INTU › analyze
INTU · Analyze
A portfolio-specific passage was removed from the public build.
🆕 FIRST FULL ANALYSIS. The watchlist has flagged this name three times without running it.
INTU has been carried in 🏛 Evergreen Compounders with no conviction score, a stale add zone of $370–390 that sits above the current price, and a
/healthflagged 2026-07-14 that was never run. It is held in both accounts. This file closes that gap.⚠️ 21 DAYS TO A MAJOR CATALYST — Q4/FY2026 prints Monday, August 25, 2026 (confirmed).
That single event delivers FY2026 results, first FY2027 guidance, and the annual dividend raise — the highest-information event of Intuit's year. Investor Day follows Sept 17.
Manager Dispatch Log
| Agent | Status | Focus |
|---|---|---|
| Fundamentals Analyst | Complete | FY22–FY25 statements, FCF CAGR, capital allocation, per-share, dividend coverage |
| Sentiment Analyst | Complete | Reconstructed the −68% drawdown timeline, tax season 2026, insider reconciliation |
| Moat Analyst (qual + quant) | Complete | Adversarial test of the Goldman AI-disruption thesis, ProAdvisor channel, evergreen rating |
| Valuation Analyst | Complete | FCF/earnings multiples, Bogle, DYT (with pitfall applied), Graham, DDM |
| Portfolio Strategist | Complete | Dual-account position math, sizing into the Aug 25 binary |
Knowledge check performed. kb.py find INTU returned only Knowledge/Portfolio/self-state.md
(fragmentation, ~59 positions at ~$435 average) — no INTU-specific note or prior report exists in
the repo. Knowledge/Themes/software-saas.md lists INTU nowhere in its swept-name roster despite
12+ software screens; this name was inside the portfolio the whole time and never got covered.
Pitfalls applied: pitfall-yahoo-insider-purchases-counts-rsu-grants (confirmed), and critically
pitfall-dyt-inverts-when-price-caused-the-yield — see Valuation §Model 4, where it changes
the answer.
Phase 1A — Fundamentals Analyst
Income statement (FY ends July 31)
| Metric | FY2022 | FY2023 | FY2024 | FY2025 | 3yr CAGR |
|---|---|---|---|---|---|
| Revenue | $12.73B | $14.37B | $16.29B | $18.83B | +14.0% |
| Gross profit | $10.32B | $11.22B | $12.82B | $14.98B | +13.2% |
| Gross margin | 81.1% | 78.1% | 78.7% | 79.6% | Stable |
| Operating income | $2.56B | $3.14B | $3.85B | $4.94B | +24.5% |
| Operating margin | 20.1% | 21.9% | 23.6% | 26.2% | ↑ +610bps in 3 years |
| Net income | $2.07B | $2.38B | $2.96B | $3.87B | +23.3% |
| Diluted EPS | $7.28 | $8.42 | $10.43 | $13.67 | +23.4% |
| Diluted shares | 284M | 283M | 284M | 283M | −0.1%/yr |
| R&D | $2.35B | $2.54B | $2.75B | $2.93B | +7.6% (15.6% of rev) |
Operating leverage is the standout. Operating margin expanded 610bps in three years while revenue compounded at 14%. Net income and EPS grew at ~23% — a business converting growth into profit at an accelerating rate. Note the divergence from ADBE: Intuit grew faster on the top line (14.0% vs 10.5%) and much faster on the bottom (23.3% vs 14.5%).
Free cash flow
| Metric | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|
| Operating cash flow | $3.89B | $5.05B | $4.88B | $6.21B |
| CapEx | $229M | $260M | $250M | $124M (0.66% of rev) |
| Free cash flow | $3.66B | $4.79B | $4.63B | $6.08B |
| FCF margin | 28.8% | 33.3% | 28.4% | 32.3% |
FCF CAGR (3yr): +18.5% — comfortably ahead of revenue's 14.0%, and well ahead of ADBE's 10.0%. CapEx at 0.66% of revenue is lower than Adobe's. FCF is lumpy year to year (FY24 dipped) because of tax-season working capital; the trend is what matters and it is up.
⚠️ TTM FCF has fallen to ~$5.23B from the FY2025 $6.08B. Two plausible causes — the $300–340M restructuring charge (mostly landing in Q4 FY26) and tax-season seasonality — but this is not yet resolvable and must not be waved away. The August 25 print settles it. Valuation below runs both the $6.08B and $5.23B cases.
🚩 SBC not retrieved — flagged, not invented
fin.py does not surface SBC for INTU and I did not obtain it from a primary source. I am not
going to estimate it. The available proxy is the GAAP/non-GAAP gap, which is informative:
| GAAP EPS | Non-GAAP EPS | Gap | |
|---|---|---|---|
| INTU Q3 FY26 actual | $11.09 | $12.80 | +15.4% |
| ADBE FY26 guide, for reference | $17.95 | $24.40 | +35.9% |
Intuit's non-GAAP flatters less than half as much as Adobe's. Part of Intuit's gap is
amortization of Credit Karma/Mailchimp intangibles rather than SBC. Given
Knowledge/Themes/software-saas.md records SBC-vs-FCF as "the single most common disqualifier"
across the March screen series, this is a meaningful pass — but it is a proxy, and the real
number should be pulled at the next /health.
Capital allocation (FY2025) — balanced, and about to change
| Use of cash | Amount | % of FCF |
|---|---|---|
| Share buybacks | $2.77B | 45.6% |
| Dividends | $1.19B | 19.6% |
| Acquisitions | $184M | 3.0% |
| CapEx | $124M | 2.0% |
| Total returned to shareholders | $3.96B | 65% |
| Retained for reinvestment | ~$2.1B | 35% |
Cleaner than Adobe's, which returned 114% of FCF and borrowed to do it. Intuit returns two-thirds and keeps a third.
🔺 This is changing materially in FY2026. On 2026-03-16 Intuit filed an 8-K announcing that the founder and executive leadership team terminated all outstanding Rule 10b5-1 plans and that it would accelerate buybacks to deploy the $3.5B remaining authorization. H1 FY26 repurchases were $1.8B, +40% YoY; full-year buybacks are tracking to roughly double the prior year (~$5.5B). At a $87.5B market cap that is ~6% of shares in one year — a real change from the flat-share-count history, and the clearest management signal of belief in the price.
