Grade: F — Major Red Flags
Framework: Schilit *Financial Shenanigans* + Beneish M-Score + forensic accounting principles
Data: SEC EDGAR 10-K (Fiscal year ended January 31, 2026) + Yahoo Finance
Auditor: Deloitte & Touche LLP — Clean opinion (1 critical audit matter: inventory valuation)
One-line verdict: Marvell's F grade is the most concerning in this batch — five fail flags, five watch items, and a Beneish M-Score of -1.59 that breaches the "elevated manipulation risk" threshold. DSO surged 32 days in one year, AR grew 113% against 42% revenue growth, CFFO consistently trails net income, and inventory is rising while margins expand. The data center AI boom is real and drove revenue up 42% to $8.2B, but the financial profile shows significant strain. The M-Score flag deserves serious investigation — while the individual components can be explained by rapid growth and a business mix shift, the aggregate signal is the strongest red flag we screen.
| Metric | Result |
|---|---|
| Red Flags | **5** (DSO surge, inventory vs COGS, CFFO < NI, goodwill/intangibles, M-Score) |
| Watch Items | **5** (AR growth, CFFO vs revenue, gross margin swing, cash vs debt, soft assets) |
| Checks Completed | **17/18** |
| Beneish M-Score | **-1.59** (ELEVATED MANIPULATION RISK zone) |
| Auditor | Deloitte & Touche LLP — Unqualified opinion |
The AI Data Center Play
Marvell Technology is a semiconductor company that designs and sells infrastructure solutions for data center, communications, enterprise, and automotive markets. The company underwent a dramatic business mix transformation in FY2026.
Note: Marvell's fiscal year ends in late January/early February. FY2026 ended January 31, 2026.
| Metric | FY2023 | FY2024 | FY2025 | FY2026 | Trend |
|---|---|---|---|---|---|
| Revenue | $5.9B | $5.5B | $5.8B | $8.2B | +42% |
| Net Income | ($163.5M) | ($933.4M) | ($885.0M) | $2,670.1M | Turn to profit |
| Gross Margin | 50.5% | 41.6% | 41.3% | 51.0% | +9.7pp |
| Net Margin | -2.8% | -16.9% | -15.3% | 32.6% | Turn to profit |
| ROE | -1.0% | -6.3% | -6.6% | 18.7% | Turn to profit |
From the 10-K revenue disaggregation:
| End Market | FY2026 | % of Total | FY2025 | % of Total | Growth |
|---|---|---|---|---|---|
| Data Center | $6,100.3M | 74% | $4,164.2M | 72% | +47% |
| Communications & Other | $2,094.3M | 26% | $1,603.1M | 28% | +31% |
| **Total** | **$8,194.6M** | **$5,767.3M** | **+42%** |
The data center segment now represents 74% of revenue, up from 40% just two years ago (FY2024). This concentration is both the growth engine and a key risk factor. Per the 10-K: "risks related to our dependence on a few customers for a significant portion of our revenue, particularly as our major customers comprise an increasing percentage of our revenue, as well as risks related to a significant portion of our sales being concentrated in the data center end market."
The company sold its automotive ethernet business to Infineon Technologies for $2.5 billion in cash (August 2025), recording a pre-tax gain of $1.8 billion. This gain is included in "interest income and other, net" and significantly boosted reported net income.
Net Income: Unpacking the $2.67B
The turn from ($885M) net loss to $2,670M net income requires careful decomposition:
| Item | Impact |
|---|---|
| Operating income improvement | ~$2.0B (from revenue growth + margin expansion) |
| Gain on auto ethernet sale | +$1.8B |
| Restructuring charges reduction | +$338M (FY2025: $353.9M, FY2026: $15.5M) |
| Higher interest expense | -$13M |
| Higher tax provision | -$386M |
The $1.8B gain on the automotive ethernet sale is a one-time item. Excluding it, net income would be approximately $870M — a meaningful improvement from the loss years but far less dramatic than the headline $2.67B suggests.
