Forget Tech Stocks: This Construction Stock Is Building Tomorrow's Data Centers

Fluor is highlighted as a leading data center builder, with a new agreement to construct a 480-megawatt data center in Kentucky for TeraWulf, a project estimated at $3 billion to $4 billion. Management said it is pursuing additional U.S. and European data center opportunities, and the article cites $25.5 billion of backlog at year-end 2025 despite a $51 million net loss and a $643 million litigation payment. The piece is largely an investment thesis on AI infrastructure exposure rather than a new material catalyst, so near-term stock impact should be limited.

Analysis

The market is likely underappreciating how much of the data-center boom accrues to the picks-and-shovels layer rather than the headline AI compute names. For FLR, the key second-order effect is backlog visibility: once hyperscalers and neoclouds commit to power-and-shell capacity, construction spend tends to be less cyclical than semiconductor capex and can re-rate the stock even before revenue is recognized. That makes FLR a cleaner way to own AI infrastructure demand without taking direct GPU obsolescence risk.

The bigger winner may be WULF rather than the better-known AI landlords, because every large buildout validates power-hungry, colo-adjacent capacity and improves financing optionality for operators with contracted load. The risk is that these projects are capital intensive and milestone-based, so any permitting, interconnect, or customer-credit delay can push cash flow out by 2-4 quarters and compress near-term multiples. NVDA benefits only indirectly here; if data-center buildouts slow, the supply chain can see a lagged inventory unwind even if demand headlines stay strong.

The contrarian angle is that the market may be too focused on total addressable spend and not enough on execution bottlenecks: grid access, transformer lead times, and local permitting can turn a $3-4B project into a multi-year working-capital drag. That favors industrials with pricing power and balance-sheet resilience over pure-play operators levered to completion risk. The litigation overhang on FLR also matters less as a one-time noise item than as a reminder that the stock can gap on non-operating headlines, making entry discipline important.

Key Decisions for Investors

• Long FLR on pullbacks over the next 1-2 weeks; target a 6-12 month hold. Risk/reward favors a re-rating as data-center awards convert to backlog, with downside cushioned by diversified infrastructure exposure and backlog support.
• Pair trade: long FLR / short a basket of high-beta AI infra names (e.g., CRWV or NBIS via liquid proxies if available) for 3-6 months. Thesis: construction monetization is earlier and more visible than operator economics; risk is a faster-than-expected funding cycle for neoclouds.
• Add to WULF on confirmed project financing or offtake milestones; use a 3-9 month horizon. High upside if buildout proceeds, but size smaller because execution and power-availability risk can delay EBITDA conversion.
• Stay tactically neutral NVDA for now versus infrastructure beneficiaries. Data-center capex can continue to support demand, but near-term upside is less asymmetric than for contractors capturing the spend.
• If FLR rallies on headlines without contract follow-through, sell upside calls against a core long. This monetizes event-driven volatility while recognizing that actual revenue recognition from megaprojects can lag sentiment by multiple quarters.