2 Stocks to Watch From the Challenging Computer-Services Industry

The computer-services industry ranks #187 of more than 247 Zacks industries, in the bottom 24%, with 2026 aggregate earnings estimates revised down 0.4% since Jan. 31, 2026; the group fell 3.6% over the past year versus gains of 15.2% for the S&P 500 and 21.8% for the technology sector. Despite cautious client spending and public-sector procurement delays, AI-led demand supports selected names: PDF Solutions' 2026 EPS estimate rose 5.7% to $1.30, implying 38.3% growth, while CACI targets fiscal 2027 revenue of $10.65-$10.85B, up 11.3%-13.4%, supported by more than $32B in backlog. The industry's 13.62x forward P/E is below the S&P 500's 19.78x and the sector's 20.87x.

Analysis

The actionable distinction is recurring, mission-critical data/software versus labor-intensive discretionary services. PDFS has operating leverage if semiconductor customers convert AI-process experimentation into broader production deployments: incremental software/data revenue should carry materially higher gross margins than consulting, but its elevated momentum leaves little tolerance for a booking slowdown. The more relevant read-through is to semiconductor process-control and yield ecosystems—KLAC and CAMT—rather than broad hyperscalers, whose spend does not automatically translate into PDFS revenue.

CACI offers a different earnings path: funded work and classified national-security exposure make revenue less correlated with commercial IT budgets, while AI-enabled engineering can raise program margins before it produces a separately monetizable AI revenue line. The principal near-term risk is cash-conversion and award timing, not demand; continuing resolutions, procurement pauses, or a mix shift toward lower-margin acquired revenue can defer the expected earnings realization by one or two quarters. Leidos (LDOS) and Booz Allen (BAH) are the cleaner liquid comparables for relative-value positioning.

Consensus may be too quick to treat discounted services multiples as a blanket AI opportunity. AI often compresses billable hours before pricing models reset, which is negative for generalist integrators; the durable beneficiaries will be vendors embedded in proprietary manufacturing data or cleared government workflows. Over the next 6-18 months, AI capex should favor PDFS only if utilization and recurring-data attach rates rise, and favor CACI only if productivity gains appear in segment margins rather than remaining an anecdotal implementation benefit.

Key Decisions for Investors

• Initiate a 3-6 month long CACI / short LDOS pair, sized beta-neutral. CACI's funded-work visibility and potential engineering-margin upside offer a clearer catalyst path; exit if funded backlog growth decelerates materially or fiscal revenue guidance falls below the low end. Target 10-15% relative return with 6-8% pair risk.
• Do not chase PDFS after its sharp re-rating; place an alert for a 10-15% pullback or evidence of two consecutive quarters of backlog/bookings growth. On confirmation, buy a 6-12 month position versus a partial short in SOXX to isolate process-data adoption; invalidate if recurring revenue growth or customer concentration deteriorates.
• Maintain an underweight in broad commercial IT-services exposure through IGV/consulting names over the next 1-3 months. Enterprise AI pilots can delay project spending and reduce utilization before implementation revenue arrives; reverse only after utilization, bookings, and forward EPS revisions improve together.
• Monitor U.S. appropriations/continuing-resolution developments and CACI cash flow at the next earnings release. A procurement disruption is the key event risk to the CACI long; use it as a trim trigger rather than extrapolating backlog into near-term revenue.