2026-07-28 · June quarter and half-year statements; shares, cash, loans, leases and credit quality. Primary financial statements control.
https://www.sec.gov/Archives/edgar/data/1633917/000163391726000082/pypl-20260630.htmWatchlist / branded profit recovery required
The investment decision
The rebuilt base case produces 16.0% five-year total-wealth CAGR at $56.10, or 13.4% without earnings-multiple expansion. The alternative owner-cash valuation produces 13.4%. This is a conditional recovery investment, with no automatic 20% base case.
Edition update. Expanded the September concise report into a full decision-focused edition and replaced rounded workbook outcomes with a reproducible October cash/share model. Reconciled annual and June filings; corrected small rounded TTM values. Actuals remain through June 2026, not a new quarter. The current price is the 14 October regular close; buyback prices and terminal cash assumptions are separately exposed.
01. The investment debate is profit quality, not payment volume
PayPal remains a profitable payments franchise with recognised brands, stored credentials, merchant reach and risk-management infrastructure. Yet merchants can use several processors and consumers can choose another wallet at checkout. The economic question is whether its higher-value branded activity retains enough relevance to fund innovation and cash returns as competitive processing grows. A low P/E is a starting valuation, not evidence of a moat repairing itself.
The mix matters. Branded checkout, Venmo and payment processing do not have identical margins or customer economics. Volume can expand quickly while transaction-margin dollars and operating income lag. The quarterly release reports $486.4bn total payment volume, $8.682bn revenue and a 16.4% GAAP operating margin. Payment volume grew approximately 10%, revenue approximately 5%, and operating income fell approximately 5%. Those facts argue for testing profit per activity rather than celebrating scale alone. [S1][S4]
PayPal should not receive Visa’s or Mastercard’s valuation merely because all operate in payments. Network rails, wallets, processing and consumer credit have different competitive structures and capital needs. Adyen and Stripe challenge merchant distribution; device-based wallets challenge checkout placement. These comparisons inform the underwriting mechanism, but no stale peer-multiple table is presented as current evidence.
The constructive path is modest revenue growth, better mix and execution, and disciplined cash-funded net share retirement. The adverse path is that shrinking relevance in branded payments is replaced with lower-margin activity, while fraud, incentives and credit costs absorb the cash supposedly available for buybacks.
02. Revenue growth is outpacing operating progress
| USD bn except EPS | FY2024 | FY2025 | H1 2025 | H1 2026 |
|---|---|---|---|---|
| Revenue | 31.797 | 33.172 | 16.079 | 17.035 |
| Operating income | 5.325 | 6.065 | 3.034 | 2.915 |
| Net income | 4.147 | 5.233 | 2.548 | 2.217 |
| Diluted GAAP EPS | $3.99 | $5.41 | $2.58 | $2.46 |
| Operating cash flow | 7.450 | 6.416 | 2.058 | 3.117 |
| Cash capex | 0.683 | 0.852 | 0.402 | 0.439 |
| Calculated FCF | 6.767 | 5.564 | 1.656 | 2.678 |
The TTM bridge is FY2025 plus H1 2026 less H1 2025: $34.128bn revenue, $5.946bn operating income, $4.902bn net income and $5.29 diluted EPS. Revenue and net income differ slightly from the rounded September presentation; this edition uses the filed numbers. TTM FCF is $7.475bn CFO less $0.889bn capex, or $6.586bn. Half-years are not annualised. [S1][S2]
July management guidance calls for approximately $5.38 adjusted FY2026 EPS and a mid-single-digit decline in GAAP EPS from the prior $5.41. These are different bases. Our valuation uses a GAAP-like operating model with compensation costs retained; it does not apply a normalised multiple to an adjusted-EPS series without a bridge. The 2026 transformation programme may improve costs, but gross savings must first pay for reinvestment, transition charges and competitive response. [S4][S6]
No third-quarter results are assumed in this October edition. Earlier external consensus and takeover commentary are excluded from valuation because a public rumour or a dated forecast is not current issuer evidence or a binding transaction.
