Anthropic economics chief talks about the jobs that could be killed by AI

Oil is trading at $109 per barrel. Anthropic reportedly has revenue surpassing OpenAI and its economics chief warns about AI-driven job losses, while private capital investment is declining per AllianceBernstein. A political deadline — Trump’s 8 p.m. cutoff for a deal with Iran — and Piper Sandler’s warning that the end of war won’t trigger a strong rally add geopolitical and sentiment risk; a new COVID mutation nicknamed “Cicada” is also highlighted.

Analysis

Elevated energy risk premia remain the principal regime driver for commodities and macro-sensitive sectors over the next 1–9 months. With upstream capex already materially below historical trough-to-recovery cycles, a sustained risk premium shock tends to translate into 0.5–1.5 mb/d of foregone supply growth over 12–24 months; that mechanical gap amplifies price persistence rather than one-day volatility, so beneficiaries are the cash-flow levered producers while midstream/refining economics diverge depending on crack spreads.

The AI plane is bifurcating markets faster than most models assume: the spend shift is from labor and IT operating budgets into concentrated compute and cloud capex, creating superlinear returns to GPU/cloud owners and compression of unit economics for high-volume labor intermediaries. Expect revenue and hiring deceleration for firms selling commoditized human capital over 6–18 months, while a handful of infrastructure suppliers capture outsized margin expansion — that concentration also concentrates systemic risk if supply of accelerators tightens.

Private capital retrenchment lengthens hold periods and increases the supply of companies forced to access public markets or structured secondaries, which creates a multi-quarter window for arbitrage in listed names that proxy private valuations. Credit strategies that finance late-stage companies should see underwriting pickier and yield-rich opportunities, but mark-to-market shocks will compress NAVs for firms sitting on unexitable stakes.

Catalysts to watch in the near term are (1) any abrupt regional escalation that tightens seaborne flows within days, (2) compute inventory and supplier cadence announcements that change GPU lead times over weeks, and (3) scheduled policy/administrative deadlines that can flip realized volatility quickly. The consensus that fundamentals will “snap back” immediately after headline risk fades understates the asymmetric lag from capex pullbacks and re-deployment of labor/capital into compute-heavy stacks.

Key Decisions for Investors

• Long NVDA call-spread (buy 3-month 1.15x ATM, sell 3-month 1.4x ATM) sized 1% NAV — thesis: capture continued AI infra squeeze and cloud replatforming; target 3x payoff if NVDA rallies 20%+, stop-loss if spread basis deteriorates to 50% of premium.
• Pair trade: long Pioneer Natural Resources (PXD) 6–12 month exposure (2% NAV) funded by short Valero (VLO) 6–12 month exposure (1.5% NAV) — rationale: upstream captures sustained crude upside and free cash flow while refiners face volatile crack spreads; target 20–40% relative outperformance over 3–9 months, stop if WTI curve flips into heavy contango.
• Event hedge around geopolitical deadline: purchase a near-term straddle on front-month crude (via USO options or short-dated WTI futures straddle) sized 0.5% NAV — objective: asymmetric payoff to capture headline-driven moves over a 2–4 week window; cap premium paid and reduce after move >25% realized vol.
• Short staffing/recruiting incumbents (e.g., RHI or MAN) vs long MSFT (pair, net 1% NAV short staffing, 1% NAV long MSFT) — rationale: secular AI-driven demand reallocation hits high-volume staffing first; target 15–30% compression in staffing multiples over 6–12 months, hedge tech beta with the cloud leader.