Balance sheet
| Metric | FY2025 | Assessment |
|---|---|---|
| Total assets | $36.96B | — |
| Total debt | $6.64B | — |
| Cash | $2.88B (FY-end) | Trough of the tax-season cycle |
| Net debt | $3.76B (FY-end) | ~0.6x FCF — trivial |
| Current, per TTM data | $6.78B cash vs $6.9B debt | ≈ net-cash today |
| Debt/Assets | 18.0% | Healthy — lower than ADBE's 22.5% |
| Current ratio | 1.45 | Comfortable |
| Goodwill | $13.98B | 37.8% of assets; 71% of equity |
| Equity | $19.71B | Growing |
| Tangible book value | ~$5.7B | Positive — unlike ADBE's negative TBV |
🚩 The goodwill line is the real balance-sheet story. $13.98B of goodwill against $25.68B of invested capital is the accounting trace of Credit Karma ($7.1B, 2020) and Mailchimp ($12B, 2021). Intuit bought a large part of its business; Adobe built more of its own. This is the single biggest quality gap between the two names and it shows up directly in ROIC (§Moat).
Per-share — the owner's-eye view
| Metric | FY2022 | FY2025 | 3yr CAGR |
|---|---|---|---|
| Revenue/share | $44.82 | $66.54 | +14.0% |
| FCF/share | $12.89 | $21.49 | +18.6% |
| Diluted EPS | $7.28 | $13.67 | +23.4% |
Per-share growth ≈ total-company growth, because the share count has been flat (−0.1%/yr). Intuit's per-share compounding has come from the business, not from financial engineering. Adobe's came ~3.5 points from buybacks. Both are legitimate; Intuit's is the more durable source, and the FY26 buyback acceleration now adds the other lever on top.
Dividend overlay (applies — dividend grower)
| Metric | Value |
|---|---|
| Forward rate | $4.80/yr ($1.20/qtr) |
| Current yield | 1.52% |
| 5-yr average yield | 0.67% |
| Yield vs its own 5yr average | 2.27x |
| EPS payout ratio | 28.3% |
| FCF payout ratio | ~25% ($1.31B ÷ $5.23B TTM FCF) |
| Last raise | $1.04 → $1.20, +15.4% (Oct 2025) |
| Consecutive annual raises | 15 years (initiated 2011) |
Dividend CAGR: 5yr 15.3% · 7yr 14.3% · 8yr 15.1%. Every raise since 2018 has landed in an +11% to +21% band, clustering tightly at ~15%. That is remarkable consistency.
Coverage verdict: comfortably covered and improving. At ~25% of FCF, the dividend survives a 75% cut to free cash flow before coverage breaks. This is not a yield thesis — 1.52% is not income — but the payout is a genuine signal of durability, and the coverage is nowhere near stressed. The next raise is due WITH the Aug 25 print and is a direct, dated read on management's own confidence: ~15% (to ~$1.38/qtr) says the FCF story is intact; anything under 8% would be the first management-signaled crack.
Fundamentals verdict: STRONG. 14% revenue CAGR, 23% net income CAGR, 18.5% FCF CAGR, 610bps of operating margin expansion, 18% debt/assets, positive tangible book, 15 years of ~15% dividend growth at a 25% FCF payout. Two flags: the TTM FCF decline to $5.23B (unresolved until Aug 25), and $13.98B of goodwill that depresses returns on capital.
Phase 1B — Sentiment & Intelligence Analyst
🔻 Reconstructing the −68% drawdown — there was no single event
The most important finding in this file: the collapse was a four-leg, ~12-month sequence in which the market moved first and the fundamentals only partly followed.
| Leg | Window | Move | Catalyst |
|---|---|---|---|
| 0 | Jul–Aug 2025 | $794 → $655 (−18%) | Post-Q4 FY25 de-rating from a peak multiple |
| — | Sep–Dec 2025 | $620–690 | Stable. Q1 FY26 (Nov 20) was strong, guidance reiterated |
| 1 | Jan 12–16, 2026 | $635 → $531 (−16% in a week) | Wells Fargo downgrade (OW→EW, PT $840→$700) on tough tax comps; Goldman initiates Neutral |
| 2 | Jan 26 – Feb 20, 2026 | $560 → $346 (−38%) | "Software-mageddon." ~$1T of software market cap erased in a week; INTU −11% on Feb 4 alone. Generic AI de-rating, zero company news |
| 3 | Feb 26 – Mar 16 | $346 → $479 (+38%) | Relief rally: strong Q2, Anthropic partnership, $3.5B buyback acceleration + 10b5-1 terminations |
| 4 | Apr 2026 | $452 → $340 (−25%) | Second software-sector leg down |
| 5 | May 20–21, 2026 | $383.93 → $307.07 (−20.0% in ONE DAY) | Q3 FY26: TurboTax guidance cut + 17% workforce cut. The company-specific break. |
| 6 | Jun 2–25 | $307 → $251.72 (bottom) | Goldman → Sell (PT $519→$276); Stifel cut; securities class actions filed |
| 7 | Jul 2026 | $252 → $320 (+27%) | Sector rotation, value buyers |
Peak-to-trough: −68.3%. Currently −59.7% off the high and −31% below the 200-day average.
| Hypothesis | Verdict |
|---|---|
| AI disruption fear | ✅ CONFIRMED — the dominant driver. Legs 1, 2 and 4 are pure narrative with no company news |
| Guidance cut | ⚠️ PARTIAL. Total FY26 revenue guidance was raised three times. Only the TurboTax line was cut |
| Bad quarter | ❌ RULED OUT. Q1, Q2 and Q3 FY26 all beat. Q3 was a beat-and-raise that fell 20% |
| IRS Direct File | ❌ INVERTED — Direct File was killed for the 2026 season. A tailwind that failed to convert |
| Competition | ✅ CONFIRMED — but from $0/$15 budget filers, not AI-natives |
| Management change / bad acquisition | ❌ Not found. Goodarzi still CEO; Aujla still CFO |
FY2026 results as reported — the guidance was RAISED, not cut
| Quarter | Reported | Headline | Outcome |
|---|---|---|---|
| Q1 FY26 | 2025-11-20 | GBS $3.0B +18%, Consumer +21%, Credit Karma +27% | Beat; guidance reiterated |
| Q2 FY26 | 2026-02-26 | Consumer $1.5B +15%, GAAP EPS $2.48 +49% | Beat; stock rallied |
| Q3 FY26 | 2026-05-20 | $8.6B +10%; non-GAAP EPS $12.80 vs $12.48 est | Beat & raise — stock −20% |
Q3 FY26 segment detail: GBS $3.3B +15% (+17% ex-Mailchimp) · Online Ecosystem $2.5B +19% (+22% ex-Mailchimp) · QuickBooks Online Accounting +22% · QBO Advanced / Intuit Enterprise Suite +38% · online payment volume +30% · Consumer $5.3B +8% · TurboTax $4.4B +7% · Credit Karma $631M +15% · ProTax $278M flat.
| FY2026 guidance | Q1 (Nov) | Q3 (May, current) |
|---|---|---|
| Revenue | $20.997–21.186B (+12–13%) | $21.341–21.374B (+13–14%) ⬆ |
| Non-GAAP operating income | — | $8.784–8.804B (+16%) |
| Non-GAAP EPS | — | $23.80–23.85 (+18%) |
| GBS | +14–15% | ~16% ⬆ |
| Consumer | +8–9% | ~10% ⬆ |
| — TurboTax | +8% | ~7% 🔻 the only cut |
| — Credit Karma | +10–13% | ~19% ⬆⬆ |
| ProTax | +2–3% | ~4% ⬆ |
The TurboTax cut was ~$40M in dollars ($5.305–5.330B → $5.277–5.282B). The stock lost ~$21B of market cap on it. That ratio is the entire debate.
