Actionable insights — Equities Extremely Complacent; De-lever and Prepare to Buy The Dip

8:40 1. Manufacture a variant perception — find where the consensus playbook contradicts recent history

1. Write down, in one sentence, what consensus expects to happen in the next crisis. Here: "yields rise → something breaks → equities fall → yields fall on a flight to safety."
2. Go back and check the actual tape at every comparable episode of the current regime — not the textbook, the prints. Count the repetitions.
3. If the observed behaviour contradicts the expected behaviour repeatedly, you have a variant perception rather than a contrarian opinion. Weight it by the count: five repetitions in six years is a regime, not a fluke.
4. Then look for the mechanism that forces the repetition (insight 2). A variant perception with a mechanism behind it is tradeable; one without is just a hunch.

• The first risk-off after this recording: if long yields stop falling inside ~3 weeks and turn up while equities keep falling, the regime is confirmed and the clock on insight 2 starts.

19:32 2. Trace the marginal buyer — and monitor the de-grossing chain as a mechanism, not a mood

1. For any market, ask who is the marginal buyer today and how they are funded. Patient, non-profit-motivated holders (central banks, pensions) behave differently from levered, mandate-driven ones.
2. Quantify the share. Use official sources — here a Fed white paper (Oct-2025) showing 37% of net note/bond issuance since 2022 was Cayman Islands hedge funds, i.e. the basis trade on leverage.
3. Write the chain out as a sequence you can watch in real time: equity vol spikes → hedge-fund risk managers force the book flat → they sell Treasuries (their biggest position) → long yields rise → higher yields feed more equity vol → repeat.
4. Note where the chain terminates: it only breaks when policymakers inject dollar liquidity. That injection is the buy signal — not the panic itself.

• Equity vol spikes paired with rising long yields (the fingerprint of de-grossing rather than flight-to-safety); then watch for the liquidity response — buyback size, issuance skew to bills, RRP/reserve moves, or explicit QE — as the moment to buy the dip.

10:53 3. Keep a policy-reaction level, not a valuation level

1. Instead of asking what a yield "should" be, record the level at which policymakers have observably changed behaviour — the tweet, the walk-back, the issuance change.
2. Update the level as tolerance shifts, and keep both numbers: the level that used to trigger a response and the level being tolerated now.
3. Adjust for the debt stock. A higher debt load makes the same yield more binding, so rising tolerance is a stretched elastic, not a durable change of regime.
4. Trade the reaction: approach to the zone means expect a policy back-down (risk-positive, briefly); a sustained break through it means the reaction function has failed (risk-negative, structurally).

• 10-year US yields entering 4.6–4.9%; a jawboning tweet or an abrupt policy softening as confirmation the level still binds — or silence above 4.9% as the sign that it no longer does.

12:04 4. Count the back-downs — credibility decays continuously and repriced all at once

1. Treat every policy retreat under market pressure as a data point, not a relief rally. Keep a running count.
2. Identify the first-principle the retreats are eroding — here, that the US military ultimately backs the Treasury market and the dollar.
3. Look for physical corroboration of the same erosion outside markets (a capability being neutralised cheaply), which tells you the deterrent itself, not just the rhetoric, is weakening.
4. Do not expect a linear price response. Model it as an option: nothing, nothing, nothing — then a repricing "all at once." Size positions to survive the flat period (see insight 8).

• Each retreat on tariffs/war under yield pressure; cheap asymmetric weapons neutralising expensive platforms; foreign official demand for Treasuries continuing to fall while private levered demand rises.

22:40 5. Judge a central bank by its revealed mandate, not its stated one

1. Collect the episodes where the stated mandate (price stability) and an unstated one conflicted, and record which one won.
2. Name the unstated mandate in the institution's own words — it usually has an official euphemism you can search for in the minutes and press conferences.
3. Check whether the arithmetic makes the unstated mandate unavoidable (debt/GDP, deficit, collateral centrality) rather than a matter of the chair's personality.
4. Then forecast the duration, not the outcome: assume the unstated mandate wins and ask only how long the new chair can hold out — that interval is the drawdown you must survive.

• Warsh's first real Treasury-dysfunction test; how many days/weeks of dysfunction he tolerates before intervening — the length of that hold-out is the size of the opportunity.

24:27 6. Map your creditors — then watch the weakest one's bond market as your leading indicator

1. List who actually funds the debt you're exposed to, in order of size, and check each creditor's own fiscal condition. A creditor with your disease can't be a shock absorber.
2. Overlay the two yield series. If they trade in lockstep, the smaller/weaker market is your early-warning screen — it breaks first and tells you where yours is going.
3. Find the one market not conforming; the outlier identifies who benefits from the stress and often who is engineering the alternative system.
4. Assume correlation at the break: "once one of them breaks, they're all going to break in very short order."

