22 near 0.5–1.5% are rolling onto reinvestment rates management describes as “just shy of 4%.” Hedge income runs £3.4bn (2023) → £4.2bn (2024) → £5.5bn (2025) → c.£7.0bn (2026) → c.£8.0bn (2027), roughly 47% of guided 2026 NII. Against guided 2026 NII growth of ~£1.3bn, guided hedge growth is >£1.5bn: by management’s own numbers, more than 100% of 2026 net-interest-income growth is hedge reinvestment, with every other moving part — mortgage spreads, deposit repricing, volume — netting negative. Lloyds is writing mortgages at ~70bps completion margins against a back book maturing near 100bps, so each roll destroys ~30bps. Ex-hedge residual NII has fallen ~25% from its 2022 peak, and net non-interest income after operating-lease depreciation is only 1.4% above 2021. The hedge is worth roughly 8–9 points of RoTE and is ~90–95% contracted for 2026 — genuine, bankable cash — but the ladder fully reprices by around 2029, at which point income plateaus and growth stops.

Competitively, Lloyds is the biggest UK bank rather than the best one. The moat is real but shared: a legally protected five-firm oligopoly holding ~75% of UK current accounts, high state-erected entry barriers, and habit-based captivity evidenced by a system-wide switching rate of just 1–2% per year sustained across thirteen years of mandated, free, seven-day switching. In Greenwald’s taxonomy this is economies of scale combined with customer captivity — but it belongs to the industry, and NatWest holds a more profitable version of it. Lloyds converts the largest deposit base, mortgage book and app estate in Britain into a 58.6% cost:income ratio against NatWest’s 48.6% and a 12.9% group RoTE against NatWest’s 19.2%. A firm with the greatest scale in a scale business should have the lowest cost ratio; it does not. The one genuinely Lloyds-specific edge is distribution economics — direct-to-bank mortgage share up 3pp to ~26%, ~85% of current accounts opened through a seven-minute mobile journey — which is real, narrow, recent and copyable. Meanwhile Nationwide, a mutual with no shareholders to serve, is out-recruiting all four of the Big Four combined on net current-account switching, and the Halifax brand — a persistent switching loser — is being retired, removing precisely the mass-market coverage Greenwald prescribes defending.

Earnings quality and capital are both weaker than the headline ladder suggests. The reported 17bp asset-quality ratio is roughly 8bp flattered by model calibrations and debt sales (management concedes the underlying figure is “closer to 25 basis points”); Q1-2026’s 17.0% RoTE carries ~1.2pp of benefit from a 23.2% effective tax rate against ~27% guidance; and 38% of the 2025 PBT increase came from a favourable swing in “volatility and other items” driven by disposal and acquisition gains. Normalised, clean FY2025 RoTE is ~13.5–14.0%. On capital, the surplus story is finished: pro-forma CET1 of 13.2% sits ~20bp above the 13.0% target, absolute CET1 capital is lower than in 2021 while RWAs have grown 20%, and sustainable distribution capacity is roughly £3.4–4.0bn a year from run-rate generation. Most tellingly, TNAV per share has risen 0.7% in four and a quarter years — 57.5p at December 2021 to 57.9p at March 2026 — despite £9.45bn of buybacks retiring 17% of the share count. All of the shareholder return has been distribution; none has been book compounding.

Against that, the price has done all the work. The ADS has more than tripled from its 2022 low and doubled since end-2024; P/TNAV has gone 0.83x → 0.94x → 0.93x → 1.04x → 1.73x → 1.96x on last-reported tangible book, with the price-to-book percentile at roughly the 80th of the stock’s own decade. On a Gordon-growth basis at an 11% cost of equity and 3% growth, 1.96x embeds a sustained ~18.7% RoTE — above the >16% 2026 guide, well above clean delivery, and above anything this franchise has printed. Lloyds trades ~30% above NatWest on tangible book while earning less, and at ~1.87x forward TNAV against a UK peer average nearer 1.35x. Buybacks, accretive at 0.8–1.0x book in 2021–24, are now dilutive to TNAV per share at ~2x.

Verdict in brief: a genuinely improving, genuinely de-risking bank — the motor-finance provision held flat in Q1-2026 for the first time since 2023, Basel 3.1 will release £6–8bn of RWAs in January 2027, credit is benign and distributions run near 6% of market capitalisation. But the competitive advantage is the industry’s rather than the company’s and is monetised less efficiently than the best peer; the earnings uplift is a finite, sector-wide treasury effect being capitalised as perpetual franchise improvement; and the valuation has moved from a wide margin of safety to a demanding premium. This note takes no position and sets no price target; it lays out the embedded expectations and the falsification tests for each side. (The one deliberately-labeled exception, Claude’s Take, appears above.) One further caveat governs everything below: Lloyds reports H1-2026 results and an entirely new multi-year strategy on 30 July 2026, six days after this report date.

2. Business Overview

Lloyds Banking Group traces its origins to 1695 (Bank of Scotland) and 1765 (Lloyds), and in its modern form is the product of the 2009 HBOS rescue and a decade of post-crisis state ownership, fully exited in 2017. It is today a domestically-focused, ring-fenced UK universal bank with £944,072m of total assets at 31 December 2025 (£968,125m at 31 March 2026), £489bn of gross lending, £496bn of customer deposits and 60,061 full-time-equivalent employees — of whom ~55,265 are in the UK. Unlike Barclays or HSBC, there is no meaningful international or investment-banking diversification: this is a bet on Britain, and it should be underwritten as one. (FACT: Form 20-F FY2025, filed 13-Feb-2026; FY2025 results release, 29-Jan-2026.)

2.1 Segment structure and where the profit comes from

The FY2025 Form 20-F confirms three operating segments plus a residual, unchanged through 2025 and 2026: Retail; Commercial Banking; Insurance, Pensions & Investments (IP&I); and Equity Investments & Central Items. The only 2025 presentational change moved certain divisional variable-payment costs from operating costs into divisional other income, with comparatives restated and no net P&L impact.

Retail is ~50% of underlying pre-tax profit, Commercial ~38%, IP&I ~5% and Central ~8%. Note that Retail also carries a £1,445m operating-lease depreciation charge (motor leasing, Tusker) — a genuinely industrial cost line sitting inside a bank, and the principal reason Lloyds’ 3.17% banking NIM is not comparable with NatWest’s 2.47%. Any peer NIM comparison that ignores this is wrong.

2.2 The loan book: this is a UK housing lender

Sixty-seven per cent of the loan book is UK residential mortgages. Growth is concentrated in higher-margin, higher-risk consumer lines — cards +10%, unsecured +15%, motor +7%, Retail Europe +21% — while the UK SME book shrank 5% on government-backed-loan repayments. Mortgages grew 3%, but at compressing spreads that management has flagged in every single quarterly release.

2.3 How the money is actually made — the structural hedge

The single most important mechanism in this business, and the one most often mis-described as “net interest margin expansion,” works as follows.

1. The raw material — rate-insensitive balances. Roughly £103bn of personal current accounts paying essentially nothing, plus ~£28bn of tangible equity. These balances do not reprice when Bank Rate moves.
2. The monetisation — the hedge. Lloyds converts them into a rolling ladder of receive-fixed interest-rate swaps: sterling notional £244bn at 31-Dec-2025, £246bn at 31-Mar-2026, weighted-average life ~3.75 years. That life implies roughly £65–70bn of notional matures and reinvests every year. The income is therefore mechanical and lagging — decoupled from spot Bank Rate, and driven by the gap between what maturing tranches were struck at and what today’s swap curve pays.

†at 31 March 2026. Management guidance is verbatim: “The Group expects sterling structural hedge earnings to be c.£7.0 billion in 2026, to be c.£8.0 billion in 2027, with earnings growth from the structural hedge expected to continue thereafter.” On the Q1-2026 call management confirmed ~90–95% of 2026 hedge income and ~80% of 2027 is already locked, with Q1-2026 hedge income alone at £1.6bn.

3. The asset spread — currently a negative. Lloyds’ Q1-2026 mortgage completion margin was ~70bps against a maturing back book at just under 100bps: every mortgage that rolls destroys ~30bps. Management expects that headwind to be “petering out at the beginning of next year… certainly by the first half of 2027.”
4. The genuinely high-margin books: cards (3.4% ECL coverage, +10% balances), unsecured loans (+15%), motor finance, and Commercial Banking, whose banking NIM rose to 4.93% from 4.51% — though on a shrinking average asset base, which makes it a deposit-margin outcome rather than a lending-franchise one.

The arithmetic that governs the whole report: guided 2026 NII growth is ~£1.3bn (£13.6bn → >£14.9bn), against guided hedge income growth of >£1.5bn. Subtracting hedge gross income from underlying NII gives a residual that has fallen from £10.6bn (2022) to £8.1bn (2025) and is guided lower again in 2026. (Mandatory caveat: this residual is not a clean franchise-margin measure — hedge “gross income” is the gross return on hedged notional while the funding cost of those deposits sits in the residual, so the level is not meaningful. The direction and the attribution are.)

2.4 The deposit franchise, and the tell in Q1-2026

The Q1-2026 movement is the revealed preference that proves the mechanism: Lloyds deliberately let £3.0bn of fixed-term savings run off — the IMS cites “Group participation decisions in the fixed term deposit market,” and management described price discipline in a market that is “increasingly competitive and, at times, negative margin” — while current-account balances rose £0.6bn. Management is explicitly buying franchise deposits and refusing commodity ones, because only the former feed the hedge. Loan-to-deposit ratio 98%.

2.5 Franchise scale and distribution

Market position (sourced, honestly labelled): #1 in UK mortgages with c.19% flow share in 2025 and 18.5–19% in Q1-2026 (disclosed); #1 in direct-to-bank mortgage flow at c.26%, +3pp y/y (disclosed); #1 in UK current accounts at an estimated ~20–21% (estimate — Lloyds does not disclose a PCA unit share); #1 UK motor-finance lender by book (£17.4bn); top-three in credit cards, SME, workplace pensions and individual annuities; >14% of new home-insurance policies and a protection share up from 5.8% to 7.8%.

Verdict. A very large, very domestic, deposit-rich lender whose profit is ~50% retail and ~67% collateralised by UK housing, and whose current earnings engine is a treasury reinvestment mechanism rather than a lending franchise. The business is comprehensible, the disclosure is good, and the concentration — one country, one currency, one housing market, one rate path — is the defining structural fact.

3. Industry Dynamics

3.1 Structure: concentration rising where the profit is

The UK’s ring-fenced Big Five are Lloyds, NatWest, Barclays UK, HSBC UK and Santander UK, alongside Nationwide, a mutual now ranked #2 in both mortgages and retail deposits. The Big Four hold roughly 75% of UK current accounts. The supply side has consolidated hard in twenty-four months:

1. Nationwide / Virgin Money — integration completed in FY2026; underlying PBT £2.0bn (+9%); mortgage balances £286.3bn = 16.3% balance share; retail deposits £270.8bn = 12.2% share.
2. NatWest / Sainsbury’s Bank — retail assets absorbed.
3. Santander UK / TSB — completed 30 April 2026, £2.65bn all-cash, creating ~28m customers, #3 in current accounts and #4 in mortgages, with ≥£400m of cost synergies and the 215-year-old TSB brand to be retired by H1 2027.

Against that, the digital challengers are winning accounts rather than balances: Monzo has passed 11m customers (deposits £11.2bn, +88%) with roughly half using it as a primary account; Starling has 4.6m accounts but growth has slowed sharply to +10%; Revolut has >50m customers globally and >£1bn of profit; Chase UK continues to scale. More than 30m UK adults now hold at least one digital-only account, and Chase, Starling and Monzo occupy the top three service-quality rankings, ahead of every incumbent.

INTERPRETATION: concentration is increasing where the profit pool is — deposits, mortgages, SME — and fragmenting where it is not: low-balance transactional accounts. The challengers have taken primacy at the young and low-balance end without yet taking the interest-earning deposit pool or the mortgage book. That distinction is the entire argument for why incumbent returns have held up, and it is also the thing to monitor, because primacy today is balances tomorrow.

3.2 UK mortgages: a commoditised refinancing market

Roughly 1.8 million fixed-rate mortgages mature in 2026, and product transfers at £261bn are ~87% the size of the entire gross new-lending market. Around 85–90% of new lending is broker-intermediated. This is a refinancing-and-retention market, not an acquisition market.

The pricing evidence is unambiguous. A documented 2026 price war has Nationwide cutting up to 16bps with a lowest two-year fix at 3.54%, best deals near 3.5% — the lowest since 2022 — with trade commentary noting the cuts reflect “lenders trimming margins where competitive pressure is sharpest, not a wholesale repricing on the back of cheaper funding.” Lloyds itself cites “deposit churn and asset margin compression” every quarter.

This is a commoditised, price-transparent product with no pricing power. The borrower’s comparison set is every lender’s rate on a broker’s screen, the product is identical, and switching at maturity is free. Lloyds’ ~70bps completion margin against a ~100bps maturing book is the arithmetic proof: the incumbent’s back book is worth more than anything it can write today. Mortgages are a scale-and-cost-of-funds game, not a franchise.

3.3 Deposits: the current account is the only real asset

UK household deposits stood at roughly £2.19 trillion in early 2026 (+4.3%). The effective rate on new deposits was 3.76% in March 2026 against a 3.75% Bank Rate, with market-leading easy-access rates at 3.6–3.9% AER — that is near-100% pass-through at the margin, and deposit betas on shopped money of effectively 1.0.

Which is precisely why Lloyds walked away from £3.0bn of it. The franchise value is the ~£103bn of current accounts plus equity, not the £322bn of retail deposits in aggregate. The remainder is a commodity funded at market. Any argument that Lloyds’ deposit base is a moat must be made about the current-account slice specifically, and must survive the switching data in 4.3.

3.4 The structural hedge is an industry phenomenon, not a Lloyds one

(Barclays figures from Barclays PLC FY2025 results and Q1-2026 results.)

Sell-side consensus expects UK banks to deliver ~10% NII growth in 2026 — roughly double the European rate — explicitly because of the hedge, with 2027 maturing tranches yielding ~2.1% against materially higher reinvestment.

How much of the sector’s 2024–2028 earnings uplift is just the hedge rolling? For Lloyds, hedge income rises £4.2bn → £8.0bn+ between 2024 and 2027 — +£3.8bn — against total NII rising £12.8bn → ~£15.5bn, or +£2.7bn. The hedge increase exceeds the entire NII increase. Every other moving part nets negative. The same shape holds at Barclays and NatWest. Substantially all of the sector’s 2024–2028 NII growth, and roughly 8–9 points of Lloyds’ RoTE, is a treasury reinvestment effect.

What happens when it stops. With a ~3.75-year weighted-average life, the ladder fully reprices to the prevailing swap curve by roughly 2029. Once the sub-1.5% vintages are exhausted, hedge income becomes a function of current swaps: if Bank Rate settles near Lloyds’ own 3.5% terminal assumption, income plateaus near £8bn rather than collapsing — but the growth stops, and with it the earnings bridge that underwrote the re-rating. If Bank Rate falls materially below the ~4% reinvestment level, the disclosed downside sensitivity begins to bite through both the hedge and the deposit book.

Disclosed rate sensitivity (FY2025; 50% deposit pass-through, 100% asset pass-through), banking-book NII:

The hedge dampens but does not eliminate rate sensitivity, and Lloyds notes the sensitivity is greater on the downside because of deposit-pricing lags.

