XYZ The DOJ Accrual More Than Doubled — So Why Did Profit Come Back?

The Loss Reversed, and the Legal Liability Grew Anyway

One quarter after a $308.7 million GAAP net loss built almost entirely from restructuring and legal charges, Block posted a $88.5 million GAAP profit. Gross profit grew 25% year-over-year to $3.17 billion, Adjusted Operating Income hit a record $864 million (27% margin, an all-time high), and Adjusted Diluted EPS grew 65% to $1.02 — Block itself calls the quarter's growth "strong financial performance" and raised full-year guidance on the back of it. Reading only the headline numbers, this looks like the workforce cut and the AI-native rebuild described last quarter are paying off exactly as management said they would.

The footnotes tell a less tidy story. Block's accrual against its unresolved Department of Justice investigation — first disclosed at $240 million in Q1 2026 — is now $526 million, more than double in a single quarter (see Beyond the Usual below). That increase shows up quietly inside a 84% year-over-year jump in G&A expense that management attributes to "increased accrued legal contingencies," not as a headline the Shareholder Letter calls out on its own. A reader who stops at the "Q2'26 Highlights" table would have no way of knowing the company's largest unresolved legal exposure got meaningfully bigger in the same three months it returned to profit. Both facts are true at once: the underlying business is genuinely growing faster and more profitably than a year ago, and the DOJ matter is not converging toward resolution — it's getting more expensive, with Block's own disclosure language unchanged from last quarter: "it is reasonably possible that the ultimate resolution could result in losses in excess of the amount accrued and such losses could be material."

The workforce reduction itself is now complete. The Workforce Plan that cut headcount by more than 40% "concluded during the second quarter of fiscal 2026," with Q2 restructuring charges "immaterial" against the $495 million already booked in Q1 — meaning this quarter is the first real read on whether the AI-driven engineering efficiency story holds up without a one-time restructuring charge distorting the comparison. Product development expense fell 16% year-over-year on reduced personnel costs, and management's framing continues to lean on Builderbot, its internal AI code-review tool, which it says now reviews and writes "nearly all" production code changes as of June — an escalation from Q1's "over 90% of production code changes" and "15% nearly fully autonomously." Whether that's durable engineering capacity or a smaller team running faster on borrowed time is still the open question this series flagged last quarter, and one more quarter of clean data doesn't settle it either way.

The Prescription

Block should keep pointing to what's now two consecutive quarters of evidence: Cash App gross profit growing faster than Square's (31.5% vs. 13.0% year-over-year this quarter), driven by a credit business — Cash App Borrow, Afterpay Post-Purchase — that keeps scaling origination volume (Consumer Lending origination volume up 59% year-over-year to $18.9 billion) while management continues to describe risk loss rates as "healthy." That's the actual engine of this business, and it's the one segment of the letter that reads as unambiguously good news without a footnote attached.

What Block should stop doing is separating the DOJ accrual's growth from the quarter's profitability narrative. The Shareholder Letter frames the $88.5 million GAAP profit and the raised guidance as validation of the AI/hardware "capability compounding" thesis that opens the letter, without once naming that the same quarter's G&A growth was driven in meaningful part by a legal contingency that just got $286 million more expensive. That's not a disclosure failure — the number is in Note 16 of the 10-Q, plainly stated — but leading with "strong financial performance" while burying a doubling legal liability in a footnote is the same sequencing choice this series flagged in Q1, made again in Q2. A reader tracking Block quarter to quarter deserves to know the DOJ matter is trending worse, not just that it exists.

Key Financial Metrics

Q2 2026 vs. Q2 2025, all figures in USD as reported

Metric	Q2 2026	Q2 2025	YoY	Note
Total net revenue	$6,617.7M	$6,054.5M	✅ +9.3%	Bitcoin ecosystem revenue fell 12.8% on a strategic fee cut; both other revenue lines grew faster than the total
Gross profit	$3,166.1M	$2,536.5M	✅ +24.8%	The metric Block itself features; matches the 25% figure in the Shareholder Letter after rounding
Operating income	$446.9M	$484.3M	⚠️ -7.7%	Q2 2025 had no comparable legal/regulatory accrual growth; see Beyond the Usual
Adjusted Operating Income	$864M	$550M	✅ +57.1%	Record; 27% margin on gross profit, up from 22% a year ago and 25% last quarter
Net income attributable to common stockholders	$88.5M	$538.5M	⚠️ -83.6%	Q2 2025 included a $212.2M bitcoin remeasurement gain; Q2 2026 had an $88.5M remeasurement loss instead — a roughly $300M swing on that line alone
Adjusted EBITDA»	$1,169M	$891M	✅ +31.2%	Record; 37% margin on gross profit, up from 35% a year ago
Diluted EPS	$0.15	$0.87	⚠️ -82.8%	Same bitcoin-swing and legal-accrual dynamics as net income
Adjusted Diluted EPS	$1.02	$0.62	✅ +64.5%	Record
Free cash flow» (operating cash flow less capex)	$966.6M (operating cash flow ~$1,019.8M, capex ~$53.2M)	Not separately disclosed in this filing (six-month figures only)	—	Trailing-twelve-month FCF of $3,882M, per the Shareholder Letter's own reconciliation
Cash and cash equivalents (quarter-end)	$6,427.3M	Not disclosed in this filing (10-Q balance sheet compares only to Dec 31, 2025)	—	Total liquidity, including restricted cash/customer funds/marketable securities and credit-facility availability, was $8.8B per the Shareholder Letter

