1. The tactical breakout trade — buy strength, pause when stretched, add on the event
The repeatable method
- Treat a leadership name as a trade, not a marriage: buy the breakouts as they print new highs (adding on strength, not weakness), and say so — the framing keeps you disciplined about the exit.
- When near-term momentum gets "stretched," stop adding and sit back — don't force size into an extended move.
- Pre-commit the next add to a catalyst (the earnings print), so the pause has a defined end rather than becoming indecision.
Here: Terranova on AAPL at a new all-time high (~$320) — "it's a tactical trade," bought breakouts last ~$310 Thursday, won't buy further with momentum stretched, "after earnings I'd be interested in buying more." His "All Star of 2026."
Watch for
- A name making fresh highs you're buying up into; a momentum reading that's stretched (pause signal); a scheduled catalyst to re-engage on.
2. Read low analyst targets + a low buy-rating as embedded upside, not a warning
The repeatable method
- Pull the 12-month average price target and the % buy-rating for a strong name — and compare the target to the current price.
- When the average target sits at the price and the buy-rating is unusually low for a leader, read it as "embedded skepticism": the crowd of analysts is torn, so the bar to beat is low.
- Position for the re-rating that comes when the company executes and the skeptics are forced to raise targets — the opposite of a crowded, everyone-loves-it setup.
Here: Terranova on AAPL — a ~$320 average 12-month target (right at the price) and only a 61% buy rating vs "generous" targets on Amazon/NVIDIA/Alphabet: analysts have an "embedded skepticism… remarkable to me," which he reads as the upside setup.
Watch for
- A leader whose average target barely exceeds the price; an out-of-line-low buy-rating; analysts openly "torn on the growth rate."
3. Let the valuation come to you — initiate a never-owned name at a multi-year-cheap multiple
The repeatable method
- Keep a quality name you've never owned on a watch list and wait for the valuation — not the story — to reset to a multi-year low.
- Cross-check that the fundamentals are intact or improving as the multiple compresses (growth, margins, cash flow, market share, a moat that's hard to replicate).
- Buy when it's "out of favor" and has stopped going down ("found its footing") — and plan to keep adding on ugly days rather than chasing green ones.
Here: Link's first-ever NVDA buy — 18× forward (cheapest since 2019), ~80% revenue growth, mid-70s GM, FCF doubling, 97% GPU share + hard-to-replicate software; "really out of favor… found its footing," and she'll keep adding on ugly days.
Watch for
- A best-in-class name at its cheapest multiple in years; fundamentals steady-to-improving into the de-rating; price basing rather than still falling.
4. Stick to your style — hold through rotation, don't get wrapped up in the day-to-day
The repeatable method
- Name your style explicitly (price-aware/GARP/value or momentum) and buy only setups that fit it — don't drift between styles as the tape rotates.
- Once you own a thesis name, hold it through the fast rotations (semis↔software, large↔small, cap-weighted↔equal-weighted) — a single-day drop is not new information.
- Anchor the position to a fundamental catalyst you're waiting on, so daily volatility can't shake you out before the catalyst lands.
Here: Lebenthal on MU — bought over three weeks into a ~3% position, "not paying too much attention to today's downturn" (down ~5%); the fundamental bait is the coming Chips-Act anniversary letting Micron return capital to shareholders. "My style is not to get too wrapped up in day-to-day."
Watch for
- A named, consistent style; a thesis name mid-rotation (hold, don't flinch); a dated fundamental catalyst (a policy anniversary, a capital-return trigger) underpinning the hold.
The repeatable method
- Going into a sector's earnings, accept that the stocks usually "trade crummy" around the print even in a good environment — so don't chase them up beforehand.
- Build a ranked shopping list of which names you'd buy, with the specific line items you're underwriting (NII, loan growth, deposits, fee/IB/M&A momentum).
- Favor the laggard with an identifiable, un-priced catalyst over the crowded favorite where expectations have already run — and be wary when a group's expectations flip from low to high in weeks.
Here: Link — "the banks are a buy and they always trade crummy around earnings, so you make your shopping list": NII +9%, loans +7–8%, IB +26%, M&A +9%. Favorite laggard COF (7.5×, ~$2.6B unappreciated Discover synergies) and watch-list WFC (asset-cap lift, NII bottoming); Terranova wary that expectations flipped low→high in 30 days.
Watch for
- A sector that sells off on its own good prints; a laggard with a concrete un-priced catalyst; a group whose expectations spiked fast (caution flag).
6. Trade up in quality on the weaker name's temporary problem — funded by selling the laggard peer
The repeatable method
- Within a pair of direct competitors, identify the higher-quality one (better management, technology/robotics leadership) and treat a temporary, fixable problem there as your entry.
- Fund the buy by selling the lower-quality peer you already own — a quality upgrade, not a net-new bet.
- Separate a fixable one-off (a cyber incident, a single product misstep) from a pattern of "missteps" — buy the former, exit the latter.
Here: Sethi bought SYK (Stryker — "higher-quality, robotics leader") on its current cyber issue, funding it by selling ZBH (Zimmer); separately sold ZTS (Zoetis) as "a mistake — they kept on misstepping." He also bought RACE (Ferrari) down double digits on a fixable EV misstep, "sold out two years."
Watch for
- A quality leader hit by a temporary, fixable problem (buy); a lower-quality peer to swap out of (funds it); a repeat-offender you should exit rather than average down.
7. Fade an over-hawkish Fed — but let the bond market referee the call
The repeatable method
- When a Fed official turns explicitly hawkish, note the market's re-priced hike odds and your own view (e.g. "the high for rates is in").
- Then check the bond market against your view: if the long yield is pushing to new highs, treat that as the market telling you a hike is more real than you think — don't fight both the Fed and the tape.
- Respect which assets get hit first — the longest-duration, highest-valuation names (semis) lead the sell-off — and let the upcoming data (CPI, the Chair's testimony) resolve it before sizing up.
Here: Waller's "any serious policy rule calls for raising the rate" speech pushed September-hike odds 63%→77% (July to 47%) and the 10-yr to ~4.61. Terranova still doubts a hike but concedes "the bond market is pushing up against that… that's why equities are falling" — and the semis led down (MRVL −8%, MU −6%, INTC −6.5%) into CPI + Warsh on the Hill.
Watch for
- A hawkish surprise + jumping hike odds; the long yield confirming or contradicting your rate view; the highest-multiple group leading the sell-off; a near-term data referee (CPI/testimony).
8. Split a stock's slide into fundamental damage vs mechanical/technical pressure
The repeatable method
- When an owned name keeps sliding, separate two questions: has the business thesis broken, or is an external, mechanical force doing the selling?
- Look for a plumbing explanation — here, a glut of new corporate debt issuance widening spreads and raising borrowing costs for every heavy issuer — that hits the stock without changing its fundamentals.
- If the thesis is intact and the pressure is mechanical/temporary, hold (or trim only to a comfortable weight) rather than capitulating at the low.
Here: ORCL at a 52-week low — Sethi holds (sold half earlier), still sees the transition/earnings power, it just "gets no credit"; Lebenthal pins the recent weakness on debt-market indigestion from monumental Amazon/SpaceX offerings (7%+ yield-to-worst) rather than a broken Oracle story.
Watch for
- A persistent slide in a name whose fundamentals you still trust; a mechanical seller you can identify (debt supply, index flows, forced selling); pressure that's temporary rather than thesis-breaking.
Methods distilled from the public CNBC Halftime Report audio episode (transcript in transcript.txt) for personal study. Not investment advice. © CNBC for source material.