The adopted passive core and new benchmark — Meb Faber’s shareholder-yield weighting (dividends + buybacks + debt paydown) instead of market cap, at a 9x forward P/E vs SPY’s 20x and QQQ’s 22.5x; “little opportunity cost and a fat margin of safety,” with the performance gap expected to keep widening for years.
Valuation comparator only — at 22.5x forward the most expensive leg of the spread, and the closest listed proxy for the high-multiple “overdetermined failure” problem and the 2028 chip-supply tsunami.
The exhibit for what is wrong with cap-weighting, not a short — “comical that it has 1,226 holdings yet allocates 27.61% to three companies”; diversification that is nominal, efficient like a just-in-time supply chain right up until something breaks.
Weighting, not stock-picking, is where Ferg starts: market-cap indexing is "incredibly efficient and is hard to beat," but it is the just-in-time supply chain of asset allocation — unbeatable on cost until something breaks, at which point "resilience/redundancy is everything." The remedy he has adopted is Meb Faber's shareholder yield (EYLD) as the passive core and as the benchmark his own portfolio must beat, bought at a 9x forward P/E against SPY's 20x and QQQ's 22.5x. The valuation caution behind it is borrowed rather than asserted — Kedrosky's "overdetermined failure" statistics and Myrmikan's AI-debt/Fed-bailout history — and the trade-hunting instinct runs to structural, date-certain events (forced fund liquidations) rather than to narratives. (Based on the single processed post to date — 2026-SEP-01; will be revised as the archive grows.)
Shareholder yield over market-cap weighting. "This doesn't mean I want to take the other side of market cap weighting; I want to go long in a smarter, passive way." Weight by cash actually returned to owners — dividends, buybacks, debt paydown — which a company cannot fake. He sees "little opportunity cost and a fat margin of safety," and has made EYLD "what I plan to benchmark my portfolio against moving forward."
Concentration is measured by top weights, not holdings count. The exhibit: EEM holds 1,226 names yet puts 27.61% in three of them — "comical." Cap-weighting mechanically compounds the winners into the whole bet, and that missing redundancy is paid for all at once.
High multiples are a forecast, not a risk to be sized around. Kedrosky's framing, endorsed as "the best articulation of the danger of high multiples/valuations I've heard": at a 70 P/E there are ~20 independent 5% ways to fail, and "an overdetermined system tells you that… there's greater than a 60% chance of failure" — so what looks unpredictable "is actually highly predictable."
AI scepticism sourced from debt history and supply arithmetic. Myrmikan Research's "AI Debt Failure Will Prompt Another Wave of Fed Bailouts" supplies the arc ("nothing like a good dose of history to help understand the present"); the specific call is a chip-capacity flood — "unprecedented cash inflows into Taiwanese and Chinese chip manufacturers, with what looks like a tsunami of supply in early 2028… once you lock in supply, my friend, prices are going to zero," because fixed costs must be covered.
Hunts date-certain, structurally forced set-ups. The "KOL delisting" trade — a fund wound up at NAV on a published schedule, with mandate-constrained holders forced out into thin liquidity beforehand. Regret at missing FM's $27.23 liquidation distribution (Jan 2025) is written up as a screen, complete with "a quick backtest."
Process discipline over impulse. A mechanical reading rule ("after a book is recommended three times, I read it"), current research paired with history, and a pre-commitment to the social cost of contrarian holdings (the Ian Cassel wrong/stupid/lucky sequence).
Appearances
One dated page per appearance — each has its full stock table, talking points, and the saved transcript. Newest first.