15 million apps could disappear in the next 5 years: Palo Alto's Nikesh Arora

Palo Alto Networks CEO Nikesh Arora suggested that the roughly 15 million smartphone apps today could be supplanted within five to ten years by conversational AI agents that orchestrate tasks across services, potentially turning app-first companies in sectors such as ride‑hailing, food delivery, fintech and e‑commerce into backend providers. Regulators and industry leaders warn this shift raises competition, data‑control and liability questions—e.g., who is responsible if an agent fails or takes a commission—creating strategic and regulatory risks for incumbent app platforms and potential opportunity for agent-layer providers.

Analysis

Market structure: Conversational AI agents re-center value from front-end apps to orchestrators, cloud infra and security. Winners: AI model owners, cloud providers and cybersecurity vendors (PANW) that authenticate/orchestrate flows; losers: app-centric monetization (App Store take-rates) and small app-first marketplaces (consumer-facing aggregators). Expect a multi-year shift (3–7 years) where pricing power migrates to agent platforms and API gatekeepers, compressing margins for some app businesses by 10–30% over that horizon.

Risk assessment: Key tail risks are regulatory intervention (agent “tax” or forced interoperability) and liability regimes that could impose large fines or insurance costs—> potential equity drawdowns of 20–40% for affected incumbents. Near-term (days–months) volatility will spike on announcements; medium-term (6–18 months) depends on partnerships/SDKs from Apple/Google; long-term (3–7 years) depends on standardized APIs and model dominance. Hidden dependencies: payments rails, identity/auth, chip supply and standardized APIs—failure in any stalls adoption.

Trade implications: Tactical long exposure to cybersecurity (PANW) and selective cloud/AI infra; defensive shorts/hedges on app-store reliant cashflows (AAPL) where App Store revenue risk is highest. Use option structures to express asymmetric views: buy 9–12 month PANW calls and 12-month AAPL puts rather than outright large directional bets. Rebalance on concrete catalysts (Apple agent SDK release, major platform deals) within 60–90 days.

Contrarian angles: Consensus underestimates incumbents’ ability to embed agents (Apple could monetize agents via device lock-in), so a full AAPL destory thesis may be overdone—limit position size. Historical parallel: portal→search shift shows platform capture is reversible if incumbents control OS+wallet. Unintended consequences include a new “agent tax” and concentrated systemic risk that could spawn regulatory breakups; size positions to tolerate 30% drawdowns and use event-driven stop/review rules.

Key Decisions for Investors

• Establish a 1.5–2.0% portfolio long position in PANW within 1–4 weeks; complement with buying 9‑month PANW calls 5–10% OTM sized 0.5% portfolio to leverage catalyst-driven upside. Take profit if PANW rallies +30% or cut at -15% loss.
• Implement a hedged pair: short AAPL equal-notional to 1.0% of portfolio while keeping a long PANW 1.0% position (net market-neutral exposure). Reassess after 60–90 days or on any Apple agent/SDK announcement.
• Buy AAPL 12‑month puts 10% OTM sized 0.5–1.0% portfolio as insurance against App Store monetization loss; roll or exit on major regulatory clarity (EU/US rulings) or if AAPL outperforms tech index by >10%.
• Rotate +3% portfolio weight into cybersecurity, cloud infra and AI chip suppliers over 60 days (funded by reducing consumer app/exposure by 2–3%). Monitor three triggers within 90 days: Apple/Google agent SDK launch, major booking/payment API deals, and any regulator statements on agent commissions.