1. For a priced event, size the non-consensus outcome, not the consensus one
The repeatable method
- Read the market-implied probability (fed funds futures, prediction markets). Above ~85%, the expected outcome is mostly in prices.
- Ask what the surprise would signal, not just what it would do mechanically — a central bank skipping a priced hike implies it sees something worse.
- Hedge (or de-hedge) for the asymmetric side; trim protection that only pays on the consensus outcome.
Here: at ~93% odds, Weiss "took off a little bit of my hedges"; Harrington: a no-hike means "the Fed knows something really bad"; Lebenthal: no hike and "yields are going to spike and crater the equity markets"; Santoli: it would be "an act of hostility" on a market priced 90% for one thing.
Watch for
- Implied probability above 85% with positioning (e.g., CFTC bond shorts) already crowded on the consensus side.
2. Let the long end grade the policy move
The repeatable method
- Before the decision, note the 2-year vs the policy rate (a gap of 100bp+ says the market wants more) and the 10-year level.
- After a hike, if long yields fall, the market is saying one move is enough; if they rise, it wants a campaign.
- Compare with a historical template (e.g., December 2015's single hike then a year's wait).
Here: Liesman: "the two year running over 100 basis points higher than the Fed funds. The market's telling the Fed chair something"; Terranova: if they raise and "yields fall," the market is comfortable with one hike; Santoli: the long end "is going to be the tell."
Watch for
- The 10-year's close on decision day versus its pre-meeting level, and whether the statement uses words like "series" or "campaign."
3. Test whether the policy rate even moves the rates that matter
The repeatable method
- Measure the cumulative policy change over a cycle against the change in the 10-year and mortgage rates.
- If they diverge (cuts with rising long yields), treat policy-rate moves as signalling/credibility events rather than financing-cost events.
- Anchor the equity view on earnings guidance instead; expect range-bound trade until it arrives.
Here: Harrington: seven cuts totalling 1.75pp took the 10-year "from 3.73% to 5%"; with the S&P at ~7,600 both on June 2 and today, she expects it to "yawn and wander" until outlooks come.
Watch for
- Q3 guidance season; a break out of the multi-month index range on outlooks, not the Fed.
4. Treat the first company to quantify a macro cost as the start of a sector read
The repeatable method
- When a company puts a number on an input-cost hit (fuel, labour, tariffs), note the size and whether peers share the input.
- Don't average down on the reporter until the input price stabilizes.
- Screen the most exposed peers and the likely beneficiaries of the same price (producers, refiners, inventory rebuilders).
Here: J.B. Hunt's CFO cut earnings 5–10% on diesel and driver costs; Terranova: "$6 diesel we need to bookmark today… Is this the beginning of hearing the story about earnings degradation across the board?" — won't add JBHT, flags ODFL; Lebenthal takes the other side via XOM, CVX, LNG, FANG, RIG.
Watch for
- Other truckers, airlines and farm-equipment users pre-announcing fuel hits; diesel crack spreads; inventory rebuild once the conflict ends.
5. Add to a winner on a successful retest, not on the headline
The repeatable method
- After a post-earnings drop, wait for a series of higher lows.
- Buy when price tests a widely watched average (e.g., 100-day) and holds, and has recovered the earnings gap.
- Set the thesis as a breakout above the prior all-time high; reassess if the average breaks.
Here: Terranova's sixth AAPL buy: tested the 100-day at $309, "held exactly there," recovered the July loss, target a break above $345.
Watch for
- A close below the 100-day average; product-cycle data (lead times) confirming or denying demand.
6. Discount what AI leaders say about slowing down by what they're spending
The repeatable method
- Note each executive's commercial position (chip seller, model lab, platform) before weighing a policy statement.
- Compare statements with capex and hiring — actions reveal the real race.
- Invest on cash generation and earnings, not on the safety debate.
Here: Harrington: "why are you telling us it's dangerous, but you're still building hand over fist?"; Weiss: "everybody's got a bias… disregard what they're saying. They're all advancing"; her META thesis rests on free cash flow and earnings growth.
Watch for
- Capex guidance at the next earnings round versus public calls for slowdowns or regulation.
Methods distilled from the public CNBC Halftime Report audio episode of 2026-SEP-16 (transcript in transcript.txt). Not investment advice. © CNBC for source material.