Per the 10-K: "We recognized interest and other income, net of $1.7 billion in fiscal 2026 as compared to interest and other loss, net of $174.4 million in fiscal 2025. The change was primarily due to the $1.8 billion gain on sale of our automotive ethernet business."
Cash Flow: The Weakest Link
| Metric | FY2024 | FY2025 | FY2026 |
|---|---|---|---|
| Operating Cash Flow | $1.37B | $1.68B | $1.75B |
| CapEx | ($0.35B) | ($0.29B) | ($0.36B) |
| Free Cash Flow | $1.02B | $1.39B | $1.39B |
| CFFO / Net Income | -1.47 | -1.90 | 0.66 |
CFFO/NI of 0.66 means only 66 cents of operating cash for every dollar of reported profit. Over the past three years, the ratio has been negative or below 1.0 consistently — this is the C1 fail. The key driver: $590.8M in stock-based compensation (7.2% of revenue) inflates the gap between GAAP earnings and cash generation.
Per the 10-K: "Stock-based compensation expense declined slightly in fiscal 2026 compared to fiscal 2025" — $590.8M vs. $597.4M. This is a massive non-cash expense that boosts reported operating cash flow (as a non-cash add-back) but also represents real dilution to shareholders.
The cash flow from operations also reflects factoring activity. The filing notes: "During the year ended January 31, 2026, we generated cash from operations from the sale of certain trade accounts receivable on a non-recourse basis to a third-party financial institution pursuant to a factoring arrangement." AR factoring can artificially boost operating cash flow by converting receivables to cash faster than natural collection.
The M-Score Warning: -1.59
The Beneish M-Score of -1.59 exceeds the -1.78 "elevated manipulation risk" threshold. This is the most serious flag in our screening. Let's examine each component:
| Component | Value | Interpretation |
|---|---|---|
| DSRI | 1.496 | DSO surged — AR growing much faster than revenue |
| GMI | 0.810 | Gross margin improving — benign |
| AQI | 0.825 | Asset quality improving — benign |
| SGI | 1.421 | High revenue growth — elevated but explainable |
| DEPI | 1.102 | Depreciation slowing slightly |
| SGAI | 0.676 | SG&A declining relative to revenue — operating leverage |
| TATA | 0.041 | Positive accruals — profits outpace cash |
| LVGI | 1.094 | Slight leverage increase |
The two primary drivers of the elevated M-Score are:
Context matters: Marvell is undergoing a genuine business transformation. Data center revenue grew 47% and now represents 74% of total. These customers (hyperscale cloud providers) typically have longer payment cycles than enterprise or consumer customers. The AR factoring activity suggests management is aware of the collection timing mismatch.
However, the M-Score does not distinguish between "genuine growth with collection lag" and "aggressive revenue recognition." The aggregate score demands investigation.
The 18-Point Screening
Revenue Quality
| # | Check | Result | Detail |
|---|---|---|---|
| A1 | DSO | Fail | DSO surged 32 days (65 to 97) |
| A2 | AR vs Revenue | Watch | AR growth 112.6% far exceeds revenue growth 42.1% |
| A3 | Revenue vs CFFO | Watch | Revenue +42.1% but CFFO only +4.1% |
A1: A 32-day DSO surge in a single year is a major red flag in any screening framework. DSO went from 65 days to 97 days. The 10-K notes concentration risk: four customers represented a significant portion of accounts receivable. The combination of DSO surge, AR factoring, and data center concentration suggests Marvell is extending more generous payment terms to win or retain large AI chip customers.
A3: Revenue grew 42% but operating cash flow grew only 4%. This divergence is concerning — in a healthy growth scenario, cash flow should at least keep pace with revenue. The gap is driven by AR build-up (cash tied up in receivables) and the working capital demands of rapid growth.
