SCHW · Forward model · Bull case

The Bull case, 20 quarters out

Model as of 2026-08-29

Each segment is projected from its own operating driver, rolled up into consolidated cash flow, and discounted back to a fair value per share. The assumptions below are editable — change them and every number on this page moves with them.

READ THE FIELD NAMES WITH CARE: this is an equity-level DCF wearing an enterprise-value engine's labels, the construction proved on the BAC, JPM and GS pages. The page says 'Enterprise value' and 'Net cash'. What it means is EQUITY value and a deliberate zero - $249.7B of bank deposits, $124.0B payable to brokerage clients, $43.8B payable to brokers and dealers and $37.1B of borrowings are FUNDING at a bank and a broker-dealer, not financing to be netted, so the engine's 'Equity' IS the equity value and fair value is a true per-share number. SCHWAB IS A SINGLE-SEGMENT REPORTER, and this model does not pretend otherwise. There is one consolidated statement of income with five revenue lines - net interest revenue, asset management and administration fees, trading revenue, bank deposit account fees and other - and no segment profitability of any kind. The five verticals ARE those five disclosed lines, copied verbatim from the five-quarter Financial and Operating Highlights tables, and they sum to printed total net revenues EXACTLY in all nineteen quarters from 2021 Q4 to 2026 Q2, $7,072M in the basis quarter. Nothing is split, allocated or apportioned; every historical point is disclosed, not one is estimated. SIX SUBSTITUTIONS make the equity reading correct, all checked against the June quarter. (1) 'EBITDA margin' is the CONSOLIDATED PRE-TAX INCOME margin of 51.8807% ($3,669M on $7,072M), already net of every operating expense AND net of the provision for credit losses on bank loans, which Schwab reports inside the Other revenue line rather than as an expense. It is not pre-provision pre-tax income. The SAME margin sits on all five verticals because Schwab publishes none by line; asset-gathering economics are almost certainly better than the firm average and trading's almost certainly worse, but inventing the difference is exactly the failure this model refuses. The split earns its place through five genuinely different REVENUE drivers - two disclosed price-times-base capacity lines, a disclosed volume-times-price unit line, a contractual run-off, and one growth residual - not through differentiated economics. (2) capexIntensity is 0.2% of revenue as the premises proxy used on BAC, JPM and GS, not the true 2.62% cash-capex intensity ($185M on $7,072M — the statement-of-cash-flows figure; the release's operating-highlights line shows $792M, but $633M of that is a non-cash software-licence commitment booked to long-term debt, so the cash capex the model uses is $185M), because pre-tax income is ALREADY after $198M of depreciation and amortisation plus $142M of acquired-intangible amortisation; deducting full capex on top would double-count. Cash capex is in fact BELOW quarterly D&A, so the 0.2% is a small conservatism. It is 0.2% on every vertical rather than zero on any, because no line's projected revenue goes negative: Other revenue's minimum projected quarter is $302.3M against a nineteen-quarter historical minimum of $127M. (3) corporate.taxRate is 23.6849%, the basis quarter's own GAAP effective rate ($869M over $3,669M), on EXACTLY the same basis as the pre-tax figures. THERE IS NO FULLY-TAXABLE-EQUIVALENT ADJUSTMENT AT SCHWAB and none is imported from BAC or JPM. (4) overheadPctRevenue of 2.2049% carries PREFERRED STOCK DIVIDENDS AND NOTHING ELSE: preferred stock dividends and other were $119M, grossed up at 23.6849% to $155.9M of pre-tax equivalent, 2.2049% of net revenues. Operating expenses are already inside the margin, so a second overhead would double-count them. Without this line free cash flow would land on net income of $2,800M rather than net income available to COMMON of $2,681M, overstating value by 4.4%. (5) discountRate is a 9.25% COST OF EQUITY, derived by inverting the bank identity on Schwab's own observables. (6) exitEvEbitda of 12.9736x is a 17x P/E grossed DOWN by tax, because the engine applies it to terminal PRE-TAX earnings. CALIBRATION IS THE TEST OF ALL SIX: the first projected quarter's free cash flow is $2,691.0M against actual 2026 Q2 net income available to common shareholders of $2,681M, +0.37%; the static no-growth no-glide version is $2,670.2M, -0.40%. BANK ANCHORS, all at 30 June 2026, because a bank valued on a DCF alone is not credible: tangible book value per share $14.0968 and book value per share $25.4230; ROE 25% and return on tangible common equity 44% for the quarter, 41% for the half and a derived 41.4% trailing twelve months; consolidated Tier 1 Leverage Ratio 8.7% against a 4%-5% requirement, down from 9.3% at year-end 2025, with Charles Schwab Bank at 10.7%; adjusted Tier 1 Leverage Ratio 6.8%, down from 7.1% at year-end, at the BOTTOM of management's own 6.75%-7.00% operating objective; a $20.0B buyback authorisation under which 11.2 million shares were repurchased for $1.0B in the quarter and 35.5 million for $3.4B in the half; and a quarterly dividend of $0.32, raised 19% in January. WHICH CONSTRAINT BINDS, AND IS 'EARNINGS ARE DISTRIBUTABLE' HONEST? The binding constraint is the