Intel Loses $19B in Value as Foundry Questions Meet a 24-Year High in Yields — Hold Into Q3, Upside to $123

Intel traded at $109.51 at 11:34 a.m. ET on Thursday, October 8, down $3.61 or 3.19% from Wednesday's $113.12 close. The stock was indicated at $110.45 before the open, was at $111.06 at 10:59 a.m. and slid to $109.26 by 11:02 a.m. as the Nasdaq extended its losses. Volume reached 39.8 million shares in the first two hours, the most of any US-listed stock and 37% of the 106.7 million three-month daily average.

At $109.51 the company is valued at $575 billion on 5.25 billion shares outstanding, with an enterprise value near $595 billion. The session's decline removed $19 billion of market value. The 52-week range runs from $32.89 to $142.35, with the high set on June 30. Intel is 23% below that peak and still up 197% for 2026.

There was no company announcement behind the move. The only investor notice of the week was the scheduling of third-quarter results for October 29. The selling came from two directions that have been pressing on the stock since Monday. Semiconductors were the weakest group in a falling market, with the 10-year Treasury yield touching 5.35% and Brent crude at $105 a barrel. And the debate over Elon Musk's Terafab project, where Taiwan Semiconductor is now in talks to take a role alongside Intel, has put a question mark over the most visible outside customer for Intel's foundry.

Those two pressures land on a valuation that leaves little margin for error. Intel trades at 67 times forward earnings, 10.1 times trailing sales and 204 times trailing free cash flow. The company lost $11.29 billion over the past twelve months on a GAAP basis. Investors are paying for a turnaround in manufacturing that has shown real progress in the numbers but has yet to produce a signed leading-edge customer.

The stock has also become a headline instrument. It gained 9.1% in a single session earlier in this stretch, fell 2% on Monday when the TSMC talks were confirmed, rose 0.55% on Wednesday after chief executive Lip-Bu Tan said Intel remains in the project, and gave back 3.19% on Thursday. Over five trading days it is down 6%. Over one month it is up 18%.

A Week of Terafab Whiplash: How the Tape Got Here

The sequence began on Friday, October 2, when a report said TSMC was exploring ways to help Terafab run its planned Texas factories. The most likely arrangement, according to that report, would have TSMC own and operate a new facility with Terafab as an anchor customer. Early on Saturday, Musk confirmed the conversations on X: "Just discussions, but something may come of it."

Intel opened Monday with a premarket gain of 3% on the idea that any Terafab progress was good news, then reversed and fell more than 5% as traders worked through what a TSMC role would mean, before bouncing 1.5% off the low. It finished the session down 2%. On Tuesday the stock closed at $112.50.

Wednesday brought a response from the top. Tan told reporters ahead of an event in Tokyo marking 50 years of Intel operations in Japan that the company will remain part of Terafab. The shares traded between $111.15 and $115.30 and closed at $113.12, up $0.62 or 0.55%, on 82.3 million shares. That was a modest gain on a day when the Nasdaq fell 0.22%, and it suggested the reassurance had registered.

Thursday undid it. Premarket indications were down 2.07% to 2.44% between 7:47 a.m. and 8:33 a.m. ET, before the broader market had opened, with commentary again pointing to the TSMC discussions. The cash session added the macro layer: chip stocks sold off across the board as crude jumped 5% and long-dated yields returned to 24-year highs. By late morning Intel was the most active name on the tape and 3.19% lower.

The pattern in the price is a series of lower highs since the June 30 record. Intel peaked at $142.35, fell to $92.32 the day after second-quarter results in July, and rebuilt to the mid-$110s by early October. One trading reference had the stock at $123 during the recent run. The high this week was $115.30. Each rally has stopped short of the last.

Volume confirms that conviction is thinner on the way up. Wednesday's 82.3 million shares on a gain compares with a 20-day average of 105.3 million. Thursday's 39.8 million in two hours puts the stock on pace to exceed that average on a decline.