Expense Quality
| # | Check | Result | Detail |
|---|---|---|---|
| B1 | Inventory | Fail | Inventory growth 34.8% exceeds COGS 18.6%, margin rising |
| B2 | CapEx | Pass | CapEx growth 23.0% vs revenue 42.1%. Normal |
| B3 | SG&A Ratio | Pass | SG&A/Gross Profit = 18.3%, excellent |
| B4 | Gross Margin | Watch | Gross margin swung +9.7pp (41.3% to 51.0%) |
B1: Inventory growing 34.8% while COGS grew only 18.6%, with gross margins simultaneously expanding, is a classic Schilit red flag. The 10-K identifies inventory as a critical audit matter area: "Inventories are stated at the lower of cost or net realizable value. The Company records a reduction to the carrying value of inventory that is determined to be excess, obsolete or unsellable based upon assumptions about future demand and market conditions. As of January 31, 2026, the Company's consolidated inventories balance was $1,388.0 million."
Deloitte flagged this as the sole CAM: "We identified inventory valuation as a critical audit matter because of the significant assumptions management makes with regard to estimating the net realizable value of inventories, specifically, forecasted demand." Inventory building ahead of demand is normal in a growth cycle — but the magnitude (34.8% vs. 18.6% COGS) means Marvell is betting heavily that demand will continue.
B4: Gross margin improved from 41.3% to 51.0%, driven by "better cost absorption driven by higher revenues, partially offset by a shift in product mix" and the absence of $357.9M in impairment charges from FY2025 restructuring. The 10-K states: "The decrease in cost of goods sold as a percentage of net revenue was primarily due to impairment charges of $357.9 million for acquired intangible assets, inventories, property and equipment, and other non-current assets associated with restructuring actions during fiscal 2025."
Cash Flow Quality
| # | Check | Result | Detail |
|---|---|---|---|
| C1 | CFFO vs NI | Fail | CFFO < Net Income for 3 consecutive years |
| C2 | FCF | Pass | $1.4B, FCF/NI = 0.52 |
| C3 | Accruals | Pass | 4.1% accruals ratio. Low |
| C4 | Cash vs Debt | Watch | Cash $2.6B covers 55% of debt $4.8B |
C1: This is a persistent problem. Over three years, CFFO has been below net income — with ratios of 0.66 (FY2026), -1.9 (FY2025), and -1.47 (FY2024). The negative years reflect GAAP net losses with positive operating cash flow (the losses were driven by non-cash charges). Even in the profitable FY2026, cash generation lagged earnings.
The $590.8M stock-based compensation is the primary distortion — it's added back to CFFO but represents real economic cost (dilution). When SBC represents 7.2% of revenue, the gap between GAAP NI and CFFO becomes structurally embedded.
Balance Sheet
| # | Check | Result | Detail |
|---|---|---|---|
| D1 | Goodwill + Intangibles | Fail | $13.1B = 92% of equity |
| D2 | Leverage | Pass | Debt/EBITDA 1.1x. Healthy |
| D3 | Soft Asset Growth | Watch | Other assets grew 105.7% vs revenue 42.1% |
| D4 | Impairment | N/A | No write-off data |
D1: Goodwill of $11.1B and intangible assets of $2.1B total $13.1B, or 92% of stockholders' equity of $14.3B. This is the legacy of the 2021 Inphi acquisition (approximately $10B). The goodwill is not growing — intangibles declined 11% YoY as acquired intangibles amortize. But the size relative to equity means any significant impairment would crater book value.
D3: Soft assets grew 106% vs. revenue growth of 42%. Part of this reflects the Celestial AI and XConn Technologies acquisitions completed post-fiscal year end (February 2026): "We completed the previously announced acquisitions of Celestial and XConn in which we paid $1.3 billion (or $1.0 billion, net of cash acquired) and $280.0 million in cash, respectively. We also issued shares of approximately 24.5 million shares for Celestial and approximately 2.1 million shares for XConn."
