Fund manager lists four of his best-value unloved stocks to swerve the AI hype

Ranmore Fund Management manager Sean Peche said he’s avoiding the AI IPO hype and is buying undervalued names, highlighting Ping An (2318-HK) for a ~6% dividend yield and trading below book value in Hong Kong. He also pointed to Comcast (CMCSA) for “annuity-style” income and strong cash flow despite the stock being down ~16% over the past year, and Diageo (DGE-GB) for ongoing brand momentum plus a $1B restructure under CEO Dave Lewis. In Asia, he backed Tencent (700-HK), citing earnings that have tripled while the stock is bought at around its 2018 price, arguing China’s AI model compute costs are a fraction of the U.S.

Analysis

This is less a sector call than a signal that capital is rotating toward self-help and away from duration narratives. The cleanest beneficiaries are businesses with hard cash flow and credible balance-sheet optionality: their upside comes from buybacks/dividends forcing the market to re-underwrite terminal value, not from heroic growth assumptions. The main loser is anything priced for secular acceleration where the market can easily contrast that story with low-capex, high-yield alternatives.

CMCSA is the closest thing here to a mechanical rerating: if management keeps shrinking the asset base while preserving cash generation, the stock can behave more like a financial asset than a media name. That said, cable churn is the key falsifier; if broadband losses re-accelerate, the market will ignore capital allocation and reprice the residual business lower. DEO is more fragile — premium branding helps, but moderation plus private-label trade-down can quietly compress mix and margins even when reported volumes look orderly.

TCEHY stands out because the market is paying a low multiple for earnings that do not require U.S.-style capex intensity, which is a genuine second-order advantage if AI economics stay capital disciplined. The contrarian gap is Ping An: book value and dividend yield look attractive, but those metrics only matter if the market trusts the mark-to-market quality of assets and the policy backdrop. Over 1-3 months, earnings and capital-return commentary matter; over 6-18 months, the real question is whether these are reratings or just classic value traps with a better narrative.

Key Decisions for Investors

• Long CMCSA vs short a cable-fragility proxy (CHTR) over 3-6 months: bet on self-help and asset-shedding while isolating the risk that the market keeps discounting secular cord-cutting; exit if broadband net adds or churn deteriorate for two consecutive quarters.
• Build a starter long TCEHY position on 5-8% pullbacks for a 6-12 month horizon: the setup is multiple expansion if earnings continue compounding with restrained capex; cut if China tech policy or gaming approvals turn adverse.
• Avoid chasing DEO until the next print confirms mix resilience: if volume trends keep weakening faster than price/mix can offset, the stock can remain a value trap despite the brand portfolio; reconsider only if margin guidance stabilizes.
• Treat 2318-HK (Ping An) as a watchlist trade, not a blind buy: the dividend/book discount only works if asset quality and capital return stay credible; the thesis breaks on any deterioration in reported solvency, credit marks, or regulatory pressure.
• If forced into a basket trade, favor a long CMCSA/TCEHY basket over high-multiple AI beneficiaries for the next quarter: the risk/reward is better while the market is rewarding cash flow visibility over capex intensity.