Operating income actually declined 7.7% year-over-year even as gross profit grew nearly 25% — the opposite pattern from Q1, when GAAP operating income swung to a loss on a one-time restructuring charge. This quarter's gap is smaller and structural rather than one-time: G&A expense grew 84% year-over-year, which management attributes to "increased accrued legal contingencies" — consistent with the DOJ accrual's growth from $240 million to $526 million disclosed in the same filing (see Beyond the Usual). Net income and diluted EPS both look far worse year-over-year than the operating story alone would suggest, but that's mostly a bitcoin accounting swing, not a business problem: Block's bitcoin investment posted a $212.2 million unrealized gain in Q2 2025 and an $88.5 million unrealized loss in Q2 2026, a roughly $300 million swing on a single non-operating line that has nothing to do with Cash App or Square's underlying performance.

Segment Comparison

Block reports two segments — Cash App (consumer financial services) and Square (seller commerce and payments) — plus a small "Corporate and Other" category for products not assigned to either.

Segment	Q2 2026 Gross Profit	Q2 2025	YoY	Q2 2026 GPV»	GPV YoY
Cash App	$1,973.0M	$1,500.5M	✅ +31.5%	Not separately GPV-tracked (see below)	—
Square	$1,160.2M	$1,026.8M	✅ +13.0%	$72,800M	✅ +13%
Corporate and Other	$32.8M	$9.2M	—	—	—
Total	$3,166.1M	$2,536.5M	✅ +24.8%	—	—

Cash App keeps carrying the group's profit growth by a wide margin (31.5% vs. Square's 13.0%), continuing the pattern from Q1 2026 (38.3% vs. 9.4%) — the gap has narrowed slightly quarter-over-quarter, but Cash App is still growing more than twice as fast as Square on the metric Block itself uses to run the business.

Cash App

Cash App gross profit grew 31.5% year-over-year to $1,973.0 million, a deceleration from Q1 2026's 38.3% but still the segment's second-fastest reading in the five quarters this letter discloses. Growth was broad-based: Commerce Enablement volume (Cash App Card, Cash App Pay, Afterpay, Cash App Business) grew 17% to $56.5 billion, and Consumer Lending origination volume — Cash App Borrow plus BNPL products — grew 59% year-over-year to $18.9 billion, driven largely by Borrow. Primary Banking Actives reached 9.4 million, up 17% year-over-year, while monthly transacting actives stayed roughly flat at 59 million — the same "convert existing users into paycheck-routing users, not just more users" pattern flagged in Q1, still holding a quarter later. Management describes cohort-level Borrow risk loss rates as "healthy" as origination shifts toward a six-week loan product, though — unlike Q1's letter, which disclosed specific cohort loss-rate percentages (3.16%, 3.01%, 2.67%) — this quarter's letter doesn't repeat those numbers, an omission worth watching if it continues next quarter.

Square

Square gross profit grew 13.0% year-over-year to $1,160.2 million on GPV that also grew 13% (11.5% constant currency) to $72.8 billion — accelerating from Q1's 13.1% reported growth in dollar terms while U.S. GPV growth specifically hit its strongest rate since Q2 2023. Management flagged that roughly two points of this quarter's Square gross-profit growth came from a one-time tariff reimbursement, which happened to roughly offset the lapping of last year's network remediation payment — two offsetting one-time items in the same quarter, netting to something close to the segment's underlying growth rate. International GPV grew 28% (25% constant currency), continuing the geographic-diversification trend already visible in Q1. New Volume Added from Independent Sales Organization partnerships — the newer distribution channel flagged last quarter — isn't broken out with a specific growth rate this quarter, but ISO partnerships are cited again as a contributor to new-customer acquisition.