03. A high headline cash yield needs an owner-cash bridge
| TTM June 2026 cash bridge | USD bn |
|---|---|
| CFO less cash capex | 6.586 |
| Less stock compensation expense / economic award reserve | (1.001) |
| Less net held-for-investment loan funding | (0.653) |
| Conservative owner-cash proxy | 4.932 |
This bridge treats employee awards as an economic cost and funds loans before describing cash as available to owners. Net award-settlement tax cash less issuance proceeds was $216m, but it is not deducted again: the withheld shares overlap the full economic award cost already reserved through SBC. This corrects the prior edition’s more punitive bridge. Held-for-sale loan timing already passes through reported CFO, so it is not deducted again. The $653m additional loan deduction covers the net funding shown in investing cash flows. [S1][S2]
The result is a useful economic proxy, not a new GAAP metric. On June common shares of 862m, the October price implies roughly $48.36bn equity value: headline FCF yield is 13.6%, while this owner-cash proxy yields 10.2%. Neither is a cash distribution promised to investors. Six-month cash taxes fell to $225m from $837m, illustrating why one period’s cash conversion should not simply become a perpetual assumption. [S1]
Our new forward owner-cash model is more cautious. It starts from GAAP-like net income and never adds SBC back, effectively reserving the economic cost of offsetting employee issuance. D&A is assumed at 2.8% of revenue, capex 2.6%, and net loan funding 1.9%. No separate settlement-cash deduction is made after the full award reserve. Thus forward owner cash equals net income less 1.7% of revenue. This is a disclosed normalisation assumption, not management FCF guidance.
Net retirement is then funded only from this after-cost cash. We do not subtract the SBC reserve and also insert a second full award-dilution charge. Equally, share retirement already increases EPS, so no separate buyback yield is added to shareholder returns.
04. Corporate cash, credit and the share denominator
June eligible corporate liquidity of about $13.530bn comprises cash, short-term investments and non-strategic long-term investment balances. Funded debt is $13.400bn. Customer account assets are excluded from corporate liquidity; they are not an available buyback budget. The model gives no extra terminal value to existing corporate liquidity and does not assume it funds forecast repurchases. [S1]
Operating lease liabilities were $810m, with $11m finance leases disclosed separately. Rent and lease expenses remain in operating earnings. The forward model uses $300m annual other net expense and a 20% tax rate; both are house normalisations rather than predictions of investment marks, currency effects or precise refinancing costs. A large debt-funded acquisition would require a fresh financing bridge.
Credit deserves its own monitor. Consumer net charge-offs were 4.8% and merchant net charge-offs 7.6% on the issuer’s annualised Q2 definitions, above the prior-year 3.6% and 6.1%. A lower reported credit-loss expense can coexist with higher charge-offs when reserve movements differ. We therefore monitor losses, delinquencies and funding needs together, rather than using one expense line as proof of improving credit. [S1]
The starting 882m diluted-share proxy is the Q2 weighted average, not the June legal common-share balance of 862m. Using the larger proxy in the forward model is conservative and makes dilution visible. Forecast net repurchases reduce that proxy; a detailed vesting and treasury-stock schedule is not claimed. Current institutional ownership changes, borrow rates and options positioning are unverified and are not scored.
05. New five-year scenarios: cash disciplines the EPS story
Revenue starts at the reconciled $34.128bn TTM level. Net income equals revenue times the scenario operating margin, less $300m other net expense, taxed at 20%. The model then calculates owner cash, dividends and net share retirement each year. Terminal run-rate EPS divides year-five net income by year-end shares; the annual average-share EPS shown later is a different denominator. All future operating and capital-allocation inputs are Capital Thesis assumptions.
| Assumption / output | Bear | Base | Bull |
|---|---|---|---|
| Revenue annual growth | −2.0% | 4.0% | 6.5% |
| Terminal operating margin | 12.5% | 19.0% | 22.0% |
| Cash after dividends allocated to net buybacks | 40% | 80% | 85% |
| Terminal shares, million | 820.6 | 671.9 | 605.6 |
| Terminal run-rate EPS | $3.47 | $9.04 | $13.19 |
| Exit earnings multiple | 8× | 12× | 15× |
| New retained cash per ending share | $8.59 | $6.20 | $6.46 |
| Cumulative dividends | $2.00 | $3.09 | $3.29 |
| Terminal wealth | $38.32 | $117.73 | $207.62 |
Bear five-year total-wealth CAGR: -7.3%.
Base five-year total-wealth CAGR: 16.0%.
Bull five-year total-wealth CAGR: 29.9%.
Base five-year total-wealth CAGR without rerating: 13.4%.
