Actionable insights — The Momentum Reset and How to Trade It

1. Read the sector-ETF internals to time a momentum-reset entry

1. When a hot group whipsaws violently, don't guess the bottom off price — read the internals of its ETF: RSI (is it back to a neutral ~50 from overbought?), the average constituent's short-window rate-of-change (how deep was the flush?), and the count at fresh highs vs knocked off them.
2. Treat a "cooled but bouncing" reading (RSI ~49, average 10-day change bottomed near −15%, zero new highs but green today) as a reset that hands sidelined buyers an entry — not a breakdown.
3. Accept you won't nail the low; the point is that the risk/reward improves once the froth is out, so a name you liked higher is more ownable lower.

• Sector-ETF RSI resetting toward 50; a deep average rate-of-change trough; the count of constituents reclaiming their moving averages.

2. Price in the leveraged-single-stock-ETF distortion — and sell the premium it creates

1. Recognize a structural fact: a large pool of leveraged 2× single-stock ETFs turns a given dollar of assets into ~2.5× that in notional exposure, so daily rebalancing amplifies moves regardless of any fundamental view.
2. Don't react to the amplified move — treat the wild ±7% swings as noise from holders "with no meaningful point of view."
3. Monetize the volatility instead: the same flows inflate options premium, so sell calls against a name you own to get paid to wait (and lower your effective cost basis).

• Elevated single-stock implied vol / call premium on the popular momentum names; days where the derivative flow, not news, drives the move.

3. Stay true to your knitting — don't switch styles mid-rotation

1. In a fast, day-to-day rotation, the biggest error is abandoning the style you're good at (a momentum investor turning fundamental, or vice versa) to chase the group leading today.
2. Anchor buys to your own discipline: if you're a fundamental buyer and the thesis is unchanged, a lower price is a better entry, not a reason to wait for a signal you can't define.
3. Allow that styles can intersect — the same name can be a value stock and a momentum play at once — but let your framework, not the rotation, drive the decision.

• The urge to buy whatever's green today; a thesis still intact while the price is lower — your cue to add, not wait.

4. Buy best-in-breed on a rare valuation discount (relative-underperformance screen)

1. Screen dominant compounders for the rare window when the stock lags badly while the fundamentals improve — measured by underperformance vs its group and a multiple at a multi-year low.
2. Confirm the business is still winning (market share, margins, free-cash-flow trajectory, earnings revisions still rising) so you're buying a discount, not a value trap.
3. Buy the leader rather than the cheaper also-ran; you're paying up in quality, down in price.

• A dominant name at a multi-year-low multiple with rising earnings revisions; underperformance vs peers as the entry trigger, not the deterrent.

5. In a CapEx super-cycle, ask where the value accrues — own the customer, not the spender

1. Frame the whole theme as one question: of all this capital being spent, where does the profit finally land?
2. Downgrade the spenders — companies handing >100% of cash flow to CapEx get no valuation premium (the market discounts them like industrial conglomerates; they're in the building, not harvesting, phase).
3. Favor the owner of the end-demand / customer relationship, who can layer the new capability on an installed base and collect fees without funding the build-out. Flag the "double-benefit" cases (spender that also owns the product).

• CapEx as a share of operating cash flow; the shrinking Mag-7-vs-493 premium; who owns the customer at the far end of the spend.

6. Answer "is it a buy?" by naming the holding period first

1. Refuse the buy/no-buy question until the time frame is specified — the same stock can be a "no" for one month and a clear "yes" for three-months-plus.
2. For the longer horizon, sanity-check with valuation that already prices the pessimism: a PE below the market multiple and a favorable PEG (price/earnings-to-growth) means you're paid to wait out the CapEx-drought-to-free-cash-flow-reversal.
3. Separate the manager's own turnover from the call — a short-term rebalance out of a name isn't a long-term bear case.

• PE/PEG vs the market and the name's own history; the point where CapEx-heavy free cash flow inflects positive.

7. After an indiscriminate "apocalypse" selloff, stratify the group — real disruption vs "wrong place, wrong time"

1. When a whole sector is sold as a basket on a narrative (here, "AI kills software / cyber"), separate the names with a genuine business-model threat from those just caught in the ETF-level dumping.
2. Look for the company that flips the narrative with data — proving the disruptor is actually a tailwind (more AI → more of what it sells) and backing it with hard metrics (ARR growth, record free cash flow, a rule-of-40 score).
3. Own the proven winner through the fear; use the poster-child laggards as the contrast, not the buy.

• A company reframing the disruptor as a demand driver; rule-of-40, ARR growth and FCF margin separating winners from the sold-off basket.

8. The "best stock in the market" screen — sector waking up + intact long-term trend + still-cheap-for-growth

1. Require the sector to be turning in your favor (financials waking up ahead of earnings) as the tailwind.
2. Prefer a genuine turnaround with self-help catalysts still ahead (spin-offs, exiting regulatory overhangs, "shrinking to grow profitably"), and a valuation still cheap on the new, higher earnings.
3. Manage risk off a durable trend line the stock has "obeyed all year" (e.g. the 200-day) rather than trading the earnings event itself.

• A lagging sector turning; a turnaround with catalysts still pending; a name honoring its 200-day all year as the risk line.

9. Name the enterprise loser — who gets bundled out?

1. In a murky layer where winners are hard to pick, invert the question and identify the clearer losers — a discipline that's often easier and just as actionable (avoid/short).
2. Flag any standalone product whose "good enough" substitute comes bundled free inside a platform customers already pay for — that's structural pricing pressure, not a fixable execution gap.
3. Separately, dock a giant spender whose flagship AI product is mediocre and whose spend shows "no output" — execution risk on top of the CapEx discount.

• A point product duplicated free inside a dominant suite; a heavy AI spender whose own product underwhelms — both flags for the loser side.