1. After a scheduled event, check whether the crowded bet failed to pay 2:02
The repeatable method
- Before a known catalyst (a Fed meeting), note the consensus speculative position (e.g. short Treasuries for a yield spike).
- After the event, check whether the market gave that trade what it wanted. If not, treat it as exhaustion of the crowded side.
- Confirm with price: a gap higher and groups that usually diverge (semis and software) rising together.
- Pair it with sentiment and positioning: defensive positioning plus bearish surveys means buyers are under-invested.
Here: Terranova: speculators went into the Fed short Treasuries expecting a 5¼% 10-year — "you didn't get that" — and the S&P gapped higher Friday into Monday with semis and software up together: "2 words risk on."
Watch for
- Whether the gap holds; the 10-year's reaction to the next data print; CFTC Treasury positioning.
2. When analysts pile on, check the fundamentals they are cutting 10:50
The repeatable method
- Track the survey sentiment (AAII bears) and the sell-side direction (targets and multiples being cut).
- Check the hard data against it: GDP nowcast, jobless claims, retail sales, earnings revisions, credit spreads, breadth of sector earnings growth.
- If the data still rises while opinion falls, buy the volatility; name in advance what would change your mind (the economy rolling over, a Fed that hikes more than expected).
Here: Link: AAII bears at a year high, targets being cut, yet GDPNow 5.1%, claims historically low, estimates rising, multiples down from 22× to 19×, 8 of 11 sectors with double-digit earnings growth — the pile-on is "absolutely the wrong decision." Her exit condition: a Fed that goes "three times" this year.
Watch for
- Forward EPS revisions, GDPNow, weekly claims, high-yield spreads.
3. Plan the seasonal drawdown in advance and buy it 7:59
The repeatable method
- Know the historical pattern for this point in the calendar (here: the drawdown between the August high and the midterm-year low).
- Size the expected drop from history (7% typical, 25% extreme) and decide in advance what you will buy.
- Hold both halves of the market (cap-weight and equal-weight) so you can add to whichever leads.
Here: Talkington: a drawdown every midterm cycle for 6½ decades, followed by the strongest stretch (November–February) — she will "take those draw downs to add to positions on both RSP and your Q holdings."
Watch for
- The S&P's distance from its August high into November; oil, which she treats as the driver of yields.
4. Re-buy a trimmed name when the market starts pricing a new business 18:32
The repeatable method
- Trim into a parabolic move; note the price.
- Re-buy when two things line up: price momentum (a strong one-month move) and a segment the market has not priced (data center, auto, IoT, selling spare compute).
- Accept paying above the low — the signal is the market recognizing the new story, not the cheapest price.
Here: Lebenthal trimmed half of QCOM at ~$230 in May, admits he could have re-bought at $160, and doubles it now on +19% in a month and a data-center business — the same pattern he sees in META (selling excess compute). He then picks MU as another "unloved" chip coming back.
Watch for
- Segment disclosure of data-center revenue; whether the momentum holds through the next earnings report.
5. Measure mega caps against their own multiple history 20:28
The repeatable method
- Compare the current forward P/E with the stock's 10-year (or long-term) average.
- Check that the business is still growing (cloud growth rates) so the discount is not a value trap.
- Add when the multiple is at or below half its history and growth holds.
Here: Link: AMZN at 20× vs. a 44× 10-year average with AWS +37%; NVDA at 17× vs. 35× — she "just recently bought NVIDIA when it was really almost at the lows."
Watch for
- Forward P/E versus the 10-year average; quarterly cloud growth (AWS, Azure, Google Cloud).
6. Sell calls on a range-bound holding until it breaks out 16:53
The repeatable method
- Identify the range and the ceiling the stock keeps failing at.
- While it stays under the ceiling, sell covered calls above the range to earn income on a stock you keep.
- Stop selling calls once it closes through the ceiling — that is when the upside you would give away is largest.
Here: Talkington on NVDA: range-bound 210–230 under a ~232 ceiling; "until it can break out, I would just sell calls" — a pattern that "has been very, very solid for the last six months." Her UBER final trade uses the same range logic (buy 70, target 78).
Watch for
- A close above ~232 on volume.
7. Judge a rally by what leads it 40:37
The repeatable method
- On an up day, check whether cyclicals and equal-weight join in, or only mega caps with defensive qualities lead.
- Check what else explains it (oil and yields falling).
- Treat a narrow, macro-driven rally as tactical: respect the low it made, but reserve judgment on a new trend.
Here: Santoli: the leadership doesn't "represent genuine risk" — it is the Mag 7 plus semis coming back "first in, first out," lifted by oil and yields; last week's post-Fed low is "meaningful on a tactical basis." Terranova adds that summit optimism may already be priced.
Watch for
- RSP vs. SPY; the MAGS ETF's record highs; oil and the 10-year into the Xi–Trump summit.
Methods distilled from the public CNBC Halftime Report audio episode (Spotify transcript) for personal study. Not investment advice. © CNBC for source material.