Grade: C — Some Red Flags, Investigate
Framework: Schilit *Financial Shenanigans* + Beneish M-Score + forensic accounting principles
Data: SEC EDGAR 10-K (filed Mar 3, 2026, fiscal year ended Jan 31, 2026) + Yahoo Finance
Auditor: Ernst & Young LLP — Clean opinion (unqualified)
One-line verdict: Autodesk has two legitimate red flags — AR outpacing revenue for two consecutive years and goodwill at 156% of equity — plus three watch items. However, the M-Score is clean at -2.68, CFFO/NI of 2.18 shows extraordinary cash conversion, FCF of $2.4 billion vastly exceeds net income, and the accruals ratio of -10.7% indicates earnings are significantly understated relative to cash. The AR growth is explained by Autodesk's transition to a new direct transaction model that shifts revenue recognition timing. The Z-Score at 0.48 (distress zone) reflects software-company balance sheet characteristics (negative retained earnings from historical buybacks), not financial distress. We set the grade at C: the AR pattern and goodwill deserve monitoring, but the cash generation profile is excellent.
| Metric | Result |
|---|---|
| Red Flags | **2** |
| Watch Items | **3** |
| Checks Completed | **17/18** |
| Beneish M-Score | **-2.68** (clean) |
| Altman Z-Score | **0.48** (distress zone — misleading for software) |
| Piotroski F-Score Prob | **0.56%** (low) |
| Auditor | Ernst & Young LLP — Unqualified opinion |
The Business: Design and Make Platform
Autodesk is the dominant provider of design software for architecture, engineering, construction (AEC), and manufacturing. Its products — AutoCAD, Revit, Fusion, and Autodesk Construction Cloud (now Forma for Construction) — are industry standards. Revenue is 97% recurring through subscriptions, making this one of the most predictable revenue streams in enterprise software.
Critically, Autodesk is in the middle of a major business model transition: it launched a "new transaction model" where Solution Providers (formerly resellers) provide quotes to customers but the actual transaction occurs directly between Autodesk and the customer. The 10-K warns that "the change in recognition of sales incentives to Solution Providers from contra revenue to operating costs under the new transaction model" will "continue to positively impact calculated revenue growth, while being broadly neutral to calculated operating profit and free cash flow dollars."
This transition is important context for reading the AR and revenue growth numbers.
Profitability: High-Margin Software at Scale
| Metric | FY2023 | FY2024 | FY2025 | FY2026 | Trend |
|---|---|---|---|---|---|
| Revenue | $5.0B | $5.5B | $6.1B | $7.2B | +17.5% YoY |
| Net Income | $823M | $906M | $1.1B | $1.1B | +1.1% YoY |
| Gross Margin | 90.4% | 90.7% | 90.6% | 91.0% | Stable at ~91% |
| Net Margin | 16.4% | 16.5% | 18.1% | 15.6% | Declining |
| Recurring Revenue | — | — | $6.0B | $7.0B | +18% |
Revenue grew 18% to $7.21 billion. The 10-K attributes this to "an increase in subscription revenue" and notes that "recurring revenue as a percentage of net revenue was 97% for both fiscal years ending January 31, 2026 and 2025."
Gross margin of 91.0% is among the highest in all of technology — almost pure software economics. There is virtually no cost of goods sold.
Net margin declined from 18.1% to 15.6% despite the revenue growth. This requires investigation: net income barely grew (+1.1%) on 17.5% revenue growth. The explanation lies in the new transaction model — sales incentives previously deducted from revenue are now recognized as operating expenses, inflating both revenue and SG&A. The 10-K confirms the model shift is "broadly neutral to calculated operating profit and free cash flow dollars."
Key performance indicators: Net revenue retention rate (NR3) was above 100-110% on a constant currency basis. Remaining Performance Obligations (RPO) grew 20% to $8.3 billion, with current RPO up 23% to $5.5 billion. This backlog visibility is a strong signal of future revenue.
Cash Flow: The Real Story
| Metric | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|
| Operating Cash Flow | $2.1B | $1.3B | $1.6B | $2.5B |
| Net Income | $823M | $906M | $1.1B | $1.1B |
| **CFFO / Net Income** | **2.52** | **1.45** | **1.45** | **2.18** |
| Free Cash Flow | $2.0B | $1.3B | $1.5B | $2.4B |
CFFO/NI of 2.18 is extraordinary — for every dollar of reported net income, Autodesk generates $2.18 in operating cash flow. This gap exists because of deferred revenue: customers pay upfront for annual or multi-year subscriptions, but revenue is recognized ratably over the contract period. The cash arrives before the revenue is recognized.
