CLS surged 12%: Is the 40 billion target really a dream?

NEXT PICK · Market Insights

CLS surged 12%: Is the 40 billion target really a dream?

Tuesday, October 27, 2026

On October 27, Celestica ($CLS) closed up 12.12% at $276.54, with trading volume expanding to 3.1 times the daily average over the past three months, and its market value returning to $34.87 billion. Before the earnings report, this AI hardware star had already lost more than 30% of its stock in October.

What ignited the market wasn't rumors, but the Q3 earnings released after the market closed the previous night: quarterly revenue of $5.83 billion, up 83% year-over-year, an upward revision of full-year guidance to $21 billion, and for the first time the $40 billion revenue target for 2027, far exceeding Wall Street's consensus of about $35.7 billion.

But problems followed

What makes a manufacturing service provider with a profit margin below 9% convince the market that it can double its revenue within a year? Is today's big bullish candlestick the starting point of value rebound, or is it a rebound in sentiment after an oversold price?

Latest price $276.54 ▲ +12.12%

Q3 revenue $5.83 billion ▲ +83%

2027 target 40 billion USD ▲ +90%

Not just a foundry, but the "final assembly workshop" of an AI data center

Celestica's business is divided into two main segments: Connectivity and Cloud Solutions (CCS) for data centers and networks, and Advanced Technology Solutions (ATS) covering aerospace and defense, industry, medical technology, and capital equipment.

In Q3, the weight of the two was worlds apart: CCS revenue was $4.87 billion, a year-over-year surge of 102%, while ATS revenue was $960 million, up 23%. In other words, today's CLS is essentially a hyperscaler and a capital expenditure stock for cutting-edge AI labs.

Its chain is much longer than that of typical contract manufacturers

From design and development, engineering services, and component procurement, to manufacturing, assembly, testing, system integration, precision machining, logistics, and after-sales maintenance and IT asset disposal, we cover almost the entire lifecycle of a single piece of hardware—from blueprints to retirement.

The reason customers find it difficult to replace them easily is that AI servers and high-speed switches require lengthy design implementation and quality certification. Switching suppliers means re-certification, lost yield, and delivery delays—this stickiness is more valuable than any single order.

The most recognizable chip is its collaboration with AMD: in March this year, both parties announced a joint advancement of the Helios rack-level AI platform, with CLS responsible for the R&D, design, and manufacturing of scale-up network switches for connecting AMD's next-generation Instinct MI450 GPUs, with the platform planned for customer launch by the end of 2026.

If we compare AI data centers to a city rapidly expanding, NVIDIA and AMD design engines, while CLS is more like the general contractor that installs engines, pipelines, and circuits into buildings and handles acceptance. The report gives a non-fungible score of 7/10: Authentic, verified by revenue, but not patent monopoly.

24x forward P/E ratio: expensive or not?

Based on closing prices, CLS's rolling P/E ratio is 28.72 times, corresponding to a 12-month EPS of $9.63; If converted to the company's newly raised 2026 adjusted EPS guidance of $11.55, the forward P/E is about 23.9 times, while the 2027 target of $21.50 is only about 12.9 times.

The valuation conclusion of the research report is "reasonable"

With a 2026 guidance of 22 to 28 times, the current stock price is right in the middle range; The price-to-sales ratio (P/S) is less than 2 times, PEG is only 0.30, which looks cheap, but the enterprise value multiple (EV/EBITDA) is about 20 times, and the price-to-book ratio (P/B) is 5.61 times, so the absolute valuation is not low.

Today's gains reflect the "credible 2026 guidance," not the "40 billion will definitely be achieved in 2027." If the 2027 target is truly achieved, even if only 18 to 22 times, valuation space will clearly open up; If the target is postponed, the PEG logic supporting valuation will lose effect overnight.

Cash flow is the most important aspect of this financial report to be viewed with calm. According to the company's official standards, Q3 free cash flow was only $127.5 million, about 2.2% of revenue. The full-year free cash flow guidance was instead revised down from $600 million to $460 million, citing early allocation of working capital for 2027 growth.

More importantly, the company completed a $3.45 billion share issuance in August, issuing 11.1 million shares, and plans to reach $1.5 billion to $2 billion in capital expenditure in 2027. The report classifies the cash situation as "self-fundraising," but judging from the pace of the issuance and expansion, this growth machine is fueled by at least some new money from shareholders.

With an annual free cash flow of $460 million compared to a market cap of $34.87 billion, the free cash flow yield is only about 1.3%, with almost no cash flow discount protection. So, are investors buying cash now, or profits two years from now? The answer is obviously the latter.

Revenue has broken through, but the stock price is still below the moving average

The report defines CLS as the "eve of a breakthrough" (pre_breakout): revenue, operating profit, and net profit have all been solidly a surge in volume, with a 69% growth target for 2026 and a target of 90% growth in 2027. The Helios partnership provides a clear product focus, no longer just a concept story.

However, the conditions for confirming a breakout are not yet in place. Even with today's sharp rise, the $276.54 stock price is still about 15.5% below the 50-day moving average of $327.15 and about 16.9% below the 200-day moving average of $332.92. Fundamentals are accelerating, but technicals have yet to return to an upward channel.

The confirmation signals listed in the report are very specific

Quarterly revenue progressed according to guidance, operating margin remained stable at 8% to 9%, cash flow conversion improved, and Helios transitioned from collaborative development to identifiable revenue. Today is just the first item to get the answer.