3.5 UK macro as of mid-2026 — and Lloyds’ notably bearish own view

Lloyds’ own ECL base case is materially more bearish than consensus, and it is a stagflation case. Between December 2025 and March 2026 the bank cut its 2026 GDP assumption from 1.2% to 0.5%, raised CPI from 2.6% to 3.4% (peaking 3.9% in Q4), pushed exit-2026 unemployment from 5.1% to 5.6%, cut 2026 house-price growth from +1.6% to +0.7%, and moved to assuming no Bank Rate cuts at all in 2026 (3.75% flat, first cut deferred to Q3-2027). Probability-weighted ECL £3,223m; severe-downside £5,263m.

Read-through: higher-for-longer is good for near-term NII — the hedge reinvests at ~4% and deposit spreads hold — and bad for credit and volumes. Lloyds is explicitly underwriting that trade-off, and the market is currently paying for the first half of it without discounting the second.

3.6 Regulation: deregulatory on process, unchanged on conduct outcomes

(a) Ring-fencing reform — real and favourable. HM Treasury published the conclusions of its Ring-Fencing Review on 18 May 2026 (“Safeguarding Stability, Enabling Growth”). The primary core-deposit threshold rises from £25bn to £35bn, reviewed every three years from 2028. Implementation runs through the Financial Services and Markets Bill 2026-27, a PRA consultation on relaxing shared operational services, an HMT consultation on the growth allowance and wider product permissions for ring-fenced banks, and a PRA/FPC review of the interaction with the Basel 3.1 output floor and leverage ratio. Implementing consultations launched July 2026.

(b) Basel 3.1 — delayed to 1 January 2027, and accretive for Lloyds. UK Basel 3.1 now applies from 1 January 2027 (FRTB-IMA further delayed to 2028 per CP9/26, June 2026). Lloyds guides a Day-1 RWA reduction of £6–8bn — releasing roughly £0.8–1.0bn of CET1. Unlike trading-heavy peers, Lloyds is a beneficiary.

© The November 2025 Budget — the bank-tax question, answered. Chancellor Rachel Reeves delivered the Budget on 26 November 2025 with no increase to the bank levy and no new windfall or bank tax. Banks continue to pay the 3% corporation-tax surcharge on top of the 25% main rate (28% effective) plus a balance-sheet levy of up to 0.1%, together forecast to raise ~£2.5bn. The City had priced a raid; it did not materialise. INTERPRETATION: this removed a live overhang and is part of why UK banks re-rated through H1-2026 — but it is a political, not structural, reprieve, renewable at every fiscal event.

(d) FCA motor-finance redress — live, and currently suspended.

(e) Other. FCA Consumer Duty supervision is intensifying with explicit focus on the price-and-value outcome; the FCA is simplifying mortgage rules to widen access; HMT/FCA/PRA are reviewing SM&CR with the aim of halving its burden; the FCA has absorbed the PSR and is reforming the FOS redress framework in H2 2026. The direction of travel is deregulatory on prudential and conduct process, but not on conduct outcomes — the motor scheme proves the redress machine runs at full capacity even while the “growth agenda” does.

3.7 Profit pools and the capital cycle

Every major player is raising return targets simultaneously. That is not evidence of durable individual advantage; it is evidence of a common exogenous driver. Returns are not being competed away at the reported average level — but they are at the margin: ~70bps mortgage completion spreads, near-100% deposit pass-through at the top of the savings tables, and a mutual paying cash switching bonuses and Fairer Share payments out of member surplus.

Marathon capital-cycle placement: LATE-RECOVERY / EARLY-BOOM. The positive signals are genuine — capacity is still being removed (Lloyds closing 245 branches; the TSB and Halifax brands both being retired; three mid-tier consolidations in 24 months), balance sheets are repaired, capital is being returned rather than raised, asset growth is modest (L&A +5%, RWAs +5%), and Basel 3.1 is releasing RWAs. The warnings are equally clear: returns at multi-decade highs and publicly targeted higher by every major player at once, uniformly bullish sell-side, a hard re-rating, and capital arriving at the low-balance end via digital entrants with effectively unlimited funding. Marathon’s core warning applies precisely — high returns are attracting capital; it is simply arriving as price competition (mortgage spreads, savings rates, switching bonuses) rather than as new bank licences, while the reported average return is temporarily flattered by a treasury effect. Marathon’s own caveat also applies: ring-fencing, MREL, FSCS and licensing are barriers erected by the state, which is why the returns persist and why the cycle does not clear normally — and the same state hand produced PPI (~£50bn sector-wide) and is now producing the ~£9.1bn motor scheme. A periodic conduct levy is the price of a protected profit pool.

Verdict — a structurally good industry earning a cyclically and mechanically inflated return

Good, because: a legally protected five-firm oligopoly holding ~75% of current accounts; high state-erected entry barriers (licence, capital, MREL, ring-fencing, FSCS); a near-necessity product; structurally low switching despite thirteen years of mandated free switching; and a consolidating, not fragmenting, supply side. Five separate institutions earning mid-to-high-teens RoTE simultaneously is prima facie evidence of a real barrier.

Inflated, because: mortgages are commoditised and losing ~30bps on every roll amid a live 2026 price war; savings deposits price at ~Bank Rate; and ~47% of Lloyds’ 2026 NII — and more than 100% of its 2026 NII growth — is a structural-hedge reinvestment effect that mechanically exhausts by ~2029.

Net: a good industry whose present returns must not be capitalised as the through-cycle rate. The honest through-cycle number for a UK ring-fenced bank is low-teens RoTE, not the high-teens being printed and targeted in 2026.

4. Competitive Position

4.1 Naming the moat in Greenwald’s taxonomy

Type 3 — economies of scale combined with customer captivity. The captivity leg is habit and inertia — Greenwald’s most common and, per unit, weakest form of demand advantage, but the most robust in aggregate — plus modest search friction on the current account. The scale leg is a fixed cost base (technology, compliance, brand, a shrinking branch estate) spread across ~26–28m customers, a £496bn deposit book and a £489bn loan book.

There is no supply/cost advantage from proprietary technology. Everything in Lloyds’ £3–4bn programme — cloud migration, GenAI, app rebuilds, data platform — is bought from third parties available to every competitor. This is Greenwald’s explicit case that “information technology innovations created by third parties confer advantages on none.”

The decisive qualification: this is a shared four-to-five-firm oligopoly advantage, not a differentiated one — and NatWest holds a stronger version of it.

4.2 Greenwald’s two diagnostic tests

Market-share-stability test (<2pp drift over 5–8 years = formidable barriers; >5pp = none). UK mortgage share has moved from ~19–20% (balance share ~18.9% in 2023) to 18.5–19% flow share in Q1-2026 — drift of roughly 1–1.5pp. Retail-deposit share is broadly stable. Passes at the aggregate level.

But run the test at segment level and it is less comfortable. Nationwide has climbed to 16.3% of mortgage balances and #2 in retail deposits; Santander UK’s gross mortgage lending jumped from £16.1bn to £25.3bn in a single year (+57%) and it has just absorbed TSB; and Lloyds is running its savings book down by choice. The share is stable because Lloyds is choosing price over volume in savings and defending hard in mortgages — stability bought, not stability given.

ROIC test (sustained after-tax 15–25%+ = advantage present; 6–8% = absent; RoTE is the banking analogue). FY2024 ~12.3% → FY2025 12.9% reported / 14.8% ex-motor → Q1-2026 17.0% → 2026 guided >16%. Against a UK bank cost of equity conventionally 10–12%, Lloyds earns a modest through-cycle spread and a large current one. Passes — but only across the rate-elevated window, and the current pass is substantially attributable to the hedge (~47% of NII), not to franchise pricing power. Through the 2015–2021 low-rate era the identical franchise earned far less. The advantage is real, but its expression is rate-dependent — which is exactly Greenwald’s requirement that a moat tie to a financial outcome that would deteriorate without it.

4.3 Head-to-head: biggest, not best

*Not like-for-like with NatWest’s 2.47%: Lloyds’ NIM carries motor operating-lease income and a larger unsecured/cards book. Do not read this row as an efficiency comparison.

The most damning competitive fact in the file: Lloyds has the largest deposit base, the largest mortgage book, the largest customer base and the largest app estate in Britain — and the worst cost:income ratio and the lowest group RoTE of the domestic majors. The FY2025 cost:income gap to NatWest was ~10 percentage points (58.6% vs 48.6%), narrowing to ~5pp on Q1-2026 run-rates. A firm with the greatest scale in a scale business should have the lowest cost:income ratio. It does not. That is the empirical refutation of a differentiated scale advantage.

4.4 Switching costs: low in mechanism, high in practice — and shared

Britain has had mandated, guaranteed, free seven-day current-account switching (CASS) since 2013, which in theory eliminates switching friction entirely. It did not.

INTERPRETATION: the captivity is habit and inertia, not contractual lock-in. It is real, measurable, economically valuable — and available to every incumbent equally.

Lloyds’ own switching performance is the single most revealing competitive datapoint, because it is brand-split:

The Lloyds brand is a modest net winner; the Halifax brand is a persistent net loser. The group runs a net-flat-to-slightly-positive switching position only because one brand offsets the other. In a mandated-free-switching market, a genuinely advantaged franchise should be a consistent and large net gainer. Lloyds is not — and a mutual with no shareholders to serve is beating all four of the Big Four combined. (Label: the brand-level read is inferred from CASS flow data, not stock shares — INTERPRETATION.)

4.5 The technology transformation: advantage, or table stakes?

Inputs: ~£3bn of strategic investment across 2022–2024, roughly two-thirds aimed at revenue growth and diversification, extended into a ~£4bn programme; ~9,000 technology and data hires; data centres cut ~50% since 2021; legacy applications cut ~17.5%; 50 GenAI use cases launched in 2025 for ~£50m of P&L benefit.

Outputs: strategic-initiative revenue of £1.4bn annualised delivered by 2025 against an original c.£1.5bn-by-2026 target, subsequently raised to c.£2bn by end-2026; 21.5m app customers (+~45% since 2021); ~85% of current accounts opened through a seven-minute mobile journey; direct mortgage applications +32% y/y and direct-to-bank mortgage share +3pp to ~26%; Lloyds Premier launched for the mass affluent; 750k+ workplace-pension app users (+75%); protection share 5.8% → 7.8%; >14% of new home-insurance policies; agreed acquisition of the digital wallet Curve.

Assessment — partially yes, and mostly in distribution, not cost. The 3pp gain in direct-to-bank mortgage share structurally reduces broker procuration fees on the highest-volume product, which is a real and measurable unit-economics advantage on the one product where everyone else pays an intermediary. Protection share nearly doubling is real diversification into a capital-light line. Against that, £50m of GenAI benefit is a rounding error against a £9.8bn cost base, and the acid test fails: after £3–4bn of spend, FY2025 cost:income was still 58.6% against NatWest’s 48.6%, and the sub-50% target only arrives in 2026 — the year the hedge does the heavy lifting on the denominator. Cost:income improving because income is exploding is not an efficiency advantage. On Greenwald’s framing this is operational effectiveness — necessary, emulable, table stakes. Every UK bank is running the identical app-first, branch-closure, cloud-migration playbook, and Chase, Starling and Monzo out-rank all four incumbents on service quality.

4.6 The Halifax retirement, read competitively

On 3 July 2026 Lloyds announced it will consolidate to a single England/Wales/Northern Ireland retail brand: ~190 Halifax branches rebranded or absorbed through 2027, Halifax stops opening new accounts, Bank of Scotland retained in Scotland, with no job losses or additional closures announced. The cost logic is defensible — one marketing budget, one app, one product set.

But it removes a distinct market position — Halifax as the challenger-priced, mass-market, ex-building-society brand — and multi-brand coverage is precisely the defence Greenwald prescribes for a scale incumbent: do not leave niches undefended. Lloyds is choosing cost over coverage in the same year Nationwide is winning the switching war on exactly that mass-market, value-led positioning. This is a genuine strategic risk, not a clean cost win.

4.7 Where Lloyds is losing and gaining

Losing: mortgage flow share drifting from ~20% to 18.5–19% while Nationwide consolidates 16.3% of balances and Santander UK’s gross lending grew 57% in a year; Halifax as a persistent CASS net loser; primary-account share at the young and low-balance end to Monzo and Chase; service-quality rankings behind Chase, Starling and Monzo; and savings deposits — deliberately (−£3.0bn in Q1-2026).

Gaining: direct-to-bank mortgage share (+3pp to ~26%); protection (+2.0pp to 7.8%); home insurance (>14% of new policies); cards (+10%) and unsecured loans (+15%); Retail Europe (+21%); commercial deposits (+5%); IP&I open-book AuA £232bn (+15%).

Verdict — a real but shared and undifferentiated moat: a scale player in a commoditised market

The moat exists. Greenwald type-3 — economies of scale plus habit-based captivity — evidenced by stable ~19% mortgage share, ~1–2%/yr system-wide switching despite thirteen years of mandated free switching, ~£103bn of near-zero-cost current accounts, and a five-firm oligopoly earning mid-to-high-teens returns simultaneously. That is a genuine barrier to entry and it is why UK banks earn what they earn.

But it is not Lloyds’ moat — it is the industry’s moat, held more profitably by NatWest (RoTE 19.2% vs 12.9%; cost:income 48.6% vs 58.6%) and at higher divisional returns by HSBC UK (21.6%) and Barclays UK (19.7%). Lloyds converts the largest deposit base, the largest mortgage book and the largest app estate in Britain into the lowest reported group return of the domestic majors. In Greenwald’s terms it has the scale but not the scale economics — the empirical signature of size without advantage. The one genuinely Lloyds-specific edge is distribution economics, and it is narrow, recent and copyable.

5. Growth History and Forward Opportunities

5.1 The historical record: no revenue growth for five years, then a rate cycle

*FY2022 restated for IFRS 17; there is a series break between the FY2021 and FY2022 columns.

Two observations govern the growth verdict. First, net income grew 27.1% across five years — about 4.9% a year — and almost all of it arrived in the single step from 2021 to 2022 when Bank Rate rose. Between 2023 and 2025 underlying NII actually fell £130m. Second, underlying profit is lower in 2025 (£6,777m) than in 2021 (£8,040m), because 2021 flattered itself with a £1,207m impairment release and because costs have since risen 28%.

Volume growth is modest and mix-driven. Loans went £440bn (2020) → £481.1bn (2025), about 1.8% a year — below nominal GDP. Deposits went £451bn → £496.5bn. Inside that, the growth is in consumer credit (cards +10%, unsecured +15%, motor +7%) and Retail Europe (+21%), offset by a shrinking SME book (−5%).

Organic vs. acquired: overwhelmingly organic. The acquisitions of the period — Embark, Tusker, Cavendish Online, the residual 49.9% of Schroders Personal Wealth, and the agreed purchase of Curve — are small, bolt-on and capability-driven rather than scale-driven. There is no roll-up here; nor is there a large integration risk.

5.2 Other income: flat once the leasing gross-up is netted

The headline “other income +9%” is gross of an operating-lease depreciation charge that itself grew 10%. Netting it:

Net non-interest income in 2025 is only 1.4% above 2021. The two-year trend (2023→2025, +12%) is genuinely positive and reflects real progress in protection, home insurance and wealth; the five-year trend is flat. Operating-lease depreciation itself is worth watching as an earnings risk: it rose from £373m (2022) to £1,454m (2025) on fleet growth, higher-value vehicles and declining used electric-car prices — a residual-value exposure that already carries £243m of provisions inside the motor book.

