BARC BARC

Key data

The business

Barclays is a universal bank in two halves. Barclays UK is the ring-fenced retail and business bank plus Barclaycard UK, deposit-funded and stable. Barclays International holds the Investment Bank, the US Consumer Bank cards book, and the corporate and private banks. The franchise that sets the price is the Investment Bank: it runs roughly 40% of group income and is the single reason the stock sits below its domestic peers. Strip the IB out and you have a high-teens-return UK bank bolted to a US cards book; bolt the IB back on and earnings swing with the global markets and banking-fee wallet, which the market pays less for. In H1 2026 the IB earned a 16.0% RoTE and crossed £4B of quarterly income in Q1; Barclays UK returned 20.4%, the UK Corporate Bank 21.3%, the Private Bank and Wealth 26.9%, and the US Consumer Bank 30.2% reported (10.5% underlying, the gap being the reserve build on the cards book).

The engine by income is the Investment Bank at ≈40% of group income; the engine by return is Barclays UK at a 20.4% RoTE on a deposit-funded book, with a UK net interest margin that widened to 3.7% from 3.55% a year ago. What the financials do not show: the whole re-rate rests on Barclays proving the IB can hold a low-to-mid-teens return through a cycle rather than the high-single-digits it printed in 2022-2023. The 2026-2028 plan leans on exactly that, with group RoTE targeted to step from more than 12% in 2026 to more than 14% by 2028. What changed this half: 2026 income guidance was raised to about £31.5B, and credit costs turned up.

The numbers

H1 2026 total income rose 11% to £16.5B, attributable profit reached £4.19B (from £3.52B a year earlier), and statutory RoTE was 14.8%; on the 524p local price against 423p of tangible net asset value per share, you are paying 1.24x TNAV for a bank now earning a mid-teens return on that tangible equity.

Forward attributable profit is not separately guided; consensus tracks EPS and RoTE, and the buyback shrinks the share count into those EPS lines. The credit stack is where the print got worse: the H1 2026 group loan loss rate was 62bps (Q2 51bps), above the low-50s the market had penciled in, on a £1.4B impairment charge that included a £228M single-name charge in the Investment Bank and a higher UK motor-finance redress provision, now £430M cumulative. Group coverage held broadly stable at about 1.2%; the most granular Stage 3 disclosure (FY2025 basis) showed roughly 24% coverage on wholesale and 82% on cards, the cards reserve being the buffer against the US consumer book.

Peer set: LLOY, NWG, HSBA, STAN (UK large-cap banks).

At 1.24x TNAV and about 10x forward earnings, the market is renting you the Investment Bank's cyclicality: Barclays compounds TNAV per share (409p to 423p in six months, up 3.4%) and plans to return more than £15B across 2026-2028, over a fifth of the current market cap, while lifting RoTE toward more than 14%, yet it trades below Lloyds and NatWest purely because those two carry no global IB. Is the cheapness real or a trap? The trap case has teeth: the Q2 16.1% RoTE flatters the half on a strong 2026 markets wallet, and credit is deteriorating, with the loan loss rate at 62bps and the motor-finance provision climbing. Normalize the IB to a mid-teens through-cycle return rather than the Q2 peak, and credit to a 50-60bps run-rate, and forward earnings still support a low-double-digit RoTE; the cheap does not fully disappear, which is what separates this from a melting-book value trap. This is a re-rate-plus-distribution story, not a long-duration compounder: the return comes from the gap between 1.24x TNAV and peers at ≈1.4x closing as the IB proves consistency, plus roughly 7% a year of buyback and dividend shrinking the count. Bulls are paying for the RoTE step to more than 14% and the capital return; bears see the IB's earnings quality and a credit cycle the reserves may be lagging. Closest comps are Lloyds (≈1.4x TNAV, pure UK domestic, no IB), NatWest (≈1.5x, higher RoTE, UK-focused), HSBC (Asia-weighted), and Standard Chartered (emerging-market, sub-TNAV); Barclays is the only UK large-cap where a global IB both caps the multiple and supplies the earnings upside.

Management

Venkat's team hit all its FY2025 financial targets and lifted the 2026 income guide at the half, which reads as realist-to-conservative rather than sandbagging given the RoTE beat. Capital return is the clearest signal: the diluted share count fell from 15.3B to 14.6B across FY2024-2025 and keeps falling, so buybacks are genuinely retiring stock rather than offsetting compensation. The three-year plan of more than £15B is over a fifth of the market cap and is being executed near 1.2x TNAV, which is accretive while the stock trades below peers. They are earning double-digit returns on tangible equity that sit above the cost of capital, so the buyback is a use of genuine excess capital, not a lever to flatter a fading return. The one item to watch is the motor-finance redress: the £430M cumulative provision is an estimate under an FCA scheme, and provisions of this kind have a history of climbing.

The linchpins

Motor-finance redress is the most dangerous near-term variable because it is a live, uncapped estimate. Barclays' £430M cumulative provision covers commission redress on its partner-financed UK motor book under an FCA-led industry scheme whose final scope is still settling. It matters because an increase drops straight through to attributable profit and dents the very TNAV the re-rate is priced against. Falsification: a materially higher provision at Q3 2026 or FY2026 results, or an FCA scheme design broader than the sector has reserved for, breaks the "credit is a bump, not a cycle" read.

The Investment Bank's return consistency, paired with credit normalization, is the second linchpin. The re-rate needs the IB to hold a low-to-mid-teens RoTE as the 2026 markets wallet softens, while the group loan loss rate settles back toward the low-50s from H1's 62bps. The IB printed 16.0% in H1 but leaned on a strong Q2, and a single £228M name already showed up in impairment. Falsification: an IB RoTE sliding toward high-single-digits in H2, or a loan loss rate that keeps climbing past 65-70bps into 2027, would say the earnings base is more cyclical and the credit turn more structural than the bull case allows.

The re-rate gap itself is the third linchpin, and the most forgiving. The thesis pays 1.24x TNAV expecting a drift toward the ≈1.4x domestic-peer multiple as RoTE delivery accumulates. It matters because roughly half the expected return is multiple, not earnings. Falsification: RoTE holding above 12% for several quarters while the P/TNAV discount to Lloyds and NatWest fails to narrow would signal the market is permanently discounting the IB, leaving only the ≈7% distribution yield as the return.

Closing

You are paying 1.24x tangible net asset value and about 10x forward earnings for a bank stepping its return on tangible equity toward more than 14% and handing back over a fifth of its market value by 2028; the discount to domestic peers is the market renting you the Investment Bank's cyclicality. The strongest linchpin gets its next read soon: the motor-finance provision and the 62bps H1 loan loss rate both update at Q3 2026 in late October, with FY2026 results in February 2027 the fuller checkpoint. This works if the IB holds a mid-teens RoTE and credit settles near 50-60bps as the redress provision stops climbing, and fails if motor-finance redress balloons or a credit cycle outruns the reserves while the IB gives back its 2026 markets gains. A cyclical re-rate with a capital-return floor, not a long-duration compounder.