Beyond the Usual

The DOJ accrual more than doubled, from $240 million to $526 million, in a single quarter

Block's accrued estimate of loss tied to its ongoing Department of Justice inquiry rose from $240 million as of Q1 2026 to $526 million as of Q2 2026 — an increase of $286 million in three months. The underlying matter is unchanged from last quarter: the inquiry began after a March 2023 short-seller report, the SEC closed its own parallel investigation in March 2026 with no enforcement action, and the DOJ presented settlement terms in March 2026 that Block says it "disputes... the basis and methodology" of, with a Block counterproposal and ongoing negotiations as of this filing. What's changed is the size of the number Block itself now believes it's more likely to owe. Block's disclosure language is identical to last quarter's — "it is reasonably possible that the ultimate resolution could result in losses in excess of the amount accrued and such losses could be material" — but an accrual growing by more than 100% in one quarter, without a stated reason for the increase beyond ongoing negotiation, is the opposite of a matter trending toward resolution. This is a bigger, not smaller, unresolved liability heading into Q3.

The securities class action and derivative litigation remain unresolved, with no new developments this quarter

The federal securities class action alleging Block made misleading statements about its anti-money-laundering and compliance programs (covering share purchases between February 2020 and August 2024) and the related shareholder derivative suits — both of which survived motions to dismiss on January 6, 2026 — show no material procedural developments in this quarter's filing beyond what was already disclosed in Q1: the board's special litigation committee, formed in April 2026 to investigate the derivative claims, is still active, and Block still states it cannot estimate a loss or range of loss for either matter, only that a loss is "reasonably possible" and "could be material." The absence of new developments isn't itself a finding, but a matter this consequential going a full quarter without visible progress toward resolution is worth continuing to track.

Block settled a multistate Attorneys General complaint over Cash App's complaint-handling — for an amount small enough that it isn't quantified

In July 2026, after the quarter closed but before this filing, Block entered into a resolution with Attorneys General from multiple states to settle claims related to Cash App's handling of customer complaints and disputes, which Block describes as "for an immaterial amount." This is a new disclosure not present in the Q1 2026 filing — a multistate AG action reaching settlement is a real regulatory data point about how Cash App's consumer-complaint process was being scrutinized, even though the dollar amount itself doesn't move the needle financially.

Block bought back $701 million of stock this year while its bitcoin investment lost value

Block repurchased 11.6 million shares of Class A common stock for $700.9 million in the first half of 2026, under a program the board expanded by $5 billion in November 2025 (to $9 billion total authorization); $4.6 billion of that authorization remains. Over the same period, Block's bitcoin investment — a deliberate treasury holding since 2020 — grew in bitcoin count (from 8,883 to 9,117 coins) but fell in fair value from $777.5 million at year-end 2025 to $533.9 million at quarter-end, as bitcoin's price declined faster than Block kept adding to the position. Both are ordinary, disclosed capital-allocation choices, not red flags on their own — but a company simultaneously buying back its own equity and adding to a volatile treasury asset that's currently underwater on a mark-to-market basis is worth a reader's attention as two different bets running in parallel.

Block's unrecognized share-based compensation cost grew to $2.5 billion (from $1.8 billion at the end of Q1 2026), now expected to be recognized over a weighted-average 3.0 years rather than Q1's 2.4 years — a longer runway of future dilutive expense than a quarter ago, not a shorter one, despite the headcount reduction that concluded this quarter. Non-cancelable cloud-computing purchase commitments stand at $2.30 billion as of Q2 2026 (down slightly from $2.24 billion at Q1 2026, with $280.1 million due in the remainder of 2026), a real multi-year fixed obligation running in parallel with the AI-infrastructure story management keeps telling. Square Loans cohorts have had loss rates "of less than 4% through every cycle we've seen," per this quarter's letter — the first time management has framed the loan book's loss performance across cycles rather than quarter to quarter, worth watching as a benchmark against future quarters' actual cohort data.

Management's Framing of "Compounding" Doesn't Mention the DOJ Number Getting Bigger

Block's Q2 2026 Shareholder Letter opens with an extended narrative about hardware and AI as the "two capabilities compounding fastest right now" — the Square card reader as the original example of "a hard capability owned end to end that converts directly into growth," extended through Bitkey, Cash App Cards, and this quarter's launch of Cash App Tags (NFC chips linking physical objects to a Cash App Card). On AI specifically, management describes staying "model agnostic" through internally built tools (goose, Builderbot, Moneybot, Managerbot, and a newly launched internal tool called Buzz for "agent collaboration, communication, and code repositories"), with Builderbot now handling "over 200,000 operations per day" as of last quarter and, this quarter, helping "write and review nearly all" production code changes in June.

That strategic framing — proprietary capability as durable moat against commoditized AI models — occupies roughly the first third of the letter before a single financial metric appears. What it doesn't mention anywhere is the DOJ accrual's growth from $240 million to $526 million, disclosed later in the same filing's legal-proceedings footnote. Nothing about that omission is unusual for a shareholder letter, which is a marketing document by design, not a legal disclosure — but the contrast between an extended, confident growth narrative up front and a doubling legal liability mentioned nowhere in that narrative is the same sequencing pattern this series flagged in Q1, repeating in Q2.

