Advanced Micro Devices’ Margins Rose, but Cash Flow Slipped

AMD’s consolidated growth accelerated in Q2 FY2026. Revenue reached $11.54 billion, up 50.3% from $7.68 billion in Q2 FY2025 and 12.6% from $10.25 billion in Q1 FY2026. With no company press release or prior guidance supplied, the appropriate report-card reference is the year-earlier quarter rather than consensus or a guidance midpoint.

Profitability changed more dramatically than revenue. Gross margin rose to 53.8%, compared with 39.8% a year earlier and 52.8% in the immediately preceding quarter. Q2 FY2025 was an unusually weak comparison, but the sequential improvement is also meaningful: gross margin added 1.0 percentage point while revenue increased by $1.29 billion.

Operating margin reached 17.3%, up from negative 1.7% in Q2 FY2025 and 14.4% in Q1 FY2026. Operating income was $1.99 billion. This was the strongest operating margin in the supplied 12-quarter history and continues a recovery from 10.8% in Q1 FY2025, 13.7% in Q3 FY2025 and 17.1% in Q4 FY2025. The result suggests that incremental revenue is currently carrying attractive operating leverage.

EPS increased to $1.38 from $0.54 a year ago and $0.84 in the prior quarter. Net income was $2.30 billion. The earnings result was therefore stronger than both the year-earlier and sequential comparisons, although the supplied data do not provide enough detail to separate operating improvement from below-the-line effects.

Free cash flow was the exception. It declined 9.8% year over year to $1.56 billion from $1.73 billion and fell from $2.57 billion in Q1 FY2026. The quarterly table defines free cash flow as cash from operations less capital expenditure. Because the supplied XBRL cash-flow point values appear to cover a different duration than the standalone quarter, the quarterly table is the controlling reference here.

The balance sheet remains a source of resilience rather than immediate concern. AMD ended the period with approximately $5.09 billion of cash and $2.35 billion of debt, equivalent to net cash of roughly $2.73 billion. That cushion limits financing risk, but it is small relative to the company’s $837.2 billion market capitalization and does not materially alter the valuation conclusion.

The quarter strengthens the operating side of the thesis. Revenue growth of 50.3%, a 53.8% gross margin and a 17.3% operating margin demonstrate that AMD can combine rapid expansion with improving profitability. FY2025 revenue had already grown 34.3% over FY2024, and the latest quarter extends that momentum rather than interrupting it.

Margin quality matters because the current valuation cannot be supported by revenue growth alone. At $504.64 per share, AMD trades at 129.7x trailing EPS of $3.89. A business at that multiple must sustain high growth while converting an increasing share of revenue into profit and cash. Q2 provided evidence for the first two requirements, but the free-cash-flow result was less conclusive.

Trailing free cash flow is $8.4 billion, equal to a 1.0% yield on the current market capitalization. Q2 free cash flow of $1.56 billion was positive and substantial in absolute terms, but it moved in the opposite direction from revenue and EPS. One quarter does not establish a structural cash-conversion problem, yet the divergence deserves attention because free cash flow ultimately supports intrinsic value.

The longer financial history also argues against treating the latest margin as permanently secured. Annual operating margin was 22.2% in FY2021, fell to 5.4% in FY2022 and 1.8% in FY2023, then recovered to 7.4% in FY2024 and 10.7% in FY2025. That record shows both the company’s earnings potential and the cyclicality or execution variability inherent in the results.

Return on invested capital remains another constraint. FY2025 ROIC was 4.5%, well below its FY2021 peak of 38.4%. The latest quarterly ROIC measure improved to 9.0%, but the valuation already anticipates a much more durable improvement. The thesis therefore turns on whether recent operating leverage develops into sustained free-cash-flow growth and higher returns on capital rather than remaining a favorable phase in a volatile history.

There is no prior SageNoodle fair value to carry forward, so this quarter establishes the initial valuation. Because forward guidance and consensus forecasts were not supplied, the scenarios use the disclosed trailing EPS of $3.89 and explicit earnings multiples. This avoids silently constructing a revenue or earnings forecast that the source material cannot support.

The bear case is $120 per share, based on 30.8x trailing EPS. It receives a 30% probability and assumes that the latest growth rate moderates, operating margins retreat from the Q2 level, and free-cash-flow conversion remains uneven. A 30.8x multiple would still recognize AMD as a growth semiconductor company while allowing for cyclicality and execution risk.

The base case is $220 per share, based on 56.6x trailing EPS, and receives a 50% probability. This is the snapshot fair value. The multiple assumes that Q2’s margin improvement is directionally durable, that earnings continue to grow, and that the company retains a substantial growth premium. It is deliberately generous relative to the current 4.5% annual ROIC and 1.0% trailing free-cash-flow yield.

The bull case is $400 per share, based on 102.8x trailing EPS, with a 20% probability. This outcome assumes that Q2’s 50.3% revenue growth and 17.3% operating margin signal a lasting earnings inflection, free cash flow catches up with reported profit, and the market continues to assign an exceptional multiple. Even this scenario is 20.7% below the current price.

The probability-weighted value of the three scenarios is $226 per share, close to the $220 base case. Under the required verdict rule, the shares are Overvalued because base fair value is only 43.6% of the current price. Put differently, the stock would need to decline approximately 56.4% to reach the base estimate, or trailing earnings would need to rise substantially without meaningful multiple compression.

EV/EBITDA is not disclosed in the supplied data because EBITDA is unavailable. The valuation therefore does not rely on that measure. The quarter justifies a strong operating assessment, but it does not justify marking fair value anywhere near $504.64 without assumptions extending beyond the provided evidence.

The bearish valuation conclusion could prove too conservative if Q2 marks the beginning of a prolonged earnings and cash-flow inflection. Revenue is already expanding rapidly, and operating margin has risen from negative 1.7% a year ago to 17.3%. If that operating leverage persists while free cash flow rebounds, trailing EPS could become a poor representation of the company’s earning power.

The supplied headlines point to enthusiasm around AI-optimized processors and a potentially expanding addressable market, but those articles are not used as financial evidence because the underlying company disclosures were not provided. Product mix, segment growth, backlog and customer concentration are also absent from the supplied filing extract. A stronger-than-visible business mix could therefore make the consolidated historical numbers understate future growth.

Conversely, the operating thesis would weaken if gross margin falls back below 50%, operating margin reverses from its 17.3% level, or free cash flow continues to lag net income. The annual record shows that AMD’s profitability can move materially across cycles. A premium valuation amplifies the share-price consequence of even a moderate slowdown.

The most important test is cash conversion over the next several quarters. Sustained quarterly free cash flow above the Q1 FY2026 level of $2.57 billion, accompanied by continued revenue growth and stable margins, would support a higher earnings base and could justify revisiting fair value. Continued free-cash-flow declines despite rising EPS would indicate that the headline earnings improvement is less valuable than it appears.

Valuation itself remains the clearest risk to the market price. At 129.7x trailing earnings, favorable execution is not enough; execution must exceed already demanding expectations for years. Q2 moved the operating evidence in the right direction, but the price still discounts a result beyond even the explicit bull case used here.