Terafab and TSMC: What Is Actually at Stake for 14A

Terafab is the chip manufacturing venture Musk is building in Texas to supply Tesla, SpaceX and xAI. Intel joined in April, and Musk said at the time that Tesla would use Intel's forthcoming 14A process for chips made there. That made Tesla the first major announced outside customer for 14A, and until last week 14A was the only process technology publicly attached to the project.

The significance for Intel goes beyond one contract. In its annual report the company states that 14A is its first node designed from the start for external customers, and that if it cannot secure a significant external foundry customer for it, it "may pause or discontinue" pursuit of next-generation leading-edge processes. Tan has said the same in plainer terms. A buyer with three of the most capital-intensive AI and automotive programs behind it was the anchor tenant the roadmap needed.

Two versions of a TSMC arrangement have been described. In one, TSMC owns and runs a new Texas facility and Terafab is its anchor customer. In the other, SpaceX holds a majority stake and TSMC supplies technology and operating expertise. In either case TSMC would be expected to bring its own manufacturing process, which is the job 14A holds today. Musk has indicated TSMC could supplement Intel, and no agreement has been signed.

The softer reading leaves room for both. Terafab's stated ambitions are large enough that a single foundry partner may be unable to supply them, and Intel offers things TSMC does not at the same scale in the US: co-design of chips and process, advanced packaging and domestic production on a process designed in America. SpaceX is reported to be raising $40 billion in debt to buy Nvidia chips, which shows the scale of spending in the Musk companies.

The harder reading is that the most prominent name on Intel's 14A list is now shared, and that TSMC, with 62% of the global foundry market and decades of experience building chips for outside customers, is the default choice whenever it is available. Intel's foundry record with third parties is short.

Two near-term markers will show which reading is closer. Intel's 14A process design kit is due in October. And Tan said in January that customers would begin making firm 14A supplier decisions in the second half of 2026, with more through mid-2027. The first of those decisions should be visible by the January earnings call. So far, no major outside company has publicly committed to having Intel build its chips on 14A at Intel's own fabs.

Foundry Economics: $5.77 Billion of Revenue, $293 Million From Outside

The numbers behind the foundry explain why one customer matters so much. Intel Foundry reported second-quarter revenue of $5.77 billion, up 31% from a year earlier and 6% from the first quarter. Of that, $293 million came from external customers, 5.1% of the segment total. The rest was Intel manufacturing its own processors. The foundry posted an operating loss of $2.09 billion for the quarter, an annualized run rate above $8 billion.

There is progress inside those figures. External revenue was $22 million in the second quarter of 2025, so it has grown thirteenfold in a year, though most of it is tied to Altera, the programmable-chip business Intel partly divested. Wafer output on the 18A process rose more than 50% from the first quarter to the second, and management said yields were ahead of internal targets. The 18A-P variant entered risk production on the schedule shared with customers a year ago.

Intel also became the first chipmaker to put ASML's High-NA EUV lithography into high-volume manufacturing, using it for a subset of Core Ultra Series 3 processors, code-named Panther Lake. That is the tool set 14A is designed around. Getting it into volume production on an 18A-family product reduces one of the larger technical risks on the next node.

The specifications Intel has published for 14A are competitive on paper: a 15% to 20% performance gain over 18A, a 30% increase in density and a 25% to 35% reduction in power consumption. Volume production is planned for 2028. A further variant, 18A-PT, aimed at higher interconnect density for stacked designs, is also slated for 2028.

The investment required is heavy. Intel raised its 2026 capital spending outlook to more than $20 billion on the July call. Its Ohio complex, budgeted at $28 billion, is scheduled for 2027 to 2028. It expanded assembly and test capacity in Penang, Malaysia, to meet demand for advanced packaging.

What the foundry lacks is the thing that makes the economics work: outside volume at the leading edge. A fab is a fixed-cost asset, and filling it only with internal products leaves it below the utilization that TSMC achieves by pooling demand from Apple, Nvidia, AMD and hundreds of others. Intel's annual filing names no leading-edge external customer. The $2.09 billion quarterly loss narrows meaningfully only when that changes.

