Actionable insights — breadth, seasonality and the two RSI clocks

1:18 1. Use technicals to define where you are wrong — the falsifiability test

1. State the view as a level, not an opinion. Before entering, write down the price at which the thesis is disproven — the breakout that failed, the trendline that broke, the moving average lost.
2. Note that this is precisely what a fundamental view cannot supply. "Too high" or "too low" has no invalidation price; a chart level does. That is Newton's stated case for the discipline: people upset about fundamentals "should utilize technicals more to understand exactly where we are and where we could go and where you're wrong."
3. Match the tool to the horizon and say the horizon out loud — his is "3 to 6 weeks at a minimum, but hopefully longer." A call without a horizon cannot be scored.
4. Keep the inputs few and consistent across every asset: price action, momentum, volume, sentiment, seasonality, cycles — applied identically to equities, commodities, currencies and treasuries so the process transfers.
5. Separate your work from the house macro view rather than blending them. Newton runs the tactical clock; Tom Lee's fundamental, top-down work runs the intermediate one. Two independent processes disagreeing is information; one process contaminated by the other is not.

• Your own calls that lack an invalidation price — those are opinions, not positions; a thesis that survives every price is untestable; the horizon quietly stretching after a call goes against you (the classic tell that a trade has become an investment).

8:59 2. Run breadth as a standing early-warning system — the index cannot tell you it is hollow

1. Accept the premise first: leadership by a handful of names is normal, not a defect — "that's been the case in US markets along with most foreign markets for the last 100 years," Exxon and DuPont then, hyperscalers now. So narrowness alone is not a sell signal; deteriorating narrowness is.
2. Maintain a fixed dashboard and read it in the same order every time: % of stocks above their 20-, 50- and 200-day moving averages (the primary), the advance-decline line, the McClellan oscillator and the summation index.
3. Treat the % above the 200-day as a trend variable, not a level: the number matters far less than its slope. Rising through 50 toward 60 from a 40 base is the bullish configuration.
4. Ask the one question the index cannot answer: is the number of stocks fuelling this move rising or falling while the index makes new highs? "Just because the S&P is moving to new highs does not mean that the broader market is."
5. Cross-check with the broadest advance-decline line available (he uses the Russell 3000, ~98% of investable US market cap). An A/D line at new highs alongside the index is confirmation; an A/D line diverging while the index rises is the warning.
6. When breadth rolls over ahead of price, cut risk before the index confirms — that is the entire purpose of the dashboard: "that will give you an absolute early warning as to when stocks should start to falter and fall."

• The % above the 20-day rolling first (it leads the 50 and 200); the Russell 3000 A/D line failing to confirm a new index high — that divergence is the actual sell trigger; the summation index turning down while the S&P grinds up; breadth deteriorating specifically while the mega-cap weights carry the index, which is the 2021 template.

5:21 3. Run momentum on two timeframes so a bullish trend and a cautious entry can coexist

1. Compute RSI on both the weekly and the daily chart before forming a view. They answer different questions: weekly = is the intermediate trend exhausted? daily = is the entry stretched?
2. Map the four combinations and act only on the diagonal. Weekly not overbought + daily overbought = own it, but do not add here — the trend is intact and the entry is poor. Weekly overbought + daily oversold = the opposite: don't chase strength, the trend itself is tired.
3. Convert the reading into a calendar statement, not a directional flip: "we're nearing levels where I think we will start to slow down and pause. That likely happens in the month of August."
4. Use the same split to resolve the classic contradiction of a strategist being "bullish but not buying" — the two clocks let both be true without hedging language.
5. Confirm with the corresponding trend structure: momentum "sloped to the upside" on the weekly is the permission to stay long through a daily-timeframe pause.

• Weekly RSI actually crossing into overbought — that is the flip from "pause" to "trim"; a daily reset (RSI back toward the middle) while the weekly holds up, which is the add signal; monthly momentum for the assets where he uses it (crypto), where the sign was the deciding vote.

