Intel Is Down 32% From Its Peak — What the Analyst's $92 Target Actually Says

Intel was 2026's comeback story until it wasn't. The stock ran more than 400% in a year to a high near $141 on June 22, then slid about 32% through the summer. On September 4, Mizuho's Vijay Rakesh — a five-star-rated analyst — cut his price target to $92 from $109 while keeping a Hold rating. The headline makes it look like one more bearish call piling on. It is really the opposite: an analyst looking at the current price and saying the turnaround is now priced in. That gap between the story and the numbers is the whole investment case in a single move.

So before you read "down 32%" as a sale, ask what, exactly, the market is discounting — and whether the loudest version of that story has already been paid for.

What actually broke

The pullback was not an Intel-specific failure. It started in early July, when Bank of America and Morgan Stanley warned that AI semiconductor valuations had run ahead of near-term demand. That touched off a sector-wide rotation: the chip index fell more than 20% and entered bear-market territory even as the underlying demand story stayed intact. Intel, as the most shockingly-revalued name in the group after a 400% run, gave back the most. That is the signature of a sentiment correction in a liquidity- and narrative-driven market — the macro wind that lifted everything turned, and the highest-beta names bore the brunt. Nothing about the company's operating momentum changed because of the analyst note.

In fact the quarterly numbers, on their face, look strong. Intel reported $16.1 billion of second-quarter revenue, up 25% year over year — its fastest growth in roughly fifteen years — with non-GAAP earnings of $0.42 a share, more than double the $0.20 it had guided. The 18A manufacturing process that the whole turnaround rests on is reportedly ahead of internal production targets, up more than 50% from the first quarter.

What the $92 target says

Now the rub. Rakesh did not cut because he thinks Intel will collapse, or even because he thinks it is expensive in the abstract. He cut to $92 while calling the stock "fairly valued" at current levels, meaning shares were trading right around where he thinks they should be. A Hold at $92 — with the stock at roughly $95 — is an analyst saying "the cheap-comeback phase is over; the market and I now agree on what this is worth."

Look at the consensus and the same conclusion stares back. The average Wall Street target sits near $99, marginal upside from the current price, with ten Buys, eighteen Holds, and two Sells across thirty covering analysts. When a fast mover's analyst community clusters tightly around the price rather than far above it, that is the market describing equilibrium, not upside. The re-rating already happened.

The economics underneath

Here is where the Persona's discipline — check the data, not the narrative — matters most. The market is now paying roughly 60 times next year's expected earnings. That multiple only makes sense if you believe the foundry transformation has already succeeded. The economics are not there yet.

Intel Foundry grew 31% in the quarter to $5.8 billion of revenue, but it still lost $2.1 billion. Those losses are narrowing — down from $3.2 billion a year earlier, with the loss per dollar of revenue cut roughly in half — but the deeper problem is who actually buys from it. Only about $293 million of that $5.8 billion, roughly 5%, came from external customers; some 95% came from Intel's own product groups. Management has it boxing above its weight on external revenue of only about $1.2 billion a year, targeting foundry profitability only in 2027 and needing "a few billion dollars" more in outside orders to get there.

The dilution adds its own drag. In August, Intel launched a $15 billion equity offering to fund its factory build-out, upsized to $20 billion as the shares kept climbing. At roughly $100 a share the original deal alone would have added about 150 million shares, around 3% dilution. That is the capital the foundry ambition requires — equity, not debt — and it is the price existing holders pay so the turnaround can be funded.

Put it together and you get the clean version of what happened: the liquidity cycle and the exponential-age AI thesis re-rated Intel's stock from a broken chipmaker to a foundry champion in a matter of months, and the market — via the analyst cut — is now telling you the re-rating is complete. You are no longer buying a transformation at a discount. You are paying full price for a transformation that is only about halfway to proof, with the external-customer revenue and the 2027 break-even still ahead of it.

None of this makes Intel a bad company or a falling knife. It makes it a cyclical question, and the right question is the honest one: at 60 times forward earnings, priced for a completed turnaround, is the risk-reward the same as it was a year ago when nobody believed? The analyst's $92 says the rerating happened. The remaining question — whether execution catches up to the price — is now the only one that matters, and it is measured in quarters, not headline percentages.