Second-Quarter Results: The Strongest Revenue Growth in 15 Years

The case for the stock rests on the quarter Intel reported on July 23. Revenue was $16.13 billion, up 25% from $12.86 billion a year earlier and the company's fastest growth in more than 15 years. That was $1.8 billion above the midpoint of management's own guidance and 12% above the $14.45 billion consensus. Adjusted earnings were $0.42 a share, double the $0.21 expected and a swing from a $0.10 loss in the prior-year period.

Margins did the most to change the narrative. GAAP gross margin was 40.4%, up 12.9 percentage points from 27.5% a year earlier. Non-GAAP gross margin reached 41.8%, against 29.7%, and came in 280 basis points above the 39.0% guide. Non-GAAP operating margin moved from negative 3.9% to positive 17.2%. Operating income was $1.80 billion against a $3.18 billion loss, with research and development spending down 8.6% to $3.37 billion.

The Data Center and AI segment led. Revenue rose 59% to $6.26 billion with a segment operating margin of 40%, and custom chip revenue nearly tripled. Tan said on the call that "AI is driving unprecedented demand for compute" and that demand for Intel's products continued to outpace its growing supply. The Client Computing Group grew 13% to $8.88 billion.

The GAAP bottom line was a different story. Intel reported a net loss of $11.03 billion, or $2.16 a share, far wider than the $2.92 billion loss a year earlier. With operating income positive, the loss came from items below that line. It is the reason trailing twelve-month net income stands at negative $11.29 billion and the trailing price-to-earnings ratio is not meaningful.

The stock's reaction is instructive for anyone positioning into October 29. Intel closed at $100.23 the day before the report and at $92.32 the day after, a decline of 7.89% on what was by any measure a blowout quarter. The shares had run into the print, and a beat of that size was already in the price. In April, by contrast, the stock jumped 23.60% from $66.78 to $82.54 after first-quarter results. In January it fell 17.03% from $54.32 to $45.07.

Three reports, three moves averaging 16% in absolute terms. Applied to $109.51, a move of that size implies a post-earnings range between $92 and $127.

October 29: Why 42% Gross Margin Is the Number That Moves the Stock

Intel reports third-quarter results after the close on Thursday, October 29, with a conference call at 2:00 p.m. Pacific time. Guidance issued in July calls for revenue of $15.8 billion to $16.8 billion, GAAP gross margin of 41.0% and non-GAAP gross margin of 42.0%, with GAAP earnings of $0.31 a share and non-GAAP earnings of $0.38.

The revenue midpoint of $16.3 billion would be a 19% increase from a year earlier and 1% above the second quarter. Consensus has already moved past it. Published estimates range from $16.42 billion to $16.72 billion on revenue and $0.39 to $0.40 on adjusted earnings, against $0.23 in the third quarter of 2025. Intel has exceeded its own revenue forecast for seven consecutive quarters, so the market expects an eighth.

That is why revenue may be the least important line in the release. A stock at 67 times forward earnings and 55 times 2027 estimates is priced on where margins are going. The 42.0% guide is 20 basis points above the 41.8% delivered in the second quarter. Flat margins on higher revenue would raise the question of whether supply constraints and the cost of ramping 18A are absorbing the benefit of volume.

Three things would move the stock up. A non-GAAP gross margin of 43% or better would show operating leverage arriving faster than guided. A fourth-quarter outlook above $17 billion would extend the growth run. And any named 14A customer commitment, or a clear statement on the structure of Intel's role at Terafab, would address the largest open question directly.

Three things would move it down. A margin at or below 42% with cautious commentary on substrates and memory supply would feed the view that the easy gains are done. A capital spending figure moving further above $20 billion without a matching customer announcement would pressure free cash flow. And a foundry operating loss that fails to narrow from $2.09 billion would revive doubts about the segment's path to breakeven.

Management flagged on the July call that industry-wide shortages of wafers, memory and substrates were the dominant near-term constraint. Those shortages cap how much demand Intel can convert to revenue, and they raise input costs. A historic surge in memory prices has been a feature of the second half across the hardware sector. How much of that Intel passed through in pricing will show up in the margin.