5.3 Where forward growth is supposed to come from

1. The structural hedge (the dominant contributor). +£1.5bn in 2026 and +£1.0bn in 2027, ~90–95% and ~80% contracted respectively. This is the growth. It is high-confidence, it requires no execution, and it stops.
2. Mortgage-spread normalisation. The ~30bps destroyed on each roll should stop being a headwind “certainly by the first half of 2027” per management. That converts a negative into a neutral, not into growth.
3. Strategic-initiative revenue. £1.4bn annualised delivered by 2025 against an original c.£1.5bn-by-2026 target — subsequently raised to c.£2bn by end-2026. The direction of travel is credible and the target was upgraded rather than missed, which is a genuine mark in management’s favour.
4. Capital-light fee businesses. IP&I open-book AuA £232bn (+15%), total AuA £279.6bn (+13%), net flows £7.9bn (+39%); protection share 5.8% → 7.8%; >14% of new home-insurance policies; Lloyds Premier for the mass affluent (roughly twice the relationship depth). This is the most attractive growth vector in the business — genuinely capital-light and genuinely under-penetrated relative to a 26–28m customer base — but it is small: IP&I is 4.9% of group underlying profit.
5. Consumer credit. Cards and unsecured are growing 10–15% off a small base at high margins. This adds margin and credit risk into a management base case that assumes 5.6% unemployment.
6. Basel 3.1 (January 2027). A £6–8bn RWA reduction is not revenue growth, but it releases ~£0.8–1.0bn of CET1 and therefore supports distributions.

5.4 What is not going to grow

UK GDP is running ~0.5–1.4%; the mortgage market is a refinancing market growing ~4% in gross terms with 87% of it product transfers; the SME book is shrinking; and the deposit market is one Lloyds is actively withdrawing from at the margin. There is no volume story here. Nor is there a geographic one — Retail Europe (+21%) is a Dutch/German mortgage book of £20.4bn, useful but immaterial to the group return.

Verdict — low-quality growth: mechanically-sourced, finite, and not franchise-driven

The growth on offer over 2026–27 is high-confidence and largely contracted, which is worth something. But it is not the kind of growth that deserves a growth multiple: it is a treasury reinvestment effect with a defined endpoint around 2029, layered on a loan book compounding below nominal GDP, in a market where the flagship product loses 30bps on every roll. The genuinely attractive growth — capital-light insurance, protection and wealth — is real, is being executed competently, and is too small (4.9% of profit) to change the group’s character within the forecast horizon. Strip the hedge and this is a low-single-digit-revenue-growth business. That is a perfectly respectable thing to be; it is not a thing to pay ~2x tangible book for.

6. Financial Quality

6.1 The RoTE ladder, and what is actually inside it

Lloyds presents three return figures and the market has anchored on the highest: 12.9% statutory FY2025, 14.8% excluding the motor charge, and 17.0% in Q1-2026. All three overstate the clean run-rate. Three adjustments matter:

(a) Impairment is flattered by ~8bp. FY2025 underlying impairment was £795m, an asset-quality ratio of 17bp. Management’s own words: “The asset quality ratio excluding the model and debt sale benefits is considered to be closer to 25 basis points, both for the full year and the fourth quarter. The Group expects the asset quality ratio to be c.25 basis points in 2026.” On £481.1bn of loans, the 8bp gap is roughly £385m pre-tax (~£280m post-tax) of non-recurring benefit. Q1-2026’s AQR was already 25bp — i.e. the benefit has already gone.

(b) Q1-2026 carries a ~1.2pp tax benefit. The Q1 effective tax rate was 23.2% (£470m on £2,025m) against ~27% medium-term guidance. Normalising to 27% cuts attributable profit from £1,413m to ~£1,336m and the reported 17.0% RoTE to ~15.8%.

© 38% of the 2025 PBT increase was a swing in “volatility and other items.” The line moved from −£332m (2024) to −£70m (2025), a £262m favourable swing against a £690m total increase in statutory PBT. The release attributes it to “the gain on sale of the Group’s bulk annuities portfolio to Rothesay Life plc … and the gain following the full acquisition of Schroders Personal Wealth.” Those are disposal and bargain-purchase gains — non-recurring by definition.

Normalised FY2025 RoTE build (INTERPRETATION):

Normalised FY2025 EPS is roughly 8.0p per ordinary share (32.0p per ADS) against 7.0p reported. The honest characterisation is: a bank earning a clean ~13.5–14% RoTE, guiding to >16% on a contracted hedge tailwind, that printed 17.0% in a single tax-flattered quarter.

6.2 Costs — and the risk hiding inside the 2026 guide

Operating costs rose 28.7% between 2020 and 2025 (£7,585m → £9,761m, ~5.2% a year) against net income up 27.1%. Positive jaws appeared only in 2025 (income +7%, opex +3%). FY2025 cost:income was 58.6% — 53.3% excluding remediation. Gross cost savings of £1.9bn have been delivered since 2021.

RISK FLAG. The 2026 guide is cost:income <50% with operating costs <£9.9bn. Getting from 58.6% to <50% in a single year requires income of roughly £19.9bn and total costs including remediation below ~£9.95bn — which implicitly assumes remediation of roughly £0.1bn or less, against £899m (2024) and £968m (2025). Q1-2026 remediation was £11m, so the run-rate is consistent so far. But the <50% target has essentially zero room for another conduct charge, and roughly half the improvement comes from the denominator (income) rather than the numerator.

6.3 Credit quality: genuinely benign, with one open question

FY2025 impairment of £795m splits into a £721m pre-MES charge (Retail £734m, Commercial a £14m credit) and a £74m updated-MES charge — against a £394m MES credit in 2024. The ECL allowance fell 8% to £3,353m underlying.

Stage mix at 31 December 2025 (statutory): Stage 1 £430,193m; Stage 2 £42,679m (8.8%); Stage 3 £6,526m (1.3%); POCI £5,076m. Coverage: Stage 1 0.2%, Stage 2 2.7%, Stage 3 15.9%, total 0.7%. By book: UK mortgages 0.2%; credit cards 3.4% (Stage 3 coverage 44.2%); UK unsecured 4.1% (Stage 3 58.0%); UK motor 2.5% (Stage 3 56.0%, including £243m of residual-value provisions); Commercial 1.0%. At 31 March 2026 both Stage 2 (£41,444m, 8.5%) and Stage 3 (£6,463m, 1.3%) improved quarter-on-quarter. Credit is benign and improving, not deteriorating.

ECL by scenario is essentially unchanged across the two most recent balance-sheet dates (31-Dec-25 probability-weighted £3,228m vs 31-Mar-26 £3,223m; severe downside £5,275m vs £5,263m). Sensitivity: a 1pp rise in unemployment phased over ten quarters adds £148m of ECL (2024: £203m). Downside scenario: unemployment peaks 7.8%, house prices −3.3%/−5.8%; severe downside: unemployment 10.5%, house prices −7.6%/−12.6%.

OPEN QUESTION — is the book slightly under-reserved? The allowance was held effectively flat between December 2025 and March 2026 despite a visibly worse base case (exit-2026 unemployment 5.6% vs 5.1%; 2026 house-price growth +0.7% vs +1.6%). Management offset the deterioration with the release of a £50m tariff/geopolitical overlay and quarterly model calibrations. That is defensible — judgemental adjustments totalled +£242m at year-end against −£15m a year earlier — but the direction of the offset deserves scrutiny at the H1 print.

6.4 Balance sheet, funding and liquidity

Total assets £944,072m (Dec-25) → £968,125m (Mar-26); total equity £47,867m → £48,231m, of which ordinary shareholders’ equity £41.8bn, AT1 £5.9bn and non-controlling interests £0.2bn at December. Loan-to-deposit ratio 97–98%, which is conservative and self-funded.

Funding shows the one genuine soft spot. Wholesale funding rose to £99.4bn (2024: £92.5bn), with sub-one-year wholesale up 18% to £37.0bn and money-market funding under one year up 57% to £26.6bn following the repayment of £13.1bn of Bank of England TFSME drawings. Debt securities in issue rose from £78,271m to £91,884m on £13.6bn of Q1 issuance. LCR 145% (2024: 146%) and NSFR 124% — but the NSFR fell 5pp year-on-year. None of this is alarming for a bank with £62bn of cash at central banks and a 98% loan-to-deposit ratio, but the mix of short-dated wholesale funding is drifting the wrong way and should be tracked.

6.5 Capital: the surplus story is over

The FY2025 capital-generation bridge, in basis points from 13.5% pro forma: banking build +228; insurance dividend +9; RWAs −54; other +14; Retail secured CRD IV −19 = +178 before motor; motor provision −31 ⇒ 147bp generated. Ordinary dividend −97 and buyback accrual −79 ⇒ 13.2% pro forma.

The requirement stack is Pillar 1 4.5%, Pillar 2A ~1.4% (reduced by the PRA in Q3-2025), CCyB ~1.8%, CCB 2.5% and O-SII 1.6% at group level — roughly 11.8% disclosed, plus a confidential PRA buffer. Headroom at 13.4% is ~160bp (~£3.9bn); at the 13.0% target ~120bp (~£2.9bn) before the PRA buffer.

INTERPRETATION — the excess-capital story is finished. Absolute CET1 capital is lower than four years ago (17.3% × £196bn = £33.9bn in 2021 vs 14.0% × £235.5bn = £33.0bn in 2025) while RWAs grew 20%. From 13.2% pro forma to the 13.0% target there is ~20bp of surplus left. All future distributions must come from run-rate generation (>200bp ≈ £4.8bn on £240bn of RWAs) less RWA growth (2025’s £10.9bn absorbed ~£1.4bn). Sustainable distribution capacity is roughly £3.4–4.0bn a year — about 5.1–6.0% of market capitalisation — not the 5.8% delivered in 2025, which included surplus release. Basel 3.1’s £6–8bn RWA reduction in January 2027 adds a one-off ~£0.8–1.0bn.

6.6 The number that indicts the last four years: TNAV per share

*IFRS 17 restated.

Tangible net asset value per share is +0.7% over four and a quarter years — 57.5p at December 2021 to 57.9p at March 2026 — despite roughly £9.45bn of buybacks that cut the share count 17%. The 2022 cash-flow-hedge-reserve hit, the IFRS 17 restatement, a near-100% payout ratio and £1.95bn of motor provisions consumed the retained earnings. Including dividends, 57.5p plus cumulative 2021–25 DPS of 13.98p reaches 71.9p — +25.0% over four years, about 5.7% a year of total book-value return, against a 12–15% reported RoTE. The gap is the payout ratio: essentially all earnings were distributed rather than retained.

This is the single most important framing correction in the report. A bank earning a mid-teens RoTE that pays out essentially all of it does not compound book value; it converts earnings into a dividend-and-buyback stream. That is a legitimate model, and for an income holder it may be exactly the right one — but it must be valued as a yield instrument, not as a compounder. TNAV/ADS is currently 231.6p = US$3.08 at GBP/USD 1.3315.

6.7 The insurance arm (Scottish Widows): small, volatile, and quietly weakening at the margin

FY2025 IP&I underlying profit was £330m (+50%) — but only £303m (+38%) excluding the full acquisition of Schroders Personal Wealth, and the division is just 4.9% of group underlying profit (2024: 3.5%). The recent history is volatile within a narrow band: £281m (2021), £391m (2022), £220m (2024), £330m (2025).

IFRS 17 quality flag. Deferred profits (contractual service margin plus risk adjustment) were £5.2bn at December 2025 after a £413m release to income, of which only £93m was added from new business (2024: £126m, −26%). New business value recognised was £80m (2024: £111m, −28%), including −£13m of losses on initial recognition. New business is replacing only about 23% of the CSM being released — the CSM balance grew via experience and assumption effects and the SPW acquisition, not underwriting. Life and pensions PVNBP sales rose 15% to £21,047m and the general-insurance combined ratio improved 8pp to 89%, which is real but cyclical.

Open question: the insurance Solvency II ratio fell from 158% to 144% pre-dividend between December 2024 and December 2025 (regulatory view 148% → 140%) on only £200m of dividends up to Group. That 14pp decline is not explained in the results release and should be asked about.

Verdict — economics that are good, improving, and materially less good than the headline

Do the economics improve with scale? Only partially, and less than they should. The positives are real: benign and improving credit; a self-funded 98% loan-to-deposit balance sheet; positive jaws in 2025 for the first time in years; £1.9bn of delivered cost savings; a capital-generative model producing >200bp a year; and a franchise throwing off £3.4–4.0bn of sustainable annual distributions. The negatives are equally real: a clean RoTE nearer 13.5–14% than the 17% headline; a cost:income ratio ten points worse than the best domestic peer despite the largest scale; an excess-capital cushion that is spent; short-dated wholesale funding drifting up; an insurance arm whose new business is replacing less than a quarter of the profit it releases; and — decisively — four and a quarter years of essentially zero tangible book-value-per-share growth. The business generates cash well and retains almost none of it. Value it accordingly.

7. Capital Allocation

7.1 The distribution record — genuinely good, and the best thing management has done

Four consecutive years of ≥15% dividend growth; a 16.2% DPS CAGR from 2021; the pre-COVID 3.21p peak reclaimed in 2024. £18.3bn returned in five years, roughly half via buyback. The FY2019 final dividend was cancelled at the PRA’s request in March 2020 and the FY2020 final of 0.57p was “the maximum allowed under the PRA’s temporary framework” — history that matters, because it establishes that the distribution is regulator-permissioned, not contractual.

Current policy, verbatim: “a progressive and sustainable ordinary dividend policy whilst maintaining the flexibility to return further surplus capital through share buybacks or special dividends.” One genuine improvement announced at FY2025: the Board will now review excess distributions every half year rather than annually — the 30 July 2026 print is the first test, with consensus expecting a ~£1.1bn interim buyback.

Yields at 24 July 2026: dividend 3.22%; FY2025 buyback 2.65%; total shareholder yield ~5.9%. Payout on FY2025: dividend alone 52.1% of EPS; including buyback, £3.91bn against £4,196m attributable = 93.2%. Lloyds distributes essentially everything it earns.

The killer statistic, however, is that the absolute return has been flat at £3.4–3.9bn for five years while the market capitalisation roughly trebled. Lloyds’ own framing tracks the collapse: “c.14% of market cap” (FY2023) → “c.9%” (FY2024) → “c.6%” (FY2025). That is a pure denominator effect. The distribution has not grown; the price has. An investor buying today is buying a 5.9% yield where a 2023 buyer bought 14%.

7.2 The buyback: price-insensitive by design, and now destroying tangible book

The 2026 programme was 74.3% deployed (£1,300.3m of £1.75bn) as at 24 July 2026, having retired 1,307,854,149 shares. Monthly VWAPs run January 108.10p, February 104.02p, March 94.77p (445m shares — the largest and cheapest month), April 97.99p, May 97.00p, June 103.31p, July-to-24th 112.63p.

Lloyds has run a mechanical ~£2bn-a-year buyback straight through a 2.5x re-rating without ever changing the size or the framing. At a 99.43p average, each share retired costs 99.4p to acquire 57.0p of tangible book — roughly 42p of TNAV surrendered per share. This is arithmetic, not opinion. It is defensible only if the >16% guided RoTE proves durable, in which case a justified P/TNAV of ~1.8–2.0x is arguable. But Lloyds has never conditioned buyback size on P/TNAV, never offered a special dividend as the alternative, and never disclosed a price at which it would stop. It is a CET1-drawdown tool executed in the market, not a capital-allocation decision.

Worse, the FY2025 release attributes part of the 4.6p rise in TNAV/share to “a reduction in the number of shares following the share buyback programme.” At 1.47–1.74x book that attribution is wrong-signed — the buyback subtracted from TNAV/share in that period, it did not add to it.