7:24 4. Put the buying window on the calendar — midterm-year seasonality and the 4-year cycle

1. Before forecasting, locate the year within the 4-year political cycle and use the matching seasonal pattern rather than the generic one. "It's a tricky time usually in the third quarter, specifically of a midterm election year."
2. Identify the historically weak stretch (Q3 into October of a midterm year) and treat it as a scheduled drawdown to be bought, not an unscheduled risk to be feared.
3. Name the window in advance and size for it: any pullback "from October into November is probably going to be one of the better buying opportunities we've seen throughout the entire 4-year cycle." A named window forces the cash to exist when it arrives.
4. Set the expectation for the interim honestly — "backing and filling," a market where "everything works all at once" only "after the midterms." That framing prevents a chop-driven exit.
5. Layer the asset-specific seasonals on top rather than substituting for them: gold's August–October window, and the crypto four-year drawdown rhythm below.
6. Never let seasonality override price. It sets the when to expect, while breadth and momentum decide whether it is happening.

• The seasonal weakness failing to appear — a market that refuses to correct on schedule is stronger than the model and forces a higher-priced entry; the pullback arriving with breadth deterioration (that is a different, worse setup than a seasonal dip); the post-midterm broadening actually showing up in the A/D line.

10:50 5. Screen sectors on the ratio chart — a multi-year relative downtrend breaking is the entry

1. Chart the sector divided by the index, not the sector alone. In a rising market almost everything rises; only the ratio tells you where money is actually rotating.
2. Look specifically for a ratio that has been falling for years and has just broken its downtrend — the longer the base of underperformance, the larger the pool of underweight investors who have to buy back in.
3. Identify the identifiable, dateable cause of the underperformance and ask whether it is spent. Here: drug pricing pressure and the removal of ACA subsidies — policy shocks that hit earnings once and then stop hitting them.
4. Demand internal breadth before believing the breakout: the move must run across sub-industries with different business models. Biotech, pharma and HMOs rising together is rotation into the sector; one sub-group moving is a story about one sub-group.
5. Prefer sectors with a structural argument the flows can keep feeding (demographics, in healthcare's case) so the relative trend has somewhere to run.

• The ratio falling back below the broken downtrend (a failed relative breakout is a fast, clean exit); breadth inside the sector narrowing to one sub-industry; a fresh policy shock that re-dates the cause; and, on the other side, the same test turning up on a laggard region (China) or a rolled-over one stabilising (Latin America).

13:27 6. Diagnose a sector decline by its sequence — one-group-at-a-time is repair, all-at-once is a break

1. When a large sector weakens, do not ask "how far did it fall?" — ask "did it fall together or in turn?" A simultaneous decline across every sub-group means a systemic repricing; a rolling decline means capital rotating within the sector.
2. Write down the order of the dominoes and the date each one topped. Newton's tech sequence: software (late last year) → semiconductors and semi-cap equipment → memory ("the final shoe to drop").
3. Then track the recovery in the same order and use the earliest group as the leading indicator: software "has since bottomed. We've seen software start to push off the lows" long before memory did.
4. Declare the repair complete only when the last domino turns — including the mega-cap complex: "one by one all of these areas have stabilized including the mag seven, the hyperscalers."
5. Confirm the final leg on a relative basis against an equal-weighted version of the sector, which strips out the index-weight distortion: memory "has broken back out versus equal weighted technology."
6. Then find the cleanest listed expression of that last leg and buy the vehicle rather than the individual name — here the country ETF whose index is dominated by memory manufacturers.
7. Temper the timeline: a group that fell this far repairs slowly — "it's not going to be a straight shot back to the highs but I think it will take time."

• Memory losing the relative breakout versus equal-weighted tech — that would put the last domino back down; the Nasdaq stalling at 27,100/30,700 as forecast (a normal pause) versus breaking back below the triangle (a failure); the sequence restarting from software, which would mean a genuine sector-wide repricing rather than rotation.

6:22 7. Trade the consolidation resolution — a month-long triangle that finally gives way

1. Identify a triangle consolidation: a range whose highs and lows compress toward each other over weeks, meaning buyers and sellers are converging on a price and volatility is coiling.
2. Note how many times the boundary has been "attacked." Repeated tests exhaust the sellers sitting on that level, so the eventual break carries more force.
3. Take the break in the direction of the prevailing trend as the confirmation, and require the other major index to break at the same time. A single-index break is noise; the S&P and Nasdaq clearing together is a market event.
4. Cross-check against the macro tape at the moment of the break — Newton's confirmation was benign CPI/PPI, retreating yields, and participation from non-tech sectors.
5. Then immediately re-apply the timeframe test (insight 3) and the resistance map (insight 6): "short-term very constructive" does not mean the move runs unimpeded to the old highs.

