Stockchase InsightsCelestica IncCLS.TOPARTIAL BUYApr 21, 2025

It remains Canada's best way to play the data centre build out, though it's volatile. Revenue last quarter was up 83% and they raised their 2026 outlook. They supply network and hardware platforms to the hyperscalers and Frontiers AI Labs and that demand remains strong. She sees more upside, though shares have dipped.

He's very cautious on the AI theme. Really well run. But very thin profit margins. He's been steadily reducing tech exposure; need to see more favourable action before putting new $$ to work.

He likes Canadian tech. The chart is consolidating with lower highs, but also higher lows in a triangle shape. It could break out, up or down. You need proof that it is breaking out.

Broadening out customer base. Broad exposure to the increasing AI trade. Likes it at this price.

AI buildout expected to last until 2028-29, when we'll probably also see a replacement cycle begin. And CLS is right there to capitalize on that. Continues to have amazing numbers in the 20-30% range. Multiple's not cheap, but growth rate has been fairly good. No dividend.

Doesn't pay a dividend, so his firm doesn't own it. Admires the business. If you're setting up a data centre, one of your first phone calls might be to this company. Very well positioned -- data centres might slow, but infrastructure is picking up.

Bottomed out. He just wrote a put on this recently. One of the cheapest on PEG that you're going to get in the space. Growing ~76%, trading at a much lower multiple than that. Highly cyclical play. Trades at 16x PE for 2027. Strong, visible growth out to 2029. 

Put it in a non-registered account, where you don't mind taking some risk. 

Competes with CSCO and ANET. Very solid balance sheet liquidity. High capital returns to support growth. Elevated valuation of ~31x PE. Bit expensive for this section of the AI ecosystem (the enablers). Once it gets above $455-460, he'll be selling some calls.

Think of it as a Canadian DELL. Largely tied to GOOG's capex, the TPU program, and networking. Benefiting from the super-cycle. Not sure inside the AI trade is the best place to park capital. As supply starts catching up with demand, chances to earn outsized margins start compressing.

Not expensive today. Don't sell. Keep an open mind to ASML or TSM (for more stability).

It's had a run-up in the last year and needs to see some consolidation the coming year. Will bounce between $250-350 until there's a catalyst. CLS keeps beating and raising.

Peak in June around $600. Starting to slowly move below the 200-day MA, which is one of the key lines in the sand for him. Definitely trim a bit here. Long-term price momentum stalled, RSI has come off.

Moves in tandem with the semiconductor space. Have to see what happens with NVDA tonight.

Competitors have better cost of capital and scale. He'd prefer one of the Asian suppliers. Better opportunities elsewhere.

Very much aligned with the entire platform being built out, not just individual components. Direct exposure and concentration to the hyperscalers as customers. He thinks those capex numbers are going up, so there's still upside in this name. But with any narrative shift, this will get sold off first.

In a massive growth segment, with the AI infrastructure buildout. Strategically in a very good position. If you have your heart set on this one, valuation is very reasonable at below 20x PE. Yet FCF is fairly meagre, which indicates valuation is still stretched. Watch and wait.

Instead, he uses TSM for client portfolios. Valuations in the space tend to be fairly rich. So this name isn't a buy today. Watch and wait on this one too.