FCF of $2.4 billion on $1.1 billion of net income (FCF/NI of 2.11) means free cash flow is more than double reported earnings. This is not a red flag — it is the hallmark of a subscription software model with upfront cash collection.
Operating cash flow surged 52.6% year-over-year, from $1.6B to $2.5B, significantly outpacing the 17.5% revenue growth. Deferred revenue grew 14% to $4.7 billion.
Balance Sheet: Software Accounting Quirks
| Item | FY2026 | Notes |
|---|---|---|
| Cash | **$2.6B** | Up from $1.9B |
| Total Debt | **$2.7B** | Stable |
| Cash minus Debt | **-$0.1B** | Nearly neutral |
| Goodwill | $4.3B | From acquisitions (Payapps, Construction Cloud) |
| Intangibles | $0.5B | Declining via amortization |
| **Goodwill + Intangibles / Equity** | **156%** | Fails threshold |
| Debt/EBITDA | **1.4x** | Healthy |
| Interest Coverage | **N/A** (minimal interest) |
The goodwill at 156% of equity requires context. Autodesk has negative retained earnings — not because of losses, but because of aggressive share buybacks that reduced equity below the goodwill level. The $4.3 billion in goodwill comes from strategic acquisitions (Payapps for construction payments, PlanGrid, BuildingConnected, and others). These are product acquisitions in the AEC space that are generating recurring subscription revenue.
The Z-Score of 0.48 ("distress") is entirely misleading for a software company. The model penalizes negative working capital (deferred revenue exceeds current assets) and negative retained earnings (from buybacks), both of which are features, not bugs, of a profitable subscription software business.
The 18-Point Screening
Revenue Quality
| # | Check | Result | Detail |
|---|---|---|---|
| A1 | DSO | ⚠️ | DSO increased by 13 days (60 to 73 days) |
| A2 | AR vs Revenue | ❌ | AR outpaced revenue for 2 consecutive years |
| A3 | Revenue vs CFFO | ✅ | Revenue +17.5%, CFFO +52.6%. Cash outpaces revenue dramatically |
A1 and A2 are the most notable findings. DSO jumped 13 days and AR outpaced revenue for two years running. This requires explanation.
The 10-K provides the answer: Autodesk's new transaction model shifts billing from distributors to direct customer relationships. The 10-K notes: "With the continued growth of our online Autodesk branded store and our new transaction model, we are transacting directly with more end customers, rather than through distributors." Revenue from TD Synnex (the largest distributor) dropped from 39% of revenue in FY2024 to 33% in FY2025 to just 14% in FY2026.
When you shift from selling through distributors (who pay quickly) to billing end customers directly (who pay on standard net-30/60 terms), DSO naturally increases. Additionally, the 10-K notes the "transition to annual billings for multi-year contracts impacted the timing of our billings and cash collections."
A3 provides the counterweight: CFFO grew 52.6% vs 17.5% revenue growth. If AR growth were masking fake revenue, cash would not be growing 3x faster than revenue.
Expense Quality
| # | Check | Result | Detail |
|---|---|---|---|
| B1 | Inventory | ✅ | No material inventory. Pure software |
| B2 | CapEx | ✅ | CapEx -25.5% vs revenue +17.5%. Capital-light model |
| B3 | SG&A | ✅ | SG&A/Gross Profit = 46.8%. Normal for enterprise software |
| B4 | Gross Margin | ✅ | 91.0%, +0.4pp YoY. Stable at exceptional levels |
Software purity: no inventory, declining CapEx, 91% gross margins. The SG&A ratio at 46.8% of gross profit is typical for an enterprise software company with a large sales force and R&D organization.
Cash Flow Quality
| # | Check | Result | Detail |
|---|---|---|---|
| C1 | CFFO vs NI | ✅ | Ratio 2.18. Extraordinary cash conversion |
| C2 | FCF | ✅ | $2.4B, FCF/NI = 2.11. FCF doubles net income |
| C3 | Accruals | ✅ | -10.7%. Deeply negative — earnings heavily understated vs cash |
| C4 | Cash vs Debt | ⚠️ | Cash $2.6B covers 95% of debt $2.7B. Nearly even |
Cash flow quality is excellent across the board. The -10.7% accruals ratio is one of the lowest in our coverage universe, meaning reported earnings are significantly more conservative than cash earnings. This is the opposite of a manipulation signal.
C4 is a borderline watch: cash nearly covers debt. With $2.4B in annual FCF, Autodesk could pay off its entire debt balance in slightly over one year.