• A failed break — price falling back inside the triangle within days is a reversal signal, not a pause; one index breaking without the other; the break happening on deteriorating breadth, which would make it a trap rather than a start.

21:48 8. The yield-curve steepening playbook for financials — and the growth-vs-term-premium test

1. Watch for the specific configuration: front end anchored, long end rising. Here the anchor was a policy change — Warsh "basically eliminated the prospects for forward guidance," so the short end stopped moving while the long end drifted up.
2. Translate it into the business model rather than the narrative: "banks borrow short, lend long… it's very helpful for the net interest margins." The trade works through arithmetic, not sentiment.
3. Check the historical analogues before sizing — 2013's taper tantrum, post-2016 — "whenever we see a big steepening in the yield curve and rates start to creep up, that's generally been very constructive for financials."
4. Run the decisive quality test: is the long end rising on growth expectations or on term premium? Growth-driven steepening is a genuinely good environment for banks (loan demand and credit quality both improve); pure term-premium steepening carries a fiscal or inflation worry that eventually hits everything. "It's not just because of term premiums rising… I think it's also because of growth expectations."
5. Sequence the exposure within the sector: capital-markets banks lead (deal and trading activity turns first), then the deposit-funded lenders break out as the margin actually widens.
6. Use an international read as corroboration — European banks "on an explosive tear for more than a couple of years" is the same mechanism running ahead of the US.

• The curve flattening again (a Fed that resumes forward guidance, or long yields falling on growth fear) — that removes the entire thesis; steepening that turns out to be all term premium, which is a warning not a tailwind; credit quality deteriorating as 6.7% mortgages and higher corporate borrowing costs bite; the breakouts in the lender names failing.

18:40 9. Classify a commodity spike as supply or demand before letting it change your macro view

1. When a commodity moves violently, attribute the move before forecasting anything downstream from it. Ask: is a physical route/production blocked (supply), or is consumption genuinely growing (demand)?
2. A supply shock is self-limiting — it reverses when the physical constraint clears, so it should not be extrapolated into a permanent inflation regime: "any sort of move in crude is 100% supply shock. I don't sense it is going to prove to be long-lasting."
3. Forecast the path, not the level, and hold both legs at once: $82 → ~$100 over one-to-two months, then ~$50 by year end once the strait reopens. A two-way path is a legitimate answer; forcing a single direction is not.
4. Now propagate the classification through every dependent variable: temporary crude → term premium up then down → long yields up then down → the Fed should not hike ("I don't think the Fed should be hiking") → the headwind on gold and crypto lifts on the same schedule as the buying window.
5. Cross-check the market's own inflation forecast rather than the commentary: "most break evens have been plummeting" — if breakevens are falling while the commodity spikes, the market already agrees the shock is temporary.
6. Add the policy lag as a final filter: a Fed move takes "about 12 to 18 months to filter through the system," so responding to a shock that resolves in three months is guaranteed to be mistimed.

• The strait deal failing to materialise — a supply shock that persists stops being temporary and re-rates the whole chain; breakevens turning up (the market disagreeing with the temporary read); crude holding above 100 past the one-to-two month window; the Fed hiking anyway, which would tighten into a shock that is already resolving.

24:14 10. Metals: trade the seasonal window, respect real rates, and treat RSI-90 plus late enthusiasm as the top

1. Start from the seasonal window — "normally, August through October, we get very good gains" in precious metals — and check whether the larger cycles have bottomed (his did, in June).
2. Then apply the single biggest offsetting variable: real interest rates. Gold pays no income, so a high real yield on safe bonds is a direct competing offer. "Historically, it's been a tricky time to own precious metals when [long rates creep higher]… we're seeing real rates at almost former highs."
3. Where seasonality and rates conflict, forecast a capped rally rather than picking one: to ~4,600, then "sort of a final shoe to drop."
4. Refuse the first bounce off a low as an entry — "I really don't trust this first little bounce off the lows, and I'm much more of a buyer on weakness." Buy the retest, not the rebound.
5. For tops, use the two-part signal that called the January peak: an RSI near 90 (a rarity, "one of its more overbought levels of all time") arriving together with broad public enthusiasm after years of the move being ignored. Late-arriving optimism is the confirming half — "very few people recognized that metals were in a big bull market up until about 3 to 6 months prior to the peak."
6. Locate the position in the cycle before making the long-term call: typical ~6-year cycles, "almost a 4-year rally" against only "at least a 7-month pullback" — an incomplete correction, hence patience.
7. Discount central-bank buying as a timing input. It is present and supportive ("the central banks have started to re-engage and are buying now") and it did not prevent the 5,000 → 3,500 bust.

