CLS · Celestica, Inc. · Stock research

Current Thesis

Celestica is the mid-cap electronics-manufacturing/ODM partner that assembles the racks, 800G/1.6T switches, and optics hyperscalers order by the data-center row, and the fundamental line is still climbing three months past its Q1 crush — but the tape spent June unwinding its first real momentum reset of the cycle and now sits one session from a binary Q2 print. The 2026-04-27 Q1 result put revenue at $4.05B (+53% YoY) and adjusted EPS at $2.16 (from $1.20), and management lifted FY2026 guidance in a single move to $19B revenue (from $17B) and $10.15 EPS (from $8.75). Price ran to a $474 high on 2026-06-02, then dropped 12.6% on 2026-06-05 to $371.71 — about -21% off the high in three sessions — slicing the $410/$390 support shelf. Into the 2026-07-27 print, consensus EPS $2.10 sits below the company's own $2.14–$2.34 guide, so a number that merely matches the Street is a guide miss. The structure is mid-repair and the event is binary; the cleaner entry is the post-print reaction, not a swing taken the day before.

Bullish and bearish views on Celestica, Inc.

The model's bull view on Celestica, Inc. (CLS), in brief: Q1 2026 (2026-04-27) was a beat-and-raise of size: revenue $4.05B (+53% YoY), adjusted EPS $2.16 vs $1.20 a year earlier, with FY2026 guidance lifted to $19B revenue (from $17B) and $10.15 EPS (from $8.75). The bear view: -12.6% on 2026-06-05 to $371.71, about -21% below the $474 high (2026-06-02) in three sessions: the first distribution event of the cycle, breaking the $410.93/$390.11 shelf. Both cases follow in full.

Bull Case

• Q1 2026 (2026-04-27) was a beat-and-raise of size: revenue $4.05B (+53% YoY), adjusted EPS $2.16 vs $1.20 a year earlier, with FY2026 guidance lifted to $19B revenue (from $17B) and $10.15 EPS (from $8.75). Raises of that magnitude rarely land at a fundamental peak.
• CCS segment +76% YoY to $3.24B at 8.6% margin (Q1 2026, up from 8.0%): the hyperscaler-facing line compounds while margin expands, cutting against the usual "EMS adds revenue at no incremental margin" objection.
• HPS (Hardware Platform Solutions, the rack-scale line inside CCS) ~$1.7B, +63% YoY (Q1 2026): the custom-integration dollars most levered to a single hyperscaler build cycle are still growing off a large base.
• Alphabet's ~$80B capex raise (reported June 2026) maps to roughly $10–16B of incremental networking opportunity — the demand pool underwriting the ~70% CY26 CCS growth bulls model.
• AMD Helios AI-rack design win (cited by BofA at its $430 target) widens the customer set beyond Broadcom/Google ASIC-fabric work into the AMD GPU-rack ecosystem.
• Analyst targets rose only after the print landed: CIBC $480 (from $425), BMO $450 (from $370), BofA $430 — each raised on delivered June-2026 numbers rather than front-running them.

Bear Case

• -12.6% on 2026-06-05 to $371.71, about -21% below the $474 high (2026-06-02) in three sessions: the first distribution event of the cycle, breaking the $410.93/$390.11 shelf. Names that shed a fifth of their value that fast usually base for weeks before re-running.
• Technical downgrade to Hold plus ~$1.4M insider selling (early June 2026): fast money marking valuation stretched near the highs even where the operating story holds.
• Consensus Q2 EPS $2.10 sits below the company's own $2.14–$2.34 guide: a print that only meets the Street reads as a guide miss, and the setup carries full binary risk with the event one day out.
• Category multiple ceiling: an EMS/ODM integrator carries a structurally lower terminal multiple than the silicon designers it serves; near $370 on ~$10.15 FY EPS the forward multiple is already full, and any CCS deceleration compresses it fast even on a headline revenue beat.
• Customer concentration: hyperscaler demand is top-heavy, and a single program push-out — the kind that took ~18% out of the stock in a week in 2024 — is an air-pocket the diversified guide cannot cushion in the moment.

Setup & Price Structure

The chart is in repair mode. From the $474 high (2026-06-02) it lost 12.6% on 2026-06-05 to $371.71 and cut clean through the $410.93 and $390.11 support shelf, which now acts as overhead. A clean read requires a daily close reclaim above $390 and, ideally, a higher low built in the $340–$360 zone to prove the June sell-off was a correction rather than the first leg of a top. Below that sits the pre-May breakout base near $330 — the level separating a deep pullback in an uptrend from a completed top. The 2026-07-27 print lands before any of that structure has confirmed, so a fresh entry at current price is a bet on the event, not on the setup. The theme behind the name — hyperscaler capex and datacenter networking — is still accelerating on the capex line, but CLS's own tape is the slowest-moving part of the story right now.

Catalyst Calendar (next 30 days)

• Late July 2026 — hyperscaler capex commentary (GOOGL / META / AMZN / MSFT calls): the demand-side read-through. A 2026 capex trim on any of these calls pressures the whole datacenter-infrastructure complex and CLS with it; a further raise reinforces the $10–16B incremental-networking case.
• ~Early August 2026 (est.) — post-print analyst revisions: whether the CIBC $480 / BMO $450 / BofA $430 cluster holds, extends, or reverses will confirm or break the chase-the-tape phase the June target escalation put the name in.

Elapsed catalysts

• 2026-07-27 — Q2 2026 earnings (confirmed, tomorrow): the binary. Watch CCS YoY growth and CCS margin first; a total-revenue beat means little if CCS decelerates below ~50% or margin slips under the 8.6% Q1 mark. Any FY2026 guide revision off $19B / $10.15 EPS moves the stock more than the quarter itself. (passed 13d ago)

What Would Change Our Mind

The operating thesis breaks on a weekly close below $330, which loses the pre-May breakout base and reframes the June correction as a completed top. Short of price, the read inverts on a 2026-07-27 print showing CCS YoY growth under ~50%, CCS margin back below the 8.6% Q1 level, or any cut to the $19B FY2026 revenue / $10.15 EPS guide — any one of which turns a deceleration scare into confirmed deceleration. A hyperscaler trimming 2026 capex on its late-July call would remove the demand pillar underneath the entire picks-and-shovels frame. On the other side, a daily close reclaim above $390 with a higher low in $340–$360 restores the setup and shifts the read toward a post-print gap-and-go entry.

Correlation Notes

CLS trades as a high-beta expression of the hyperscaler-capex trade, tightly correlated with the AI-networking complex — ANET and AVGO on the switching/silicon side, and the capex principals (GOOGL, META, AMZN, MSFT) on the demand side. It is the under-owned, sharper-drawdown version of that trade: a Canadian-listed EMS that momentum funds rotate into for torque and out of fastest when the group wobbles, which is why its June drawdown ran deeper than ANET's over the same window. The single largest exogenous swing factor before the print is hyperscaler capex headlines; USD/CAD translation is a second-order input on the guide given the Canadian listing. When the datacenter-infrastructure theme is bid, CLS outperforms; when the group de-rates, it leads to the downside.