The one genuine piece of counter-evidence, and it deserves credit: Lloyds bought most heavily in the March-2026 drawdown, retiring 445m shares at the programme’s cheapest 94.77p. Whether that was opportunism or calendar coincidence is not disclosed.

7.3 The dilution leak nobody discloses

Shares in issue excluding own shares: 70,996m (Dec-21) → 66,944m → 63,508m → 60,491m → 58,799m (Dec-25) → 58,268m (30 June 2026).

• Gross shares repurchased across all five programmes: c.16.13bn
• Net reduction Dec-2021 → Jun-2026: −12,754m, or −17.96%
• Implied re-issuance to employee share plans: c.3.4bn shares — roughly 21% of all buyback volume.

The mechanism is documented: Lloyds files “Notification of Issuance of Equity Securities… to satisfy awards under the Company’s share plans,” and total voting rights rose in January 2026 (58,886m → 59,008m) and again to 59,029m by 27 February 2026 while the buyback was running. Roughly £1 in every £5 of buyback spend offsets employee dilution rather than shrinking the base. This net figure is disclosed nowhere and must be reconstructed from monthly voting-rights filings — a disclosure-quality criticism as much as a capital-allocation one. Note also that shareholders approved the removal of the 5% “inner” dilution limit from discretionary share plans at the 2025 AGM (96.47% for), a quiet loosening that passed without comment.

7.4 Capital: the glide path is nearly exhausted

The Board’s stated target, verbatim from the Q1-2026 IMS: “The Board’s view of the ongoing level of total CET1 capital required… remains c.13.0%. This includes a management buffer of c.1%. The Board intends to pay down to the CET1 capital target of c.13.0% by the end of 2026.”

Stock surplus: 13.4% against c.13.0% on £240.8bn of RWAs = only ~40bp, or ~£1.0bn, of surplus left. Flow: >200bp ≈ £4.9bn gross, less an ordinary dividend of ~£2.4bn ⇒ ~£3.5bn of theoretical 2026 buyback capacity against £1.75bn announced — consistent with consensus of ~£3.0bn of 2026 buybacks. Post-2026, distributions are capped by generation less RWA growth, and RWAs grew £10.9bn in 2025 and £5.3bn in Q1-2026 alone. The swing factor is Basel 3.1’s Day-1 RWA reduction of c.£6–8bn (c.35–45bp) on 1 January 2027.

And the hardest target was missed twice. Capital generation guidance of c.175bp for 2024 produced 148bp actual; 2025 produced 147bp. Only on an ex-motor basis (177bp and 178bp) does the company come close — and the company now reports the metric primarily on that ex-motor basis. The live guide is >200bp for 2026.

7.5 M&A: small, cheap, and a persistent failure to buy what it actually needs

Roughly £820m of disclosed cash consideration across five years against £18.3bn returned — the right ratio, and no ego deal. But three criticisms are real.

First, Curve was bought out of a litigated process and has not closed. IDC Ventures, Curve’s largest external shareholder at ~12%, petitioned the High Court on 21 November 2025 — two days after the Lloyds announcement — alleging “intentional concealment of material information,” breaches of contract and directors’ duties, and claiming the process “wiped out more than £670m in potential value.” The £120m price is roughly half the capital Curve raised over a decade. The sum is immaterial to a bank distributing £3.9bn a year; the signal is not, for a company whose digital narrative rests on execution credibility and which cited Curve in the CEO’s FY2025 statement as evidence of digital leadership.

Second, Lloyds keeps reaching for the thing it is structurally short — non-interest income — and keeps losing or picking wrong. It sat out TSB (correctly; competition clearance was implausible at ~20% mortgage and current-account share). It bid for Aegon UK and lost — Standard Life/Phoenix won at £2.0bn in April 2026, taking 3.8m customers and £160bn of AUM. It is now exploring a bid for Aldermore, which FirstRand is selling because of the motor-finance scandal, with Metro Bank reported at ~£2bn against RBC’s £1.35bn central valuation. Lloyds’ own house broker publicly questioned the logic, arguing the deal “would add scale rather than new capabilities,” that the main benefit is cheaper funding, and that it would do “little to reduce Lloyds’ dependence on net interest income” — recommending wealth management instead. Buying a conduct-tainted specialist lender, requiring a motor-finance indemnity, on top of Lloyds’ own £1.95bn motor provision, while trading at 1.96x tangible book, would be a textbook late-cycle mistake.

Third — and least scrutinised — Lloyds Living. Launched as Citra Living in 2021, it is now one of the UK’s largest private landlords: c.8,850 homes, portfolio value past £2bn, occupancy 95%, and a stated ambition of 50,000 rental homes by 2030, with 980 homes contracted from Barratt Redrow and ~1,000 more added in July 2026. It sits inside “Equity Investments and Central Items” — a division whose net income fell 6% in 2025 to £707m — and receives roughly four sentences in an 80-page results release, with no disclosed return metrics. A ring-fenced, deposit-funded bank is building a large, levered, illiquid, politically-exposed, RWA-consuming residential property book while paying out 93% of earnings. Those are two contradictory statements about its own cost of capital.

One further accounting note: roughly half the “£3–4bn strategic investment” is capitalised, not expensed — the cash flow statement shows “Purchase of other intangible assets £1,252m (2025), £1,259m (2024)”, with goodwill and intangibles rising to £8,593m and the CET1 deduction to £5,996m. The programme therefore flatters cost:income and RoTE (via a smaller tangible-equity denominator) while consuming CET1 through the deduction.

7.6 Scoring management against its own February-2022 targets

A genuinely mixed but net-respectable scorecard: two clear beats, two misses, two on track. The honest caveat that governs it all is the industry point made earlier — the dominant earnings driver is the structural hedge, a rate-cycle windfall available to every UK retail bank, not a Lloyds execution win.

7.7 Remuneration: the structure improved and the quantum ratcheted

Charlie Nunn’s single-figure pay: £3.681m (2023) → £6.169m (2024, restated) → £7.407m (2025). That is +101% in two years and the highest single figure for a Lloyds CEO in the entire ten-year disclosure table. The CEO-to-median-UK-colleague pay ratio moved 80:1 → 114:1 → 141:1 (+76% in two years) against a median colleague on £52,638. William Chalmers reached £4.976m.

The plan that actually paid was not a performance plan. One hundred per cent of the long-term incentive in the 2024 and 2025 single figures came from the Long Term Share Plan — a restricted share plan whose vesting turns on three pass/fail underpins (CET1 above target; RoTE above the UK peer average; a rising dividend) that a capital-generative UK bank clears almost by construction. In 2024 the RoTE underpin was cleared by 70bp. The LTSP has vested at 100% of maximum three cycles running. By contrast, no LTIP — the genuinely performance-tested plan introduced with the 2024 grants — has ever vested; the first cycle completes on 31 December 2026 and pays in 2027. Separately, Lloyds itself discloses that £1,272k of Nunn’s £2,976k of 2025 long-term incentive (43%) was pure share-price appreciation, and the Committee considered and rejected a windfall adjustment despite ~75% share-price growth over the period.

The 2026 policy, approved at the 14 May 2026 AGM with 94.97% for. The framing is a 44% cut in fixed pay — true, and the smaller number. The same policy:

800% variable against ~110% fixed is roughly 7.3:1 — Lloyds has now consumed most of the 8:1 headroom it obtained in 2024, two years after telling shareholders the executive opportunity “cannot be increased as a consequence.” The structure is genuinely better (financial weighting 50% → 75%, relative TSR 20% → 30% against eighteen European banks, the RoTE range stretched to 15–18% against 12.9% statutory in 2025, and the soft Committee-judged “strategic delivery” block cut from 35% to 15%). The quantum is much larger. Both are true; only the first leads the narrative.

Accountability for motor finance is thin and getting thinner. In 2025 the motor charge was excluded by design from the bonus financial metrics; the Committee then applied a discretionary −7.9pp haircut, worth £152,117 to Nunn — 2.05% of a £7.4m package — against an £800m in-year charge and £1.95bn cumulative. In the same year the LTSP was explicitly ruled immune from motor finance and paid him £2.98m. No malus or clawback has ever been applied to a Lloyds executive director, through DCA/motor finance, HBOS Reading, or the wider conduct history. From 2026, all “exceptional items” are excluded from the bonus and LTIP metrics entirely, replacing a mechanical channel with pure Committee judgement.

AGM votes show no revolt in any year — the worst pay vote in four years is 5.77% against. The one telling datapoint is the 2025 vote on Cathy Turner, the Remuneration Committee chair, at 3.95% against, the highest of any director that year — the conventional UK channel for pay displeasure — registered immediately after the January-2025 fixed-pay reset, and reverting to 0.71% in 2026.

Board: Sir Robin Budenberg remains Group Chair (re-elected 14 May 2026, 99.05% for), with no successor announced; his nine-year limit falls in October 2029, and his fee rose from £654,500 (2024) to £750,000 (2025) to £850,000 (2026). Nunn and Chalmers are both in post — an unusually stable pair. Two of four customer-facing divisional CEOs were replaced with external hires in early 2026 (Amanda Murphy from HSBC; John Langley from Wells Fargo), both on fresh multi-million LTIP grants immediately ahead of the strategy launch. Danuta Gray joined as a NED and Chair of Scottish Widows Group on 1 July 2026 — and was a NED of Aldermore Bank from 2014 to 2021, including Senior Independent Director and Interim Chair, which is either coincidence or diligence.

7.8 Insider dealing and ownership: no bullish signal, and one mildly negative one

The complete population was examined — all 87 relevant SEC filings for 1 January 2024 to 24 July 2026, yielding 297 individual MAR Article 19 notification blocks. Stripping out the mechanical categories (monthly Share Incentive Plan payroll deductions, quarterly Fixed Share Awards which are salary delivered in shares net of tax, SAYE option exercises at 24–39p strikes, dividend reinvestment, and nil-cost LTIP grants) leaves:

Three genuine open-market cash purchases in thirty-one months, totalling ~£521,000, none by an executive director:

Against £15.41m of discretionary on-market selling — 17,786,028 shares across 39 disposal lines by nine executives — and that selling is accelerating in both size and realised price:

Sellers include the Chief Risk Officer (2,250,000 shares, £1.58m), the Group COO, the CEO of IP&I, the Chief Legal Officer and the Chief People Officer. There is no UK analogue to a Rule 10b5-1 plan here — every one of these is a discretionary, on-market disposal. Three of the sellers are described in their own filings as merely “on track to achieve” the Group’s shareholding policy — i.e. selling while still below their required holding — materially weaker language than the “continues to comply” used for others.

The buy-to-sell ratio is roughly 1:30 by value. Nobody at Lloyds has ever bought a share above 84p. Budenberg bought roughly every 12–24 months whenever the stock sat between 28p and 47p, accumulating 2.5m shares for £1.06m — and has bought nothing since 23 February 2024, through a period in which the stock roughly doubled from 45p to ~113p. He is not leaving; he was re-elected in May 2026. The absence of a fifth purchase is itself the datapoint.

Nunn and Chalmers are neutral by revealed preference. Both hold large stakes — Nunn 10.1m shares outright (656% of salary), Chalmers 10.7m (1,128%) — but every share arrived via remuneration. Nunn has never been the subject of a discretionary purchase or sale notification in his entire tenure. High absolute alignment; zero incremental conviction. (Honest counter-argument: these are windfall sales out of very large nil-cost award stacks by people paid overwhelmingly in stock, and diversification is rational. But the symmetry test fails — if diversification explains the selling, nothing explains the complete absence of buying by the same cohort at any price.)

One governance number is indefensible. Nathan Bostock — the board’s most experienced commercial banker, former CEO of Santander UK, former CFO and CRO of RBS, and Chair of Lloyds Bank Corporate Markets — owns 430 ordinary shares, worth roughly £450, against £359,000 of annual fees. Lloyds imposes no shareholding requirement on non-executive directors at all; only executives have one. That is a genuine governance gap.

On the register, the single most important ownership event of the period. The FY2025 Annual Report states, verbatim: “It is understood that Harris Associates L.P. disposed of their holding during the course of 2025.” Harris (Oakmark; David Herro) had held 3,546,216,787 shares — 4.99% of the company — and had been a top-five holder since the mid-2010s. It was the register’s only large, active, non-index, value-orientated shareholder, and it sold out completely into the 2025 rally. That removes the one holder whose presence was itself a valuation opinion, leaving a register that is essentially passive and index-dominated (BlackRock’s DTR-notified 5.14% has been unchanged since 2015; Norges 3.02%; no state stake since 2017; free float effectively 100%; insider ownership ~0.11%). The blunt read: the long-standing value holder rang the register at 65–95p while the executive committee sold into the same tape and the Chair stopped buying.

(Methodological caveat, and it matters: Lloyds published exactly two TR-1s in the whole thirty-one-month window, both from Norges in March 2024. With 58.9bn shares out, 1% is ~589m shares, and investment-manager aggregation exemptions let large holders report on a delayed or exempt basis. The Harris exit had to be disclosed by the company in prose because no TR-1 was ever filed for it. Do not infer ownership stability from the absence of TR-1s — the substantial-shareholder table is close to useless as a live signal.)

Verdict — a qualified pass with two real criticisms

Good, and it should be said clearly: £18.3bn returned in five years; a 17.96% net reduction in the share count; four consecutive dividend rises of ≥15%; a CET1 glide path actually executed; £24bn of RWA optimisation since 2021; a move to semi-annual excess-capital reviews; 178bp of ex-motor capital generation; no ego M&A, and in Schroders Personal Wealth one genuinely excellent deal that cost no cash at all. Management has also beaten two of its six 2022 targets and delivered £1.9bn of cost savings. This is not a capital-destroying board.

Criticism one: the buyback is price-insensitive by design. Run at 0.77x book and at 1.74x book with identical size and identical framing, mis-described in the FY2025 release as accretive to TNAV/share when at those prices it is the opposite, with no disclosed hurdle and no special-dividend alternative. Roughly 21% of the spend leaks straight back out to employee share plans.

Criticism two: capital leaks in contradictory directions. A bank distributing 93% of earnings and telling shareholders it has no better use for the money is simultaneously building a £2bn-plus build-to-rent portfolio toward a 50,000-home ambition, capitalising ~£1.25bn a year of software, bidding for a wealth manager it lost and a specialist lender its own broker says adds nothing, and buying a fintech out of a High Court dispute that eight months later has not closed. Those are two irreconcilable statements about its own cost of capital.

And the insider file offers no support for the current price. Three purchases totalling £521k against £15.4m of accelerating executive selling, no executive-director buying ever, no NED shareholding requirement, and the register’s only large active value holder gone.

8. Changes and Headwinds — Last Two Years

8.1 Motor finance / discretionary commission arrangements — the dominant story

Where Lloyds sits. Black Horse is the UK’s largest motor lender by book (£17.4bn at Q1-2026, +7% y/y). Lloyds’ £1.95bn is ~21% of the £9.1bn industry estimate — a plausible market-leader share, with no evidence of under-provisioning relative to peers (Barclays’ comparable Q1-2026 top-up was £105m; FirstRand raised its motor provisions to £750m and is exiting the UK entirely).

What management actually claims, verbatim (Q3-2025 call): “£1.95 billion… represents our best estimate… the FCA proposals, as currently proposed, represent the heaviest weighting in our overall scenario analysis… i.e., all DCAs, most of the commission… gets handed back, and it is a very high response rate… even if the FCA proposals come out exactly as they are today, then our overall position is not going to move by that much. So we are not far off then, in short.” Chalmers refused to disclose the scenario weightings or any numerical sensitivity across three separate attempts by two analysts.