Target Valuation Range

Modestly overvalued against a base-case DCF^» given the stock's move over the quarter, with the DOJ accrual's growth (see Beyond the Usual above) an underweighted risk in a price that rallied 26% off the Q1 close. A DCF using Block's disclosed trailing-twelve-month Free Cash Flow of $3.88 billion implies an equity value of roughly $57-$270 per share across bear/base/bull scenarios (base ~$140), against an actual June 30, 2026 close of $76.00 — inside the range, between the bear and base cases and meaningfully above where the stock sat at the end of Q1 ($60.18).

Using 600.76 million shares outstanding (540.78 million Class A plus 59.98 million Class B) and the June 30, 2026 close of $76.00:

Market cap → enterprise value	Q2 2026
Share price (period-end)	$76.00
Shares outstanding	600.76M
Market capitalization	~$45.66B
Plus: total liabilities	$17.15B
Less: cash and cash equivalents	$6.43B
Enterprise value	~$56.38B

The multiple expansion since Q1 2026 is real, and it's running faster than the underlying fundamentals grew:

Peer-multiple sanity check	Q1 2026 (TTM)	Q2 2026 (TTM)	Change
Market capitalization	~$35.82B	~$45.66B	✅ up 27.5%
Enterprise value	~$47.27B	~$56.38B	✅ up 19.3%
Revenue (TTM)	$24,478M	$25,042M	✅ up 2.3%
Adjusted EBITDA (TTM)	$3,664M	$3,942M	✅ up 7.6%
EV/Revenue»	~1.93x	~2.25x	⚠️ up
EV/EBITDA»	~12.9x	~14.3x	⚠️ up
P/S Ratio»	~1.46x	~1.82x	⚠️ up

A stock re-rating 20-28% higher on trailing fundamentals that grew only single digits is either the market pricing in a genuinely stronger forward outlook (the raised full-year guidance, two consecutive quarters of Cash App acceleration) or a market getting ahead of itself on a quarter whose GAAP profit recovery partly reflects a bitcoin-accounting swing rather than pure operating improvement — this series can't settle which from the numbers alone, but the gap between multiple expansion and fundamental growth this quarter is real and worth naming plainly.

Block's disclosed trailing-twelve-month Free Cash Flow^» (operating cash flow less capex, per the Shareholder Letter's own non-GAAP reconciliation) is $3.882 billion, up from $3.258 billion a quarter ago. As with Q1, this figure is lumpy quarter to quarter for a lending-heavy fintech — treat the DCF below as an illustrative sanity check on that base, not a precision instrument:

Scenario	Key assumption	WACC	Implied per-share value
Current (period-end close)	— actual market price, for reference	—	$76.00
Bear	8% initial FCF growth decelerating to a 2% terminal rate, reflecting the DOJ overhang (now larger) and unresolved securities/derivative litigation	12%	~$57
Base	20% initial FCF growth decelerating to a 4% terminal rate	10%	~$140
Bull	30% initial FCF growth decelerating to a 5% terminal rate, reflecting the AI-driven margin story sustaining across a second consecutive quarter	9%	~$270

Each scenario's enterprise value is converted to an equity value using the same bridge as the table above (less total liabilities, plus cash). The actual close now sits closer to the base case than the bear case — a meaningful shift from Q1, when the close sat closer to bear — even though the one incremental piece of hard information this quarter (the DOJ accrual doubling) points toward more risk, not less. That's not necessarily a contradiction — the operating business genuinely did perform better — but it means the market's current price embeds less of a discount for the DOJ and litigation overhang than it did three months ago, at the same time that overhang got measurably larger.

Block's Stock Recovered Most of Its Two-Year Round Trip

XYZ closed at $76.00 on June 30, 2026, up 26.3% from the March 31, 2026 close of $60.18 and now only 15.4% below the two-year window's January 2025 high of $90.82 — a meaningful partial recovery from the 33.7% peak-to-trough decline this series flagged last quarter. The two-year window's earliest available close, $61.88 on July 31, 2024, is close to where the stock ended this quarter — meaning most of the "round-tripping, not compounding" pattern flagged in Q1 has partially unwound over just the last three months. The move within this quarter alone accounts for nearly all of that recovery: XYZ was still at $60.43 in January 2026, climbed steadily through February ($63.70) and March ($60.18, essentially flat), then accelerated through April ($70.51), May ($75.72), and June ($76.00) — a sustained three-month rally coinciding with the guidance raise and the Cash App acceleration, not a single-day earnings pop.

---