ADJUSTED TIER 1 LEVERAGE RATIO and the AOCI position inside it, NOT credit risk-weighted capital - Schwab is not a credit bank and this model manufactures no credit story. The answer is: only partly. Schwab returned $3,377M of buybacks and $1,123M of common dividends in the first half, 88.6% of the $5,078M available to common, while the adjusted ratio fell from 7.3% to 6.8% in four quarters. Adjusted Tier 1 capital was $32,526M against $479,469M of adjusted average assets; holding 6.78% while that base compounds requires retaining 6.78% of the growth - at the +5.5% a year average interest-earning assets actually grew, $1,788M or 18.4% of trailing earnings to common; at the +12.7% total assets actually grew, $4,129M or 42.5%. That is a sustainable payout of 57.5% to 81.6% against the 88.6% being paid, so the current pace is funded PARTLY OUT OF THE BUFFER, which is why the ratio fell. TWO THINGS SEPARATE THIS FROM GOLDMAN'S CET1 SQUEEZE: the 6.75%-7.00% floor is Schwab's OWN operating objective, not a regulatory minimum, so management can choose to run below it; and the adjusted ratio is depressed by a $10,575M accumulated other comprehensive loss it deliberately includes, which amortises back over time and lifts the ratio with no capital action at all. THE THREE-WAY CROSS-CHECK AND ITS LIMIT: the DCF prints $104.30 (7.399x tangible book), the Gordon identity at the disclosed 44% ROTCE prints $107.40 (7.619x), and at the derived trailing-twelve-month 41.4% it prints $100.42 (7.124x) - three methods inside 6.9% against a $108.05 tape at 7.665x. UNLIKE GS, THESE DO NOT AGREE MERELY BY CAPITALISING ONE STRONG HALF-YEAR: the trailing-twelve-month return is within 6% of the basis quarter's, so the trailing check and the current check say the same thing. What breaks the agreement is a THROUGH-CYCLE return - at 30% ROTCE the justified multiple is 4.952x and the value $69.81, 35% below the tape - and that is what the bear case is, not a discount-rate argument. P/TBV IS NOT COMPARABLE TO THE MONEY-CENTRE BANKS WITHOUT THE AOCI ADJUSTMENT: reported TBVPS is $14.0968 and the tape 7.665x, but adding back the $10,575M accumulated other comprehensive loss - unrealised securities marks that amortise back, improving $408M in the first half alone - gives $20.2158 and a 5.345x tape. Tangible book grows mechanically over the horizon without the price moving. SEASONALITY: all five verticals and the consolidated line are ASEASONAL, derived by ratio to a CENTRED four-quarter moving average over the nineteen quarters, never to a year mean, with the gate that a vertical stays aseasonal wherever the worst window-to-window spread meets or exceeds the signal. Four fail outright at ratios of 0.31:1 to 0.90:1, including trading revenue at 0.90:1 against the strong prior that a retail broker should be seasonal. Only Other revenue clears the gate, at 1.81:1, and it is rejected on STABILITY on the same ground GS's Global Banking & Markets was rejected at a stronger 2.34:1: its Q4 window ratios run 0.892, 0.846, 0.849, 0.641, a trend rather than a repeat, on a line worth 2.4%-4.8% of revenue. There is also a mechanical reason no seasonality appears here at all: the projection engine deseasonalises only the growth driver's base and then multiplies every driver's output by the factor, so a capacity or unit vertical whose inputs are the basis quarter's own seasonal levels would be adjusted twice. STANDING LIMITATIONS. First, the page's 'Enterprise value' and 'Net cash' labels, as above. Second, THE FLAT SHARE COUNT IS CORRECT ARITHMETIC BUT MAKES IMPLIED TERMINAL EPS RUN BELOW STREET FIGURES BUILT ON A SHRINKING COUNT: at the first-half pace of 35.5 million shares repurchased, 4.13% a year, the count falls to about 1,657M by end-2027 and 1,400M by 2031, so this model's 2027 EPS of $6.34 becomes $6.62 and its terminal $7.10 becomes $8.77 - a 23.5% gap by the end of the horizon. For scale, our stored analyst ratings record a $133 target on 2027 EPS of $8.32, 31% above even the shrinking-count figure; this base case is materially more conservative than the sell side and says so rather than hiding it. Third, SHARES ARE PERIOD-END BASIC (1,728,119,192), matching BAC, JPM and GS; weighted-average DILUTED is 1,739,000,000, just 0.63% higher, and substituting it moves fair value from $104.30 to $103.65 and the headline from -3.47% to -4.07%. It does not flip the sign, unlike Goldman's 2.2% gap which did. Fourth, the 2026 Summer Business Update deck that carries the guidance is captured locally as a business-update filing and has no captured canonical URL; every guidance figure quoted here was read from that document. Fifth, the base case's full-year 2026 revenue growth of 17.30% sits just below management's 17.5%-18.5% guide, a gap the page should own rather than tune away.

Where each case comes from

From cash flow to fair value

The projected path

The published model's 20 quarters. The horizon and sliders above do not change this table.

Every row is projected. A year-over-year change is shown only where the quarter it compares against exists — an em dash means there is no comparable quarter, not a flat year.

Model revisions

Assumptions are marked to reality as each quarter prints. Every change is appended here, with the fair value the model produced at the time, so the model's own history stays visible.