Intel: Priced Like the Foundry Already Won

The Setup

A quick word about timing, because it matters. Intel reports second-quarter earnings after the market closes today, and the options market is pricing a move of roughly 13.6% in either direction. Publishing an avoid thesis hours before a binary event might look like a prediction about that event. It is not. Whatever tonight's numbers say, they will not answer the question this article is about, because that question is not whether Intel can beat a quarterly estimate. Intel has beaten six in a row. The question is whether the stock's price, after a run of more than 400% in a year, leaves any room for the multi-year foundry bet at its center to merely go well instead of perfectly.

Fifteen months ago, Intel traded at $17.67, its lowest price since 2009, and was widely written off. Today it trades near $105 after touching $135 at the late-June peak, a resurrection that has made it the most celebrated turnaround in technology. That comeback is genuinely real, and the next section gives it full credit. But the same market that left Intel for dead at $17 is now pricing it as though the hardest, most uncertain part of the rebuild, winning the world's biggest chip designers as foundry customers at scale, is already accomplished.

The stock itself has started to show the strain. In early July, Intel fell 21% in a single week, with daily drops near 10%, before bouncing more than 8% this Tuesday as traders positioned for the print. A stock that swings like that around headlines is a stock whose price rests on expectations, not results. This piece is about measuring that gap with the tools we always use: the reported numbers, and the platform's systematic scores.

The Greatest Comeback Story in Tech

The bull case deserves to be stated at full strength, because it is one of the best stories in the market. Since taking over in March 2025, CEO Lip-Bu Tan has done nearly everything a turnaround textbook prescribes. He cut roughly 15,000 positions, restructured the foundry into its own subsidiary, and concentrated engineering on 18A, the most advanced logic process ever built entirely in the United States. Execution followed: Intel has now beaten its own guidance for six consecutive quarters. In the first quarter of 2026, revenue grew about 7% to $13.6 billion, data center and AI revenue jumped 22% to $5.1 billion as CPU demand surged alongside agentic AI workloads, and foundry revenue grew 16% to $5.4 billion.

The strategic validation is even more striking than the numbers. The United States government converted CHIPS Act support into a 9.9% equity stake, an $8.9 billion investment that has roughly quadrupled. Nvidia invested $5 billion. SoftBank added billions more. Apple is reported to be in early discussions about manufacturing some of its processors at Intel as part of a Taiwan-plus-one strategy, Tesla has signed a production agreement, and the Panther Lake AI PC platform spans more than 200 OEM designs. For the first time in a decade, the world's most important chip companies are treating Intel's factories as a serious alternative to TSMC.

If the flywheel completes, the prize is enormous: a second TSMC on American soil, with Intel collecting foundry margins on other companies' chips for decades. That is the story the stock has priced. Now set the story beside the income statement.

What the Turnaround Has Not Turned Around

For all six of those guidance beats, Intel remains a barely profitable company. First-quarter non-GAAP earnings were $0.29 per share. Full-year 2025 revenue of $52.9 billion was roughly flat with the prior year. The consensus for tonight's report is about $0.22 per share on revenue near $14.4 billion. Annualize the most optimistic of those quarterly figures and Intel earns on the order of a dollar per share, against a $105 stock. However the turnaround is going strategically, the earnings that would justify the price do not exist yet.

The reason sits exactly where the bull case lives: the foundry. Intel's foundry segment, the second-TSMC engine at the heart of the valuation, still runs an operating loss of roughly $2.5 billion per quarter. The revenue it reports is overwhelmingly Intel manufacturing chips for itself. The margin story that transforms the company requires external customers at volume, and the losses continue until they arrive. Six quarters of beats have been six quarters of beating expectations about a business that is still, on the whole, close to breakeven while its centerpiece burns billions.

None of this is a scandal. It is what a capital-intensive turnaround looks like in the middle innings, and Tan has never claimed otherwise. The scandal-free version of the problem is simply arithmetic: the market is paying north of 100 times any earnings Intel currently produces, and even the bullish path that resets forward earnings toward $1.50 per share needs a triple-digit multiple to justify targets near $150. The turnaround is real. The earnings are not here.

The One Question the Price Has Answered

Strip everything else away and Intel's valuation reduces to a single question: does a marquee fabless chip designer commit high-volume production to Intel's 18A or 14A process? Not explore it, not test it, commit to it. The analyst community has effectively admitted this is the whole ballgame. Price targets on Intel currently span from $45 to $200, one of the widest dispersions on any large-cap stock, with a consensus rating of Hold. A spread that wide is not disagreement about margins or units. It is the market assigning odds to one binary event.

Here is the current state of that event. Apple's discussions are described as early-stage. Nvidia has run multi-project wafer tests on 18A without committing volume. Microsoft's earlier 18A commitment is real but modest relative to the scale the valuation needs. The customer flywheel the price assumes is, as of today, a collection of investments, pilots, conversations, and one automaker. Meaningful, encouraging, and not yet the thing itself.

The stock, however, sits near the top of that target range, not the middle. At $105, against a bear case that credible analysts put near $90 if foundry traction stalls and a bull case near $150 if it lands, the price has already answered the question the customers have not. That is the specific, narrow thing this thesis objects to. It is not that the foundry bet is bad. It is that the buyer at today's price receives almost none of the reward if it succeeds and carries most of the loss if it stalls.

The Government Put

The strongest-sounding argument for paying any price is the shareholder register itself: the United States government owns 9.9% of the company. Washington has decided that a leading-edge American fab is strategically essential, and one commentator captured the market's conclusion precisely: the rally reflects the government deciding Intel cannot be allowed to fail, and the market deciding to believe it.

