Surpassing expectations but falling 28.58%—where did Intel go wrong?

NEXT PICK · Market Insights

Surpassing expectations but falling 28.58%—where did Intel go wrong?

Friday, October 30, 2026

INTC trended today not because of a poor financial report, but quite the opposite: Q3 revenue was $16.51 billion, up 20.9% year-on-year, with adjusted earnings per share of $0.52, far exceeding the market expectation of $0.40, yet the stock price fell 28.58% in one day, closing at $65.76.

This was the most devastating single-day drop since 1974, with trading volume expanding to 4.4 times the daily average over the past three months. During the session, it once dropped to $63.2, with its market value dropping to just $349.19 billion—nearly $140 billion lost in a single day.

What broke the stock price were three forward-looking statements

Q4 revenue guidance is $15 to $16 billion, significantly below the expected $17.07 billion; PC declines will continue at least until the first half of next year; CFOs also admit that positive free cash flow is unlikely to be produced in 2027. So, is this a misjudgment or a belated repricing?

Latest price $65.76 ▼ -28.58%

Data center revenue $7.3B ▲ +78%

Q4 revenue guidance $15-16B ▼ Expected $17.07B

A company that both designs and chips has advantages and burdens

Intel is a rare full-chain player in the semiconductor industry, covering everything from chip design, wafer manufacturing, to advanced packaging all the way to computing product sales. Its clients include OEMs, ODMs, cloud service providers, distributors, and manufacturing service companies, while most competitors only design and do not manufacture chips.

The third-quarter ledger clearly illustrates this structure

Client business revenue was $8.1 billion, down 5% year-on-year; Data center and AI business revenue was $7.3 billion, up 78% year-on-year, with operating profit of $3.039 billion; Foundry revenue was $6.7 billion, up 58% year-on-year, but posted a loss of $1.853 billion.

Even more glaringly, of Foundry's $6.7 billion, only $411 million comes from external customers, with the vast majority still coming from Intel itself. This business currently feels more like an internal factory and is not yet a true external foundry platform.

The report gave it a score of 6 out of 10 for irreplaceability. The scores come from the combined optimization of design and process through internal manufacturing, capacity scheduling rights during tight supply, the strategic value of advanced domestic manufacturing, and capabilities like self-developed multi-beam mask writing equipment that are close to the upstream of the industry chain.

Patents, customer relationships, and scale are also bargaining chips, but each has boundaries: existing materials lack patent count, remaining term, or litigation outcomes, so it cannot be asserted that blockages form; Cloud vendors are accustomed to multiple vendors; switching platforms may incur verification and migration costs but lack exclusivity. If yield and capacity utilization lag behind, scale actually amplifies depreciation pressure.

The multi-layer AI fabric in cooperation with Infosys aims to connect infrastructure, models, data, applications, and workflows into an agent-oriented ecosystem, but its revenue contribution has yet to be disclosed. A costly factory—is it a moat, or a fixed cost burdened by it? The answer depends on whether 14A can win over external customers.

After dropping nearly 30%, why is it still not considered cheap?

Opening the valuation snapshot, the price-to-earnings (P/E) ratio has lost its significance due to the latest earnings per share of -$2.12, the price-to-sales ratio (P/S) at 5.83x, the enterprise value multiple (EV/EBITDA) at 162.81x, and the free cash flow yield at -0.31%. Technically severely oversold does not mean the valuation is cheap.

Adding the revenue from the first three quarters to the fourth-quarter guidance, the full-year forecast is only $61.2 billion to $62.2 billion, below the lowest estimate of $64 billion in the research report. Based on the guidance range, the current price-to-sales ratio remains above 5.6 times, which still bets on sustained profitability and successful manufacturing transformation.

Therefore, the inverse valuation framework in the report requires a more cautious interpretation

A price-to-sales ratio of 3 to 3.5 times corresponds to a market cap of $192 billion to $231 billion, 4 to 5 times to $264 billion to $350 billion, and 5.5 to 6.5 times to $385 billion to $494 billion. This is only sensitivity analysis, not a target price.

Current prices are close to the upper end of the mid-range scenario, which itself requires revenues of $66 to $70 billion. The high-end scenario requires maintaining an operating margin of about 15.3%, foundries securing verifiable design orders, and mass production of 14A units simultaneously, with any delay potentially pushing valuations back to the mid-to-low range.

