The Argan Stock Rally Still Isn't Over, Even After the Stock Has Nearly Doubled This Year

Argan more than doubled fiscal 2026 backlog to $2.9 billion from $1.4 billion, with backlog growth across all three segments and management expecting AI infrastructure demand to remain strong beyond 2030. Fiscal 2027 Q1 revenue rose 50% year over year to $291 million, EPS nearly doubled, and the company ended the quarter with no debt. The article frames the stock as benefiting from sustained data center and power infrastructure spending, reinforcing a strong multi-year growth outlook.

Analysis

AGX is effectively a levered call on the duration of the AI power bottleneck, but the more important second-order effect is that its backlog converts the AI capex cycle from a semiconductor narrative into a utility/infrastructure narrative. That shifts the bottleneck from chips to interconnection, generation, and grid completion, which tends to extend cycle life because those projects are harder to cancel once site work starts. The market is likely still underestimating how much pricing power sits with contractors that can actually execute on constrained power projects rather than just sell picks-and-shovels exposure to AI demand.

The near-term debate is not whether demand exists, but whether margin expansion can persist as competitors chase the same backlog. A business with this much momentum typically attracts lower-quality bidders, which can pressure future gross margins and backlog quality over the next 2-4 quarters even if reported revenue stays strong. That makes execution the key variable: if AGX keeps converting backlog faster than peers while preserving margins, the stock can stay de-rated upward; if project delays, labor inflation, or permitting friction emerge, the multiple will compress quickly because the market has already priced in a pristine compounding story.

The contrarian miss is that this is not just an AGX story; it is a signal for an extended capex wave across power equipment, engineering, and grid-adjacent services. The beneficiaries should include select industrials with balance-sheet capacity and long-cycle EPC exposure, while pure-play AI enablers with valuation tied mostly to terminal growth may actually underperform as the capital intensity of AI becomes more visible. In other words, the trade may be moving from "own AI revenue" to "own the bottlenecks that make AI revenue possible."

Key Decisions for Investors

• Stay long AGX on pullbacks over the next 1-3 months, but treat it as a momentum/execution name rather than a core hold; target a 15-20% upside leg with a hard stop if backlog conversion or gross margin sequentially decelerates for two prints.
• Add a basket long in grid/power execution beneficiaries versus expensive AI hardware winners over 3-6 months; pair AGX or industrial EPC exposure against a short in a high-multiple AI beneficiary where growth is easier to narrate than build.
• Buy call spreads in AGX with 3-6 month tenor to express continued backlog monetization while limiting downside from a valuation reset; structure for roughly 2:1 reward/risk if the market keeps rewarding order visibility.
• Fade any post-earnings gap if the stock rerates ahead of visible margin evidence; take profits into strength because backlog-driven names often peak before revenue inflects into consensus models.