🔻 Tax season 2026 — three separate problems, and only one is Intuit's fault
Management's own words on the Q3 call: "We did not have the overall tax season we expected" and they "faced pressure among the most price-sensitive DIY filers."
- The market shrank. Goodarzi: total IRS filers declined ~30bps, a gap of ~2 million units versus expectations, "the most significant industry-wide contraction since the post-COVID tax season" — against an expected +1%. The shortfall was concentrated in manual (non-e-file) filers who simply did not file.
- They lost the low end on price. Share loss among DIY filers earning under $50k — management admitted they "lost on price." The CFO clarified it was "a segment of the under-$50k, not the entirety." That slice is ~12% of TurboTax TAM.
- The mix saved the revenue but not the story. TurboTax Live revenue +36% to $2.8B (~53% of TurboTax revenue), Live customers +38%, ARPU +11% — while total online units −2% and e-file share −1 point.
⚠️ The inversion that matters most. The 2026 season was the first in which IRS Direct File did not exist — it was killed in Nov 2025 (Treasury cited $41M cost and $138/accepted return). Intuit got the exact policy outcome it lobbied for and still lost share — to private budget competitors, not the government. That makes the competitive problem harder to dismiss, not easier: the policy tailwind is now spent and cannot be repeated.
Management explicitly denies AI caused this. Goodarzi: "This is all about being price-right for customers… None of this has anything to do with AI." The market did not believe him.
The strategic response is a real business-model change, not spin: pricing shifts from complexity-based to value-based — free SKUs for simple W-2 filers under $50k, then "monetize beyond tax" via Credit Karma and money services. Truist flagged the obvious risk on Aug 3: the transition itself creates further revenue softening. Stifel now models TurboTax growth at 4–6% vs management's 6–10% target.
The rest of the business is fine — and it is most of the business
- Credit Karma: +27% → +23% → +15% across FY26; FY26 guidance raised from 10–13% to ~19%. Cross-sell working: "ARPU is ~30% higher for customers using both TurboTax and Credit Karma."
- GBS (QuickBooks): QBO Accounting +25% → +24% → +22% (mild deceleration on higher effective prices and mix); Intuit Enterprise Suite +38%; QuickBooks Live customers +50%+; FY26 guidance raised from 14–15% to ~16%.
- 🚩 Mailchimp is the sore spot — a ~2–3 point drag on GBS growth (15% reported vs 17% ex-Mailchimp; 19% vs 22% on Online Ecosystem). Management "rightsized Mailchimp's cost structure relative to its growth profile" in the layoffs. This is the worst asset in the file and a $12B acquisition that is not earning its keep.
⚠️ ARPC not verified — Intuit disclosed no explicit ARPC metric in FY26. Only "higher effective prices" language and TurboTax ARPU +11%. Do not cite an ARPC number for INTU.
The 17% workforce cut
~3,000 roles announced 2026-05-20; charges $300–340M, mostly Q4 FY26; headcount now 18,200. Framed as an AI-enabled restructuring: fewer management layers, eliminated PMO/business- ops coordination, TurboTax/Credit Karma platform de-duplication, Mailchimp rightsizing. Read it as margin defense, not growth — but note it lands with raised revenue guidance, which is an unusual and constructive combination.
Analyst consensus — eroding steadily, and still eroding
| 3mo ago | 2mo ago | 1mo ago | Now | |
|---|---|---|---|---|
| Strong Buy | 6 | 5 | 5 | 5 |
| Buy | 22 | 22 | 20 | 19 |
| Hold | 6 | 6 | 8 | 9 |
| Sell | 0 | 1 | 1 | 1 |
| Strong Sell | 0 | 0 | 1 | 1 |
| % Buy-or-better | 82% | 79% | 74% | 69% |
Mean PT $454.65 (+42%) · MEDIAN PT $430 (+34%) · High $921 · Low $250 · Consensus 1.86 = BUY.
⚠️ The "+42% to a Buy consensus" framing overstates it. The $921 high is a stale outlier inflating the mean by ~$14; most live targets cluster $275–$500. Use the median: +34%. And buy-share has fallen 82% → 69% in three months.
Notable actions — the sell side is capitulating on a lag, into a rising stock:
| Date | Firm | Action | PT |
|---|---|---|---|
| Jan 2026 | Wells Fargo | OW → EW | $840 → $700 |
| Mar 10 | Rothschild | Neutral → Buy | $670 → $700 — the last upgrade, badly timed |
| May 21 | 17 firms cut PTs in one day | mostly maintained ratings | Citi 649→591, Jefferies 650→550, RBC/Mizuho →500, Barclays →443, Truist →410, Evercore →400, UBS/WF →360 |
| Jun 2 | Goldman Sachs | Neutral → SELL | $519 → $276 |
| Jun 18 | Stifel | Buy → Hold | $375 → $275 |
| Jul 14 | Piper Sandler | assumes Underweight | $250 |
| Jul 21 | Morgan Stanley | OW → EW | $580 → $335 — AI debate unresolvable until Q3 FY27 |
| Jul 28 | TD Cowen | Buy → Hold | $504 → $304 |
| Aug 3 | Truist | Buy → Hold | $410 → $350 — J-curve, DIY headwind, flat Mailchimp, unexpected margin compression |
A downgrade arrived yesterday. There has been no upgrade since March. Truist stayed constructive long-term on the "Three Big Bets" (assisted tax, money, mid-market) growing >30%.
Insider activity — ⚠️ PITFALL APPLIED; the honest picture is uncomfortable
pitfall-yahoo-insider-purchases-counts-rsu-grants applies. Yahoo's summary would have counted
the quarterly Value: null rows (Oct 1 / Jan 1 / Apr 1 / Jul 1 scheduled RSU vestings) and
"Conversion of Exercise of derivative security" rows as purchases. They are not.