Has the negative impact already been reflected in the stock price? Our judgment is that only part of the price has been realized: the weakening guidance in Q4 has basically been reflected, but there is no specific figure for capital expenditure in 2027, and negative free cash flow has not yet appeared in the financial statements, so this uncertainty remains.

With $9.7 billion in cash flow, why can't there be any cash left?

The $9.7 billion in fiscal year 2025 is often treated as free cash flow, but after checking the financial report, it is actually operating cash flow. Adjusted free cash flow by company caliber is -$2.228 billion in fiscal year 2024 and -$1.612 billion in fiscal year 2025, both negative.

The three quarters of 2026 are -2 billion, -8.4 billion, and +1.639 billion USD respectively, with a net outflow of 12.2 billion USD from partners in the second quarter. Operating cash flow is positive year after year, but capital expenditures are included but no money remains—this is the true face of a heavy asset model.

Fortunately, there is no short-term shortage of funds. In August, the company issued a new share at $95 per share, raising $23 billion; At the end of Q3, cash and short-term investments totaled $54 billion, slightly above the $50.13 billion in debt. The conclusion is not that "cash is about to run out," but rather that "it is still far from self-sustaining," and investors in the additional share have now posted a 30% unrealized loss.

During the validation period before the breakout, the market had already voted against it

The report places INTC at the "pre-breakout (pre_breakout)" stage, based not on technical oversold conditions but on the fact that operating data has begun to recover but has not yet been verified. As a company, it has long matured, but the transformation expectations corresponding to stocks are still halfway from PC cyclical stocks to AI and foundry platforms.

Looking at the moving average, $65.76 is 24% lower than the 200-day average of $86.78 and 35% lower than the 50-day average of $101.63. The market clearly no longer prepays a premium at this stage. RSI (14) at 24.5 and Chan -51 only indicate that trading is crowded and oversold, and cannot replace a breakout in operations.

The confirmation signal is that subsequent quarters will maintain positive operating profit, data center growth will continue to translate into profit, 14A customer approvals, and foundry external revenue and capacity utilization will rise. The disproved signal is a negative single-quarter profit and delayed milestones. The clause "free cash flow remains positive" has been personally excluded from 2027 by the CFO.

Under a five-pronged attack, how does Intel hold its ground?

Intel is not facing a single competitor but a combination of pressures along the value chain. NVIDIA uses CUDA software, accelerators, and network platforms to form system-level lock-ins. If AI workloads continue to migrate to dedicated accelerators, even if Intel benefits from inference and infrastructure expansion, it may not achieve the same unit value.

AMD's EPYC competes head-on with Xeon, with energy efficiency and open platforms helping it win orders from cloud vendors and OEMs. Intel has fought back, with management saying Xeon demand is "exceptionally strong." The current challenge is actually a shortage of capacity and substrates, but market share is shifting, and there is a lack of verifiable shipment and pricing data.

The manufacturing competitor is TSMC, and the gap can be directly quantified: TSMC's Q3 revenue was $46.7 billion, up 41.2% year-on-year, with a gross margin of 66.4%; Intel's revenue during the same period was $16.51 billion, with an adjusted gross margin of 44.1%. The former is a mature foundry ecosystem, while the latter is still busy pursuing the first batch of major orders from external clients.

Marvell is competing for cloud vendor budgets in custom chips, interconnects, and optical communications; The Arm camp, represented by Qualcomm, is eroding x86. There are clues suggesting Arm's server CPU share may rise from 10% to 30%, but like the "server CPU shipments expected to increase by over 40% in 2027," these have not been verified by raw data.

The moat is real, but it still hasn't collected enough 'tolls'

Breaking it down, Intel's verifiable barriers have five layers: patent and process accumulation, switching costs from platform migration, x86 ecosystem inertia, R&D and manufacturing scale, and the strategic scarcity of domestic capacity created by export controls and geopolitics.

But these barriers do not equal pricing power. Patents may not be unavoidable; long-term relationships only increase the chances of success. Ecosystem evidence remains in cooperation announcements, with no developer numbers or software revenue; Regulation protects local production capacity but cannot guarantee product competitiveness; substrate and storage procurement remain exposed to shortages and trade frictions.

14A If yield and mass production are achieved, manufacturing barriers will be upgraded into sources of growth and pricing; Conversely, the larger the asset, the heavier the depreciation.

