2 AI Stocks Down 48% to 70% to Buy Now and Hold for 10 Years

Ciena reported fiscal Q3 revenue up 37% year over year to $1.7 billion and an $8.5 billion order backlog, which management expects to exceed $10 billion by fiscal year-end; it projects its addressable market will roughly double to $52 billion by 2029. Astera Labs' Q2 revenue more than doubled to $392 million, and its Q3 revenue guidance of $540 million-$560 million implies sequential growth of about 40%. The article sees both companies as long-term AI connectivity beneficiaries, while noting shares are well below their highs and citing AI-spending and competition risks.

Analysis

The investable distinction is where each company sits in the AI buildout. CIEN monetizes network capacity that must connect facilities and systems; ALAB depends more directly on continued adoption of particular server-level connectivity architectures. If AI budgets pause, project timing can defer both, but architecture changes or in-house/custom alternatives are a more acute risk to ALAB. Broadcom (AVGO) is a potential beneficiary if customers favor broader or internally integrated connectivity offerings, though this article alone does not support an outright AVGO short or long.

For CIEN, the key near-term test is conversion of reported demand into shipments, recognized revenue, and margins—not backlog size by itself. A backlog that fails to convert, or customer capex deferrals, would undermine the growth case and expose the stock to multiple compression. For ALAB, watch whether product expansion translates into durable revenue per deployed accelerator and whether customer or platform concentration limits bargaining power. Management’s product timing claims need confirmation in reported results.

Over 1–3 months, earnings and guidance should distinguish demand from order timing; over 6–18 months, hyperscaler capex and connectivity architecture choices are decisive. The contrarian opportunity is that a broad AI-risk selloff may discount infrastructure demand indiscriminately, but the article’s growth projections are not independent verification. CIEN offers the cleaner conditional exposure; ALAB has higher execution and substitution risk. Avoid treating the two forward multiples as directly comparable because their earnings bases differ.

Key Decisions for Investors

• Prefer a staged long in CIEN on weakness, adding only as backlog converts into revenue without margin deterioration. Falsify the thesis on guidance cuts, sustained order-to-revenue slippage, or evidence that major customers are delaying network deployments.
• Keep ALAB on a catalyst watch rather than chase the forecast ramp. Reassess after results confirm sequential growth and broader product adoption; reduce exposure if deployments or revenue per accelerator disappoint, or customers shift toward competing/custom connectivity solutions.
• Do not initiate an outright AVGO position solely as an ALAB hedge: its diversified business exposure makes the linkage imperfect. Monitor customer design wins and platform choices for evidence that AVGO is taking share from specialized AI connectivity suppliers.
• Use hyperscaler capex updates and company guidance over the next 1–3 months as near-term risk controls; the 6–18 month thesis depends on sustained data-center investment and realized infrastructure orders, not announced addressable-market estimates.