The timing adds a macro variable. The Federal Reserve's rate decision lands on October 28, the day before the report.

Data Center and Client: Where the Earnings Power Sits

Intel's two product businesses are carrying the company while the foundry absorbs losses. Data Center and AI generated $6.26 billion in the second quarter at a 40% operating margin. Client Computing generated $8.88 billion. Together they produced $15.1 billion of the quarter's revenue before intersegment eliminations.

The data center recovery is the more surprising of the two. Server revenue had been in decline for years as AMD took share and as spending shifted to GPUs. The 59% growth rate reflects two things. Demand for general-purpose compute has risen alongside AI accelerators, because every AI server still needs host processors, and inference workloads run on CPUs in large volumes. And pricing has improved: in the first quarter, server average selling prices rose 27% from a year earlier, accounting for most of that period's $926 million segment increase.

Custom silicon is the newer line. Revenue from custom chips nearly tripled in the second quarter. These are processors designed with and for specific cloud customers, a business where Intel's ability to pair design with its own manufacturing and packaging is a selling point. Intel expanded a cloud AI infrastructure partnership with Google in April.

Client Computing is the larger and slower business. Growth of 13% to $8.88 billion came on a product cycle that includes Core Ultra Series 3 on 18A, Core Ultra 200S Plus and 200HX Plus for desktop and mobile, and Xeon 600 for workstations. The first-quarter 10-Q showed client revenue up just $98 million from a year earlier, so the second-quarter acceleration was a step change.

Client is also where the risk sits heading into the fourth quarter. Premarket commentary on Thursday referenced reports of weakening PC demand pressuring chipmakers. Memory prices have surged, which raises the cost of finished PCs and can suppress unit volumes. Amazon said this week it is repricing its device lineup upward in part because of memory costs. A soft PC market would hit Intel's largest segment while leaving the data center business intact.

Supply is the binding constraint in both. Tan said in July that Intel's data center operations could not keep up with orders. That is a better problem than weak demand, but it means revenue growth depends on substrate and wafer availability more than on sales execution, and it puts a ceiling on how far any single quarter can beat.

The mix shift toward data center is favorable for margins. A segment growing at 59% with a 40% operating margin lifts the corporate average each quarter it outgrows the rest.

Balance Sheet and Cash Flow: $29.7 Billion in Cash, $20 Billion-Plus in Capex

Intel ended the second quarter with $29.73 billion in cash and equivalents and $50.54 billion in total debt, a net debt position of $20.56 billion or $3.91 a share. Book equity is $103.14 billion, or $17.36 a share. The current ratio is 1.60 and the quick ratio 1.16. Debt to equity stands at 0.49 and debt to EBITDA at 2.96 times. Interest coverage is 3.88 times.

Those are manageable figures for a company of this size, and they give management time. They are not strong for a business about to spend more than $20 billion a year on capital equipment. Over the trailing twelve months Intel generated $14.94 billion in operating cash flow and spent $12.11 billion on capital expenditures, leaving $2.83 billion of free cash flow, or $0.54 a share. With the 2026 capex budget now above $20 billion, free cash flow goes negative unless operating cash flow rises by more than $5 billion.

Dilution has been part of the funding. Shares outstanding increased 13.28% over the past year to 5.25 billion. For existing holders that is a cost: each share's claim on future earnings is 12% smaller than it was. The share count rose a further 0.41% in the latest quarter. Intel pays no dividend, having suspended it in 2024.

Returns on capital show how far the turnaround has to run. Return on invested capital is 3.42%, return on equity negative 10.72% and return on assets 1.41%. One estimate of the company's weighted average cost of capital is 15.33%, inflated by a beta of 2.23. A business earning 3.42% on capital against a double-digit cost of capital is destroying value on current numbers. The investment case is that 18A and 14A fabs earn far more once filled.

