Block (XYZ): The Most Misread Name in Crypto

The market’s two impressions of Block — cheap on revenue, risky on credit — both misread the income statement. Reported revenue overstates the business (a third is low-margin bitcoin pass-through), and the tripled loss line is mostly accounting timing. Read gross profit, read the provisions correctly, and a reasonably priced compounder appears — on one condition.

TON618 Capital Research · XYZ $78.72 · BTC $62,352 · as of 13 Jul 2026 · Download PDF

1Executive Summary

Block screens cheap and reads risky, and both impressions come from the same place: its income statement is dominated by two things that aren't what they look like — bitcoin revenue that is almost pure pass-through, and a tripled credit-loss line that is mostly an accounting artifact. Strip both away and a clearer business appears: a high-margin gross-profit engine growing in the mid-teens, split between a fast-scaling consumer neobank (Cash App) and a mature merchant acquirer (Square), with a genuine but low-economics Bitcoin identity attached. On our probability-weighted scenario model we estimate fair value at ≈$83 per share (bear $39 / base $81 / bull $134) — roughly 6% above the market, i.e. approximately fairly valued with a modest positive skew. That modest upside is real but it is conditional: it depends almost entirely on one question, which we answer with the primary filings in Section 4 — whether the alarming jump in Block's loss provisions is the front end of a credit problem or the mechanical cost of a lending book growing 175% a year. Our read of the disclosure is that it is predominantly the latter. If that read holds, Block is a reasonably priced compounder; if the credit book is quietly deteriorating, the base case is too high.

The Current State of Block's Story

Block is two businesses wearing a Bitcoin costume. The first business is Square, the little white card reader that lets a coffee shop or a hair salon take payments; over a decade it has grown into a full suite of software and lending for small merchants. The second is Cash App, which began as a way to send friends money and has quietly become something much bigger — a bank account for tens of millions of mostly younger Americans, complete with a debit card, direct deposit, short-term loans, buy-now-pay-later, and, yes, the ability to buy Bitcoin. The Bitcoin costume is the third thing people see: the company renamed itself Block, its founder Jack Dorsey is one of the most vocal Bitcoin advocates in business, and it builds Bitcoin mining hardware, a self-custody wallet, and funds open-source Bitcoin development.

Here is the trap the costume sets. Block's headline revenue — about $24 billion a year — looks enormous, and roughly a third of it is "Bitcoin revenue." But that Bitcoin revenue is almost entirely the price customers pay when they buy Bitcoin in Cash App, which Block books as revenue and then pays right back out to buy the coins. It keeps only a sliver — a few cents on the dollar. Data shows the real measure of the company is not revenue but gross profit — what's left after those pass-through costs — and on that measure Block earned about $11 billion over the past year, growing 16%. Bitcoin trading contributes only about 4% of that gross profit. So the "Bitcoin company" makes almost none of its money from Bitcoin. We believe the market has largely figured this out — which is why Block's stock is near its 52-week high even as Bitcoin has fallen 50% — but the costume still distorts every revenue-based valuation ratio, and any honest analysis has to start by taking it off.

The harder question is the one underneath. In the most recent quarter, Block's provision for loan and receivable losses tripled, and the company swung to an operating loss. On its face that looks like a consumer-credit business cracking at a fragile moment in the economic cycle. But most of that increase, we find, is not losses being incurred — it is losses being reserved for in advance, because Cash App's lending book grew 175% year-over-year and, under the accounting rules, Block must book the full expected lifetime loss on every new loan the moment it is made. The revenue from those loans arrives later; the reserve arrives now. That timing mismatch, plus a one-time legal charge, explains most of the ugly quarter. Whether it explains all of it is the single most important judgment in this note, and we take it up directly in Section 4.