• September weakness actually arriving (that is the stated buy trigger); real rates rolling over, which would flip the metals environment early — "if rates start to come back down, gold will certainly rally"; a 4,600 print followed by momentum not deteriorating, which would break the "final shoe" thesis; the same RSI-90-plus-euphoria combination reappearing at the next high.

29:11 11. Price crypto off the 4-year drawdown cycle and liquidity — then buy the sentiment flush, not the level

1. Anchor on the cycle before the chart: drawdowns of this magnitude "normally you do have every 4 years… 2022 or 2018, 2014." A halving from the high is the historical norm, not evidence of a broken asset.
2. Check monthly momentum, not daily — the timeframe that matches the cycle. Still "very negative" means the bottom is not in regardless of any weekly bounce.
3. Test the primary macro driver: liquidity. "Bitcoin tends to trade well when you have an increase in liquidity, and if anything, we're seeing almost the opposite right now" as long rates rise. Tie it explicitly to the rate call so the two forecasts stay consistent.
4. Dismiss an unconvincing rally as evidence in itself — the bounce off the June lows "has been very uninspiring to me." A weak rally inside a downtrend is information, not a bottom.
5. Check positioning and the crowd's alternative: enthusiasm has migrated ("a few years ago it was right to buy Bitcoin. Now they're buying memory stocks"). Money leaving for a hotter theme is a late-stage, not a bottoming, condition.
6. Discount the pending catalyst honestly — the Clarity Act is "probably unlikely to be signed before the midterm election," so the non-technical case has no near-term trigger.
7. Define the buy by sentiment and time, not by price alone: a bottom "in about 2 months," at a moment "when sentiment should start to turn a lot more negative" — the flush is "really something you want to buy into."
8. Split the recommendation by holding period explicitly: with "a 2-year 3-year perspective, then sure you can own crypto here, but if you have a 1-month perspective, I think it's probably better to hold off."

• Monthly momentum turning positive — the single condition he says is missing; liquidity improving as long rates peak (his own rate call is the trigger); sentiment surveys and funding turning decisively negative, which is the buy signal not the risk; a hold above the 57k prior low that refuses the undercut, which would mean the flush arrived early and the cycle low is already set.

33:39 12. Trade the dollar through its components, and follow the repatriation chain to its end

1. Never forecast "the dollar" as a single thing. It is a basket — decide which legs are doing the work. Newton expects the yen to strengthen and the dollar index to rise anyway, which forces the conclusion: "pretty good weakness in the euro and pound sterling as the dollar bounces."
2. Find the policy divergence driving each leg. Japan: intervention plus a prime minister publicly pressing the central bank, so "about a 75% chance that the Bank of Japan hikes rates in September" — near-term yen strength to ~150 on dollar-yen.
3. Then separate the tactical move from the structural one. Japan's "process of gradualism following almost a couple decades of deflation" keeps its rates far below America's, so after the hike "it's going to be right to sell the yen again" — 175–180 dollar-yen.
4. Follow the flow-of-funds consequence to other markets, which is where the real information is: rising JGB yields make domestic Japanese assets competitive again and pull Japanese savings home. "There's a huge rate repatriation theme happening… it really has global implications" — a large, decades-old bid leaving foreign bond markets.
5. Close the loop with the equity implication: a dollar that bounces to 102.5 and then "starts to roll over probably into next year" is "a good sign for emerging markets" — which is what makes the EM rotation (insight 5) actionable on a specific timeline.

• The BOJ not hiking in September (the 75% failing) — dollar-yen would run rather than dip; DXY failing well before 102.5, which would pull the EM rotation forward; JGB yields rising far enough to accelerate repatriation, which would pressure US and European long bonds independently of domestic data; euro/sterling strength breaking the "which leg" assumption.