UBS upgraded its rating to “buy”! Computing power hardware company Celestica (CLS.US) has received many Wall Street group songs, and the compound EPS growth rate is expected to increase by 50% in the next two years

The Zhitong Finance App learned that due to the acceleration of global artificial intelligence infrastructure construction, computing power and data center hardware manufacturing giant Celestica (CLS.US) has recently once again become the focus of major Wall Street banks. As hyperscale customers accelerate data center upgrades to higher throughput architectures, the bank's profit outlook for this key “water seller” has increased dramatically.

UBS analyst David Vogt directly raised Celestica's stock rating from “neutral” to “buy” and raised the target price from $410 to $430 in the latest research report.

UBS pointed out that there is a strong demand for Ethernet switches and AI/ML computing hardware in the market. More importantly, Celestica's customized 1.6T rack-level solution for leading AI giants such as OpenAI will be the core engine driving the company's medium- to long-term revenue and profit growth. UBS conservatively predicts that under strong order guarantees, the company's compound annual earnings per share (EPS) growth rate (CAGR) for the 2026-2028 calendar year (CY26-CY28) will reach nearly 50%, and EPS is expected to rise from $10.83 in CY26 to $24.72 in CY28.

Investment banks sing intensively within a month

In addition to UBS, many major Wall Street banks have also recently released research reports, showing Celestica an overall pattern of “optimistic about long-term growth and keeping an eye on short-term capital operations.” According to PriceTargets data, 20 of the 23 covered institutions currently offer “buy”, 1 “strong buy”, 2 “hold”, and none sell; the average target price is 438.86 US dollars, the highest forecast is 510 US dollars, the minimum is 300 US dollars, and the consensus rating is “buy”.

Institutions such as J.P. Morgan Chase and Wolfe Research maintain high target prices ($475 and $485), stressing that cloud vendor capital expenditure on 800G and 1.6T network infrastructure exceeds expectations, and Celestica directly enjoys industry dividends as a core manufacturing partner.

Barclays analyst Tim Long maintained an “gain” rating, but lowered the target price slightly from $430 to $406 in early August. The agency believes that this fine adjustment mainly reflects the market's rational assessment of the chip supply chain pace and short-term industry expenditure digestion period in January, and does not affect its fundamental logic.

Look at the core of the consensus: AI infrastructure order fulfillment

The underlying logic that Wall Street collectively favors is highly convergent — the multi-year explosion of AI data center capital expenditure is being transformed into CLS's substantial orders and profit improvements:

The first is the amount of high-speed network hardware. Demand for Ethernet switches from hyperscale cloud vendors such as Meta and Amazon continues to be strong, and CLS directly benefits as a core hardware agent.

The second is a breakthrough in AI-specific computing hardware. The 1.6t rack-scale solution supplied by the company for OpenAI is seen as a key driver for the accelerated growth of EPS after 2027. UBS expects revenue to grow at a compound annual rate of about 39% until 2029, and “there is still room for growth.”

Third, financial flexibility has increased. TD Cowen pointed out that after CLS recently completed equity financing, its balance sheet was significantly strengthened to prepare for the challenging supply chain environment in FY2027. Management had enough room to deal with potential disruptions and compete effectively with well-funded original design manufacturers (ODMs) in Asia.

Celestica completed a $3 billion common stock issuance plan in early August to expand global production capacity. Underwriting teams such as Citibank and TD Securities analyzed that the increase in issuance brought about equity dilution and stock price shocks in the short term, but the combination of self-financing and equity financing could lock in long-term orders from large-scale customers for the next few years and accelerate the market's revaluation from traditional OEM (EMS) to a high-barrier “AI network and computing power solution provider.”