2Investment Context

• What the security is: XYZ is a claim on two gross-profit streams — a growth consumer-fintech (Cash App) and a mature merchant acquirer (Square) — plus a small, high-visibility Bitcoin sleeve. Its earnings variance comes not from the crypto cycle but from the consumer credit cycle, because its fastest-growing product is lending.
• The ticker changed; the company didn't. Block moved from SQ to XYZ in early 2025. Financial history is continuous.
• Price position: XYZ trades at $78.72, near its 52-week high ($82.50), nine months into a 50% Bitcoin drawdown — the market's clearest statement that it no longer prices Block as a crypto proxy. Its crypto-correlation era (2020–21, when it traded with Coinbase) is over; its rolling 90-day beta to Bitcoin has collapsed to ≈0.41, the lowest of the regulated crypto-adjacent complex (see the synthesis note).
• Capital returns are now material: Block repurchased ≈$2.3B of stock in FY2025 and $636M in Q1'26, shrinking the share count from ≈620M to ≈595M in a year — a real per-share tailwind and a signal of management's own value view.

3Business & Revenue Architecture

The one adjustment that governs the whole analysis: value gross profit, not revenue.

*\FY2024 net income is not a clean anchor: it includes a ≈$1.5B one-time income-tax benefit — from a deferred-tax valuation-allowance release and a restructuring-related deferred-tax-asset recognition (≈$1.9B of gross one-time benefits, net of ordinary tax) — against pretax income of only $1,357M. Do not anchor on it.

The table tells the story revenue hides. FY2025 revenue grew 0.3%; gross profit grew 16.5%. Q1'26 revenue grew 5%; gross profit grew 27%, and Cash App's gross profit grew 38% on flat segment revenue — because low-margin bitcoin volume shrank while high-margin financial services grew. Every revenue multiple on this company is noise; the note prices gross profit and what survives below it.

Bitcoin: the identity is not the economics. Bitcoin generated $8.5B of revenue in FY2025 but only ≈$419M of gross profit — about 4% of the company's total, at a ≈4% margin (and only ≈2% of gross profit in the seasonally quieter Q1'26). On the balance sheet Block holds 9,032 BTC (≈$0.56B at spot, cost basis ≈$0.31B), still dollar-cost-averaging ≈150 BTC/quarter, plus the Proto mining-hardware venture, the Bitkey wallet, and Spiral's open-source funding. This is the deepest philosophical Bitcoin commitment in large-cap fintech attached to a P&L that is ≈96% non-Bitcoin. The total value at risk from a further 50% BTC decline — treasury mark plus a slice of trading gross profit — is low-single-digit percent of the equity. Block is the control group of the crypto complex: proof that the market can fully de-correlate a "Bitcoin company" when the economics say to.

Becoming a bank — the growth vector and the risk, which are the same thing. Financial Solutions revenue (Cash App Borrow, Afterpay/BNPL, Square Loans) grew from $875M (Q1'25) to $1,322M (Q1'26), +51%, originated increasingly through Square Financial Services, Block's FDIC-supervised industrial bank. Cash App has ≈59 million monthly actives generating ≈$107 of gross profit each per year (FY25 Cash App gross profit ÷ actives) — a small fraction of what a bank extracts from a primary-account customer with deposits, lending, and interchange. Closing that ARPU gap is the entire bull case. But the instrument of the migration is lending, which is why the growth vector and the credit risk are the same fact — and why Section 4 is the pivot of the whole note.

4Financial Statement Analysis & Quality of Earnings — the pivot

We screened the statements against a standard earnings-quality checklist. Three items dominate, and the first is the most important judgment in this note.