INTERPRETATION. Lloyds carries the largest motor provision of any UK lender and has explicitly bought certainty at the adverse end of its own modelled range — a defensible, even conservative, choice. But “we can’t be far off” rests entirely on undisclosed scenario weights and an FCA-assumed 75% opt-in rate that is now untested and untestable: no communications have gone out, so no response-rate data can exist. The 2 July suspension defers the cash outflow into 2027 and beyond — and defers the moment the provision can be marked against reality. Residual exposures sit outside the scheme too: >30,000 claimants seeking ~£66m (reported 27 March 2026), a stayed Competition Appeal Tribunal claim, and an April-2026 Court of Appeal listing on whether an omnibus claim form can carry multiple unfair-relationship claims. The key near-term disclosure is the first opt-in data — potentially visible at the 30 July 2026 results.

8.2 Tax and regulation — three favourable outcomes, none of them earned

The November 2025 Budget. A restoration of the bank surcharge to 8% (~£2bn a year) was widely expected; a QE levy was the other tipped measure. Neither appeared. Banks continue to pay 25% corporation tax plus a 3% surcharge above a £100m allowance (28% effective) and a balance-sheet levy of up to 0.1%, together raising ~£2.5bn. LYG rose 4.3% on 25 November and 3.3% on 26 November. Asked directly about bank taxation on the Q3-2025 call, Chalmers deflected entirely — “those are really decisions for the government” — asking only for “a stable and a predictable tax regime.” As the most UK-domestic large bank, Lloyds has the highest beta to this, and a Treasury surcharge review remains live. This is a recurring annual binary, not a settled question.

Basel 3.1 was delayed by the PRA to 1 January 2027 (FRTB-IMA to 2028), with near-final rules read as softened for mortgages and SMEs, and Lloyds guiding a Day-1 RWA reduction of c.£6–8bn. (A note of scepticism: much of that relief comes from moving off foundation IRB in Commercial. Chalmers repeatedly caveated the related CRD IV model outcomes as “subject to PRA approval and therefore, of course, risk of modification.” Denominator relief is not the same thing as de-risking.)

Ring-fencing reform conclusions were published on 18 May 2026, raising the core-deposit threshold from £25bn to £35bn, with implementing consultations launched in July 2026 and a PRA/FPC review of the interaction with the Basel 3.1 output floor. The FCA growth agenda is running in parallel: Consumer Duty simplification, and a Mortgage Rule Review (CP26/18, open to 28 July 2026) removing the advised-sale interaction trigger and full affordability assessment for term reductions and lightening remortgage assessments. Lloyds is the single largest beneficiary of mortgage-rule liberalisation — UK mortgages are £323.1bn of a £481.1bn book.

8.3 Operational, restructuring and the March-2026 incident

On 12 March 2026 an IT fault let customers see one another’s transactions (−3.9% on the day). The Treasury Committee demanded an explanation on 17 March and disclosed on 27 March that up to 447,936 customers’ data was exposed, including account details and National Insurance numbers. No ICO or FCA sanction had been announced as at 24 July 2026 — an unquantified open exposure, occurring mid-migration and mid-offshoring.

Restructuring continues at pace: ~136 branches closed and ~1,600 roles cut in the 2025 programme, ~300 IT roles offshored to Hyderabad, ~233 further branch closures during 2026 plus 12–13 in early 2027 (roughly 1,600 branches closed since 2015, more than any UK network), offset by 1,300 new digital roles, >1,000 AI roles, and 300 more AI/technology hires announced in June 2026. On 3 July 2026 Lloyds announced the retirement of the 173-year-old Halifax brand (discussed under Competitive Position). AI is being pushed hard — ~50 live GenAI use cases delivering ~£50m in 2025, targeting >£100m in 2026, and on 21 April 2026 Lloyds became the first UK lender with an AI investment-guidance tool, in the FCA’s second AI Live Testing cohort.

Credit ratings improved: Fitch upgraded Lloyds subsidiaries to ‘AA’ from ‘AA−’ on 12 May 2026, with the resolution debt buffer at 18.3% of RWAs.

The legacy conduct tail has not closed. The remediation provision balance stood at £2,276m at December 2025 (2024: £1,600m), and remediation charges of £899m (2024) and £968m (2025) each ran at ~5% of net income. HBOS Reading compensation runs up to £3m per victim for a total of ~£600m, largely settled through the Foskett Panel. And the Dobbs Review — commissioned in April 2017 to determine what Lloyds knew at the time of the HBOS acquisition, whether it reported properly, and whether there was a cover-up — remains unpublished after nine years (the House of Lords was told on 26 March 2026 that drafting is under way and findings will be shared with the FCA). Low likelihood, high impact, and entirely absent from consensus.

8.4 What management will not discuss

Across the FY2025, Q3-2025 and H1-2025 calls, management deflected: anything beyond 2026 (four analysts, all routed to July — Nunn, verbatim: “You’re close to getting us to talk about beyond 2026, which we are vehemently against because that will be July”); motor scenario weightings (three refusals); the payout ratio; whether Basel 3.1 relief gets distributed (“far too early”); the 2026 cost number (“I won’t kind of confirm or deny”); the hedge maturity schedule (refused outright); wealth M&A; and bank taxation. On the hedge, by contrast, management is expansive and precise.

INTERPRETATION: management is voluble about the mechanically visible and evasive about everything requiring a forward judgement. That asymmetry is itself evidence about where the confidence genuinely lies — and it is a further reason to treat the 30 July strategy update as the real information event.

Verdict — net strengthening, but the strengthening is exogenous

The three biggest positives of the last two years — the Supreme Court rejecting fiduciary duty, the FCA scheme landing at £9.1bn rather than £18bn-plus, and the Chancellor sparing banks in November 2025 — were binary judicial and political outcomes, not management achievements. Add a delayed and softened Basel 3.1, loosening ring-fencing and a mortgage-friendly FCA, and the bulk of the 2024–26 re-rating is the removal of tail risk, not the emergence of competitive advantage.

The endogenous positives are real but narrower: hedge-driven NIM expansion, 5% loan growth, cost discipline, £1.4bn of strategic revenue, £1.9bn of cost savings, a Fitch upgrade to AA, and a zero-cash wealth acquisition. The negatives are £1.95bn of permanent motor-related capital loss (31bp of 2025 generation), a serious data incident with no regulatory outcome yet, net current-account switching losses at the Halifax brand, an unclosed and litigated fintech acquisition, and build-to-rent strategy creep.

Do these developments strengthen or weaken the thesis? They strengthen the business and weaken the investment case, because they have already been paid for. An investor buying today is buying after the de-risking, not before it — and buying it at 1.96x tangible book rather than the 0.90x at which the same franchise, with the same hedge and a larger unresolved conduct exposure, was available in December 2024.

9. Risk Analysis

Risk verdict. The distribution here is unusual and worth naming precisely: the near-term risks are modest and the medium-term risks are structural. Credit is benign and improving, capital is adequate, funding is self-generated, and the single largest historical overhang — motor finance — is genuinely de-risking. What the risk matrix says is that the downside is not a blow-up; it is a de-rating. Risks 1, 2 and 9 all point the same way: the earnings engine is finite, the rate environment is the only thing keeping it running, and the capital return is being executed at a price that destroys book value. A holder is far more likely to lose money here through multiple compression against flat-to-modestly-growing earnings than through a credit event.

10. Valuation Discussion — embedded expectations

No price target and no recommendation appears in this section. The single labelled exception is Claude’s Take at the top of this note.

10.1 Where the stock trades

Market data at 2026-07-24: LYG ADS $6.02; LLOY.L 113.45p; GBP/USD 1.3315. Cross-check: 113.45p × 4 × 1.3315 = $6.04 ≈ $6.02 ✓. Market capitalisation 58,799m ordinary shares × 113.45p = £66.7bn (~$88.8bn).

The re-rating in one row. Year-end price to year-end TNAV: 2021 ~0.83x · 2022 ~0.94x · 2023 ~0.93x · 2024 ~1.04x · 2025 ~1.73x · now 1.96x. On an independent basis, ROIC.ai’s (AT1-inclusive) price-to-tangible-book series reads 0.59x (2020), 0.72x, 0.86x, 0.78x, 0.90x, 1.50x (2025) — different absolute level, identical shape. The multiple has roughly doubled in eighteen months.

Own-history percentile context. The AZI valuation index at 2026-07-23 puts LYG at the ~80th percentile of its own decade on price-to-book and the 75th on P/E, with a 57.4th composite. (Methodological warning, and it matters: AZI mixes a USD price with GBP per-ADS earnings and book value, so the absolute P/E of 21.2x and P/B of 1.88x are inflated by roughly the GBP/USD rate. The percentile ranks survive because the convention is constant across history, with FX-drift noise. Discount the P/E percentile specifically, because FY2025 GAAP EPS is motor-depressed. The signal that survives: LYG’s price-to-book sits near the richest of its own decade.)

10.2 The comparable set — the sharpest fact in the file

(Barclays and HSBC figures from those companies’ FY2025 and Q1-2026 results releases.)

Lloyds trades at roughly a 30% higher price-to-tangible-book than NatWest while earning a lower return on tangible equity, with a cost:income ratio ten points worse. There is no framework in which that is explicable as a quality premium. The plausible explanations are (a) the market is pricing Lloyds’ 2026 guidance while pricing NatWest’s delivered results, (b) the motor-finance de-risking is being re-rated as an event rather than valued as an earnings stream, or © Lloyds’ higher retail-deposit gearing is being credited with more hedge duration. None of the three is a durable competitive-advantage argument; all three are timing arguments.

10.3 Embedded expectations — what the price requires

Using the standard justified-multiple identity for a bank, P/TNAV = (RoTE − g) / (CoE − g):

At an 11% cost of equity and 3% growth, today’s price requires a sustained ~18.7% return on tangible equity in perpetuity. Set that against the evidence: 2026 guidance is “>16%”; the best single quarter ever printed under this management is 17.0% (≈15.8% normalised for tax); clean FY2025 delivery was ~13.5–14%; and the five-year average RoTE is roughly 11.4%. Lloyds has never earned 18.7%. The market is not capitalising what the bank earns; it is capitalising the top of the guidance range, on the assumption it becomes permanent, in a year when roughly 8–9 points of that return is a treasury reinvestment effect with a defined endpoint.

The reverse test — what each RoTE level is worth (at 11% CoE, 3% g):

(ASSUMPTIONS, stated plainly: CoE 11% for a domestic UK bank — LYG’s measured beta is 0.94 with 28.7% twelve-month realised volatility and 18.0% idiosyncratic volatility; g 3% nominal UK. Both are judgement calls. A 10% CoE lifts every justified multiple by roughly 15%; a 12% CoE cuts it by roughly 12%. The table is a sensitivity frame, not a forecast, and carries no price target.)

10.4 Scenario analysis

Bear (~25% probability). The Bank of England eases faster than Lloyds assumes; the hedge reinvests at ~3% rather than ~4%; mortgage spread compression persists into 2028; the Upper Tribunal upholds the FCA scheme in an expansive form and a further motor top-up of £0.5–1.0bn lands; remediation returns and the <50% cost:income guide is missed. Sustained RoTE settles at 12–13%, justified P/TNAV ~1.13–1.25x, and TNAV grows only with retained earnings. This is a −35% to −45% outcome on the multiple, partially cushioned by a ~3–5% dividend.

Base (~55%). Guidance is broadly met: 2026 RoTE >16%, cost:income just under 50%, hedge income c.£7.0bn then c.£8.0bn, motor finance closes at the £1.95bn provision, Basel 3.1 releases £6–8bn of RWAs in January 2027, and distributions run £3.4–4.0bn a year. But the hedge tailwind fades from 2028 and the durable RoTE settles at 15–16%. Justified P/TNAV 1.50–1.63x, or roughly 87–94p against 113.45p — a −17% to −23% multiple adjustment, offset over three years by ~5–6% annual distributions and modest TNAV growth. Roughly flat-to-modestly-negative total return from here.

Bull (~20%). The 30 July strategy update sets a credible ≥16% through-cycle RoTE framework with cost:income in the mid-40s; the capital-light insurance, protection and wealth businesses scale enough to change the mix; the mortgage headwind turns to a tailwind in 2027 as the back book fully reprices; motor closes cleanly and releases provision; and rates settle at 3.5–4% permanently, so the hedge plateaus at £8bn rather than declining. Sustained RoTE 17–18%, justified P/TNAV 1.75–1.88x — i.e. roughly today’s price is justified, and the return is the ~6% distribution plus TNAV growth.

The asymmetry is the point: the bull case gets you approximately today’s price plus a yield; the bear case costs you a third or more. That is the shape of a fully-valued security, not a mispriced one.

10.5 What the market is underwriting correctly, and incorrectly

Correctly: that the hedge income through 2027 is real, contracted and high-confidence (~90–95% locked for 2026); that credit is benign and improving; that motor finance is de-risking, with the first non-increase in the provision since 2023 landing after final rules; that Basel 3.1 is a January-2027 capital release for Lloyds specifically; that the UK Budget spared banks; and that distributions of ~6% of market capitalisation are sustainable from run-rate generation.

Incorrectly, or at least optimistically: that a >16% RoTE is a through-cycle rather than peak-cycle number; that a treasury reinvestment effect deserves a franchise multiple; that Lloyds warrants a ~30% premium to a higher-returning, more efficient NatWest; that buybacks remain value-accretive at ~2x tangible book when they are now dilutive to TNAV/share; and that four and a quarter years of zero TNAV-per-share growth is compatible with a compounder’s rating. Above all, the market appears to be pricing the guidance rather than the delivery, six days before the guidance is due to be replaced.

11. Variant Perception

11.1 The consensus belief

Consensus holds that Lloyds is a de-risked, high-quality UK domestic bank in the middle of a durable step-change in returns: motor finance is behind it, the structural hedge underwrites a mechanical earnings bridge through 2027, cost:income falls below 50%, RoTE steps from 12.9% to >16% and stays there, and the ~6% distribution yield plus a growing book justifies a premium multiple. On this view, forecast RoTE rising from 12.9% (FY25) to ~16.7% (FY26) is a strong trajectory for a domestic lender and the market is right to pay for it in advance.

11.2 The strongest bull case

It is better than the sceptics allow. The hedge is not a trick — it is the monetisation of a genuinely valuable, genuinely durable low-cost deposit franchise that took three centuries to assemble, and it is ~90–95% contracted for 2026 and ~80% for 2027. Management has done what it said: £1.9bn of cost savings delivered, strategic-initiative revenue at £1.4bn annualised with the target raised to £2bn, direct-to-bank mortgage share up 3pp, protection share up from 5.8% to 7.8%, and a 17.0% RoTE printed in Q1. Motor finance was the great overhang and the provision held flat after final rules — the single most important positive datapoint on the issue in three years. Credit is improving quarter-on-quarter, Basel 3.1 releases capital in January 2027 while penalising trading-heavy peers, ring-fencing reform is loosening, and the November Budget removed the tax tail. Distributions run near 6% of market cap and are covered by run-rate generation. And the factor evidence says this is not a crowded trade: Momentum is zeroed in all four models, the dominant loading is DividendYield in a regime where DividendYield is the strongest-performing style factor of the last twelve months (z = +1.71), and LYG has lagged BCS, HSBC, NWG, SAN and DB over two, three and five years. If the motor discount has not fully closed, the laggard is the opportunity.

