Actionable insights — Trading the Tech Wreck: NASDAQ's Rough Week

1. The second-/third-derivative beneficiary screen — buy the supplier, not the spender

1. When a theme's "obvious" stocks get crowded and punished, stop asking "who is the theme?" and ask "who gets paid by the theme?" — the second-derivative beneficiary (the supplier), then the third (the supplier's suppliers, and the back-office processes the theme transforms).
2. Confirm the misallocation with a relative-performance tell: the spender down, the supplier up, on the same news. If the market is already rewarding the supplier, the rotation is real, not a forecast.
3. Look past the headline use-case to the whole value chain the theme touches — including the boring operational savings (fewer engineers, faster processes) that drop straight to margin.

• Spender-down / supplier-up pairs on shared news; supplier order backlogs and "queues"; opex-collapse disclosures (process time, headcount) inside the customer base.

2. Own the index when you can't pick the winner — let the basket rotate for you

1. When the leaders of a theme are stumbling but you don't know which (if any) will recover, buy the cap-weighted index instead of guessing — its weights shift automatically toward whatever inside it is working.
2. Decompose the index's return: if it's up while its famous members are down, something less obvious has grown into a bigger weight and is doing the heavy lifting — own that drift rather than fighting it.
3. Use the active-manager failure as confirmation: if stock-pickers without that hidden exposure are badly trailing the index, the basket is the edge.

• Index up vs marquee members down; rising index weight of a non-obvious sub-group; active-vs-index dispersion.

3. Separate the business from the stock — own the lender, sell the manager

1. When a sector is hit by a sentiment/flows scare, split the question: are the underlying assets impaired, or only the stock's sentiment? Treat those as two different trades.
2. If the cash-flows are sound ("money-good") but the equity is being sold on fear, keep the cash-flow vehicle (the high-yield instrument) and exit the sentiment-sensitive equity (the manager whose multiple de-rates on flows).
3. Respect the tape: even with great fee growth, a broken chart + negative flows means "more sellers than buyers" — don't fight it on fundamentals alone.

• Redemption headlines; the loan portfolio's credit marks vs the manager stock's chart; flows into the traditional banks as the rotation destination.

4. The marginal-buyer test — when everyone "knows," the risk is who's left to buy

1. For a consensus, multi-year theme (e.g. "the buildout lasts forever"), don't debate whether it's true — ask who is the marginal buyer once the whole market already believes it.
2. If the base case is in "everybody's thinking, every conversation," demand is largely spent; that alone "adds to the risk substantially," independent of fundamentals.
3. Reframe the timing question correctly: it isn't when the cycle ends, it's how far in advance the market discounts the end — and that lead time is unknowable, so trim into strength rather than wait for the top.

• Universally-held theses with no skeptics; valuation re-rating far above history; the index already de-rating the leaders while sell-side targets keep rising.

5. Discount conference bullishness — one data point isn't a signal

1. Treat sell-side conferences as a biased sample: no CEO gets on stage to caution on their own stock, so unanimous bullishness is the base rate, not information.
2. Don't upgrade conviction on "the queues are long" alone — it's one self-interested data point; weight it against price, positioning and the marginal-buyer test.
3. Extract the verifiable takeaways (backlog figures, process/headcount changes) and discard the mood.

• Unanimous management optimism; cross-check the bullish anecdote against hard backlog/TAM numbers before sizing up.

6. The capital-starvation screen — buy what was starved, then layer a catalyst

1. Hunt for sectors that have been starved of capital — years of no new supply ("no new builds") — because under-investment tightens future supply and lifts the returns on existing assets.
2. Quantify it: rising cap rates (the yield a property/asset throws off) from ~4% to 6–8% means you're paid more to own the same asset — the math improved while nobody was looking.
3. Then layer a fresh catalyst on top (here, AI cutting operating/documentation burden) — but temper the upside if the group has already moved; a starved sector that's re-rated gives steady gains, not a pop.

• Sectors with multi-year capex droughts; cap-rate expansion; a new efficiency/demand catalyst arriving after the starvation — and whether the move has already happened.

7. Exit when the hedge stops hedging — judge an asset by its job, not its story

1. Own a "hedge" only as long as it does its job. Define the job (rise when inflation/fear rises) and check it against behavior, not narrative.
2. If it behaves perversely — falling while inflation ran, rising only as rates fell — the thesis is broken; a no-yield, sentiment-driven asset you can't value has no claim on the portfolio once it fails.
3. Use the flows/level as the exit trigger: when the momentum money that drove it in starts coming out (a key level breaks), step aside and raise cash rather than average down.

• A hedge correlating the wrong way with its driver; a breached level where momentum buyers turn into sellers; rising cash as the patient stance.

8. Don't chase a hot first-year IPO — let lock-ups and a few quarters clear

1. For a richly-valued name that "succeeded mostly in the private market," anchor on what already happened privately: a 6× move since the last tender vs only +33% revenue means the easy gains were captured pre-IPO.
2. Respect the first-year-IPO base rate: many great companies (Facebook, Uber, Airbnb) drew down ~50% in year one; you rarely need to own day one.
3. Plan for the lock-up overhang — VCs starved of liquidity will sell when allowed — and simply watch how it trades for a few quarters before committing.

• Pre-IPO valuation vs revenue growth; lock-up expiry dates; VC selling pressure; the first few quarters of public trading before sizing up.

9. Use a single flow gauge as a binary risk-off trigger

1. Pick one clean, observable proxy for a crowded theme's fund flows and define a hard level that flips you risk-off — so the decision is mechanical, not emotional.
2. Pair the level with the flow data behind it (record outflows after record inflows) to confirm the level isn't noise.
3. Below the line, default to defense for the group (raise cash, cut overweight) until it reclaims.

• The proxy ETF's trigger level; outflow records following inflow records; the equal-weight-vs-cap-weight spread as corroboration.