✅ Genuine open-market purchases — exactly ONE:
| Date | Insider | Shares | Price | Value |
|---|---|---|---|---|
| May 22 & 26, 2026 | Vasant M. Prabhu, Director | 1,750 | $309.45 / $309.72 | $541,655 |
Bought 1–4 trading days after the May 21 crash, taking his direct holding from 0 to 1,750 — a from-scratch position at the bottom. Real conviction, immaterial capital, and no purchase by the CEO, CFO, CTO or founder.
🚩 The far larger story is founder selling into the top:
| Insider | Window | Approx. proceeds | Price range |
|---|---|---|---|
| Scott D. Cook (Founder/Director) | Dec 3–30, 2025 | ~$276M (~438k sh) | $632–678 |
| Scott D. Cook | Feb–Jun 2025 | ~$241M | $576–772 |
| Sasan Goodarzi (CEO) | Jan 7, 2026 | $26.7M (41,000 sh) | $650–651 |
| Sandeep Aujla (CFO) | Dec 2025 – Jan 2026 | ~$1.6M | $627–675 |
~$276M of founder selling in December 2025 — the last month before the collapse began. These were 10b5-1 plans, so this is not evidence of foreknowledge. It is still the most uncomfortable fact in the file, and it sits behind the securities class actions.
The March 16 gesture, and its caveat. The 8-K terminating all founder/executive 10b5-1 plans and accelerating the buyback is a genuine and rare signal. ⚠️ But Director Richard L. Dalzell kept selling after the announcement (Jun 11, 16, 23 — ~1,572 shares total). This looks like a surviving director-level plan outside the "founder and executive leadership team" scope, so probably consistent rather than a breach — flagged as unresolved; check the Form 4 detail before citing the halt as unqualified.
Other positioning: short interest 15.82M shares = 5.78% of shares out / 6.63% of float, up +28.6% MoM — shorts were still building into the July rally. Fundsmith exited INTU entirely in Q2 2026.
⚠️ Not verified: the "probes" reported in June are plaintiff-side securities class actions (BFA Law et al.) over the May 20 disclosures. No SEC, FTC, DOJ or state AG action was found. Do not conflate the two.
Phase 1C / 2A — Moat Analyst
Quantitative base
| Metric | INTU FY2025 | ADBE FY2025 |
|---|---|---|
| Gross margin | 79.6% (stable) | 89.3% (expanding) |
| Operating margin | 26.2%, +610bps/3yr | 36.6% |
| ROIC (NOPAT/invested capital) | ~15% | ~37–39% |
| ROE | 22% | 63% |
| ROA | 9% | 20% |
| FCF margin | 32.3% | 41.4% |
| Share count | flat | −3.2%/yr |
🚩 15% ROIC vs Adobe's ~39% is the cleanest quality gap between these two names, and it traces directly to $13.98B of goodwill. Intuit bought Credit Karma and Mailchimp; Adobe built more of its own. On invested capital of $25.68B for $4.94B of operating income, Intuit earns a good return — well above cost of capital — but not an elite one. Anyone calling INTU "higher quality than ADBE" is wrong on this measure and should say so.
But note what is NOT present: no margin compression. Gross margin is stable; operating margin expanded 610bps in three years. The quantitative moat-erosion signature is absent — exactly as it is at Adobe.
Moat sources, rated
| Source | Rating | Evidence |
|---|---|---|
| Regulatory intangible — IRS e-file authorization + 41 state engines + liability underwriting | Stable, arguably ↑ | Requires EFIN + ETIN, a suitability check up to 45 days incl. criminal background and credit history, and annual ATS re-certification (IRS Pub 3112). AI reduces none of this cost. |
| Switching costs — data continuity + the 7-year guarantee tail | Stable at complex, ~zero at simple | The accuracy/audit-support guarantee runs the full 7-year life of the return. But a sub-$50k W-2-only filer has essentially no continuity to lose — precisely where Intuit lost. |
| Network effects — the accountant channel (ProAdvisor) | ↑ The strongest asset in either company; stable | 200,000+ QuickBooks ProAdvisors vs Xero's ~30,000 US certified partners. QBO ≈ 80% of the US online SMB accounting market; Xero ~8.9%, with only ~18% of its accounting customers in the US. Genuinely two-sided: accountants standardize a whole book on QBO; SMBs adopt because the accountant requires it |
| Brand | Stable — but no longer supports a price premium | TurboTax still tops 2026 software rankings "but at a cost." Management itself said it "lost on price" |
| Efficient scale — assisted tax | Partial ↑ | The $37B assisted market is 88% of TurboTax TAM, with real scale economics in preparer supply — and AI lowers Intuit's cost to serve it |
| Cost advantage | None | Not a source |
🔻 Adversarial test: destroying, then partly rebuilding, the Goldman thesis
Goldman's claim (Jun 2, Sell, PT $276): an AI processes a return for ~$0.12 vs TurboTax's ~$162 average revenue per return — a ~1,300x gap meaning entrants need almost no funding. Base case: TurboTax revenue ~18% below FY2025 levels by 2030, on 20% of US filers moving fully to AI. TurboTax is ~25% of revenue and operating income.
🔴 The number is broken twice.
- Wrong numerator. $162 is blended average revenue per return across the whole TurboTax base — and that base now includes TurboTax Live at ~53% of TurboTax revenue with far higher ARPU. It is not the DIY price. Actual DIY list is $69–$129 federal + $39/state. Comparing an inference cost against a figure loaded with human-assisted revenue overstates the gap materially.
- Wrong denominator. $0.12 is the cost of drafting. It is not the cost of a compliant filing business, which must fund: annual ATS re-certification, 41 separate state income-tax engines with their own conformity rules, EFIN/ETIN suitability and Pub 4557/1345 security safeguards, refund-transfer banking, seasonal-peak support, and — the load-bearing item — the reserve behind a 7-year accuracy and audit guarantee.
The moat is in the liability wrapper and data continuity, not the software. The software was commoditized in 2005, not 2026. Supporting evidence: 2026 benchmark testing found every generic model used tax-bracket math instead of the IRS-mandated Tax Table for income under $100k (close, not compliant), and made structural errors on Form 8962 and Schedule 2. Whoever wants TurboTax's revenue must underwrite TurboTax's liability — a capital-and-insurance problem, not a compute problem.