11.3 The strongest bear case

Every point above is true and none of it addresses the valuation. More than 100% of 2026 NII growth is the hedge; ex-hedge NII has fallen ~25% from its 2022 peak; net non-interest income is 1.4% above 2021; and underlying profit is lower than in 2021. The franchise is not improving — the treasury book is repricing, and it will finish repricing by ~2029. Clean RoTE is ~13.5–14%, not 17%: the AQR is 8bp flattered, Q1’s tax rate was 23.2% against 27% guidance, and 38% of the 2025 PBT increase was disposal gains. TNAV per share is +0.7% in four and a quarter years. The capital surplus is spent — ~20bp above target — and the buyback now destroys ~2.8% of TNAV per share per programme. Competitively, Lloyds is the biggest and the least efficient: 58.6% cost:income against NatWest’s 48.6%, 12.9% RoTE against 19.2%, losing the switching war to a mutual, retiring the very brand that covered the mass-market niche, and ranked behind three digital challengers on service. And it trades 30% above NatWest on tangible book. At 1.96x, the price requires ~18.7% in perpetuity from a bank whose five-year average is ~11.4%. The lifetime record — −94.8% maximum drawdown, negative twenty-year Sharpe, still 58% below the 2007 peak — is the reminder of what happens when a levered, single-country, rate-dependent balance sheet meets a cycle turn.

11.4 The 3–5 assumptions that actually matter

1. Is >16% RoTE through-cycle or peak-cycle? Everything hinges here. If through-cycle, ~1.6x tangible book is defensible and today’s price is ~20% rich. If peak-cycle, fair value is ~1.3–1.4x and today’s price is ~30% rich.
2. Does the hedge plateau or decline after ~2029? At a 3.5–4% terminal Bank Rate it plateaus near £8bn (bull); materially below, it declines and the disclosed downside sensitivity compounds.
3. Is £1.95bn enough for motor? The provision is a scenario-weighted best estimate, not a cap, and the scheme is suspended pending a Dec-2026/Feb-2027 hearing with four challengers.
4. Can cost:income hold below 50% on a flat income line? In 2026 the denominator does the work. The test is 2028, when hedge growth stops.
5. Does the 30 July strategy update reset the framework? New multi-year targets could legitimately change the analysis in either direction — and they land six days after this note.

11.5 The variant perception, stated plainly

Consensus is right about the bank and wrong about the security. The market has correctly identified that Lloyds is de-risking and that its 2026 earnings are contracted; it has incorrectly concluded that this makes Lloyds a quality compounder deserving the highest tangible-book multiple in the UK sector. The variant view is that Lloyds is a well-run, structurally-advantaged-by-regulation, low-growth domestic utility that converts a treasury position into a ~6% distribution yield, has grown book value per share by 0.7% in four years, and is the least efficient of the UK majors — and that it is currently priced as the most deserving of them. The mispricing is not in the earnings forecast; it is in the multiple applied to it, and specifically in the relative multiple versus NatWest.

The factor-positioning read supports this framing without confirming a catalyst. LYG is a re-rated income cyclical, not a crowded momentum trade and not a falling knife: price above all three EMAs with the 50 above the 200, within 3.7% of the 52-week high, a 3-year Sharpe of 1.41 — but Momentum zeroed in all four nested models, DividendYield the dominant style loading, and the nearest factor neighbours international value and high-dividend baskets. Two facts cut against the bearish read and must be stated: the Financials sector factor is the single most out-of-favour factor in the model (−12.6% over 252 days, z = −1.66) while LYG rose 48%, so this is a UK-domestic idiosyncratic move rather than a sector-beta ride; and LYG has lagged every European bank ADR peer on two-, three- and five-year horizons, which is the strongest tape-based argument that the motor discount has not fully closed. Cutting the other way: the Feb–Mar 2026 −22.7% drawdown in six weeks came with no adverse company news, which quantifies how violently this name re-prices on macro alone. (All loadings and returns are FACT; the regime persistence inference is INTERPRETATION and is regime-caveated.)

12. Fact vs. Interpretation

13. Open Questions

1. What does the 30 July 2026 strategy update contain? The 2022–26 plan culminates this year and next-phase targets are unpublished. A credible through-cycle ≥16% RoTE framework with a mid-40s cost:income target would materially change the analysis; a vaguer “continued progress” framing would confirm it.
2. Is the ECL allowance adequate? The allowance was held effectively flat between December 2025 and March 2026 despite a visibly worse base case (exit-2026 unemployment 5.6% vs 5.1%; house prices +0.7% vs +1.6%), offset by releasing a £50m overlay and by model calibration. Is that prudent or convenient?
3. Will £1,950m prove sufficient for motor finance? Response rates, operational costs, the omnibus-claim question and the outcome of the Dec-2026/Feb-2027 Upper Tribunal hearing are all open, and the company explicitly declines to cap the exposure.
4. How much severance is inside the cost base? FY2025 costs include an “increased severance charge” and Q1-2026 costs fell 3% “partly due to lower severance,” but the amount is never separately quantified. It matters for judging the sustainability of the <£9.9bn cost guide.
5. Why did the insurance Solvency II ratio fall 14pp (158% → 144% pre-dividend) on only £200m of dividends to Group? Unexplained in the results release.
6. Is the buyback price-sensitive? The Board now says it will “review excess capital distributions in addition to the ordinary dividend every half year.” Does that review consider the price paid relative to tangible book, given repurchases are now dilutive to TNAV/share?
7. What is the terminal hedge income? Management says growth “is expected to continue thereafter” beyond 2027 but gives no 2029+ framework. The whole valuation debate turns on whether £8bn is a plateau or a peak.
8. What replaces Halifax’s mass-market coverage? Retiring a brand that was a net switching loser is defensible; leaving the value-led niche to Nationwide is not obviously so.
9. Does Lloyds disclose a current-account market share? It does not, which makes the most important franchise metric in the business unverifiable from primary sources.

14. What Must Be True

14.1 For the bull case — and how to falsify it

14.2 For the bear case — and how to falsify it

15. Source Appendix

The full, itemised source appendix — every primary filing, regulatory document, transcript, market-data pull and third-party source relied on, with URLs and access dates — appears as Appendix B below.

Primary sources of record for this note:

• Lloyds Banking Group plc, 2025 Results news release, 29 January 2026 — the principal financial source for FY2025. LSE RNS mirror: https://www.rns-pdf.londonstockexchange.com/rns/8158Q_1-2026-1-28.pdf
• Lloyds Banking Group plc, Form 20-F for FY2025, filed 13 February 2026, SEC CIK 0001160106 — reproduces the Annual Report and Accounts, including the Directors’ Remuneration Report and Item 7.A major shareholders: https://www.sec.gov/Archives/edgar/data/1160106/000116010626000010/lyg-20251231_d2.htm
• Lloyds Banking Group plc, Q1 2026 Interim Management Statement, Form 6-K, 29 April 2026: https://www.sec.gov/Archives/edgar/data/1160106/000116010626000018/lbg6-kimsxq12026.htm
• FY2024 Results (20 February 2025), FY2023 Results (22 February 2024), FY2021 Results (24 February 2022) news releases
• Earnings-call transcripts (via ROIC.ai): FY2025 (29 January 2026), Q3-2025 IMS (23 October 2025), H1-2025 (24 July 2025) — read in full. Transcript-quality caveat: the feed is machine-generated with systematic million/billion unit errors; every figure quoted in this note was reconciled to the corresponding RNS.
• Complete SEC 6-K corpus, 1 January 2024 – 24 July 2026: 85 “Director/PDMR Shareholding” filings (297 MAR Article 19 notification blocks), 2 “Holding(s) in Company” TR-1s, 476 “Transaction in Own Shares” filings, and all Board Change and results announcements
• Financial Conduct Authority, PS26/3 Motor finance consumer redress scheme (30 March 2026): https://www.fca.org.uk/publications/policy-statements/ps26-3-motor-finance-consumer-redress-scheme; CP25/27 (October 2025); and “Motor finance scheme partially suspended” (2 July 2026): https://www.fca.org.uk/news/statements/motor-finance-scheme-partially-suspended
• UK Supreme Court, Johnson v FirstRand Bank Ltd [2025] UKSC 33, 1 August 2025
• HM Treasury, Safeguarding Stability, Enabling Growth: the Ring-Fencing Review, 18 May 2026; Autumn Budget 2025, 26 November 2025
• Bank of England, Monetary Policy Summary and Minutes, June 2026: https://www.bankofengland.co.uk/monetary-policy-summary-and-minutes/2026/june-2026; PRA CP9/26 on Basel 3.1 market-risk timing, June 2026
• UK Finance mortgage-market forecasts; Current Account Switch Service (CASS) quarterly switching data; Office for National Statistics labour-market, CPI and house-price releases
• Peer disclosures: NatWest Group Q1-2026 results (May 2026); Barclays PLC FY2025 and Q1-2026; HSBC Holdings FY2025 and Q1-2026; Santander UK; Nationwide Building Society FY2026
• Market data: AZI Trading price history for LYG and peers (full history to 24 July 2026); AZI valuation-index own-history percentiles (23 July 2026); FactorsToday factor loadings, leaderboard, stock-info and factor-returns endpoints (24 July 2026); ROIC.ai statements, ratios, enterprise value and valuation multiples (24 July 2026)
• Analytical framework sources: Deutsche Bank Securities, “Banking 101 — Large Cap Bank Primer” (11 May 2011) and Deutsche Bank AG/London, “Global Banking Sector — Credit quality in a deleveraging world” (26 September 2011) — third-party sell-side primers used for analytical framework only; all data in them is fifteen years stale and the incurred-loss provisioning regime they describe was superseded by IFRS 9 in 2018.

Sourcing note. lloydsbankinggroup.com blocks scripted access from this environment; primary documents were therefore obtained via the SEC EDGAR 6-K/20-F corpus (Lloyds mirrors every RNS as a 6-K) and the London Stock Exchange RNS PDF mirror. Where a fact rests only on press reporting it is labelled as such in the body.

Standing disclaimers. Sections 1–15 of this note are deliberately position-free and contain no recommendation and no price target; the single labelled exception is the Claude's Take block at the top. Third-party aggregated data (ROIC.ai, AZI, FactorsToday) is not primary and has been reconciled to the company’s own filings, which govern where they differ. Management commentary is treated throughout as hypothesis, not evidence. This note is general information and not investment advice. The author holds no position in Lloyds Banking Group plc.

APPENDIX A — Standard Diligence Questionnaire

Supplemental appendix to the research note dated 2026-07-24. Figures in GBP and per ordinary share unless flagged (1 ADS = 4 ordinary). Where a question does not map to a bank’s business model, the correct sector analogue is given. No recommendation and no price target appears in this appendix.

General

What thoughtful questions have other investors asked about this company?

The sell-side questions on the FY2025, Q3-2025 and H1-2025 calls were unusually good, and three stand out.

Jonathan Pierce (Jefferies) asked the single best question in the file — the structural-hedge underearn. He observed that Lloyds’ guidance implies a ~2.7% hedge yield against a materially higher swap curve, leaving “an underearn versus the current curve of 4 percentage points of RoTE.” Chalmers conceded “a touch above your 2.7%… clearly still materially below where swap rates are,” and dated the catch-up as “'27, and I would include kind of two-thirds, three-quarters of '28… by the time you are at the end of '28, you’ve got most of it.” He then refused to disclose the hedge maturity schedule. That exchange establishes both the size of the mechanical tailwind and its endpoint.

Pierce also ran the surplus-capital arithmetic publicly — >200bp ≈ £5bn, plus ~20bp of CET1 paydown ≈ £500m, plus Basel 3.1 ≈ £1bn, so “that’s £6.5 billion” — and got only “I can see how you get to them. Maybe that’s the best way of putting it.” Nunn’s coda was “this is the problem we wanted to have.”

On motor finance, two analysts pressed three times for the scenario weightings behind the £1.95bn provision and were refused each time. Chalmers would say only that the FCA’s proposals “represent the heaviest weighting in our overall scenario analysis” and that “we are not far off.” That refusal is the most important unanswered question on the name.

Beyond those: whether the sub-50% cost:income target is real (management conceded “it is not going to be sub-50% by much” and, revealingly, “we will make sure that we meet it”); what happens beyond 2026 (four separate analysts, all deflected to the July strategy update — Nunn: “You’re close to getting us to talk about beyond 2026, which we are vehemently against”); whether Basel 3.1 relief gets distributed (“far too early”); and whether Lloyds will buy a wealth manager (“I obviously shan’t comment”; analysts named Evelyn Partners and the Aegon UK workplace book — Lloyds subsequently bid for Aegon UK and lost).

The questions investors are not asking, and should be: what return Lloyds Living earns on the £2bn-plus of residential property it is accumulating; why roughly 21% of buyback spend leaks back out to employee share plans; and why the buyback size has never once been conditioned on the price paid relative to tangible book.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low?

At a cyclical high, and the composition makes this unusually clear. Reported FY2025 RoTE was 12.9% (14.8% ex-motor) and Q1-2026 printed 17.0%, against a five-year average nearer 11.4% and a 2020 trough of 2.3%. Three separate flattering effects are identifiable and quantified in this note: the asset-quality ratio of 17bp against management’s own “closer to 25 basis points” underlying figure (~£385m pre-tax); a Q1-2026 effective tax rate of 23.2% against ~27% guidance (~1.2pp of RoTE); and a £262m favourable swing in “volatility and other items” driven by disposal gains, which accounted for 38% of the entire increase in FY2025 pre-tax profit. Normalised, clean FY2025 RoTE is ~13.5–14.0%.

More fundamentally, ~47% of guided 2026 net interest income is structural-hedge income, worth roughly 8–9 points of RoTE, and the hedge is rolling 2021–22 vintages struck near 0.5–1.5% onto reinvestment near 4%. That is a rate-cycle position at a favourable point in its own repricing, not a franchise at a normal operating level.

Driven by the external environment or internal actions?

Overwhelmingly external. Guided 2026 NII growth is ~£1.3bn against guided hedge-income growth of >£1.5bn — by management’s own numbers, more than 100% of the growth is treasury reinvestment, with mortgage spreads, deposit repricing and volume all netting negative. Meanwhile the three biggest positive developments of the last two years were a Supreme Court judgment, an FCA policy statement and a Chancellor’s Budget decision — judicial and political outcomes, not management achievements.

Internal actions are real but second-order: £1.9bn of cumulative cost savings, £1.4bn of annualised strategic-initiative revenue against an original £1.5bn-by-2026 target (since raised to ~£2bn), a 3pp gain in direct-to-bank mortgage share, and protection share up from 5.8% to 7.8%.

How stable are revenues?

Stable in the sense that matters for a bank — 98% loan-to-deposit funding, £496bn of customer deposits, 67% of lending secured on UK residential property, and ~90–95% of 2026 hedge income already contracted. Unstable in a different sense: net income moved £14.4bn (2020) → £17.5bn (2022) → £17.1bn (2024) → £18.3bn (2025), and essentially the entire increase arrived in the single step when Bank Rate rose. Revenue is not volatile; it is rate-determined.

Outlook for products/services?

Structurally low-growth. Loans compounded ~1.8% a year over five years — below nominal GDP. The mortgage market is a refinancing market: ~1.8m fixed rates mature in 2026 and product transfers at £261bn are ~87% the size of the entire gross new-lending market. Lloyds writes at ~70bps completion margins against a back book maturing near 100bps, so every roll destroys ~30bps until roughly H1-2027. The SME book shrank 5%. The genuine growth is in cards (+10%), unsecured (+15%), Retail Europe (+21%) and capital-light insurance/protection/wealth — the last of which is the most attractive vector and, at 4.9% of group profit, too small to change the character of the business within the forecast horizon.

