Actionable insights — Time To Start Getting REALLY Bullish?

02:29 1. Build the presidential-cycle composite yourself — and trade the dance steps, not the slope

1. Cut the index into four-year chunks starting each November 1 (election time); rebase each chunk to the same starting value.
2. Average the chunks into one composite path; overlay the current term on an offset scale.
3. Judge the fit by the timing of turns ("dance steps"), not by the overall slope — exogenous shocks bend the slope but the steps persist.
4. Note how far ahead/behind schedule the market is running and shift your expected turn dates by that amount.

• The composite's late-Sep/early-Oct low, confirmed by the washout signals in insight 3.
• Shocks on the scale of 1931/1939 — the only conditions that have broken year 3.

09:47 2. Cross-check with a second, independent-looking cycle (the decennial)

1. Repeat the composite with 10-year chunks (he uses the Dow) and read the upcoming year-digit.
2. Treat a "screwy" pattern as a hypothesis to refute with data, not to dismiss by intuition.
3. Recognize overlap: year-6s are often presidential year 2, so the two cycles partly double-count — agreement is supportive, not independent proof.

• Whether both composites keep tracking after the October low.

05:52 3. Inside a seasonal bottoming window, wait for the washout checklist

1. Once the calendar says "bottoming zone," look for: a day with a high put/call ratio, a high VIX, and breadth / breadth-momentum divergences.
2. Require a higher low in the McClellan Oscillator (price makes a low, breadth momentum doesn't).
3. Don't wait too long for perfection — the climb starts from the last of the multiple bottoms.

• A higher oscillator low on a retest of the September lows.

13:22 4. Use a new 3-year A-D high to cap the downside

1. Track the NYSE advance-decline line; flag every new 3-year high.
2. From each flag, measure the worst drawdown in the next 3 months; the typical ceiling is ~10% (average ~4%).
3. List the exceptions and their causes (a pandemic, an abrupt end to QE) — those are the only conditions to fear while the signal is live.

• A sudden Fed liquidity withdrawal (QE5 ending abruptly) — the exception class that breaks the 10% cap.

15:37 5. Keep a separate A-D line for junk bonds as a liquidity canary

1. Maintain a daily advance-decline line for corporate high-yield bonds; they draw on the same liquidity as stocks.
2. A bearish divergence (prices up, A-D down) is a reason an uptrend could end.
3. Run the McClellan Oscillator on that A-D line: the deepest oscillator reading usually comes before the final price low, so an extreme reading signals the worst is being wrung out.
4. Require junk to start improving as confirmation of the next up-leg.

• Junk-bond breadth turning up in October/November.

20:48 6. Normalize margin debt by GDP, then time it with the seven-year cycle

1. Take FINRA monthly margin balances and divide by nominal GDP (the raw arithmetic chart exaggerates).
2. Mark prior peaks; they tend to come shortly before important stock tops.
3. Measure peak-to-peak spacing (~7 years; confirm on longer quarterly data back to 1945) and project the next peak date.
4. A record level before the projected date is a warning, not a sell signal.

• The ratio peaking and rolling over as 2028 approaches.

24:44 7. Read the Fed's balance-sheet slope, not the press conference

1. Plot the Fed's combined Treasury + MBS holdings; track the slope, not just the level.
2. A flattening slope with no announcement is stealth tightening; a turn negative is QT — the fly in a bullish liquidity backdrop.
3. Remember his counterintuitive bond corollary: QE rounds have coincided with falling bond prices.

• Holdings turning down outright (full QT).

29:06 8. Time-shift gold to forecast yields and oil

1. Plot gold and shift it forward ~20.5 months against the 30-year Treasury yield, and ~19.8 months against crude oil.
2. Align the turns; allow slanted alignment lines where timing drifts.
3. Use it for direction and turn timing, not magnitude (gold doubling does not mean yields double).
4. Project the next peak: gold's top date + the lead.

• Whether central-bank-driven gold buying weakens the signal — his open question.

36:54 9. Read producers' hedging in the COT report

1. Pull CFTC Commitments of Traders net positions for commercials in crude futures (mostly producers hedging output).
2. Scale the net short against its own history: a very high net short = producers happy to lock in = topping condition; a low net short = bottoming condition.
3. Compare with price: if price is high but commercials refuse to hedge, they expect higher.

• Commercial net short surging back to extremes — the topping tell.