Department of Labour data shows Apple Meta, Google and Amazon have slashed H-1B petitions by almost half: Here are the numbers

Amazon certified H-1B applications fell to 3,057 in Q1 FY2026 from 4,647 a year earlier (‑34.2%), while Meta and Google recorded ~50% year‑over‑year declines; Nvidia was an outlier, rising from 369 to 434 (+17.6%). The decline reflects new H-1B policy changes — including a ~ $100,000 fee on new abroad petitions and lottery tilting toward higher‑paid applicants — combined with broad tech layoffs and hiring freezes (Amazon cut 16,000 corporate roles in January after 14,000 in October). These are Labor Department certification counts (not final visa approvals) and signal a material cooling in Big Tech demand for skilled foreign workers, which could pressure talent pipelines and hiring-sensitive cost trajectories.

Analysis

The immediate market read — weaker filings concentrated outside AI-hardware players — understates where the talent demand is migrating. Higher marginal compliance costs and a lottery biased to higher wages make sponsoring junior foreign hires less attractive, effectively compressing the entry-level engineering funnel and shifting demand toward senior/AI-specialist roles that domestic labour markets are thinner on; expect this to depress feature velocity for consumer-facing products on a 6–24 month horizon as replacement pipelines slow.

Second-order winners are those owning AI stack leverage and global hiring footprints: chip/IP vendors and places that can onshore compute rather than headcount. Conversely, early-stage software vendors and boutique product teams that relied on lower-cost foreign junior engineers will face rising unit labour costs and delayed roadmaps, increasing the probability of M&A by larger incumbents seeking talent-consolidation rather than organic hires.

Key catalysts and timeframes: the market will price headline visa-policy tweaks within days, but earnings cycles will reflect the labour gap over quarters and product/innovation shortfalls over years. Reversals could come from policy rollback, a surge in domestic hiring incentives, or concentrated re-hiring by AI leaders; downside tail risk includes further rule tightening or a faster-than-expected wage spiral at the senior end that squeezes margins.

The consensus — that Big Tech is uniformly net-negative — is simplistic. The structural effect is a reallocation of human capital toward capital-intensive AI players; stocks exposed to raw engineering velocity are hurt near term, but hardware/IP owners and stable enterprise outsourcers are underappreciated beneficiaries. Monitor LinkedIn job ads, vendor RFP activity, and visa-rule headlines as high-frequency signals for position timing.

Key Decisions for Investors

• Long NVDA (6–12 months): buy NVDA Jan-2027 calls or a 650/900 call spread to play concentrated AI hiring and infrastructure spend. R/R: asymmetric — limited premium vs high upside if AI-driven hiring remains concentrated; stop loss at 30% of premium.
• Pair trade (3–6 months): short META equity (or buy 6-month puts) vs long MSFT shares dollar-neutral. Rationale: disproportionate slowdown in consumer-facing product hiring vs resilient enterprise/cloud cashflows. Target 15–25% relative outperformance; stop-loss 8–10% absolute on either leg.
• Reallocate small cap risk: reduce exposure to mid/early-stage software names and redeploy to IBM or CRM (6–12 months) to capture outsourcing spend and steady services revenue as firms substitute contractors for headcount. Target total return 10–15% with lower volatility than growth peers.
• Tactical hedge (days–weeks): buy short-dated puts on AMZN and GOOGL ahead of next earnings if visa-rule headlines accelerate risk-off flow. Keep allocation <2% portfolio; payoff if sentiment-driven gap widens, cut if no policy follow-through.