Trailing figures also understate where profitability is heading. Trailing gross margin is 38.87% against 41.8% non-GAAP in the latest quarter. Trailing operating margin is 7.82% against 17.2% non-GAAP in the second quarter. EBITDA over twelve months was $16.84 billion on $57.03 billion of revenue, a 29.5% margin, with depreciation and amortization of $12.38 billion reflecting the size of the installed fab base.

Credit conditions matter more than usual here. With the 10-year Treasury at 5.35% and credit spreads wider, refinancing $50.54 billion of debt and funding new fabs costs more than it did a year ago. An Altman Z-score of 2.23 sits below the 3.0 threshold that marks low distress risk. The Piotroski score of 6 out of 9 points to improving fundamentals.

Institutions hold 67.65% of the shares. Short interest is 158.81 million shares, 3.02% of shares outstanding and 1.74 days of average volume, up from 152.25 million a month earlier. That is a light short position for a stock that has tripled.

Valuation: 67 Times Forward Earnings for a Company Still Posting Losses

Intel's valuation has moved from distressed to demanding in under a year. At $109.51 the stock trades at 67 times forward earnings. On 2027 estimates the multiple was 55 at a $114 share price, which works out to 53 at the current quote. Measured against adjusted earnings over the last four quarters, the ratio is above 100. There is no trailing GAAP multiple because trailing earnings are negative $2.12 a share.

On sales the stock is at 10.1 times trailing revenue and 8.5 times forward revenue. Enterprise value is 10.5 times sales, 35.5 times EBITDA and 134 times operating income. Price to book is 6.5 and price to tangible book 8.9. Price to operating cash flow is 38.6. Price to free cash flow is 204, for a free cash flow yield of 0.49%.

For context on how far the rating has traveled, Intel began 2026 near $37 and traded at $45.07 in January after a poor earnings reaction. At that point it was valued below book. The market value has since risen by more than $380 billion. Over the same period trailing revenue grew to $57.03 billion and the company remained loss-making on a GAAP basis. The rerating is a bet on 2027 and 2028 earnings.

Comparisons within the sector show what kind of bet. Nvidia, with a market value of $5.70 trillion, trades at 30 times trailing earnings. A leading memory producer generating $31.6 billion in annual net income trades at 11.5 times. Intel at 67 times forward is priced richer than both on an earnings basis, with lower margins and a foundry losing $2 billion a quarter. The PEG ratio of 0.97 looks reasonable only because consensus assumes earnings per share grow 94% annually over three years.

The sell-side has not kept up with the price. The average target across 49 analysts is $118.05, which is 7.8% above the current quote, with a range from $80 to $200. One neutral-rated firm lifted its target to $114 from $92 on October 6, and at least one target sits at $150. When a stock trades within 8% of the average target after tripling, the Street is chasing price, and the consensus rating of buy carries less information than usual.

Two valuation frames are in tension. On current earnings and cash flow, Intel is expensive by any standard measure. On replacement value and strategic position, as the only US-headquartered company with leading-edge logic manufacturing at a time of record AI capital spending and domestic supply-chain policy, a premium is defensible. Which frame wins depends on whether the fabs fill.

The practical read is that the price already assumes the turnaround succeeds. Good news supports the level. It takes better-than-expected news to lift it.

Sector and Macro: Chips Are the Market's Rate-Sensitive Leadership

Intel's decline on Thursday was part of a broad move. The S&P 500 was down 0.38% at 7,772.40 and the Nasdaq Composite 0.56% at 27,384.05 by late morning. Nvidia slipped 0.56% to $236.15. Super Micro Computer fell 4.48%, Skyworks Solutions 4.22%, Nokia 4.57% and Applied Optoelectronics 5.22%. Taiwan Semiconductor's US shares lost 1.31%. Intel's 3.19% drop made it one of the weaker large-cap chip stocks, consistent with a beta of 2.23.

The cause was outside the industry. Brent crude rose 4.89% to $105.10 after President Trump said he does not want a deal with Iran. The 10-year Treasury yield touched 5.35%, its highest since 2002, and the 30-year reached 5.70%. Minutes of the Federal Reserve's September meeting showed all 19 officials backed the last rate hike, and some warned that AI investment "could cause aggregate demand to outpace aggregate supply."