• A 10-year gilt / 10-year UST overlay as a daily screen; JGB long-end stress; and Chinese 10-year yields staying low as confirmation of who is on the other side of the trade.

29:44 7. Re-price the asset in a non-printable numéraire to strip out the currency

1. Before judging whether an asset is expensive or in a bull market, divide it by something that can't be issued — gold — and by the asset it's supposedly competing with (long bonds).
2. Use total return (dividends included) so the comparison is honest, and anchor to several start dates rather than one.
3. Read the divergence between the nominal and the gold-denominated series as the currency component of the "return."
4. Sanity-check with a historical analogue where the numéraire was official policy, so you can see what the same chart looked like under a gold standard.

• Nominal index highs coinciding with new lows in the index/gold ratio — resilience that is currency, not earnings; the same test applied to any market claiming a bull run.

36:22 8. Read the buyer's cadence — bigger buys at lower prices means accumulation, not trading

1. During a correction in an asset you own, stop watching price and start watching the largest structural buyer's volume per price level.
2. Line up the sequence: at each lower price, did they buy more or less? A rising quantity into a falling price is a desk working a size order, not an opinion.
3. Scale the flow against the buyer's income to see whether it's a position or a policy. Convert tonnage to dollars and compare it with the surplus that funds it.
4. If the flow is a fixed share of income rather than a price bet, it is mechanical demand — treat drawdowns as digestion and add.

• Monthly Chinese gold import tonnage vs the trade surplus (the ratio, not the tonnage); central-bank purchase resumption after a scare; the same cadence test applied to any asset with a known structural buyer.

43:33 9. Follow the settlement architecture — infrastructure reveals intent years before prices do

1. When a country states a monetary goal, don't argue about whether it's achievable — go look for the physical plumbing that would be required, and check whether it has quietly been built.
2. Ask what the stated goal's binding constraint is (here: a currency can't be a reserve currency with a closed capital account) and find the workaround the builder has chosen.
3. Follow the two legs separately: what makes the currency worth accepting (a goods base people want), and what lets the surplus holder get out (a convertible asset at a physical hub).
4. Value the consequence in the settlement asset, not the currency: if surpluses get settled in a finite asset, demand is mechanical and price-insensitive.

• New offshore yuan clearing licences in bullion centres; yuan-denominated oil/gas contracts; gold's share of settlement in bilateral trade data — infrastructure announcements as the leading indicator of the price.

45:21 10. Compare reserve assets on issuance and face value, not on yield

1. Reduce each candidate reserve asset to four properties: yield, issuance (finite or infinite), face value (finite or infinite), and who controls the issuance.
2. Under fiscal dominance, treat yield as the least important of the four — an infinitely-issuable claim yielding 4% loses to a finite one yielding 0% once issuance accelerates.
3. Test the implication on the incumbent's holders: who is being recapitalised and who is being expropriated by the shift?
4. Then re-derive the discount rate. If the backstop asset changes, the risk-free rate changes with it — and everything priced off it reprices.

• Net issuance growth of the incumbent asset vs the challenger's stock; official encouragement of household/bank gold holdings; any move that would make gold, not sovereign paper, the collateral of record.

2:13 11. Bottleneck investing — find the constraint, then buy whoever sells through it

1. Aggregate publicly available data to identify developing economic bottlenecks by sector — the choke points a growing demand has to squeeze through. (This is FFTT's stated process, run twice a week, 46 weeks a year.)
2. Prove the bottleneck with a physical series rather than a narrative one — a quantity that has been flat while the claim was growth.
3. Sort the sector by which side of the constraint it sits on: sellers through the bottleneck earn excess returns; buyers of the constrained input don't.
4. When you can't underwrite single names, point at a focused ETF and use its holdings as a ready-made screen for "the type of company I mean."
5. Where no domestic supplier exists, run the same test across countries by elimination — capability, willingness, and whether their index is already something else.

• Physical capacity series that have been flat through a "growth" decade (generation, refining, transformer lead times, machine-tool capacity); ETF holdings lists as pre-built screens for a theme you can't underwrite name by name.

52:13 12. The survival rule — size for the width of the distribution, not the base case

1. Separate the near-term path from the long-term destination and state both out loud: "very bearish in the near term, but ultimately very bullish."
2. Assess how wide the range of possible outcomes has become — not the probability of your view, but the variance around it.
3. When the range is unusually wide, leverage is the binding risk regardless of direction: the correct response is be unlevered, because the bullish decade only pays you if you're still solvent when it arrives.
4. Hold the asset that survives the transition (gold) so that de-levering doesn't mean sitting in the currency being devalued — then buy the dip that the liquidity injection in insight 2 signals.

• Your own margin/notional exposure ahead of the risk-off; the tell that the dip is buyable (the liquidity injection), not merely deep.