TSEM rose 11% in a single day: Japan made a stakes of 7 billion yen

NEXT PICK · Market Insights

TSEM rose 11% in a single day: Japan made a stakes of 7 billion yen

Tuesday, July 14, 2026

An old foundry that started operations in a mountain city in Israel in 1993 and has been doing "special craftsmanship that others refuse to do" for over thirty years, today surged 11.24% in a single day on the Nasdaq, closing at $255.49, with its market value soaring to $28.88 billion.

Those who bought it a year ago now have nearly sixfold returns on the books; What ignited today's big bullish trend wasn't a financial report, but a subsidy announcement from the Japanese government—Japan invested $1 billion to help push its capacity expansion plan in Japan to the forefront of AI optical interconnection.

So here's the question

When a mature established company is repriced by the market at a 118x P/E ratio, is this a scarce chip for AI infrastructure, or is it once again being pushed to the edge of a cliff by narrative?

Latest price $255.49 ▲ +11.24%

Market capitalization 28.88 billion USD ▲ —

December return +471.6% ▲ —

Why should AI compete to buy OEM factories that don't chase the most advanced processes?

TSEM's positioning is precisely where it is most easily misunderstood. It does not compete closely with TSMC at advanced logic nodes in 3nm and 5nm, but instead focuses on differentiated specialty processes.

Its craft platform is like a palette of multiple flavors

Silicon germanium (SiGe), silicon photonics (SiPho), RF CMOS, CMOS image sensors, mixed-signal, integrated power management—nearly ten platforms are being deployed at once. Such breadth is extremely rare among pure foundries—most peers bet only on two or three.

Our clients range from automotive, communications, and consumer electronics to healthcare and aerospace. You can think of it as a "precision customization workshop" in the semiconductor world: not producing the largest standard parts, only making those high value-added parts that would have to be scrapped and rebuilt at a new factory.

What truly excites today's market is silicon photonics. It is the key process for stuffing optical signals into chips and transporting massive amounts of data for AI data centers—as NVIDIA's GPU clusters grow larger, these optical interconnect chips are feeding them.

A price-to-earnings ratio of 118 times is shocking, but the 2028 accounts will be calculated separately

If you only look at static valuations, TSEM at $255.49 looks truly impressive: a price-to-earnings ratio as high as 118 times, with the stock price more than 60% premium above the 200-day moving average ($156.8). Such premiums have rarely lasted long in history.

But the story doesn't end with this snapshot. It is not a concept company that burns money to tell stories—its most recent quarter revenue was $413.6 million, up 15% year-over-year, net profit was $65 million, a 62% increase year-over-year, and next quarter it is set a record target of about $455 million. This is a company that makes real money and accelerates its earnings.

The more critical anchor lies in the future

The company has raised its 2028 targets to $3.6 billion in revenue and $1.2 billion in net profit. Based on current market value, the corresponding forward P/E ratio is only about 24 times.

So should the market discount it with 118x in the present, or at 24x in the future? The answer depends on one thing—whether the $1.2 billion net profit can be realized by 2028.

The market has already paid for the old factory's "second youth."

It's not contradictory to categorize TSEM as "the eve of a breakout": its underlying business has long been mature, but the market is repricing it with a completely new logic.

This round of revaluation is not about patching old businesses, but about the release of silicon photonics and advanced packaging capacity in 2027-2028, as well as the anticipated explosive demand for optical interconnects in AI data centers. The 50-day moving average ($250.88) has already climbed above the 200-day moving average ($156.8), technically confirming the strength of this trend.

In other words, the market has already bet early on its identity shift from "veteran of special contract manufacturing to core supplier of AI optical interconnects." Today's subsidy announcement pushes that expectation forward once again.

The small but beautiful moat is the phrase "can't afford to trade."

In the pure contract manufacturing landscape, TSEM is a typical example of being "small but beautiful." TSMC's market value exceeds one trillion yuan, holding about 60% of the market share. GlobalFoundries and UMC are both in the mature/specialty process layer. Its market value is less than 29 billion yuan, far smaller than the leading companies.