How big will this market be — growing, shrinking, domestic or international?

Emphatically domestic, and that is the defining structural fact. Essentially all revenue is sterling and UK-sourced; Retail Europe is a £20.4bn Dutch/German mortgage book. UK household deposits are ~£2.19 trillion (+4.3%); gross mortgage lending is forecast at £300bn in 2026 (+4%). The market is large, mature and growing roughly with nominal GDP. There is no international optionality and no geographic diversification to absorb a UK-specific shock.

Business Quality & Competitive Moat

Is the industry getting more or less competitive?

Both, in different places — and the distinction is the whole argument. Concentration is rising where the profit pool is: three mid-tier consolidations in twenty-four months (Nationwide/Virgin Money completed; NatWest/Sainsbury’s Bank; Santander UK/TSB completed 30 April 2026 at £2.65bn, with the 215-year-old TSB brand to be retired), and Lloyds itself retiring the 173-year-old Halifax brand. Capacity is being removed — roughly 1,600 Lloyds branches closed since 2015, 233 more during 2026.

Competition is intensifying at the margin and at the low-balance end. A documented 2026 mortgage price war has best two-year fixes near 3.54%, the lowest since 2022, with trade commentary explicit that lenders are “trimming margins where competitive pressure is sharpest, not… repricing on the back of cheaper funding.” Deposit pass-through is near 100% at the top of the savings tables — the effective rate on new deposits was 3.76% in March 2026 against a 3.75% Bank Rate. And Monzo has passed 11m customers with roughly half using it as a primary account, while Chase, Starling and Monzo hold the top three service-quality rankings, ahead of every incumbent.

How profitable is the business (ROIC, ROE)?

RoTE is the correct analogue; ROIC is not meaningful for a leveraged deposit-taker. Reported 12.9% FY2025 (14.8% ex-motor), 17.0% in Q1-2026, guided >16% for 2026, clean ~13.5–14%. Against a cost of equity conventionally 10–12%, that is a modest through-cycle spread and a large current one. Return on ordinary shareholders’ equity is lower than RoTE because £8.6bn of goodwill and intangibles sit outside tangible equity.

How profitable is the industry — how many competitors, what barriers to entry?

Five ring-fenced majors plus a large mutual hold ~75% of current accounts, and every one of them is earning mid-to-high-teens returns and raising its targets simultaneously: NatWest 19.2% FY2025 and 18.2% in Q1-2026 (2027 target moving toward 17%), HSBC UK 21.1%/21.6% at the divisional level, Barclays UK 20.7%/19.7%, Lloyds 12.9%/17.0%. That simultaneity is prima facie evidence of a real barrier — and equally evidence of a common exogenous driver rather than individual advantage.

The barriers are largely erected by the state: banking licence, capital requirements, MREL (Lloyds at 31.7%), ring-fencing, and FSCS deposit protection. Marathon’s caveat applies precisely — this is why the capital cycle does not clear normally here, and why high returns attract capital as price competition rather than as new entrants. The same state hand produced PPI (~£50bn sector-wide) and is now producing the ~£9.1bn motor scheme. A periodic conduct levy is the price of a protected profit pool.

Can the business be easily understood?

Yes, and the disclosure is good. The one genuine complexity is the structural hedge, which is where the earnings actually come from and which is not intuitive: a £246bn rolling ladder of receive-fixed swaps with a ~3.75-year weighted-average life, converting ~£103bn of near-zero-cost current accounts and ~£28bn of equity into lagging, mechanical income. An investor who has not understood the hedge has not understood Lloyds. Secondary complexities are IFRS 17 insurance accounting at Scottish Widows and the £1,454m of operating-lease depreciation from motor leasing sitting inside a bank’s cost line.

Can it be undermined by foreign low-cost labour?

Not the revenue — UK deposit-taking and mortgage lending are licensed, domestic and regulated. The cost base is actively being arbitraged: ~300 IT roles offshored to Hyderabad, alongside ~1,600 role reductions in the 2025 programme. That is a margin tailwind, not a competitive threat, and it carries execution risk — the March-2026 data incident occurred mid-migration.

Do brands matter?

Less than the branding spend implies, and Lloyds has just conceded the point. The CASS data is decisive: the Lloyds brand was a modest net switching winner (+12,073 in Q4-2025) while the Halifax brand was among the biggest net losers — and on 3 July 2026 Lloyds announced it is retiring Halifax. Meanwhile Nationwide, a mutual, added +64,527 net switchers, beating all four of the Big Four combined. Brands matter for trust and inertia, not for pricing power: nobody pays Lloyds more for a mortgage because it says Lloyds on it.

What is the nature of competition?

Price, on the two products that matter. Mortgages compete on a broker’s screen with an identical product and free switching at maturity; savings compete on a best-buy table. The genuine non-price competition is for the primary current account — the only balance the hedge can monetise — and there Lloyds is fighting a mutual paying cash switching bonuses and Fairer Share payments, plus digital challengers that out-rank it on service.

Customers’ switching costs?

Low in mechanism, high in practice, and shared with every incumbent. Britain has had mandated, guaranteed, free seven-day switching since 2013, which in theory eliminated friction entirely. It did not: 12.7 million switches in thirteen years against 70m+ accounts implies a sustained ~1–2% annual switching rate. The captivity is habit and inertia — Greenwald’s most common and weakest-per-unit demand advantage, and the most robust in aggregate. It is real and economically valuable. It is not proprietary to Lloyds.

Financial Condition & Balance Sheet

Assets not fully recognised on the balance sheet?

Yes, and it is the core of the business. The value of the current-account franchise — ~£103bn of balances paying essentially nothing — appears nowhere as an asset. It is monetised through the hedge and shows up only as net interest income. US bank M&A has historically paid a ~23% core-deposit premium (average P/TBV 2.9x across 920 deals), which is one way of pricing what is not on the balance sheet. Also unrecognised: the brand portfolio and customer relationships, other than where acquired.

Off-balance-sheet liabilities?

The material ones are disclosed rather than hidden. Motor finance is the live item: £1,950m provided against a scheme the FCA sizes at ~£9.1bn industry-wide, with the company stating explicitly that “the ultimate financial impact could materially differ from the amount provided.” Outside the scheme sit >30,000 claimants seeking ~£66m, a stayed Competition Appeal Tribunal claim, and an omnibus-claim question before the Court of Appeal. Total unutilised regulatory and legal provisions were £2,276m at December 2025 (2024: £1,600m). Standard banking off-balance-sheet items — undrawn commitments, guarantees, securitisation vehicles — are disclosed in the accounts. And the Dobbs Review into what Lloyds knew at the HBOS acquisition remains unpublished after nine years.

How conservative is the accounting?

Mixed, with two identifiable soft spots and one genuinely conservative choice.

Conservative: the motor provision was taken at the adverse end of management’s own modelled range, and the ECL allowance sits ~£0.4bn above the base case with a probability-weighted £3,223m against a severe-downside £5,263m.

Less conservative: (i) the FY2025 impairment charge was flattered by roughly 8bp — ~£385m pre-tax — by model calibrations and a debt-sale write-back, which the company discloses but which does not appear in the headline; (ii) the ECL allowance was held effectively flat between December 2025 and March 2026 despite a visibly worse base case (exit-2026 unemployment 5.6% vs 5.1%; house-price growth +0.7% vs +1.6%), offset by releasing a £50m overlay; (iii) roughly £1.25bn a year of software is capitalised rather than expensed, flattering cost:income and RoTE simultaneously while consuming CET1 through the intangibles deduction; and (iv) on the insurance side, new business is replacing only ~23% of the IFRS 17 contractual service margin being released (£93m added against £413m released), meaning the insurance profit line is being fed by an unwinding stock rather than by underwriting.

One presentational criticism: the FY2025 release attributes part of the rise in TNAV/share to the buyback, which at 1.47–1.74x tangible book is wrong-signed.

How CapEx-hungry is the business?

Not in the industrial sense — physical capex is modest and falling as ~245 branches close. The economic equivalents are (a) the technology programme, ~£3bn over 2022–24 extended toward ~£4bn, of which ~£1.25bn a year is capitalised, and (b) risk-weighted assets, the true capital consumption: RWAs grew £10.9bn in 2025 and £5.3bn in Q1-2026 alone, absorbing roughly £1.4bn of the year’s capital generation. A third, less-discussed call on capital is Lloyds Living, a £2bn-plus directly-owned residential portfolio heading toward a stated 50,000-home ambition, with no disclosed return metrics.

Capital Allocation & Management

How much FCF does the business generate, how does management use it, what is the philosophy?

Free cash flow is not a meaningful bank metric; the analogue is organic capital generation, and it was 147bp in 2025 (178bp ex-motor) and 148bp in 2024 — both missing the c.175bp guidance, with the live guide now >200bp. On ~£240bn of RWAs, >200bp is ~£4.9bn gross, less RWA growth.

Use of it is unambiguous: distribution. FY2025 returned ~£3.9bn — a 3.65p dividend (52.1% of EPS) plus a £1.75bn buyback — for a total payout of 93.2% of attributable profit. £18.3bn over five years, roughly half via buyback. The philosophy is a “progressive and sustainable ordinary dividend” with surplus returned via buyback or special dividend, now reviewed semi-annually rather than annually.

Sustainable capacity is ~£3.4–4.0bn a year, or ~5.1–6.0% of market capitalisation. The uncomfortable fact is that the absolute return has been flat at £3.4–3.9bn for five years while the market cap trebled — Lloyds’ own framing collapsed from “c.14% of market cap” (FY2023) to “c.9%” to “c.6%” (FY2025).

Significant acquisitions recently?

Small, cheap and capability-driven: ~£820m of disclosed cash consideration across five years. Embark (£390m plus ~£150m re-platforming) is the worst — a subsidiary reporting £5.0m of PBT in FY2024, with no goodwill impairment disclosed. Schroders Personal Wealth is the best — the remaining 49.9% acquired for zero cash, swapping a passive 19.1% Cazenove stake for full control of a business earning ~£180m. Tusker (~£300m) brought EV salary-sacrifice leasing but also the operating-lease depreciation now growing 10% a year on falling used-EV prices. Curve (~£120m, November 2025) was bought out of a litigated process — the largest external shareholder petitioned the High Court two days after the announcement — and had not completed as at 24 July 2026 against a stated H1-2026 expectation.

The pattern matters more than the sums: Lloyds keeps reaching for non-interest income and keeps losing or picking wrong. It sat out TSB (correctly, on competition grounds), bid for Aegon UK and lost to Standard Life/Phoenix at £2.0bn, and is now exploring Aldermore — an asset its own house broker publicly said “would add scale rather than new capabilities” and would do “little to reduce Lloyds’ dependence on net interest income,” and which requires a motor-finance indemnity on top of Lloyds’ own £1.95bn.

Buying back shares?

Yes — and this is the sharpest capital-allocation criticism in the file. Programmes of £2.0bn (three times), £1.70bn and £1.75bn have retired ~16.13bn shares gross, cutting the count from 70,996m (Dec-2021) to 58,268m (Jun-2026), −17.96%.

But the price paid has moved from strongly accretive to heavily dilutive, and the programme size never changed:

At a 99.43p average, each share retired costs 99.4p to acquire 57.0p of tangible book. Lloyds has never conditioned buyback size on price-to-tangible-book, never offered a special dividend as the alternative, and never disclosed a level at which it would stop. The one point in its favour: it bought most heavily in the March-2026 drawdown, at the programme’s cheapest 94.77p.

Issuing large amounts of new shares to insiders?

Yes, and it is not separately disclosed. Against ~16.13bn shares repurchased, the net reduction was 12,754m — implying ~3.4bn shares re-issued to employee share plans, roughly 21% of all buyback volume. Total voting rights rose in January and February 2026 while the buyback was running. Roughly £1 in every £5 of buyback spend offsets employee dilution rather than shrinking the base, and the figure must be reconstructed from monthly voting-rights filings. Shareholders also approved removal of the 5% inner dilution limit on discretionary share plans at the 2025 AGM (96.47% for), without comment.

Compensation policy of directors/management?

Structure improved; quantum ratcheted. Charlie Nunn’s single figure went £3.681m (2023) → £6.169m (2024, restated) → £7.407m (2025) — up 101% in two years and the highest for any Lloyds CEO in the ten-year disclosure table. The CEO-to-median-colleague ratio moved 80:1 → 114:1 → 141:1 against a median colleague on £52,638.

Critically, the plan that actually paid was not a performance plan: 100% of the long-term incentive in the 2024 and 2025 single figures came from the restricted Long Term Share Plan, which vests on three pass/fail underpins a capital-generative UK bank clears almost by construction, and which has vested at 100% of maximum three cycles running. No LTIP — the genuinely performance-tested plan — has ever vested; the first cycle completes 31 December 2026. Lloyds itself discloses that £1,272k of Nunn’s £2,976k of 2025 long-term incentive (43%) was pure share-price appreciation, and the Committee considered and rejected a windfall adjustment.

The 2026 policy, approved 14 May 2026 with 94.97% for, leads with a 44% cut in fixed pay — true, and the smaller number. It also raises maximum variable pay from 440% to 800% of salary and the CEO’s maximum package from £9.1m to £13.9m (+53%), taking fixed pay from a third of the package to an eighth. That is roughly 7.3:1 variable-to-fixed — consuming most of the 8:1 headroom obtained in 2024, two years after shareholders were told the executive opportunity “cannot be increased as a consequence.” Genuinely better: financial weighting 50% → 75%, relative TSR 20% → 30% against eighteen European banks, a RoTE range stretched to 15–18%, and the soft Committee-judged block cut from 35% to 15%.

Accountability for motor finance is thin. In 2025 the charge was excluded by design from the bonus metrics; the discretionary haircut applied cost Nunn £152,117 — 2.05% of a £7.4m package — against an £800m in-year charge. The LTSP was explicitly ruled immune and paid him £2.98m. No malus or clawback has ever been applied to a Lloyds executive director. From 2026, all “exceptional items” are excluded from both bonus and LTIP metrics, replacing a mechanical channel with pure Committee judgement.

Motivations of management?

Alignment in absolute terms is genuine: Nunn holds 10.1m shares outright (656% of salary), Chalmers 10.7m (1,128%), requirements rise to 500%/450% with a two-year post-employment hold, no hedging is permitted and no shares are pledged. Clawback runs seven to ten years.

But incremental conviction is zero, and the revealed preference across the whole executive group is negative. In thirty-one months there were three open-market purchases totalling ~£521k — by the Chair (February 2024, at 45p) and one newly-appointed NED — against £15.41m of discretionary executive selling, a ratio of roughly 1:30 by value, with selling accelerating from £1.53m at ~56p in 2024 to £8.60m at ~105.5p in 2026 to date. Nunn has never been the subject of a discretionary purchase or sale notification in his entire tenure. Nobody at Lloyds has ever bought a share above 84p. The Chair, who bought roughly every 12–24 months between 28p and 47p, has bought nothing since February 2024 — through a doubling — and is not leaving.

One governance gap deserves naming: Nathan Bostock, the board’s most experienced commercial banker (ex-CEO Santander UK, ex-CFO and CRO of RBS), owns 430 ordinary shares — roughly £450 — against £359,000 of annual fees. Lloyds imposes no shareholding requirement on non-executive directors at all.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer?