Do the AI-native products actually exist?
| Product | Real? | Can it e-file? | States | Guarantee | Verdict |
|---|---|---|---|---|---|
| Perplexity "Computer for Taxes" (Apr 2026, $17/mo Pro) | Yes | NO — "doesn't submit them to the IRS on your behalf. You still need to file — through IRS Free File, tax software, or a professional" | Federal only | None | Not a competitor. It competes with a CPA's first draft — and feeds a filing engine |
| Prime Meridian (Mar 2026; General Catalyst + Belief Capital; ex-IRS Commissioner Danny Werfel advising) | Yes | Yes — claims IRS Authorized E-File Provider | All 50; K-1s, multi-state, investment income | Not disclosed | The only real AI-native threat. Free, funded, credibly staffed |
| Chime Tax | Yes | Yes | Fed + state, both $0 | 100% accuracy, pays IRS penalties to $10,000 | A distribution attack, not an AI attack. Monetizes deposits, not the return |
🔻 Who actually took the share? This was a PRICE attack, not an AI attack. The winners priced at $0 federal / $15.99 state (FreeTaxUSA) and $0 everything (Cash App Taxes). Neither is AI-native. Both are years old. Meanwhile H&R Block grew assisted share for a third consecutive year and deliberately shed low-monetization DIY volume toward $100k+ AGI clients.
The damning read, stated plainly: Intuit lost DIY units in the year its hardest free competitor (the US government) was removed from the field, to competitors using no new technology. That is not disruption — it is a pricing error. But it is a worse indictment in one respect: it proves the price umbrella at the low end is gone permanently, and Intuit has conceded it.
Where Intuit held: the segment where switching costs are real — TurboTax Live +36% revenue, +38% customers, +11% ARPU, now ~53% of TurboTax revenue.
Is QuickBooks a stronger moat than TurboTax? Decisively yes. 200,000+ ProAdvisors vs Xero's ~30,000; QBO ~80% of the US online SMB market. Xero has not cracked it — North America is only ~7% of Xero revenue, and it paid up to $3B for Melio in 2025 precisely because US growth stalled (Melio explicitly continues serving QuickBooks customers, so it is a payments beachhead, not a channel conversion). AI-native challengers (Digits, Basis, Puzzle) are architecturally better but the category is consolidating, not exploding — Botkeeper shut down and Bench nearly did in 2025–26.
The accountant channel is not addressable by a better model. A rival must flip the accountant's entire book of clients at once.
Disruption forecast
| Horizon | Mechanism | Confidence |
|---|---|---|
| 2026–2028 | DIY ARPU compression at the low end continues. Intuit cuts price to defend funnel volume; TurboTax growth settles at 4–6%. Mix shifts hard to Live. Already in motion | High |
| 2027–2030 | AI-native filers get e-file authorization at scale. Prime Meridian is the template. The barrier is the 45-day suitability check, annual ATS certification, 41 state engines and underwriting a 7-year guarantee. Expect 3–6 credible entrants, not 30. Goldman's 20% of filers by 2030 requires roughly the entire sub-$50k DIY population plus part of the middle; we model 8–12% | Medium |
| 2028–2033 | The assisted market reprices — in Intuit's favor. AI collapses the cost of delivering "a human checked your return," which is the $37B / 88%-of-TAM segment already growing +36%. Intuit monetizes AI on the cost side while entrants monetize it on the price side. That is the right side of the trade | Medium-high |
| 2030+ | The real kill shot: disintermediation of the QuickBooks GL by an agent layer — an AI agent maintaining the books and calling Intuit only as a compliance endpoint, commoditizing QBO into plumbing. The Anthropic deal makes Intuit both participant and potential victim | Low-medium |
Moat-defense evaluation
| Action | Assessment | Risk/reward |
|---|---|---|
| Value-based pricing shift | A margin concession dressed as strategy — but the right concession. Cheap filers are the top of the Live funnel; losing them costs more than the price cut | Good. Cost is real (Stifel: 4–6% vs 6–10% target) |
| Three Big Bets — assisted tax, money, mid-market (each >30%) | The strongest defense either company in this pair has. Attacks the $37B assisted category where AI lowers Intuit's cost to serve rather than a competitor's price. Live +36%/+38%/+11%; IES +38% | Very good — and under-appreciated by the bears |
| ~3,000 layoffs (17%) | Consistent with AI lowering cost-to-serve; also signals the growth algorithm changed | Fair. Margin defense, not growth |
| Anthropic partnership (Mar 2026) | Track 1 (Claude Agent SDK on Intuit's platform for mid-market) is straightforwardly good. Track 2 — Intuit's financial intelligence inside Claude.ai/Cowork via MCP — is the interesting risk: it puts Intuit's data where the disintermediating agent lives | Defensible. Intuit controls the compliance endpoint — the agent can reason, but Intuit files. That asymmetry makes it a distribution win rather than a giveaway |
| Buyback acceleration + 10b5-1 terminations | Rare, concrete, and costly to fake | Good |
| 🚩 Mailchimp | A $12B acquisition dragging 2–3pts off GBS growth, now being cost-rationalized rather than fixed | Bad. The worst asset in the file |
Evergreen rating and moat score: 7.0 / 10
Evergreen status: EVERGREEN. Tax compliance is a legally mandated, annually recurring, government-defined workflow with an accountable-party requirement. Someone must be the filer of record carrying the liability. Intuit's risk is price, not existence.
| Component | Score | Note |
|---|---|---|
| QuickBooks + accountant channel | 8.5 | 200k ProAdvisors, ~80% US QBO share, Xero stalled after a $3B acquisition |
| TurboTax assisted / Live | 7.0 | Growing 36% — AI helps the incumbent here |
| TurboTax DIY, >$50k / complex | 6.0 | Data continuity + 7-yr guarantee still bind |
| TurboTax DIY, sub-$50k | 3.0 | Already lost, structurally unrecoverable at a premium price |
| Credit Karma | 4.0 | A lead-gen business, not a moat |
| 🚩 Mailchimp | 3.0 | The worst asset in the file |
Weighted by revenue, GBS (QuickBooks) is the largest segment AND the best-defended one. That is what holds the score at 7.0 despite the Consumer segment's problems.