But its moat isn't about scale, it's about customers who can't afford to swap. Once a chip is selected for a foundry's process design kit, the layout, yield curve, and automotive-grade reliability certification are all deeply tied to that factory. Switching factories often means starting over and taking over a year. This switching cost is the strongest moat for special contract manufacturing.

TSMC and GFS have also been catching up on silicon photonics in recent years, and competition is indeed closing in. But in the depths of special processes like SiGe and RF, the leading company is not its direct rival. Rather than saying TSEM is competing with giants for the pie, it's more accurate to say it holds onto a niche market that giants are too lazy to bend down to pick up, and others can't even make it in.

The ceiling is high, but the market share must be gradually increased through production capacity

TSEM's target market for mature/specialty process foundries is the roughly $45 billion market left after TSMC takes the lion, and it is also the main battlefield it can directly serve.

The real imagination space lies in silicon photonics

Driven by AI clusters of 800G, 1.6T optical modules, and co-packaged optics, this segment is widely expected to see rapid growth in the coming years. TSEM already holds a $1.3 billion long-term silicon photonics contract (including $290 million in customer upfront payments) confirmed for 2027, and Japan's newly added 300mm capacity is a key chip in fulfilling its 2028 revenue target of $3.6 billion. The ceiling is not low, but every inch of market share must be gradually built up through capacity ramp-ups.

Bulls' confidence: contracts, subsidies, and volume growth have all been implemented

Unlike the initial judgment of the research report, what now supports the bull market is no longer just rumors on social media, but a series of hard facts confirmed by official announcements and financial reports.

▲ Bull Case

Bears' alarm: The good news may already be in the price

The stronger the bullish logic, the more the bears need to ask a calm question: Have these positive factors already been completely wiped out by this 471.6% annual gain?

▼ Bear Case

In the coming year, keep a close eye on these key fulfillment milestones

Today's subsidy announcement is just the starting gun; what truly determines the stock price rhythm next are the key moments when the stock price moves from "paper promises" to "actual fulfillment":

•	Japan's $1 billion subsidy fund officially allocated for the start of 300mm production lines: Q4 2026, a positive direction
•	Fab 6 (Arai) retrofit and Fab 7 (Uozu) capacity ramp-up progress: 2027, favorable direction
•	Silicon photonics capacity is expected to enter mass production readiness in Q4 2027; by the end of 2027, this is a favorable direction
•	The next quarterly financial report will verify revenue and photonic business growth: As earnings season approaches, the direction depends on whether targets are met
•	The phased fulfillment of the 2028 target of 3.6 billion in revenue / 1.2 billion in net profit: 2027-2028 will determine the success or failure of the valuation

From $150 to nearly $320, this revaluation hit the drum beats

Looking back over the past 12 months, TSEM has charted an almost steep curve, with a cumulative return of 471.6%, and the stock price once surged to an all-time high of $316.85 on June 22.

The rhythm is clear

On June 18, it co-shipped over 5 million coherent photonic chips with Marvell, solidifying its identity as an "AI optical interconnect supplier"; Today, Japan's subsidies and the 2028 target increase have added fuel to the fire. The 30-day return is only 1.2%, and the current price has pulled back from its high, indicating clear signs of high-level consolidation and a tug-of-war between bulls and bears—the market is digesting chips for this surge.

⚠️ Risk Notice • High valuations: 118x P/E ratio, over 60% premium above the 200-day moving average, significant drawdown potential • Execution risk: $3 billion expansion cycle is long, with uncertainty in both equipment and yield • Fulfilling the gap: capacity will be ready by the end of 2027, and target validation will require a long wait • Intensified competition: TSMC, GFS, and Intel all ramped up their investments in silicon photonics foundry • High-level turnover: Having retreated from the June all-time high, profit-taking pressure persists

🟡 Neutral The positive factors are clear but have already been significantly priced in; upward growth depends on the 2028 target for a relay.

💬 Discussion Judgment: The track is genuine, the catalyst is genuine, but the current price is fully fueling optimistic expectations. Its current real-time buy and sell levels and signals can be viewed via TSEM on the official WeChat account.

Data source

•	Source: NextPick real-time snapshot + Tower Semiconductor official announcement and Q1 2026 earnings report + SEC 6-K filing + mainstream financial media.