The no-rerating comparison uses $56.10 ÷ $5.29 = 10.6049× earnings, keeping the same net share path, retained cash and dividends. The valuation includes only newly retained corporate cash, which earns zero in the forecast. That cash cannot also fund an unmodelled acquisition or further buybacks. Dividends accumulate without reinvestment, and the resulting CAGR is not an IRR.
The bear remains profitable and conserves more cash; it is not a severe-loss bound. Fraud, a major security event, loan-sale disruption or structural checkout erosion could produce worse results. No probabilities are attached to the three scenarios.
06. Follow the dollars into the share count
| Year E | Revenue, $bn | Operating margin | Average-share EPS | Owner cash, $bn | End shares, m |
|---|---|---|---|---|---|
| 2027 | 35.493 | 17.0% | $5.34 | 3.984 | 837.6 |
| 2028 | 36.913 | 17.5% | $6.04 | 4.300 | 794.3 |
| 2029 | 38.389 | 18.0% | $6.84 | 4.635 | 752.2 |
| 2030 | 39.925 | 18.5% | $7.75 | 4.990 | 711.4 |
| 2031 | 41.522 | 19.0% | $8.78 | 5.365 | 671.9 |
The base model buys back $2.79bn of net shares in year one, rising to $3.90bn in year five, without new net borrowing. The repurchase-price assumption is $56.10 growing 12% annually: $62.83 in year one through $98.87 in year five. These are treasury-policy cost assumptions, not a projected stock-price chart or an expected trading path. Higher repurchase prices would retire fewer shares and lower the forecast EPS benefit.
Dividends begin at $0.56 per share and grow 5% annually. After dividends, 80% of available owner cash funds net repurchases; the remainder accumulates. By year five, newly retained cash is $4.165bn, or $6.20 per ending share. Net income reaches $6.071bn and terminal run-rate EPS $9.04. The annual average-share EPS is $8.78 because buybacks occur through the period; the report does not label both as the same EPS statistic.
This is a tighter capital-allocation case than September’s model, which projected approximately 616.6m shares and $10.24 EPS in the base case. The new model ends around 671.9m shares and does not carry the old near-20% headline forward. The revised policy sacrifices apparent upside in exchange for a transparent funding constraint. [S5]
A cash-flow DCF provides a separate present-value check. Discounting owner cash before dividends and buybacks at 11% with 2.5% perpetual growth gives about $62.74 per starting diluted share; 10–12% produces roughly $56–71. About 69% of the central value comes from the terminal period. No existing liquidity surplus is added. This is an alternative valuation, never an extra component of terminal wealth.
07. A 20% return is a hurdle, not a base-case label
The conservative owner-cash conditional entry for a 20% five-year CAGR is $42.24 per share, versus $47.31 under the earnings-multiple model. Both backsolves hold operating forecasts, future repurchase prices, net share paths, dividends and terminal retained cash fixed; changing actual entry price does not automatically re-underwrite those assumptions.
The entry price of $56.10 requires $139.59 terminal wealth for 20% annually. With the base share and cash path fixed, terminal earnings would need a 14.42× multiple rather than 12×. That would turn a large part of the desired return into a rerating requirement.
For the alternative cash valuation, terminal owner cash per ending share is capitalised at the same 12× as the earnings model, then new retained cash and dividends are added. It produces 13.4% at the reference price. This deliberately separates earnings from economic cash after compensation and loan funding. It is a cross-check; the two returns cannot be averaged or added without an explicit investment framework.
The website’s newer-price sensitivity also holds the forecast wealth fixed. A lower screen price after deteriorating checkout economics is not automatically better value. It calls for lower cash forecasts or more conservative capital allocation. Price-only sensitivity is helpful for monitoring but cannot replace re-underwriting after material results.
The September $49 owner-cash threshold and near-20% earnings return no longer describe this model. The new cash path, net buyback policy and higher entry price are all visible reasons for the revision.
08. Evidence that would earn a stronger view
| Monitor | Stronger evidence | Re-underwrite trigger |
|---|---|---|
| Branded profitability | Profitable checkout growth and better transaction-margin dollars | Flat demand or incentives offsetting revenue growth |
| Operating margin | Progress toward 17–19% without rising exclusions | Sustained sub-15.5% margin after transition timing |
| Owner cash | Cash consistently supports net retirement and loan funding | Below $3.5bn annualised through a cycle |
| Shares and funding | Net reduction with stable debt and adequate liquidity | Large buybacks fail to reduce diluted shares |
| Credit quality | Charge-offs and funding demands stabilise | Further material deterioration in consumer or merchant credit |
The next earnings release, a reconciled 2027 operating outlook and evidence from the transformation programme are the useful catalysts. The company’s cost ambition needs to become net profit after reinvestment. Venmo monetisation and processing growth matter only if their contribution outweighs acquisition, funding, fraud and incentive costs. An announced partnership is not valued as guaranteed earnings.