High-multiple stocks are the first to reprice when long rates rise, because more of their value sits in distant earnings. A company at 67 times forward earnings is close to the most rate-sensitive profile in large-cap technology. Intel also carries $50.54 billion of debt and a $20 billion capex program, so higher yields raise its funding costs directly.

Industry data this week pointed the other way. Taiwan Semiconductor reported third-quarter revenue of NT$1.49 trillion, up 51% from a year earlier, with September sales up 55%. Samsung Electronics guided to quarterly operating profit of 107.4 trillion won, more than double its total for all of 2025. AMD's chief executive said the company expects chip demand to stay sharply higher over the next few years. Demand is not the issue.

The issue is financing. SpaceX is seeking $40 billion in debt to buy chips. Applied Digital reported revenue nearly triple the consensus on Thursday and its stock fell as overhead costs rose 289%. A London strategist cut a 2027 S&P 500 target to 5,000 on the argument that hyperscaler free cash flow is depleted and borrowing costs are becoming prohibitive. If AI capital spending slows because debt gets too expensive, Intel's data center segment and its prospective foundry customers both feel it.

Policy offers some support. The Defense Department this week committed a conditional $1.5 billion loan to Wolfspeed for domestic chip materials, a sign that federal backing for US semiconductor manufacturing continues. Intel is the largest potential beneficiary of that priority.

Intel is a 5.68% weight in the iShares Semiconductor ETF, so flows in and out of sector funds move the stock mechanically. On days like Thursday, it trades as a sector proxy first.

Competition: TSMC in Foundry, AMD in Processors, Nvidia in AI

Intel competes on three fronts, and its position differs on each. In contract manufacturing the rival is Taiwan Semiconductor, which holds 62% of the global foundry market and builds leading-edge chips for Apple, Nvidia and AMD. TSMC's third-quarter revenue of $46.71 billion was nearly three times Intel's entire quarterly revenue and eight times Intel Foundry's. Its shares hit a record this month.

TSMC's advantage is ecosystem as much as technology. Its process design kits, intellectual property libraries and customer support have been refined across thousands of tape-outs. Intel's foundry is asking chip designers to take a risk on a supplier with a short external track record. The Terafab talks illustrate the dynamic: once TSMC is available, a buyer needs a specific reason to choose the alternative. Intel's reasons are US-based capacity, High-NA EUV in volume ahead of rivals, and packaging. Samsung is the third player at the leading edge and is in the middle of its own profit recovery on memory.

In processors the rival is AMD, which trades at $645.86 and has taken server and desktop share from Intel over most of the past decade. Intel's 59% data center growth suggests it is holding its own in a rising market, helped by supply it controls in its own fabs at a time when TSMC capacity is tight. AMD depends on TSMC for manufacturing. If leading-edge wafers stay constrained, Intel's integrated model is an advantage in availability even where AMD's designs are competitive.

In AI accelerators the rival is Nvidia, valued at $5.70 trillion. Intel has no meaningful share in training GPUs. Its exposure to AI comes through host CPUs, custom chips for cloud providers and, prospectively, foundry and packaging services for other companies' accelerators. Nvidia is therefore both a competitor for data center budgets and the kind of customer Intel Foundry would most like to land.

Memory makers are a fourth group competing for the same investor capital. Micron rose 4.06% to $1,088 on Wednesday, and SK Hynix earns $31.6 billion a year at an 11.5 multiple with a gross margin above 75%. For a portfolio manager choosing among AI hardware names, those companies offer earnings today where Intel offers earnings in 2027.

Intel's differentiator is that it is the only company attempting all of it at once: design, leading-edge manufacturing and packaging, in the United States. That is the source of both the premium and the risk. It needs to execute on products to fund the foundry, and on the foundry to justify the multiple.

Technical Picture: Above the 50-Day at $102.86, Below a String of Lower Highs

The longer-term trend remains up. Intel trades 6.5% above its 50-day moving average at $102.86 and 32% above its 200-day average at $82.71. Both averages are rising, and the 50-day sits well above the 200-day. A stock in that configuration is in a primary uptrend regardless of what one week looks like.