Balance Sheet
| # | Check | Result | Detail |
|---|---|---|---|
| D1 | Goodwill | ❌ | Goodwill + Intangibles $4.8B = 156% of equity |
| D2 | Leverage | ✅ | Debt/EBITDA 1.4x. Healthy |
| D3 | Soft Assets | ⚠️ | Other assets grew 63.1% vs revenue 17.5% |
| D4 | Impairment | — | No data |
D1 fails on the goodwill ratio, but as noted above, the denominator (equity) is artificially compressed by share buybacks. Goodwill + Intangibles actually declined 1% YoY, showing no aggressive acquisition activity.
D3 shows soft asset growth at 63.1% vs 17.5% revenue. This likely reflects deferred tax assets and other non-current assets growing with the business transition. The 10-K's critical accounting estimates section discusses business combination fair values and deferred tax valuation allowances, both of which create non-current asset balances.
Acquisition Risk
| # | Check | Result | Detail |
|---|---|---|---|
| E1 | Serial Acquirer | ✅ | FCF after acquisitions positive |
| E2 | Goodwill Surge | ✅ | Goodwill + Intangibles -1% YoY. Stable |
M-Score
| # | Check | Result | Detail |
|---|---|---|---|
| F1 | M-Score | ✅ | -2.68, below -2.22 threshold. Clean |
The DSRI (Days Sales in Receivables Index) at 1.215 is the highest M-Score component, reflecting the DSO increase discussed above. However, all other components are well within normal bounds, and the aggregate M-Score remains clean.
Ernst & Young's Critical Audit Matter focused on revenue recognition — "analyzing the accounting treatment for the Company's various product and service offerings" and "assessing the impact of terms and conditions in contracts with customers to determine whether products and services are considered distinct performance obligations." This is standard for a complex subscription software business but confirms that revenue recognition complexity is the area of highest audit risk.
Key Risks from the 10-K
1. New Transaction Model Transition Risk
The shift from distributor-based to direct-customer transactions is the most significant operational change. The 10-K warns it creates revenue recognition complexity and timing differences. Revenue from the largest distributor (TD Synnex) collapsed from 39% to 14% of total revenue in two years. While Autodesk says this shift is "broadly neutral to operating profit and free cash flow dollars," the transition period creates volatility in AR, DSO, and the composition of revenue that makes year-over-year comparisons unreliable.
2. AutoCAD Dependency
The 10-K warns of risk from "deriving a substantial portion of our net revenue from a small number of solutions, including our AutoCAD-based software products and collections." If customers migrate from individual AutoCAD products to collections without corresponding revenue uplift, results suffer.
3. AI Disruption
The 10-K acknowledges: "AI and machine learning are propelling advancements in technology, but if they are not widely adopted and accepted or fail to operate as expected, our business and reputation may be harmed." More pointedly: "disruptive technologies such as machine learning and other AI technologies may significantly alter the market for our products in unpredictable ways and reduce customer demand."
4. International Revenue Exposure
64% of revenue comes from outside the U.S. The 10-K warns of "tariffs, quotas, and other trade barriers and restrictions" and "fluctuating currency exchange rates, including devaluations, currency controls, and inflation."
5. Annual Billing Transition Impact on Cash Flow
The 10-K notes: "we expect the transition to annual billings for multi-year contracts to impact the timing of our billings and cash collections in fiscal 2026." This means future cash flow comparisons may show volatility as the billing frequency changes for existing multi-year customers.
Summary
Grade: C. Monitor the AR pattern but the cash profile is excellent.
Autodesk presents an interesting paradox: the automated screening flags concern (two fails, three watches, Z-Score in distress), but the cash flow profile is pristine — CFFO/NI of 2.18, FCF/NI of 2.11, accruals ratio of -10.7%, and M-Score clean at -2.68.
The AR growth and DSO increase are explained by the new transaction model that shifts billing from distributors to direct customers. This is a legitimate business transition, not an earnings quality issue — and the 52.6% growth in operating cash flow confirms that revenue is being collected, just on different terms.
The Z-Score of 0.48 is a false alarm: it penalizes subscription software balance sheets (negative working capital from deferred revenue, negative retained earnings from buybacks) that are actually signs of business health.
The goodwill at 156% of equity is a real concern but is stable year-over-year and funded by $2.4B in annual FCF. With 91% gross margins, 97% recurring revenue, and RPO of $8.3B providing multi-year visibility, Autodesk's business quality is high. The grade is C because the AR pattern deserves monitoring through the transition period, not because of manipulation risk.
**Disclaimer**: This report is based on Autodesk's FY2026 10-K (SEC EDGAR, fiscal year ended Jan 31, 2026, filed Mar 3, 2026) and public financial data. This is NOT investment advice.
Data: SEC EDGAR 10-K + Yahoo Finance
Auditor: Ernst & Young LLP (Unqualified opinion, auditor since 1983)