The critical failure mode is loss of profitable checkout relevance faster than new services replace it. A low multiple can be deserved if earnings power is permanently declining. A strong cash quarter helped by taxes or reserve timing does not disprove that risk, and shrinking share count cannot indefinitely offset a shrinking profit pool.
Actuals end June 2026; this is a 15 October evidence and model revision, not a claim that new quarterly financials exist. Annual intervals are approximate five-year holding-period proxies; no fiscal stub, full debt-maturity or legal-contingency model is supplied. Figures distinguish GAAP, issuer-adjusted measures and house assumptions. Returns are nominal USD before tax, fees and currency effects. General investment research, not personalised financial advice.
Financial exhibits
Exact observations accompany every chart. Actuals, management guidance and Capital Thesis assumptions are labelled separately; the chart shapes do not imply probabilities.
Reported cash recovered faster than earnings in the first half
The gap warrants a cash-quality check; tax and loan timing can support CFO without an equivalent operating improvement.
| Observation | Net income | Operating cash flow | Calculated FCF |
|---|---|---|---|
| H1 2025 | 2.548 | 2.058 | 1.656 |
| H1 2026 | 2.217 | 3.117 | 2.678 |
Reported six-month statements; FCF subtracts all reported cash capex. Periods are not annualised. Cash tax payments were $837m and $225m respectively. Sources: [S1]
Earnings are larger than the cash available for owner returns
The base reserves cash for capex, net loan funding and awards before funding net repurchases.
| Observation | Normalised net income | Owner cash | Net repurchase budget |
|---|---|---|---|
| Year 1E | 4.5871 | 3.9837 | 2.7918 |
| Year 2E | 4.9278 | 4.3003 | 3.0462 |
| Year 3E | 5.2881 | 4.6354 | 3.316 |
| Year 4E | 5.6689 | 4.9902 | 3.602 |
| Year 5E | 6.0713 | 5.3655 | 3.905 |
New October Capital Thesis model. SBC remains expensed; no SBC cash add-back. Repurchase budget is after owner costs and dividends, and is not a second return. Sources: [S1] [S2] [S5]
Charge-offs rose despite lower reported credit-loss expense
Reserve movements and actual charge-offs can tell different stories, so monitor both.
| Observation | Consumer net charge-off rate | Merchant net charge-off rate |
|---|---|---|
| Q2 2025 | 3.6 | 6.1 |
| Q2 2026 | 4.8 | 7.6 |
Issuer annualised quarterly net charge-offs relative to average daily loan balances, excluding fraud. Consumer and merchant portfolios have different risk profiles. Sources: [S1]
Source register and research basis
Evidence checked for this edition on 15 October 2026. Dates below identify source vintages. Company guidance and house scenarios are conditional; source links and the downloadable model allow the assumptions to be inspected.
2026-02-03 · Annual statements used to reconcile TTM results; business and capital-allocation risks.
https://www.sec.gov/Archives/edgar/data/1633917/000163391726000024/pypl-20251231.htm2026-10-14 · Regular close $56.10 at 4:00 pm EDT; public observation, not guaranteed execution or real-time delivery.
https://stockanalysis.com/stocks/pypl/statistics/2026-07-28 · Issuer operating metrics, July guidance, dividend and GAAP/non-GAAP reconciliation.
https://www.sec.gov/Archives/edgar/data/1633917/000163391726000080/pypl2q-26earningsrelease.htm2026-09-16 · Original research and rounded workbook outcomes, superseded by this explicit October model. Not external consensus.
https://capital-thesis.pages.dev/report?slug=pypl-2026-09-16-concise-7893ad6a5d8d2a73e9ad2026-07-28 · Issuer business-unit metrics and transformation framework. Strategy and targets are management claims, not model outcomes.
https://s205.q4cdn.com/875401827/files/doc_financials/2026/q2/PYPL-2Q-26-Earnings-Presentation.pdf