Shorter-term momentum has faded. The 14-day relative strength index is at 48.69, just under the midline and far from the overbought readings it carried in June. The stock is down 6% over five sessions. It has posted lower highs at $142.35 in June, in the low $120s during the recent rally and $115.30 this week.

Support is layered below. The first level is Thursday's low near $109.26, followed by the $110 round number the stock lost during the morning. The 50-day average at $102.86 is the main line. It has contained every pullback since the stock recovered from the July post-earnings drop, and a close below it would be the first since then. Under that, the $100 level carries psychological weight and coincides with the $100.23 pre-earnings close from July 22. The post-earnings low close at $92.32 is the next reference, and the 200-day at $82.71 is the level that defines the long-term trend.

Resistance starts at Wednesday's $113.12 close and the $115.30 high. Above that, the average analyst target at $118.05 and the low $120s, where the last rally stalled, form the next band. A move through $123 would break the sequence of lower highs and put the $142.35 record back in view.

Volume has been heavier on declines than on advances this week, and the stock's average daily turnover of 105 million shares makes it one of the most liquid names in the market. That liquidity cuts both ways around earnings. In the last three reports the stock moved 7.89%, 23.60% and 17.03% the following day.

The gap between price and the 200-day is itself a risk. At 32% above the long-term average, Intel is extended by historical standards, though less than it was in June. Mean reversion toward the 50-day would be routine. Reversion toward the 200-day would be a 24% decline.

For the next three weeks the chart is likely to be range-bound between $102.86 and $115.30, with macro headlines and Terafab news determining which end gets tested before the October 29 report sets the direction.

Verdict: Hold, With $102.86 the Level to Watch and October 29 the Decision Point

Intel at $109.51 is a turnaround that is working, at a price that already assumes it finishes. Revenue grew 25% last quarter, gross margin has expanded 12.9 points in a year, the data center business is growing at 59% and 18A yields are ahead of plan. Against that, the company lost $11.29 billion over twelve months, the foundry loses $2.09 billion a quarter, external foundry revenue is $293 million, and the stock trades at 67 times forward earnings with free cash flow about to turn negative.

The rating is Hold. The longer-term bias is constructive while the stock holds above its 50-day average. The near-term bias into earnings is cautious.

The bull case is that Intel beats again on October 29, shows gross margin above 42%, and gives investors a named 14A commitment or a defined Terafab role. In that outcome the stock clears $115.30, tests the $118.05 consensus target and has room toward $123, a gain of 12%. A stronger signal on foundry customers reopens the path to the $142.35 high, 30% above the current price. The second half of 2026 is when management said 14A decisions would begin, so the catalyst window is open.

The bear case is that the margin stalls at guidance, capital spending rises again, and TSMC formalizes a Terafab role that sidelines 14A. With the 10-year yield at 5.35%, a 67 multiple compresses quickly on disappointment. The first downside target is the 50-day at $102.86, 6% lower. A break there brings $100 and then $92.32, the July post-earnings close, a decline of 16%. The stock fell 7.89% the last time it reported a strong quarter.

For holders with gains, the position is worth keeping, and trimming into strength above $115 ahead of the report reduces event risk at a stretched valuation. For new money, the reward relative to risk is better on a pullback toward $100 to $103 than at the current price, where upside to the average target is 7.8% and downside to the 50-day is 6%. Buying ahead of a binary earnings event at 67 times forward earnings, with the macro tape deteriorating, is a low-odds entry.

Three things would upgrade the view to Buy: a non-GAAP gross margin of 43% or higher, an announced external 14A customer at Intel's own fabs, and a foundry operating loss narrowing below $2 billion. Two things would downgrade it to Sell: a close below $102.86 on rising volume, or a formal Terafab agreement that replaces 14A with a TSMC process.

Until October 29, Intel trades on yields, oil and Musk's next post. After it, the stock

trades on whether the fabs it is spending $20 billion a year to build have customers.

That's TradingNEWS