Argan's Data Center Business Expands: Can It Become a Growth Pillar?

Argan (AGX) is expanding data center-related industrial fabrication, led by a $125M contract awarded in Nov 2025 to fabricate thermal expansion and energy storage tanks through 2028. The Industrial segment now represents 20% of revenue with ~$225M backlog (vs. the Power segment’s $2.5B), while a new North Carolina fabrication facility is expected online in Q3 FY2027 to add capacity. The stock has surged 144% YTD, trades at a premium with a forward 12-month P/E of 53.52, and analysts have lifted FY2027/FY2028 earnings expectations to imply 29.4% and 32.2% YoY growth, respectively.

Analysis

The immediate market read-through is not that AGX has become a pure AI winner; it is that investors are paying an increasingly full multiple for a niche claim on a very large spend cycle. The core issue is timing: the new fabrication capacity and contract visibility are real, but the earnings contribution is back-end loaded, while the stock has already discounted a good portion of that upside. With the shares at a rich forward multiple, any slippage in utilization, project cadence, or gross margin on the industrial work can trigger multiple compression faster than estimate revisions can catch up.

On relative positioning, the broader beneficiaries of AI infrastructure are still EME and J, because they monetize more of the wallet share across electrical, mechanical, PM, and EPCM scopes rather than a single fabrication stream. That means AGX may see headline growth, but EME/J are better vehicles for recurring order flow and faster conversion of hyperscaler capex into revenue. Second-order, AGX’s facility build creates execution and working-capital risk: if the data center cycle pauses, the new capacity can become an underutilized drag instead of an earnings lever.

The contrarian view is that the consensus is extrapolating a strategic narrative into near-term EPS power. The industrial segment is still not large enough to materially change consolidated outcomes until FY27/FY28, so this looks more like a long-duration growth story than an immediate earnings acceleration story. Thesis would be falsified if backlog converts faster than expected and industrial margins inflect upward before the North Carolina plant is fully online; otherwise, the setup is vulnerable to a disappoint-and-de-rate reaction on the next print or any indication hyperscaler capex is being delayed by power/interconnect constraints.

Key Decisions for Investors

• Avoid chasing AGX after the rerating; treat it as a hold/trim into strength and only add on a 15-20% pullback or if next guidance confirms industrial margin accretion. Risk/reward is poor at >50x forward earnings unless execution is clearly ahead of plan.
• Enter a beta-adjusted long EME / short AGX pair trade over 3-6 months. EME is the cleaner expression of AI infrastructure demand with broader revenue capture; target 10-15% spread compression, stop if AGX raises FY27/FY28 industrial assumptions materially.
• If you want exposure to the same capex theme without single-contract risk, prefer J over AGX on any pullback. J has more diversified advisory/program-management exposure, so it should hold up better if the market rotates from narrative to cash-flow quality.
• Set an alert for AGX around the first full reporting period after the North Carolina facility ramps in FY27. Only reconsider a long if utilization and backlog conversion show clear operating leverage; otherwise the capex could dilute near-term FCF.