(1) The tripled loss line is predominantly CECL front-loading, not credit deterioration — confidence: high on the mechanics. In Q1'26 the "transaction, loan, and consumer receivable losses" line jumped from $170M to $500M (+195%), swinging operating income from +$329M to −$172M while gross profit grew 27%. From the company's own MD&A table, 91% of that increase is loan losses (+$301M), and the drivers are mechanical:

• Cash App Borrow's allowance was literally $0 at December 2024. The entire book and its $435M reserve were created during 2025 by reclassifying loans from held-for-sale (previously sold, no lifetime reserve) to held-for-investment (full expected lifetime loss booked on day-one origination). The year-over-year comparison is apples-to-oranges: in Q1'25 these loans were not on the balance sheet as held-for-investment at all.
• Originations grew 175% year-over-year. Because these are ultra-short-duration, high-velocity loans, the provision (booked on origination volume) is structurally large relative to the ending balance.
• The reserve-coverage ratio traces a textbook new-book seasoning curve — 2.71% → 9.59% → 10.69% → 13.18% across the last four quarters — climbing from near zero toward a low-teens steady state as the book matures, while quarter-over-quarter balance growth collapses from +171% to +3.7%. Seasoning, not a step-change break.
• Fraud/transaction losses — the cleanest "is the consumer cracking" read — rose only 8%. Management states loss rates and portfolio credit performance "remained stable."

The one caveat that keeps this from "certain" (confidence: medium that nothing is building underneath): within the held-for-investment regime (Q4'25→Q1'26), coverage still rose 2.5 points and the 60–89-day past-due rate nearly doubled, from 2.47% to 4.66%, on a nearly-flat book. That is consistent with a seasoning book aging holiday-vintage originations into later buckets plus Q1 seasonality — but it is the single data point not fully explained by "growth plus accounting." The tell to watch: if Q2/Q3'26 coverage keeps climbing past the mid-teens on a flat or declining balance, the read flips toward genuine deterioration. As of Q1'26 it reads as seasoning, and we build the base case on that — with eyes open.

(2) The operating loss is largely non-operating in character. Beyond the CECL timing, the Q1'26 G&A line jumped $366M (+74%) — but 69% of that ($253.8M) is a one-time accrual for litigation and regulatory settlements, per the MD&A. Normalized G&A ex-litigation rose ≈23%, roughly in line with 27% gross-profit growth. So the headline operating loss is the sum of (a) reserves booked ahead of the revenue they protect and (b) a one-time legal charge — not operational decay. Underlying operating economics improved.

(3) The balance sheet is roughly net-cash-neutral, not cash-rich. Cash $6,858M + marketable securities $617M against total debt of $7,350M (senior notes plus two convertibles) nets to only ≈+$126M; adding the $563M Bitcoin treasury and $88M of other equity gets to ≈$777M ($1.33/share). (This treats the ≈$543M drawn on Block's asset-backed warehouse facilities as matched against the specific consumer receivables that secure it, and so excludes it; were it counted as debt, the net position would be roughly −$400M.) The debt is otherwise well-structured (senior notes laddered 2030–2033 plus two deeply out-of-the-money convertibles struck at ≈$299), and the buyback is funded from cash flow — but there is no large net-cash cushion under the equity. The per-share story is carried by the shrinking share count, not a balance-sheet fortress.

• Auditor Ernst & Young; timely filings; no restatements. Block's earnings question is the credit book and one-time noise, not accounting integrity.

5Valuation

5.1 What model the business dictates

Block's uncertainty lives in the consumer credit cycle, not the crypto cycle — so we do not normalize a BTC path (as in our Coinbase note). And because reported revenue is one-third pass-through, we value gross profit and the normalized operating margin on it, not revenue or unadjusted GAAP earnings (which are polluted by CECL timing, the one-time legal charge, and BTC marks). The workhorse: consolidated gross profit → normalized operating margin on gross profit → net income at a 24% tax rate → P/E, in three scenarios keyed to the credit outcome. Cross-checks: a segment view (Cash App as the growth/multiple driver, Square as the stable base) and a reverse valuation.

5.2 Scenario valuation (normalized FY2027, $M except per share)

Probability-weighted fair value ≈ $83, roughly 6% above the market — approximately fairly valued with modest positive skew. The base case ($81) sits right at the current price and assumes the credit book seasons cleanly, Cash App holds mid-teens gross-profit growth, and the operating margin on gross profit recovers to ≈20% (from FY25 GAAP 16.5%). The bear ($39) is where the credit read is wrong.