Phase 2B — Valuation Analyst
Model 1 — Forward earnings (primary)
| Basis | EPS | Multiple at $319.97 |
|---|---|---|
| FY2026 guided non-GAAP | $23.82 | 13.4x |
| FY2027E consensus (implied by Yahoo fwd P/E 11.69) | ~$27.37 | 11.7x |
| TTM GAAP | $16.52 | 19.4x |
⚠️ The "11.7x forward" figure everywhere is on FY2027 consensus, not FY2026 guidance. On what the company has actually guided, INTU trades at 13.4x. Both are cheap against a historical 30–40x, but 13.4x is the number to underwrite.
| Scenario | FY27E EPS | P/E | Implied price |
|---|---|---|---|
| Bear — TurboTax 4%, Mailchimp bleeds, no re-rate | $25.50 | 12x | $306 |
| Base — mid-teens EPS growth, partial re-rate | $27.00 | 16x | $432 |
| Bull — Three Big Bets compound, AI fear lifts | $28.50 | 20x | $570 |
Model 2 — FCF multiple (co-primary)
| Basis | FCF/share | Yield at $319.97 |
|---|---|---|
| FY2025 ($6.08B ÷ 273.54M) | $22.23 | 6.95% |
| TTM ($5.23B ÷ 273.54M) | $19.12 | 5.98% |
| Multiple | on FY25 FCF | on TTM FCF |
|---|---|---|
| 14x (≈ today) | $311 | $268 |
| 18x | $400 | $344 |
| 21x | $467 | $402 |
| 25x | $556 | $478 |
18–21x is the defensible band for a business compounding FCF at 18.5% with an 8.5/10 moat on its largest segment — discounted from a true compounder multiple because ROIC is 15%, not 39%.
Model 3 — Bogle expected return (5yr)
| Input | Value |
|---|---|
| Dividend yield | 1.52% |
| Earnings growth | ~11% (conservative vs management's "at least mid-teens") |
| Current fwd P/E | 13.4x |
| If it re-rates to 18x by 2031 | +6.1%/yr |
| Bogle = 1.5% + 11% + 6.1% | ≈ 18.6%/yr |
| If the multiple never re-rates | ≈ 12.5%/yr |
Model 4 — Dividend Yield Theory ⚠️ PITFALL APPLIED — THIS SIGNAL IS VOID
Knowledge/Playbook/pitfall-dyt-inverts-when-price-caused-the-yield applies, and it changes the
answer. The naive read is seductive: current yield 1.52% vs a 5-year average of 0.67% = 2.27x
its own history, which looks like a screaming DYT buy.
Decompose it, as the pitfall requires: - The dividend rose +15.4% (Oct 2025, $1.04 → $1.20). - The price fell −59.7% ($794.09 → $319.97). - Price alone would have multiplied the yield by 1 / 0.403 = 2.48x. Add the dividend growth and the expected multiple is 2.48 × 1.154 ≈ 2.86x. - Observed multiple is 2.27x — less than price-decline alone would predict.
Conclusion: the yield expansion is entirely price-caused. DYT provides no independent buy signal here — it is restating the drawdown in a different unit. Weight: 0. (This is exactly the trap the pitfall was written for, and it is the second time it has paid for itself.)
Model 5 — DDM ⚠️ not the right lens. Weight: 0.
The payout is only ~25% of FCF and the 5yr dividend CAGR (15.3%) exceeds any sane discount rate, so Gordon Growth breaks (g > r) and a two-stage model would value ~25% of the cash flows while ignoring 75%. Reporting this to show it was run and correctly discarded.
Model 6 — Graham's Intrinsic Value ⚠️ Weight: 0.5 — as a reality check only
√(22.5 × $16.52 × $75.22) = **$167.20** against a $319.97 price. Graham says INTU is worth half
its price. Do not act on this — Graham was built for asset-backed businesses, and INTU's $75.22
BVPS is inflated by $13.98B of goodwill (71% of equity). Stripped of goodwill, book value is
~$5.7B and the Graham number falls further.
The useful signal is not the number; it is why it is so low. It quantifies how much of Intuit's balance sheet is acquired rather than earned — which is the same fact that produces the 15% ROIC. Carry it as a quality caveat, not a valuation.
Fair value: $380 – $470 (midpoint ~$425)
| Model | Weight | Range |
|---|---|---|
| Forward earnings (16–20x on FY26 guide) | 40% | $381–$476 |
| FCF multiple (18–21x, blended FY25/TTM) | 40% | $344–$467 |
| Bogle | 15% | Supports 12.5–18.6%/yr — consistent with the range |
| DYT / DDM / Graham | 5% (0/0/0.5) | Void — reported and discarded |
Fair value $380–470 → +19% to +47% from $319.97. Sits just below the analyst median of $430 and well below the outlier-inflated mean of $454.65 — which is the right place for a conservative house to be.
🔻 Zones — the recorded add zone is wrong and must be replaced
Watchlist.md carries "Add $370–390," which is ABOVE the current price of $319.97. That zone
was set against a much higher fair value and has caused the watchlist to record INTU as "still
below its add zone" for three consecutive scans — reading a name into its buy range as though
it were out of it. Replaced:
| Old (stale) | New | |
|---|---|---|
| Entry zone | $370–390 (above spot) | $300–345 — currently IN ZONE |
| Strong buy | — | < $280 |
| Trim | — | 17x fwd (≈ $465) |
| Conviction | none recorded | [7.0] |
Phase 3 — Tension & Debate
Tension 1: Fundamentals/Moat say "the business is fine." The tape says −68%.
Fundamentals: revenue +13–14%, guidance raised three times, operating margin +610bps/3yr, FCF CAGR 18.5%, dividend covered at 25% of FCF with 15 straight years of ~15% raises. Nothing in the reported FY26 numbers justifies a 60% de-rate.
Sentiment: the market is not pricing an earnings question. It is pricing a terminal-value question — Goldman's 2030 scenario in which one-quarter of the profit pool is structurally impaired. That is unfalsifiable today.
Resolution — both are correct, and the gap between them is the opportunity and the risk. The decisive question is whether the 2026 share loss was mechanism or noise. The Moat Analyst's answer is the most valuable finding in this file: the share went to $0 and $15.99 competitors using no new technology, in the year the US government's free option was removed. That is a pricing error, not disruption. Pricing errors are fixable — Intuit is fixing it — and the company conceded the low end while growing the assisted segment 36%.
Weighted view: the market has priced an AI disruption that the 2026 evidence does not show, using a season whose biggest single factor was a shrinking filer pool. But the bears earn a real point: the low-end price umbrella is permanently gone, and Intuit has said so.
Tension 2: Moat says 7.0/10. Fundamentals says ROIC is only 15%.
Not a contradiction — they are measuring different things, and both belong in the verdict. The moat is genuinely strong (200k ProAdvisors, ~80% US QBO share, a regulatory barrier AI does not reduce). The return on the capital Intuit deployed to assemble it is mediocre, because $13.98B of goodwill from Credit Karma and Mailchimp sits in the denominator. Both facts survive.
Practical consequence: it caps the multiple. A business with a 8.5/10 moat and 39% ROIC earns 25x+. A business with a 8.5/10 moat and 15% ROIC earns 18–21x. That is exactly the band applied above — and it is why fair value is $380–470 rather than $500+.