The market is not small, but today's problem is that the ceiling is sinking

Existing materials do not have authoritative market size figures for server CPUs, AI accelerators, clients, networks, or foundries, nor do we fabricate TAM, SAM, or SOM amounts. Logically, the total disk covers all the above demands, and the available market depends on x86 and external foundry customers, while the available share is constrained by 14A yield, performance, and capacity.

Boundaries are becoming clearer

Management expects total PC sales to decline by about 12% in 2026 and another 10% in 2027, with storage shortages driving up overall device prices; Q4 client revenue is expected to drop over 20% quarter-on-quarter, while data center sales are expected to grow by around 15%. This is a shift in turn, not a broad rise.

For the three-year scenario, optimism suggests continued high growth, a surge in volume from data centers and external foundry, and mass production from 14A clients; The benchmark is revenue growth but with volatility and capital expenditure pressure; The pessimistic scenario is growth concentrated in low value-added businesses and insufficient capacity utilization. Due to the lack of verifiable forecasts, no three-year revenue range is provided.

Bulls' confidence: AI demand and recovery are genuine

The core of the bull market isn't "too much to drop," but that operations have emerged from the trough. If this continues, high price-to-sales ratios will be gradually absorbed by profits:

▲ Bull Case

Bears' confidence: The current price is still paying for success that has not yet happened

Bears are not saying companies have lost competitiveness, but rather that current prices demand sustained profitability and successful manufacturing, yet the evidence is still insufficient:

▼ Bear Case

Next, focus on three things: financial reports, clients, and milestones

In the next two quarters, the share price will most likely have a say in the progress of deliveries, not a rally. The high growth in the 14A design suite and data center in October has already materialized, and the real test is now:

•	Around January 28, 2027, Q4 financial reports and 2027 capital expenditure guidance: High impact, direction depends on expenditure scale
•	Over the next two quarters, 14A external clients have announced that management has made expectations, and if they fail, it will be clearly bearish
•	Before Q1 2027, the 14A 1.0 design kit will be released; Risk trial production in the second half of 2027: high impact will determine the credibility of mass production in 2028
•	From Q4 2026 to Q1 2027, Israel's Fab 28 recovered: $300 million in revenue each quarter, with gross margin dropping by 200 basis points in Q4

From huge losses to wild surges, and then to a single day of falling back

Fiscal year 2024 is at its bottom: revenue of $53.1 billion, operating loss of $11.678 billion, and net loss of $18.756 billion. Fiscal year 2025 revenue is $52.85 billion, down 0.5% year-over-year, with operating losses narrowing to $2.214 billion, net loss of only $267 million, and loss per share of $0.06, as market fears begin to fade.

Volatility in 2026 will be even more dramatic: Q1 operating loss of $3.136 billion, loss per share of $0.73; Q2 operating profit turned positive to $1.8 billion, but loss per share was $2.16; Q3 net profit returned to $3.32 billion. Fundamentals are repairing, but the slope of recovery is being prematurely depleted by the stock price.

The roller coaster of stock prices is even more exaggerated

This year, it once reached a 52-week high of $142.34, but has now pulled back 53.8%, but still up 63.7% over the past 12 months. The previous snapshot showed a market cap of $459.68 billion, but today's $349.19 billion is a difference of $110.49 billion, reflecting this round of expectations compression.

Looking back, the market judged that operations had improved from the first quarter's low, but the mistake was to extrapolate a quarter's good results into a comprehensive reversal. The financial report exceeded expectations but plunged, indicating that investors were not buying in the past three months, but in the continuity of the next three years.

⚠️ Risk Notice • Profit recovery interrupted: Q4 guidance has clearly weakened; if quarterly operating profit turns negative, valuations will lose support • 14A yield, mass production ramp-up, or customer introduction may fall short of expectations, resulting in the 2028 mass production target falling short • With rising capital expenditures and insufficient capacity utilization, negative free cash flow in 2027 may exceed expectations • TSMC, AMD, NVIDIA, and the Arm camp are competing on multiple fronts, squeezing market share and profit pools • PC demand remains weak, with AI server demand dragged down by substrate shortages and the Fab 28 shutdown

🟡 Neutral The repair is real, but the cash flow turning positive has been delayed, and the current price has not been fully discounted due to failure

💬 Discussion A deep drop doesn't mean cheap; now is not the time to bottom-fish. Live buy and sell positions post INTC on the official account

Data source

•	Source: NextPick real-time snapshot + company earnings reports + SEC announcements + mainstream financial media.