5.3 Reverse valuation — the sobering counter-read

Stripping out balance-sheet extras, the market pays ≈$46B for the franchise. At 20×, that requires ≈$2.3B of normalized net income — which, on flat trailing gross profit, implies an operating margin of ≈28% on gross profit, well above the ≈20% normalized level and FY25's 16.5% GAAP. In plain terms: today's price already banks meaningful margin expansion or continued double-digit gross-profit growth. This is not a deep-value setup. Our scenario model's modest upside comes from growth plus a 21× multiple; the reverse valuation is the reminder that the price is not cheap on today's economics — it is fair if Block executes.

5.4 Relative valuation cross-check

Block's 28× normalized multiple is ≈3× PayPal's — a premium the Cash App growth vector can justify, but only if the credit book behaves and margins expand. The multiple is the swing judgment; we hold the base at 21× (below the current 28×) to embed maturation.

6XYZ Versus Direct Bitcoin Ownership

Block completes the trilogy's punchline. Three US-regulated franchises, three relationships to the same asset:

• Coinbase: crypto activity is the P&L; fair value swings ±60% across BTC scenarios, and the equity beats the coin in no modeled state.
• Robinhood: crypto is a ≈13%-of-revenue flow business; fair value moves single digits with BTC.
• Block: crypto is the founder's religion and ≈4% of gross profit plus ≈1.2% of market cap in treasury. Flexing Bitcoin across ≈$40k / $62k / $110k moves Block's fair value by nearly nothing — and the stock made new 52-week highs during the Bitcoin winter. Block is the proof that the market can fully decouple a "Bitcoin company" when the economics warrant.

One honest asymmetry: Proto (mining hardware), Bitkey, and the DCA treasury policy give Block optionality on a hyperbitcoinization tail that Robinhood lacks — worth words, not dollars, in the base case. And the tail runs both ways: were Dorsey to lever the balance sheet into Bitcoin, the beta would re-couple overnight. Today, neither is in the numbers.

7Special Situation: The Separation Question

The perennial speculation is that Block should split Square (mature, cash-generative, attractive to a strategic acquirer) from Cash App (a high-growth consumer bank that would command a richer standalone multiple). Conditional-gate read:

Estimate: ≈5–10% within 24 months. The section earns its place for the SOTP discipline it imposes — it shows Cash App is where the value and the multiple live, and Square is the stable ballast — more than for the probability. Secondary situations: a bank-charter re-rating if Square Financial Services scales deposit funding, and the (economically small, narratively seismic) tail of a Bitcoin-treasury policy change in either direction — the perfect one-paragraph illustration of this note's identity-versus-economics theme.

8Risk Factors

What makes ≈$83 too high (the bear):

1. The credit call is the note. If the 60–89-day past-due creep (2.47%→4.66%) is the leading edge of genuine deterioration rather than seasoning, provisioning stays elevated, growth must be throttled to fix the book, and both earnings and the multiple compress. This is the dominant risk and the thing to monitor every quarter.
2. Multiple compression. At ≈28× normalized earnings, Block is priced as a grower; a growth stumble or a credit scare re-rates it toward PayPal's single digits fast.
3. Consumer-cycle exposure. Cash App's base skews younger and lower-income; a genuine consumer recession hits both spending (Square GMV, Cash App inflows) and credit (Borrow, Afterpay) at once.
4. Square share loss. Its +9% could slow toward flat under pressure from Toast (restaurants) and Fiserv/Clover (SMB).

What makes ≈$83 too low (the bull):

1. Cash App network effects. 59M actives compounding ARPU toward bank-like levels is how the bull beats every multiple argument; each $10 of ARPU-gross-profit closure is ≈+$590M of gross profit.
2. CECL optics reverse. As the loan book stops growing at 175%, provisioning normalizes and reported earnings inflect up mechanically — the mirror image of today's ugly optics.
3. Buyback tailwind. ≈$2.3B/yr of repurchases at a shrinking share count is a quiet, compounding per-share lift our forward count only partly credits.
4. Bitcoin/Proto tail — a genuine hyperbitcoinization leg would make the "4% of gross profit" framing look clever-but-wrong via Proto and narrative re-rating.