In the gaps between giants, victory depends on engineering depth rather than scale

CLS's comparable competitors include Amkor Technology, Flex, Foxconn, and Quanta, but they are not simply in the same track: Amkor is deeply engaged in advanced semiconductor packaging, and CLS is more complementary than alternative in the AI server supply chain.

Flex's scale and multi-industry coverage are closest to CLS's, but its growth elasticity is more dispersed among consumer electronics, automotive, and industrials; CLS focuses its chips on AI infrastructure, networks, and hyperscalers, which allows it to run faster during boom cycles and fall harder when cycles reverse.

Foxconn and Quanta have stronger large-scale procurement and deeper vertical manufacturing capabilities. CLS cannot beat these giants with low costs; its success lies in engineering capabilities in high-speed network switches, rack-level system integration, and joint development relationships with chip giants like AMD.

The shortcomings are equally clear

Research reports point out that Q3 operating profit grew by only about 41.5%, far behind revenue growth; According to institutional statistics, its top three clients contributed about two-thirds of sales, and if these clients adjust their procurement pace, their bargaining power will be tested.

The report also mentioned rumors that Vicor had filed a patent complaint regarding vertical power supply technology for data centers, but currently lacking USITC documents and company announcements for cross-corroborating, it can only be used as a risk variable to track. This moat is real, but it resembles a canal that needs constant dredging rather than a natural barrier.

$40 billion—is it the ceiling or the new floor?

The report admits that authoritative market size data for AI infrastructure, rack-level platforms, and 1.6T networks is insufficient, so a more reliable anchor is the company itself: $21 billion in 2026, $40 billion in 2027, with $19 billion in revenue added over two years. This alone indicates that the serviceable market is rapidly being expanded by hyperscaler capital expenditures.

As for 2028, the report only does stress testing: if growth drops to 10%, 25%, or 40%, revenue will be around $44 billion, $50 billion, and $56 billion, respectively. This is not company guidance, but rather a reminder to readers that whether valuations can hold firm depends on how long the peak of AI capital spending can last.

Growth has moved from the story to the financial report

The strongest argument for the bulls is that these numbers have already been written in the financial reports and guidance:

▲ Bull Case

Bears do not deny growth, only question the quality of growth

The bears' logic is not to be pessimistic about AI, but rather to worry that the price the market pays for growth is too high:

▼ Bear Case

After the financial report is released, what should we look for next?

Today's surge has already digested the financial reports and 2027 targets—the biggest catalysts. Next, the influence of stock price will shift to the following factors:

•	October 27 Investor and Analyst Day: Management's detailed explanation of the 2027 orders, capacity, and profit margin path will determine whether the rebound can continue. The view is positive but still needs to be verified
•	Q4 2026 Delivery: Company guidance revenue is $6.275 billion to $6.575 billion, adjusted EPS $3.37 to $3.57; whether it can reach the upper bound is a short-term key factor affecting both directions
•	AMD Helios platform to launch to customers by the end of 2026: whether the first mass-produced switches can be converted into identifiable revenue is a mid-term positive
•	Introduction of optical communication and high-speed network products: CPO switches, Tomahawk 6, 102.4T, and 1.6T platform progress, catalysts for industry sentiment, with a neutral to positive influence
•	Industry conferences like DCD Connect MENA: boosting AI infrastructure topics but making it difficult to quantify order contributions, resulting in limited impact

While the fundamentals kept rising, the stock price had a rollercoaster ride

Over the past two years, CLS's report card has been nearly impeccable: 2024 revenue of $9.65 billion, net profit of $428 million, EPS of $3.61; 2025 revenue to rise to $12.39 billion, net profit of $832.5 million, and EPS of $716, nearly doubling.

Entering 2026, quarterly revenue climbed from $4.05 billion in Q1 to $4.7 billion in Q2 and $5.83 billion in Q3, with cumulative revenue for the first three quarters reaching $14.58 billion, a year-on-year increase of about 67%. However, the stock price did not keep pace: after the additional share issue in August, there was a noticeable pullback, and in October, it fell more than 30% before the earnings report.

The conclusions of the research report review are very straightforward

The market first accepted that "growth outpaces traditional manufacturers," but most easily misjudged that "high revenue growth automatically leads to profit expansion," and "low PEG equals cheap." Today's rebound shows the market is willing to retrust the former, but remains cautious about the latter two.

⚠️ Risk Notice • Doubling revenue in 2027 combined with a 9% profit margin target is extremely difficult to execute; if delayed, valuations will be quickly compressed • With downward revisions to free cash flow guidance and rising capital expenditures, growth may become more dependent on external financing • The pace of key component supply and capacity ramp-up is uncertain, which may limit deliveries • Clients are highly concentrated among a few hyperscalers, and a shift in capital expenditure from major clients directly impacts revenue • Tariffs, export controls, and unverified patent disputes pose potential compliance and supply chain disruptions

🟡 Neutral Guidance fulfillment has been factored into the stock price, but the 2027 target and cash flow conversion remain to be verified

💬 Discussion Bottom line: 2026 is stable, 2027 is not yet proven, so it's not advisable to chase highs. CLS real-time buy and sell levels, post CLS on the official account for viewing.

Data source

•	Source: NextPick real-time market snapshot + Celestica's Q3 2026 earnings report and SEC announcement + AMD's official press release + mainstream financial media coverage + NextPick research team report.