PodcastThe Monetary Matters NetworkKieran Cavanna

From Soros to Old Farm: How to Identify the Market’s Top Thematic Risk-Takers

Geopolitical shocks—from Iran‑linked events to Europe’s exposure from globalization—are reframing macro risk and prompting rapid portfolio adjustments, while Latin America emerges as a politically driven growth opportunity. Technology discussion centered on AI capex as a cross‑sector, multi‑year cycle that favors enabling infrastructure over pure‑play chip ownership, and highlighted the high bar for scale in hardware and biotech. Market themes emphasized concentrated, high‑conviction allocations and rigorous operational due diligence as paths to outsized returns, alongside structural industry flows and fee pressures. The outlook calls for a multi‑year capex tailwind but stresses short‑term nimbleness to capture the forecasting edge.

Geopolitical RiskEconomic GrowthArtificial Intelligence

Preview · 14 more points in the app

Read the Full Summary

Key Takeaways

The Iran‑linked shock at the start of 2026 flipped consensus from a non‑inflationary US boom to a higher risk of recession plus inflation, demonstrating that geopolitical events can rapidly change macro regimes and force hedge funds to reposition quickly within actively managed, drawdown‑constrained books.

Globalization has outsourced critical defense and energy capabilities, leaving the UK and Europe structurally vulnerable and uncompetitive, and Europe’s valuation advantage has not translated into performance because a weak regional tech base has driven benchmark drift and missed local winners.

Shifting politics across Brazil, Argentina, Chile and parts of Central America create a plausible, politically driven upswing in Latin America—many regional economies are lightly levered, making EM credit allocations run by regional managers an attractive opportunity amid broader reallocations in alternatives.

AI capex is the dominant cross‑sector investment theme for the next 3–4 years, with high‑end compute capacity effectively sold out; investors should prioritize AI‑enabling infrastructure (power, optical systems, memory, even nuclear exposure) over owning pure‑play chip names like Nvidia and remain nimble on near‑term execution.

The high performance bar in hardware and biotech (illustrated by missing GLP‑1) and the missed SpaceX opportunity show transformational engineering and biotech winners require scale, speed and conviction—precisely the type of deals the co‑invest model (where $20M ideas can scale to >$100M) is designed to capture.

GEO-POLITICS

Iran Shock — At the start of 2026 a consensus expected a non‑inflationary US boom, but events linked to Iran shifted the outlook toward a higher chance of recession with inflation, illustrating how geopolitical shocks can rapidly flip macro regimes and force swift portfolio repositioning.

“I mean, think of the start of 2026 as we're sitting here that we were going to have a non-inflationary economic boom in the US and then we made it…”—

European Exposure — Globalization outsourced critical defense and energy capabilities, leaving the UK and Europe structurally vulnerable and less competitive; higher industrial production costs versus China and the US create persistent security and competitiveness risks.

“I think we really screwed up on globalization in a like that you could just like the UK unfortunately is in this position where they've outsourced their defense to the…”—

TECHNOLOGY

AI Capex — Speakers identify AI capex as the dominant cross‑sector investment theme—touching energy, industrials, tech and telco—and favor AI‑enabling infrastructure (power, optical systems, memory, even nuclear exposure) over owning pure‑play chip names like Nvidia.

“a famous quote that Besson said about once a week which is make sure the main thing is the main thing and right now the main thing is AI capex…”—

Scale Edge — Hardware and biotech managers face a high performance bar in 2025–26 and uncertain persistence over the next three years; missing the GLP‑1 trade highlights that scale at large pharma, not small biotech, often captures these transformative cycles.

“I think like if like right now if you're a hardware you know tech manager or you're a biotech manager in 2025 and 2026 the bar is very high I…”—

ECONOMY

Co‑Invest Model — Co‑investments attract hungry, intellectually curious managers and can scale quickly—turning a $20M idea into >$100M—while offering better alignment than fee‑rich large managers, making co‑investing a scalable source of high‑conviction deal flow.

“I learned a great deal by having internal people as well as external risktakers and it was such a good experience because you could have people who are real domain…”—

MARKETS

Concentrated Bets — The firm concentrates capital into one or two high‑conviction manager ideas and has seen outsized outcomes—an optical strategy returned ~250–300% in a year and a repurposed data center re‑rated from ~$3 to ~$35—but such concentration requires active trimming and risk controls.

“It's been the best thing I've ever done and that's one one manager…”—

Allocator Signals — When selecting managers the guest prioritizes repeatable decision‑making over lucky outcomes, values strong risk managers and founder alignment, and treats friends‑and‑family commitments and top first investors as key reputational signals.

“I do that on hundreds of managers and then when I see those things play out I try to deduce is it luck did they get it right once or…”—

Operational Diligence — Longstanding external allocation programs (e.g., Soros) put heavy weight on operational due diligence and ongoing exposure monitoring, and large institutions can leverage reputation and separately managed accounts to secure favorable terms and distribution signals.

“so that's really what I knew there. But Soros is an amazing firm in that they've had a long history of allocating externally actually at least 20 plus years…”—

FORECASTS

Capex Cycle — The guest expects a firm capex cycle over the next three to four years with high‑end compute capacity effectively sold out, supporting durable demand for compute suppliers, but advises investors to prioritize nimbleness—being right over the next six months matters more than being years ahead.

“And it's very clear to me that the next three, four years, this capex cycle is firm and the com and compute is sold out at the high end…”—

Related Episodes

Important Disclaimers

Not Investment Advice: For education only — not financial, tax or legal advice.

AI-Generated: Made by AI from source transcripts and may contain errors.

No Guarantees: Past performance does not predict future results; investing carries risk.

Our Commentary · Request Removal