M&A Risk
| # | Check | Result | Detail |
|---|---|---|---|
| E1 | Post-Acquisition FCF | Pass | FCF after acquisitions positive |
| E2 | Goodwill Surge | Pass | Intangibles declined 11% YoY |
Beneish M-Score
| # | Check | Result | Detail |
|---|---|---|---|
| F1 | M-Score | Fail | -1.59 (> -1.78). ELEVATED MANIPULATION RISK |
Critical Audit Matter: Inventory Valuation
Deloitte identified inventory excess and obsolescence as the sole critical audit matter: "We identified inventory valuation as a critical audit matter because of the significant assumptions management makes with regard to estimating the net realizable value of inventories, specifically, forecasted demand. This required a high degree of auditor judgment and an increased extent of effort when performing audit procedures to evaluate the reasonableness of management's estimate of forecasted demand."
At $1.39B, inventory represents a meaningful portion of current assets. If data center demand slows — due to customer digestion cycles, competitive pressure, or macro conditions — Marvell could face significant write-downs. The FY2025 restructuring already included $357.9M in impairment charges on inventories and other assets.
Key Risks from Item 1A
1. Data center customer concentration. The 10-K warns of "risks related to our dependence on a few customers for a significant portion of our revenue, particularly as our major customers comprise an increasing percentage of our revenue." Four customers represented a significant portion of receivables.
2. Custom silicon risk. Hyperscale customers are increasingly designing their own chips. The filing notes "risks related to the potential impact of AI on our business model and products" and customer ability to "develop their own solutions, vertically integrate which may reduce the need for our products."
3. China trade restrictions. Per the risk factors: "risks related to tariffs and trade restrictions with China and other foreign nations including risks related to the ability of our customers, particularly in jurisdictions such as China that may be subject to trade restrictions."
4. Acquisition integration. Marvell completed the Celestial AI ($1.3B) and XConn Technologies ($280M) acquisitions post-fiscal year end. The Inphi acquisition ($11B goodwill) remains the dominant balance sheet feature. The filing notes additional contingent consideration: "With respect to the Celestial AI transaction, we may be required to issue additional shares of our common stock through fiscal 2029."
Altman Z-Score and F-Score
| Model | Score | Interpretation |
|---|---|---|
| Altman Z-Score | **4.02** | Safe zone (>2.99). Healthy |
| F-Score (Dechow) | **1.45** | Low fraud probability (0.54%) |
The Z-Score of 4.02 indicates no bankruptcy risk. The F-Score's fraud probability of 0.54% is low, though the soft asset ratio (0.83) — reflecting the goodwill-heavy balance sheet — is the highest component. Notably, the M-Score and F-Score can diverge because they measure different things: the M-Score focuses on year-over-year changes in financial ratios (and Marvell's rapid transformation triggers multiple change-based flags), while the F-Score uses level-based metrics.
Summary
| # | Check | Result |
|---|---|---|
| A1-A3 | Revenue Quality | Fail-Watch-Watch |
| B1-B4 | Expense Quality | Fail-Pass-Pass-Watch |
| C1-C4 | Cash Flow Quality | Fail-Pass-Pass-Watch |
| D1-D4 | Balance Sheet | Fail-Pass-Watch-N/A |
| E1-E2 | M&A Risk | Pass-Pass |
| F1 | Beneish M-Score | Fail |
Grade: F. Multiple flags demand investigation.
Marvell's F grade is earned through a broad pattern of concerning signals, not just one or two structural quirks:
The data center AI tailwind is real — Marvell's 47% data center revenue growth and design win pipeline are genuine. But investors should be especially vigilant about AR collection trends, inventory turns, and whether operating cash flow catches up to reported earnings in coming quarters. In companies undergoing rapid transformation, the M-Score can generate false positives — but it also catches real problems early. Monitor closely.
**Disclaimer**: This report is based on Marvell's FY2026 10-K (SEC EDGAR, fiscal year ended January 31, 2026) and public financial data. This is NOT investment advice.
Data: SEC EDGAR 10-K (Fiscal year ended January 31, 2026) + Yahoo Finance
Auditor: Deloitte & Touche LLP (Unqualified opinion, 1 critical audit matter)