Both halves of that are probably true, and neither one supports the current price. A government backstop protects Intel the company: its fabs, its employment, its national-security role. It does not protect the equity multiple. A world where Intel survives, keeps building, and muddles through with a slow foundry ramp is entirely consistent with the stock trading at half its current price. History offers a long list of strategically essential, government-entangled enterprises whose national importance never translated into shareholder returns. The July air pocket, 21% in a week on foundry doubts, was a live demonstration that the government put has a strike price far below where the stock trades.

What the Wealth Engine Scores Say

Before the valuation verdict, here is what the Wealth Engine Pro platform's systematic scoring shows for Intel right now.

The Company Strength score of 32 is the lowest of any stock this column has examined in months, and the components explain why: Financial Health of 44 reflects the thin profitability and heavy capital burden, and the Moat score of 3 out of 15 is the systematic model's blunt assessment of a company that lost its process lead for a decade and is spending billions to win it back. The fair value of $31.02 is strict and backward-looking, and it is fair to note that it cannot see the strategic assets, the government stake, or the 18A pipeline. Do not anchor on the precise figure. Anchor on what it measures: the business Intel is today, on its reported numbers, supports a price nowhere near $105. The rest is paid for the future.

These scores are systematic. They evaluate companies based on reported financials, balance sheet quality, moat characteristics, and valuation models. They measure what a company is today, not what it might become, and they extend no credit for foundry customers who have not signed.

In this case, the editorial thesis and the platform scores point in the same direction. The article argues the price has answered a question the customers have not, and the systematic data, reading the weakest strength score in months against an Expensive, Bearish valuation, lands in the same place from a different angle. When both the quantitative model and the qualitative analysis flag the same gap between price and substance, that convergence is worth paying attention to. Research Intel yourself on the platform and decide which signal matters more for your situation.

The Valuation Verdict

The arithmetic is simple and severe. At $105, Intel trades at more than 100 times the earnings it currently generates, roughly 3.4 times the platform's calculated fair value, on a business whose centerpiece loses $2.5 billion a quarter. The bull case that gets to $150 requires forward earnings to reset toward $1.50 per share and the market to pay a hundred times that. The bear case, if foundry traction stalls, sits near $90 by the bulls' own math, and the platform's systematic estimate sits far below that.

Now add tonight. Options price a 13.6% move on the report, and the stock has already shown it can shed a fifth of its value in a week when foundry sentiment wobbles. A beat tonight, the seventh in a row, would confirm that the turnaround is on track. It would not change the valuation problem, because the valuation problem is not about any quarter. It is that a stock up more than 400% in a year has pre-paid for a customer flywheel that remains, in the filings, pilots and conversations.

The buyer at $105 is not buying the comeback. The comeback already happened, and it was magnificent. The buyer at $105 is buying the next act, the external foundry at scale, at a price that assumes it arrives on schedule, from customers who have tested the machines and not yet placed the order. That is not a position in a turnaround. It is a prepayment for one that has not finished turning.

What Could Go Wrong

This steelman is as strong as any this column has written, because Intel has spent six quarters making skeptics look foolish.

The execution streak may simply continue. Six consecutive beats under Tan is not luck. It is a management team that has consistently under-promised and over-delivered, and tonight could make it seven. Momentum in execution tends to persist, and each beat pulls the earnings reset closer.

One signature could rewrite everything. The target dispersion cuts both ways. If Apple converts its discussions into a volume commitment, or Nvidia moves from wafer tests to production, the stock likely gaps toward the high end of the range overnight, and this thesis is run over in an afternoon. The probability of that event is genuinely unknowable, which is precisely why the market cannot agree on a price.

The strategic tailwinds are real and rare. A government equity partner, the only leading-edge fabs on American soil, a Taiwan-plus-one imperative reshaping every large chip buyer's supply chain, and a CPU demand cycle turning back in Intel's favor as agentic AI workloads spread. Few companies in history have had this much structural wind at their back.

CPU strength is underrated. Data center and AI revenue grew 22% last quarter, and Panther Lake spans more than 200 designs. If the core product business keeps re-accelerating, it can carry more of the valuation while the foundry matures, softening the binary.

The thesis breaks if a marquee customer commits, the foundry losses narrow on schedule, and earnings reset toward the bull case over the next several quarters. In that world Intel grows into the price and this caution ages badly. What the avoid thesis argues is not that this is impossible. It is that at $105 it is already assumed, which is the same asymmetry that has run through this series from Who Is Winning the AI Race to the Nvidia versus AMD battle: a real prize, fully priced before it is won.

The Bottom Line

Intel's comeback is real, and it deserves the admiration it is getting. A company left for dead at $17.67 rebuilt its process technology, recruited the government, Nvidia, and SoftBank as shareholders, and put itself back at the center of the most important industry in the world. Nothing in this thesis disputes any of that. The dispute is about what a buyer receives in exchange for $105 a share.

What the buyer receives today is a company earning on the order of a dollar per share, whose foundry loses $2.5 billion a quarter, whose decisive customers are still in the testing and talking stage, and whose own analyst community cannot narrow the fair price to anything tighter than $45 to $200. The platform's systematic scoring reads Weak, Expensive, and Bearish, with the lowest strength score this column has flagged in months. Tonight's report, whichever way the 13.6% swing breaks, will not resolve the question the price has already answered on the customers' behalf.

That is the entire philosophy here. Wealth Engine Pro evaluates companies on what they are and what they earn, not on the most thrilling version of the next act. The narrative says Intel is the second TSMC in the making. The data says it is a barely profitable turnaround, magnificently executed so far, priced as though the hardest part is already done. The comeback happened. The coronation has not. When a price pays for both, the data is the part worth trusting.