Framework note. As with the other two notes in this set, the disagreement worth having is about earnings — here, specifically, about credit — not about multiples. Our verdict is deliberately conditional: it is a bet that Block's disclosure means what it says about loss rates remaining stable. We have stated the exact number that would prove us wrong.

9Catalyst Calendar

10Ten Open Questions, Answered Against the Model

1. Is the tripled loss line CECL or credit deterioration? Predominantly CECL front-loading + a held-for-investment reclassification + 175% origination growth — high confidence on the mechanics, medium confidence nothing is building underneath. The coverage curve is textbook seasoning; the 60–89dpd doubling is the one caveat to watch.

2. What does Block actually earn, normalized? Stripping BTC marks, the one-time legal charge, and CECL timing, ≈$1.7B normalized net income on ≈$11B trailing gross profit — ≈$2.80 EPS, a ≈28× multiple at $79.

3. Why ignore the revenue line? Because ≈35% of it is bitcoin pass-through at ≈4% margin; gross profit ($11B, +16%) is the real scale of the business, and it grows far faster than revenue.

4. What does $79 require? On flat gross profit, a ≈28% operating margin (vs ≈20% normalized) — i.e., the price already banks margin expansion or continued mid-teens gross-profit growth. Not cheap; fair if Block executes.

5. How big is the real Bitcoin exposure? ≈4% of gross profit and ≈1.2% of market cap in treasury; fair value barely moves across BTC scenarios. The identity is not the economics.

6. Cash App vs Square — what's each worth? Cash App (≈$6.9B GP, growing ≈mid-20s%) carries the value and the multiple; Square (≈$4.0B GP, +9%) is stable ballast. A separation modestly increases the sum but not enough to clear the control gates.

7. What multiple does a lending-monetized customer base deserve? Between fintech and bank — we hold 21× base (below today's 28×) to embed maturation; the bull argues the neobank scales before the multiple mean-reverts.

8. Is Square a stable annuity or slow share loss? Base case: stable at ≈+9%, needing only "doesn't melt." The risk is Toast/Clover pressure; we do not underwrite acceleration.

9. Does the buyback matter? Yes — ≈$2.3B/yr at ≈595M shares and falling is a real per-share tailwind our forward share count only partly credits.

10. What would move us to the bull? Two clean quarters of seasoning (coverage stabilizing, dpd receding), Cash App ARPU visibly closing the bank gap, and Square holding +9%. Then fair value migrates toward $110–130 with no Bitcoin bull required — and we would update on the evidence.

11Conclusion

Block is the most misread name in the regulated crypto complex, and both directions of the misreading come from its income statement. The bears see a "Bitcoin company" cracking under credit losses; the bulls see a cheap stock on a huge revenue base. Both are looking at the wrong lines. Reported revenue overstates the business (a third is low-margin bitcoin pass-through); the crypto exposure is ≈4% of gross profit; and the frightening loss line is mostly the accounting cost of a lending book compounding 175% a year. Underneath sits a mid-teens gross-profit compounder — a scaling consumer neobank stapled to a mature merchant acquirer — priced at roughly fair value with a modest positive skew. Our ≈$83 estimate is deliberately conditional: it is a bet, backed by the disclosure, that Block's loan losses are timing and not trouble. The entire thesis reduces to a number the reader can watch each quarter — Cash App Borrow's coverage ratio and its 60–89-day past-due rate. If they stabilize, Block is a reasonably priced compounder that the market has correctly de-crypto-fied; if they keep climbing on a flat book, the base case is too high and the bear is the story. Everything else — the Bitcoin costume, the overstated revenue line, the ugly headline quarter — is noise around that one signal.