Merryn Talks Money: Causeway Capital's Sarah Ketterer (Podcast)

The piece frames AI as a new toolkit for fund management—capable of scanning filings and research quickly—while emphasizing that human judgment remains critical. Sarah Ketterer highlights value opportunities across healthcare, enterprise software, UK equities, and overlooked Chinese companies, but cautions investors against chasing the AI semiconductor boom. Overall, the article is more advisory than market-moving, with limited near-term impact.

Analysis

The market’s mistake is treating AI as a substitute for stock selection rather than a force that redistributes edge. If models compress the time needed to read filings and notes, the scarce resource becomes judgment on capital allocation, product durability, and incentive quality — which favors businesses with simpler unit economics and less narrative risk, not the most compute-intensive names. That argues for a relative rotation away from the most crowded AI infrastructure exposures and toward cash-generative compounders in healthcare and enterprise software.

In semis, the second-order risk is not demand collapse; it is multiple compression from a crowded ownership base and a capex digestion phase. The group can still print strong numbers for another quarter or two, but if hyperscaler spend normalizes even modestly, the market will punish names priced for linear upside. The vulnerable subsectors are the ones most exposed to memory and AI capex bottlenecks, where supply additions can outpace end-demand faster than consensus expects.

UK equities and select China look interesting for different reasons: both have low expectation sets, which means modest operational improvement can matter more than headline growth. The contrarian read is that investors may be overpaying for visible AI winners while underweighting boring balance-sheet repair stories that can rerate without needing perfect execution. The key falsifier is a renewed wave of AI capex upward revisions or a policy disappointment in China/UK that keeps relative performance trapped.

Key Decisions for Investors

• Long XLV / short SMH for 1-3 months: express a rotation out of crowded AI infrastructure into defensive cash-flow quality. Target 8-12% relative outperformance for XLV; stop if SMH makes new highs on materially raised hyperscaler capex guidance.
• Buy IGV on a pullback or via a 3-6 month call spread: enterprise software should benefit if investors shift from 'picks and shovels' to AI monetization. Best entry is after a weak AI-semiconductor tape; risk is that software budgets get delayed in a late-cycle slowdown.
• Accumulate EWU on weakness for a 6-12 month mean-reversion trade: UK equities are a valuation/positioning call, not a growth call. Falsifier is a renewed sterling rally without earnings breadth or another domestic policy shock.
• Keep FXI/KWEB as a conditional watch item rather than a core long: only add if stimulus data and credit impulse turn up, otherwise the policy risk premium can overwhelm cheap valuation. Use size small and expect high volatility.