Tension 3: One director bought $541k. The founder sold $276M the month before the collapse.
Weighted honestly: this is a negative, not a positive. Prabhu's from-scratch purchase 1–4 days after the crash is a real signal but immaterial capital, and no operating executive bought. Cook's ~$276M in December 2025 was 10b5-1 and therefore not evidence of foreknowledge — but the aggregate insider picture is heavy net selling into strength followed by token buying into weakness. The March 16 10b5-1 termination + buyback acceleration is the stronger management signal, because it is costly and concrete. Net: mildly negative on insiders, positive on corporate action.
Phase 4 — Manager Synthesis
Weighted verdict: ACCUMULATE — [7.0]
| Question | Answer |
|---|---|
| Is this a good business? | Yes. 14% revenue CAGR, 23% net income CAGR, 18.5% FCF CAGR, 610bps of margin expansion, 18% debt/assets, positive tangible book, a 8.5/10 moat on its largest segment, and 15 years of ~15% dividend growth at a 25% FCF payout. Caveat: 15% ROIC, not elite — $13.98B of goodwill. Moat 7.0/10. |
| Has the market priced it? | No — it has priced a scenario, not the results. 13.4x guided FY26 non-GAAP / 11.7x FY27E / ~14x FCF, against a historical 30–40x. Guidance was raised three times in the year the stock fell 60%. |
| Verdict cell | Good company + cheap = the value cell — with an explicit, dated, unresolved terminal question. |
🔄 The headline finding: the watchlist ranking of these two names is backwards
| ADBE | INTU | |
|---|---|---|
| Watchlist status (before today) | ⭐ Shortlist, [7.5], "IN ZONE" | 📋 Honorable Mention, no score, stale zone above spot |
| This file's conviction | [6.5] ⬇ | [7.0] 🆕 |
| Moat score | 6.0/10 | 7.0/10 |
| Best-defended segment | Acrobat, ~$4B — not the core | QuickBooks, the largest segment |
| Most-attacked segment | Creative Cloud — the core, $17.65B | Sub-$50k DIY — ~12% of one segment's TAM |
| Pricing power | Conceded on the flagship (deferred CC price increases) | Conceded on ~12% of TurboTax TAM; raised guidance elsewhere |
| Revenue CAGR (3yr) | 10.5% | 14.0% |
| FCF CAGR (3yr) | 10.0% | 18.5% |
| ROIC | ~39% ✅ | 15% |
| Non-GAAP flattery | +35.9% | +15.4% ✅ |
| Leadership | No CEO, no CFO 🚩 | Intact; diagnosed publicly, restructured, raised guidance |
| Dividend | None | 1.52%, 15yr streak, 25% FCF payout ✅ |
| Analyst median target | $250 — below spot | $430 — +34% |
| Upside to FV midpoint | +28% | +33% |
The structural asymmetry in one line: Intuit's largest segment is its best-defended one; Adobe's largest segment is its most-attacked one. Adobe has by far the better business today — 39% ROIC, 89% gross margins, a shrinking share count. But if you are pricing moat danger rather than current cash generation, Adobe is the more dangerous of the two, and the watchlist has it ranked the other way.
Portfolio Strategist — INTU is held in BOTH accounts
A portfolio-specific passage was removed from the public build.
(Self lots: 0.25 @ $404.76 · 0.344 @ $395.99 · 0.45 @ $301.13 · 0.003 @ $289.17.)
A portfolio-specific passage was removed from the public build.
Sizing. Combined ~$1,714 is a large position by this portfolio's standards —
Knowledge/Portfolio/self-state.md records ~59 positions at a ~$435 average, so INTU is
already ~4x the typical position (concentrated in Beta). Adding is defensible on conviction but
should be sized against the Aug 25 binary, not against the discount. Enterprise SaaS is the
portfolio's most crowded sleeve, and INTU is already one of its larger expressions.
A portfolio-specific passage was removed from the public build.
🚩 Thesis-break triggers (new — none existed)
- FY2027 revenue guidance below 8% (Aug 25) — management has signaled "at least mid-teens EPS growth"; a sub-8% revenue guide would contradict it.
- Annual dividend raise below 8% (Aug 25) — 15 straight years have landed at +11–21%. A token raise is the cheapest, clearest signal management can send that FCF is under pressure.
- QuickBooks Online Accounting revenue growth falls below 15% for two consecutive quarters (currently 22%). This is the real trigger — GBS is the moat; TurboTax is the narrative.
- TurboTax online units decline again in the 2027 season (May 2027) — one bad season is a pricing error; two consecutive is a mechanism, and would validate Goldman.
- A named AI-native filer discloses >2M returns filed in a season — this is what would convert Prime Meridian from a template into a threat.
- Gross margin falls below 77% — the classic erosion signature, currently absent.
Risk register
| Risk | Prob. | Impact | Mitigant |
|---|---|---|---|
| Aug 25: weak FY27 guidance and/or a token dividend raise | 35% | −10 to −20% | 21 days away — this is why the bulk is reserved |
| Value-based pricing transition compresses margin more than guided (Truist's "J-curve") | 40% | Multiple stays 12–14x | Live +36% is already absorbing the mix shift |
| Goldman is right — TurboTax revenue below FY25 by 2030 | 20% | Terminal; −25% to FV | TurboTax is ~25% of revenue; QuickBooks (~the largest segment) is untouched |
| Mailchimp continues to bleed / impairment charge | 35% | −2 to −3pts GBS growth; a writedown would hit GAAP only | Already being cost-rationalized; ~$12B of goodwill is the exposure |
| The AI question stays unresolvable until May 2027 | High | Dead money for ~9 months | 1.52% yield + ~6% of shares retired this year pays you to wait |
| Agent layer commoditizes the QuickBooks GL | 15% (long-dated) | Structural | Intuit controls the compliance endpoint — the agent reasons, Intuit files |
| Shorts (5.78% of shares, +28.6% MoM) press into a weak print | 25% | Amplifies a downside move | Cuts both ways — a strong print squeezes |
Summary dashboard
| Metric | Value | Assessment |
|---|---|---|
| Price | $319.97 | IN new entry zone $300–345 |
| 52-week range | $252.84 – $794.09 | −59.7% off high; +27% off the low |
| Fwd P/E (FY26 guide, non-GAAP) | 13.4x | The number to underwrite |
| Fwd P/E (FY27E consensus) | 11.7x | The number everyone quotes |
| P/FCF | ~14x (FY25) / ~17x (TTM) | 5.98–6.95% FCF yield |
| Revenue (FY25) | $18.83B | +14.0% 3yr CAGR |
| FY26 revenue guide | $21.34–21.37B, +13–14% | Raised 3x during the crash |
| Operating margin | 26.2%, +610bps/3yr | Expanding |
| FCF (FY25) | $6.08B | 32.3% margin; 18.5% CAGR |
| ⚠️ TTM FCF | $5.23B | Unresolved — Aug 25 settles it |
| ROIC | ~15% | 🚩 Half of ADBE's — $13.98B goodwill |
| Debt/Assets | 18.0% | Healthy; ≈ net-cash today |
| Tangible book | Positive | Unlike ADBE |
| Shares | Flat, but ~6% to be retired in FY26 | 10b5-1 plans terminated, buyback doubled |
| Dividend | 1.52%, 15yr streak, 15.3% 5yr CAGR | 25% FCF payout — covered 4x |
| Analyst consensus | BUY; median $430 (+34%) | Buy-share 82% → 69% in 3mo |
| Short interest | 5.78% of shares, +28.6% MoM | Shorts still building |
| Moat score | 7.0/10 | QuickBooks 8.5 · sub-$50k DIY 3.0 |
| Conviction | [7.0] 🆕 | First score ever assigned |
| Fair value | $380–470 (mid $425) | +19% to +47% |
| Entry / Strong buy / Trim | $300–345 / <$280 / 17x fwd | Old zone $370–390 was above spot |
| Next catalyst | Q4/FY26 print Mon Aug 25, 2026 | 21 days. + Investor Day Sep 17 |
Verdict
ACCUMULATE — conviction [7.0]. Staged around the August 25 print.
