The 10 Most Popular HBR Articles of 2025

2025 was marked by broad uncertainty around AI, tariff policy and the macroeconomy, and Harvard Business Review’s most-read pieces reflected those concerns. Popular articles focused on decision-making challenges, the evolving debate over DEI, and unintended consequences of AI — offering qualitative insight into executive and public sentiment but providing no direct market data or financial metrics to drive immediate trading decisions.

Analysis

Market structure: AI/cloud infrastructure and semiconductor-equipment makers are the direct winners — expect outsized revenue concentration in NVIDIA (NVDA), ASML (ASML), and cloud platforms (MSFT, GOOGL) over 12–36 months as enterprise capex clusters. Losers are tariff-exposed exporters and legacy low-margin manufacturing (large industrials, some consumer electronics OEMs) facing margin pressure if trade policy tightens; expect 5–15% EBITDA volatility for exposed OEMs in a tariff shock. Cross-asset: risk-off episodes will push shorter-duration Treasuries up and USD stronger while industrial metals (copper, aluminum) face downside; equity implied vols will spike 20–50% on headline policy moves.

Risk assessment: tail risks include a substantive AI regulatory regime (EU/US) that could slow monetization (low-probability, high-impact within 6–18 months), China export controls on advanced chips, or a tariff escalation that cuts global trade by >2% YoY. Time horizons split: immediate (days) headline-driven vol; short-term (weeks–months) earnings/tariff prints; long-term (quarters–years) structural reallocation of capex and market share. Hidden dependencies: data-center demand is highly concentrated — a single hyperscaler pause could reduce semiconductor orders by >10% sequentially. Primary catalysts: Fed decisions, CPI prints, EU AI Act milestones, and major hyperscaler capex guides over next 3–12 months.

Trade implications: favor high-quality exposure to AI infra with built-in volatility management: size 1–3% positions in NVDA and ASML with rules-based adds on >15% pullbacks over 6–12 months. Hedge macro tails with 1–2% notional SPY 3-month puts if SPY falls >5% in 10 trading days or buy 2–5% allocation to 2–5y Treasuries on Fed pivot signals. Pair trade opportunity: long ASML (ASML) 1–2% vs short CAT (CAT) 1% — structural capex for chips vs cyclical industrials exposed to tariffs and slowing Chinese infra.

Contrarian angles: consensus fears around immediate AI regulation likely overstates near-term revenue impact — monetization continues via tooling and cloud where switching costs are high; this argues for underweighting short squeezes on core infra names. Conversely, markets may be underpricing concentration risk: a single hyperscaler pause or China sanction could cascade — keep position caps and liquid hedges. Historical parallel: 2010s cloud capex acceleration concentrated winners (AMZN, MSFT) while many hardware providers never recovered; expect similar winner-take-most dynamics here.

Key Decisions for Investors

• Establish a 2–3% long equity position in NVIDIA (NVDA) over 6–12 months; add another 1–2% on any drawdown >15% from entry, and size a protective 3-month SPY or NVDA put if NVDA spikes >30% in 10 days.
• Allocate 1–2% long to ASML (ASML) and pair with a 1% short in Caterpillar (CAT) — rationale: durable semiconductor-equipment capex vs tariff/cyclical industrial exposure; rebalance if ASML outperforms CAT by >25% within 6 months.
• Buy 3–5% notional protection: either 3-month SPY puts sized to cover 1–2% portfolio drawdown if SPY declines >5% in 10 days, or shift 2–4% cash into 2–5y U.S. Treasuries on signs of Fed easing (2 consecutive CPI prints <2.5% YoY or Fed cuts signaled).
• Reduce cyclical export/manufacturing exposure by 25–50% (e.g., Industrials/Consumer Discretionary ETFs or tickers like CAT, HON) if US/China tariff headlines escalate or new tariffs remove >$20bn in trade within 90 days; redeploy proceeds into AI infra (MSFT, GOOGL) and semiconductor-equipment (ASML, LRCX).
• Monitor three specific catalysts over next 90–180 days — (1) EU/US AI regulatory milestones, (2) next two CPI prints and next Fed decision, (3) any China advanced-chip export restrictions — and reduce high-multiple software exposure by 50% within 10 trading days of any adverse outcome on these fronts.