It is an ADR. LYG trades on the NYSE at a ratio of 1 ADS to 4 ordinary shares (CUSIP 539439109, depositary BNY Mellon, 188 record holders), with 2,400,874,341 underlying ordinary shares — 4.07% of the company — in ADR form, up from 3.49% a year earlier. It is not an MLP and issues no K-1; holders receive a 1099-DIV. Two practical consequences: UK dividends carry no withholding tax, which is unusually favourable for a US holder; and the ADS carries GBP/USD exposure on top of the equity, so the dollar return can diverge materially from the sterling return. As a foreign private issuer Lloyds files 20-F and 6-K, not 10-K/10-Q/DEF 14A/Form 4 — insider dealing appears as “Director/PDMR Shareholding” RNS announcements mirrored as 6-Ks.

Dividend policy?

A progressive and sustainable ordinary dividend, with surplus capital returned via buyback or special dividend, reviewed semi-annually from 2026 (previously annual). FY2025 DPS was 3.65p (+15%) — the fourth consecutive rise of ≥15%, a 16.2% CAGR from 2021 — costing £2,160m and representing 52.1% of EPS. The dividend yield is ~3.2% and total shareholder yield ~5.9%. History matters: the FY2019 final was cancelled at the PRA’s request in March 2020 and the FY2020 final of 0.57p was “the maximum allowed under the PRA’s temporary framework.” This is a regulator-permissioned distribution, not a contractual one.

How profitable is the business?

Reported RoTE 12.9% FY2025 (14.8% ex-motor), 17.0% Q1-2026, guided >16%, clean ~13.5–14%. Cost:income 58.6% FY2025 (53.3% ex-remediation), 51.9% in Q1-2026, guided <50%. Banking NIM 3.06% FY2025, 3.17% Q1-2026 — not comparable with NatWest’s 2.47% because Lloyds’ NIM carries motor operating-lease income and a larger unsecured book.

Is net income diverging from cash from operations?

The question does not translate cleanly for a bank — operating cash flow is dominated by movements in loans, deposits and trading balances and routinely swings sign for reasons unrelated to earnings quality (ROIC’s own series shows a negative price-to-cash-flow for 2024 and a positive 10.3x for 2025 on the same business). The correct analogues are capital generation and TNAV growth, and both tell the same story.

Capital generation of 147bp in 2025 (178bp ex-motor) missed the c.175bp guide for the second consecutive year. And the decisive number: tangible net asset value per share is 57.9p at March 2026 against 57.5p at December 2021 — +0.7% over four and a quarter years — despite ~£9.45bn of buybacks retiring 17% of the count. Including dividends, the four-year total book-value return is ~5.7% a year against a 12–15% reported RoTE. The gap is the ~93% payout ratio: this business does not compound book value, it converts earnings into a distribution stream. That is a legitimate model and should be valued as a yield instrument, not as a compounder.

Risks & Downside

What factors would cause the stock to decline?

In rough order of probability-weighted impact: (1) the market re-rating >16% RoTE from a perpetual to a peak-cycle number — at 1.96x tangible book the price requires a sustained ~18.7% return that Lloyds has never earned, and a reversion to a 13–15% durable RoTE implies a ~23% to ~36% multiple adjustment; (2) faster Bank of England easing, with disclosed −50bp sensitivity of −£275m/£400m/£675m over years one to three and greater sensitivity on the downside from deposit-pricing lags; (3) a further motor-finance top-up, which would simultaneously break the <50% cost:income target, since that target implicitly assumes remediation of ~£0.1bn against £899m and £968m in the prior two years; (4) a UK macro deterioration toward Lloyds’ own downside scenario (unemployment 7.8%, house prices −3.3%/−5.8%); (5) a bank tax at any fiscal event — the November 2025 reprieve was political, not structural; (6) a disappointing 30 July 2026 strategy update; and (7) an ill-judged acquisition, with Aldermore live.

The March-2026 episode is the empirical calibration: the shares fell 22.7% in six weeks on macro and two idiosyncratic negatives, with no adverse earnings news at all.

Risk of a catastrophic loss?

Low on current evidence, but the history is not reassuring and should not be waved away. Today: CET1 13.4% against a ~11.8% disclosed requirement, UK leverage 5.1%, LCR 145%, NSFR 124%, MREL 31.7%, loan-to-deposit 98%, Fitch upgraded the subsidiaries to AA in May 2026, and the bank passed the 2025 Bank of England stress test. Credit is improving — Stage 2 and Stage 3 balances both fell quarter-on-quarter in Q1-2026 — and 67% of the book is secured on residential property with 0.2% ECL coverage.

The counterweight is what a UK recession has historically done to this balance sheet. In 1H-2011 Lloyds carried £65.5bn of impaired loans — 10.9% of the book — with 68.5% impairment in the Corporate Real Estate business support unit and 64.1% in Ireland, having booked cumulative losses above 8% of group loans against a three-year peak of ~5% in the early-1990s recession. Deutsche Bank’s 2011 stress work put Lloyds worst of the five UK banks, at −48% of 2012 tangible equity over two years of recession losses. (That is history, not a description of today’s company — the loan book, the Irish exposure and the capital position are all transformed. It is included as the correct falsification benchmark for the bear case, not as a forecast.) Softening it further: mandatory bail-in of £8.9bn of subordinated liabilities and £5.9bn of AT1 sits ahead of ordinary shareholders in resolution.

Chance of a total loss?

Very low. A total loss requires either resolution or a deeply dilutive recapitalisation. Against that: a ~£3.4–4.0bn annual capital-generation run-rate, ~160bp of CET1 headroom over the disclosed stack, a self-funded balance sheet, an AA-rated subsidiary, and a Basel 3.1 RWA release of £6–8bn arriving in January 2027. The identified tail risks — motor finance at £1.95bn provided against a ~£9.1bn industry scheme, and a hypothetical severe-downside ECL of £5,263m — are large but absorbable against ~£33bn of CET1.

The honest framing is that the risk here is not solvency; it is a de-rating. A holder is far more likely to lose money through multiple compression against flat-to-modestly-growing earnings than through a credit event. That said, the lifetime record is a standing reminder of what a levered, single-country, rate-dependent balance sheet can do: a −94.8% maximum drawdown, a negative twenty-year Sharpe ratio, and a price still 58% below its October-2007 dividend-adjusted peak after a 190% three-year run.

Recent News & Events

Has the business environment changed recently?

Yes — decisively, and almost entirely in Lloyds’ favour, which is precisely the problem for a buyer today. Four changes in twenty-four months: the Supreme Court (1 August 2025) rejected fiduciary duty and accessory liability, narrowing motor-finance exposure dramatically; the FCA’s final rules (PS26/3, 30 March 2026) sized the industry scheme at £9.1bn rather than the feared £18bn-plus, landing inside Lloyds’ existing £1.95bn provision and prompting the first non-increase since 2023; the November 2025 Budget imposed no surcharge rise, no levy increase and no windfall tax, against widespread expectation; and Basel 3.1 was delayed to January 2027 and softened, giving Lloyds a £6–8bn Day-1 RWA reduction while penalising trading-heavy peers. Ring-fencing reform (18 May 2026) and a mortgage-friendly FCA rule review add further tailwind.

All four of the biggest positives were judicial or political outcomes, not management achievements — and all four are now in the price. The one live negative: on 2 July 2026 the Upper Tribunal partially suspended the redress scheme at the instance of four challengers, with a hearing listed for December 2026 or February 2027. That defers the cash outflow into 2027 and defers the moment the provision can be marked against reality.

Significant acquisitions?

Covered above: Schroders Personal Wealth completed for zero cash in October 2025; Curve agreed in November 2025 at ~£120m and still not completed amid a High Court petition from its largest external shareholder; Aegon UK bid for and lost to Standard Life/Phoenix at £2.0bn in April 2026; Aldermore live and unresolved, with Metro Bank reported at ~£2bn against RBC’s £1.35bn central valuation. Disposals: the ~£6bn Scottish Widows bulk-annuity portfolio to Rothesay (completed Q2-2025, generating a gain that flattered FY2025 “volatility and other items”); Scottish Widows Europe to Chesnara for €110m (February 2026, at 0.64x Solvency II own funds); and the invoice-factoring business closed rather than sold.

Change in accounting policies?

No new standard adopted in the period; the last material change was IFRS 17 for insurance contracts, which restated FY2022 and cut reported TNAV/share from 51.9p to 46.5p. Two presentational points do matter. First, a 2025 change moved certain divisional variable-payment costs from operating costs into divisional other income, with comparatives restated and no net P&L effect. Second — and this is a judgement, not a policy change — the company has progressively shifted to presenting its headline metrics on an ex-motor basis (RoTE 14.8%, capital generation 178bp), and from 2026 all “exceptional items” are excluded from both the bonus and the LTIP financial metrics. That is a meaningful weakening of the mechanical link between reported outcomes and pay.

Recent changes — new markets, facilities, management?

Brands and estate: the 173-year-old Halifax brand is being retired (announced 3 July 2026), consolidating retail under Lloyds with Bank of Scotland retained in Scotland — ~190 branches rebranded or absorbed through 2027, with no job losses announced. Roughly 233 branch closures during 2026 plus 12–13 in early 2027.

Management: Nunn and Chalmers both remain in post — an unusually stable pair — and Sir Robin Budenberg remains Chair (re-elected 14 May 2026, 99.05% for) with no successor announced and his nine-year limit falling in October 2029. Two of four customer-facing divisional CEOs were replaced with external hires in early 2026 — Amanda Murphy from HSBC (Business & Commercial Banking) and John Langley from Wells Fargo (Corporate & Institutional Banking) — both on fresh multi-million LTIP grants immediately ahead of the strategy launch. Danuta Gray joined as a NED and Chair of Scottish Widows Group on 1 July 2026, having been a NED of Aldermore Bank from 2014 to 2021 including Senior Independent Director and Interim Chair.

Technology: ~50 live GenAI use cases delivering ~£50m in 2025, targeting >£100m in 2026; ~300 IT roles offshored to Hyderabad; 1,300 new digital roles and >1,000 AI roles; and on 21 April 2026 Lloyds became the first UK lender with an AI investment-guidance tool, in the FCA’s second AI Live Testing cohort. Against that, the 12 March 2026 IT fault exposed up to 447,936 customers’ data, drew a Treasury Committee intervention, and had produced no ICO or FCA sanction as at 24 July 2026 — an unquantified open exposure.

New markets: none material. Lloyds Living, however, has grown to >8,850 homes and past £2bn, with a stated ambition of 50,000 rental homes by 2030 — a substantial, barely-disclosed expansion of a deposit-funded bank into direct residential property ownership.

The event that matters most has not happened yet. Lloyds reports H1-2026 results and an entirely new multi-year strategy on Thursday 30 July 2026 — six days after this appendix is dated. The 2022–26 plan culminates this year and no 2027–2029 targets exist; management deflected every forward question on the FY2025 call to that date. The hard reported base for everything above is Q1-2026 (29 April 2026).

APPENDIX B — Source Appendix

Research date 2026-07-24. All URLs accessed 2026-07-24 unless stated. Sources are grouped by tier, primary first. Every non-obvious factual claim in this note and its appendices traces to an entry below.

Sourcing note. Lloyds is a UK foreign private issuer: it files Form 20-F and Form 6-K with the SEC, and does not file 10-K, 10-Q, DEF 14A or Form 4. Insider dealing appears as “Director/PDMR Shareholding” RNS announcements, and substantial holdings as TR-1 notifications — both mirrored into EDGAR as 6-Ks. lloydsbankinggroup.com blocks scripted access from this environment (HTTP 200 with a 24kB error page), so primary documents were obtained via the SEC EDGAR 6-K/20-F corpus — Lloyds mirrors every RNS as a 6-K — and the London Stock Exchange RNS PDF mirror. Where a fact rests only on press reporting it is labelled as such in the body.

1. Company primary filings and disclosures (highest authority)

2. Earnings-call transcripts

Retrieved via the ROIC.ai MCP (list_earnings_calls, get_earnings_call_transcript, get_latest_earnings_call), which carries 30 LYG calls back to 2016. Read in full:

⚠ Transcript-quality caveat, applied throughout: the ROIC feed is machine-generated and contains systematic million/billion unit errors (“GBP 22 million or 5%” for £22bn; RWAs as “GBP 232 million”), plus transcription errors (“FDA” for FCA; “Our losing CET1 ratio” for “closing”). Every figure quoted from a transcript in this note was reconciled to the corresponding RNS before use. No Q1-2026 transcript exists in the feed; Q1 commentary was taken from the IMS (1.3). Google Drive holds no Lloyds transcripts.

3. Regulatory, judicial and government sources

4. Industry data and peer disclosures

5. Market and quantitative data

6. Trade press and financial media (secondary — used for attribution and colour, not for primary facts)

7. Analytical frameworks

8. Evidence-quality notes and known limitations

1. The most recent hard reported period is Q1-2026 (29 April 2026). Lloyds reports H1-2026 results and an entirely new multi-year strategy on 30 July 2026 — six days after this report date. Any H1-2026 figure cited anywhere is an estimate or a consensus expectation, and is labelled as such. The 2022–26 plan culminates this year and no 2027–2029 targets exist.
2. Unit discipline. Lloyds reports in GBP per ordinary share; the ADS is 4 ordinary shares and trades in USD. Every figure in this note states its currency and share basis. Third-party feeds (ROIC, AZI) mix these conventions — see items 5.2 and 5.5.
3. P/TNAV denominator. The 1.96x headline uses the last reported TNAV of 57.9p (31 March 2026). On a projected or forward TNAV the multiple is ~1.83–1.87x. Both are shown in the valuation discussion; neither is a price target.
4. Undisclosed items relied on by management but not verifiable: the motor-finance scenario weightings behind the £1.95bn provision (refused three times on the record); the structural-hedge maturity schedule (refused); the severance charge inside operating costs (never separately quantified); Lloyds Living return metrics (not disclosed); and the UK current-account unit share (Lloyds does not disclose one — the ~20–21% figure used is an estimate, labelled as such).
5. Ownership data is structurally unreliable. Lloyds published exactly two TR-1s in thirty-one months. With 58.9bn shares in issue, a 1% step is ~589m shares, and investment-manager aggregation exemptions permit delayed or exempt reporting. The Harris Associates exit had to be disclosed by the company in prose because no TR-1 was ever filed for it. BlackRock’s DTR-notified 5.14% dates from 2015 against an 8.4% SEC Schedule 13G/A (Feb 2024). Absence of TR-1s must not be read as ownership stability.
6. Remuneration gap. The full 2025 Directors’ Remuneration Report was read via the Form 20-F (1.2). LTIP performance measures and weightings for the 2024–2026, 2025–2027 and 2026–2028 cycles were obtained and are reported in the remuneration discussion; the first performance-tested LTIP vest occurs in 2027 and no LTIP has yet vested.
7. Group customer count (~26–28 million) is an industry/press estimate, not a disclosed figure. Disclosed franchise metrics used instead: 21.5m mobile-app customers, >20.9m digitally active, >10m IP&I customers, 2,050,775 registered shareholdings.
8. Third-party aggregated data (ROIC.ai, AZI, FactorsToday) is not primary. Where it conflicts with a company filing, the filing governs and the discrepancy is recorded (5.5). FactorsToday outputs are statistical estimates: loadings, returns and drawdowns are reportable facts; persistence and mean-reversion inferences are labelled INTERPRETATION and regime-caveated.
9. No position is stated or implied. The author holds no position in Lloyds Banking Group plc. Sections 1–15 of this note carry no recommendation and no price target; the single labelled exception is the Claude's Take block.