Intuit fell 68% peak-to-trough in a year when it raised total-company guidance three times, beat every quarter, expanded operating margin 610bps over three years, grew FCF at 18.5%, and extended a 15-year dividend streak at ~15%. The market is not pricing an earnings question; it is pricing a terminal-value question — Goldman's scenario in which one-quarter of the profit pool is structurally impaired by 2030. That scenario is unfalsifiable today, which is exactly why the multiple is 13.4x.
The most valuable thing this analysis found is that the 2026 evidence does not support the mechanism the market is pricing. Intuit lost DIY units to $0 and $15.99 competitors using no new technology, in the very season the US government's free-file option was removed from the field — while the segment an LLM should theoretically have gutted, TurboTax Live, grew revenue 36% and customers 38%. That is a pricing error, not disruption. Goldman's headline $0.12-vs-$162 comparison is broken on both sides: the numerator is a blended figure loaded with human-assisted revenue, and the denominator ignores that a compliant filer must fund 41 state engines, annual IRS re-certification, and the reserve behind a 7-year accuracy guarantee. Perplexity's tax agent cannot legally e-file at all.
The bears still earn two real points, and they cap the conviction at 7.0 rather than 8. First, the low-end price umbrella is permanently gone and management has conceded it — Stifel's 4–6% TurboTax growth is probably closer to right than management's 6–10%. Second, and more damaging to the quality case: ROIC is 15%, not the 39% Adobe earns, because $13.98B of goodwill from Credit Karma and Mailchimp sits in the denominator. Intuit bought a large part of its moat. That caps the fair multiple at 18–21x FCF rather than 25x+, and it is why fair value is $380–470 and not higher.
What holds the thesis is the part nobody is arguing about: QuickBooks. 200,000+ ProAdvisors against Xero's ~30,000, roughly 80% of the US online SMB accounting market, and a two-sided accountant network that no amount of model capability addresses — a rival must flip an accountant's entire book of clients at once. Xero paid up to $3B for Melio and still gets only ~7% of its revenue from North America. QuickBooks is the largest segment and the best-defended one, growing 22%, with Intuit Enterprise Suite at +38%.
Practically: hold everything, and reserve the ammunition for 21 days from now. The combined position across both accounts is 5.356 shares at a $334.75 blended basis — essentially breakeven, not the disaster the −60% headline implies. A partial add here is defensible; the price is in the newly-corrected $300–345 zone that the watchlist has been mis-recording as "still below its add zone" for three consecutive scans. But August 25 delivers FY26 results, first FY27 guidance, and the annual dividend raise in a single release. After fifteen consecutive raises in an +11–21% band, the size of that raise is the cheapest honest read available on what management actually believes — and it is free information three weeks away. Buy the reserve after it, not before.
Sources: .mcp/fin.py (FY22–FY25 statements, derived metrics); Yahoo Finance MCP
(get_stock_info, get_recommendations, get_holder_info, get_stock_actions,
get_historical_stock_prices); Intuit Q1/Q2/Q3 FY2026 press releases and the Q3 FY26 earnings-call
transcript (Motley Fool); Intuit 8-K 2026-03-16 (10b5-1 terminations, $3.5B buyback); Intuit
press release 2026-07-30 (Aug 25 earnings date, Sep 17 Investor Day); Intuit–Anthropic partnership
release; CNBC and TechCrunch on the 17% workforce reduction; Goldman Sachs downgrade via Yahoo and
Investing.com; Morgan Stanley, Stifel, Piper Sandler, TD Cowen and Truist actions via
Investing.com/Yahoo; Forbes and Tax Notes on the IRS Direct File shutdown; IRS Pub 3112 and IRS
e-file software approval pages; TurboTax published guarantees; H&R Block Q3 FY26 results;
QuickBooks/Xero share and ProAdvisor data; Prime Meridian, Perplexity and Chime Tax primary
sources; StockTitan/MarketBeat Form 4 detail; Knowledge/Playbook/pitfall-dyt-inverts-when-price-caused-the-yield;
Knowledge/Playbook/pitfall-yahoo-insider-purchases-counts-rsu-grants;
Knowledge/Portfolio/self-state.md; Knowledge/Themes/software-saas.md.
Explicit gaps: SBC not retrieved — GAAP/non-GAAP gap used as a proxy; pull at next /health.
FY2027 guidance does not exist — any FY27 figure circulating in secondary sources is a
misattributed Q4 FY26 number. ARPC not disclosed by Intuit in FY26. Precise TurboTax e-file
share is directional only (−1 point) with no absolute figure located. Prime Meridian's 2026-season
volume, users and MeF transmitter status are not publicly disclosed; Chime Tax's filing engine
and volume unknown. The June "probes" are plaintiff-side securities class actions only — no SEC,
FTC, DOJ or state AG action found. Director Dalzell's post-March-16 selling is an unresolved
discrepancy against the 10b5-1 termination announcement. Whether the Q3 FY26 margin compression
Truist flagged is structural or restructuring-driven is not resolvable until Aug 25.**