Know the regime.
Trade only what fits.

🔄 Chapter 5. Stovall's Sector Rotation — Mapping the Economic Cycle to Leadership MARKET STRUCTURE

🎯 The core belief

Sectors do not lead the market in random order. They lead in a predictable sequence that tracks the phase of the underlying economic cycle, because different sectors respond to interest rates, credit conditions, corporate earnings, and consumer confidence at different points in the cycle. Sam Stovall's 1996 framework — refined at S&P over three decades — codified this into a repeatable rotation model. It is not a market-timing tool. It is a context tool: given where the economy is, which sectors should be leading, and if they are, that confirms the cycle read; if they are not, something is off and you should demand an explanation before you position.

🌀 The six stages of the cycle

Stovall breaks the business cycle into six stages, each of which favors a distinct group of GICS sectors. The transitions are not clean — they overlap by weeks or months — but the sequence is remarkably stable across the post-war US data.

🧠 Why this actually works — the mechanism

The Stovall rotation is not folklore. It falls out of three underlying economic mechanisms operating on different lags:

1. Monetary policy transmission — when the Fed cuts, the first sectors that benefit are those where the cost of capital or the discount rate is the dominant valuation input. Banks widen net interest margins on a steepening curve; consumer discretionary benefits from mortgage and auto financing; long-duration tech gets a valuation lift from lower discount rates. This is why Financials, Discretionary, and Tech turn first off a cycle bottom.
2. Earnings expansion — once demand actually recovers, industrial capacity gets restocked. Machinery orders (Industrials), raw material inputs (Materials), and eventually energy demand (Energy) all get funded from operating cash flow and expanding credit. This is why the "middle cyclicals" lead the mid-expansion.
3. Peak-cycle defensive rotation — as growth starts to slow but investors are still fully invested, the highest-quality stable-earnings names bid up. Investors are not selling to cash; they are rotating within equities toward Staples, Healthcare, and eventually Utilities. This is the tell that a cycle is aging.

🛰 How the Bulls N Bears Stovall Radar operationalizes this

The platform's Stovall Radar panel runs a nightly job that computes the trailing 60-day and 200-day relative strength of each of the 11 GICS sector ETFs versus SPY. The top-3 leaders are cross-referenced against the six-stage table above to infer the current cycle stage. When leadership is clean — e.g. XLF, XLY, XLK all in the top 4 — the radar prints "Stage 1: Early Recovery" with high confidence. When leadership is mixed — e.g. XLE and XLU both in the top 4 — the radar flags "transition or regime conflict" and asks the operator to reconcile with fundamentals. The point is not to be a black-box prediction; it is to force explicit reasoning about whether the tape is consistent with the macro read.

⚠ Honest limits — where Stovall breaks

• Unconventional monetary policy (2008–2020) — QE, ZIRP, and forward guidance compressed the cycle transitions and distorted normal sector responses. XLK led for a full decade regardless of stage, driven by secular software adoption and TINA (there-is-no-alternative) flows into growth. A pure Stovall model underperformed a passive S&P position over that stretch.
• GICS classification drift — Technology as defined by GICS in 2026 is a different beast than in 1996. FAANG-era communication services, cloud infrastructure, and semi cap equipment all live in different GICS boxes but behave as one economic complex. You have to reason about the underlying business, not just the ticker's sector code.
• Non-cyclical shocks — COVID (March 2020), the 2022 rate shock, and geopolitical events (energy in 2022) all produced sector moves that had nothing to do with cycle position. Stovall is silent on exogenous shocks.
• Not a timing tool alone — the model tells you which sectors should be leading; it does not tell you when to enter or exit individual names. Combine with technicals (Levels 5–9 · Technical Analysis), momentum ranking (Level 23 · Momentum), and risk management (Level 11 · Money Management).

🔗 Cross-references

See Chapter 6: The 11 GICS Sectors for the composition of each sector and what each contains. See Chapter 13: Sector Rotation Panel for the live application. See Chapter 34: Momentum & Relative Strength for the RS ranking methodology the radar uses under the hood.

🏭 Chapter 6. The 11 GICS Sectors — What Each Contains, When They Lead, What Moves Them MARKET STRUCTURE

🎯 The core belief

Every stock you'll ever trade lives inside one of 11 sector boxes, and the box matters more than most retail traders realize. Sector context tells you whether a stock is running with its neighbors (a real trend backed by the whole cohort) or against them (either you're early and brilliant, or you're wrong and don't know it yet). Sectors also rotate on predictable timescales tied to the economic cycle — this is what the Stovall framework (Level 2 · Stovall) formalizes. If you know the sector weights, the leadership order, and the key economic sensitivities cold, you'll stop trading single names in isolation and start trading them in their neighborhood.

📊 The big reference table

Weights are approximate 2025 S&P 500 sector weights per S&P Dow Jones Indices; they drift by 1-2% quarter to quarter.

🔍 Sector-by-sector deep dive

XLK — Technology (~28%)

The gorilla. AAPL, MSFT, NVDA, AVGO, ORCL alone are more than half the sector's weight. Sub-industries include semiconductors (NVDA, AVGO, AMD), software (MSFT, ORCL, CRM, ADBE), hardware (AAPL, DELL), and IT services (ACN, IBM). The economic story: tech is a long-duration cash flow story. When rates fall, tech multiples expand; when the 2Y yield spikes, tech gets discounted more aggressively than any other sector — 2022 was the textbook example, when XLK dropped 28% as the 2Y went from 0.7% to 4.7%. Tech leads early-to-mid cycle when growth is scarce and investors pay up for it. Wrinkle: GICS reclassified in 2018 — GOOGL, META, and NFLX moved from Tech to Communication Services, so XLK is now a purer semis/software/hardware bet than most retail traders assume.

XLF — Financials (~13%)

JPM, BAC, WFC, GS, MS on the banking side; BRK.B is the single biggest name; V, MA, and AXP are payments; SPGI and MCO are ratings. Financials trade on the yield curve — banks borrow short (deposits) and lend long (mortgages, commercial loans). A steep curve (10Y minus 3M) means fat net interest margins; a flat or inverted curve (like 2022-2024) crushes bank profitability. Financials also amplify credit-cycle stress: when HYG rolls over, XLF is usually right behind. Leads in early cycle when the Fed is cutting and the curve is steepening. Watch regional bank ETFs (KRE, KBE) separately — they carry more commercial real estate risk than the money-center banks.

XLV — Healthcare (~12%)

UNH is roughly 10% of the sector by itself; LLY and JNJ round out the top three. Sub-industries: pharma (LLY, PFE, MRK), managed care (UNH, ELV, CI), biotech (AMGN, GILD, VRTX), med devices (ABT, MDT, ISRG). The story is defensive — people take their statins in a recession. Drivers are pipeline news, FDA decisions, and election-year drug-pricing rhetoric. XLV historically outperforms in late-cycle and recession phases because earnings hold up when consumer discretionary craters. It has one of the lowest betas to SPX of any major sector (~0.7). Correlations to bonds are weak, so it's a real diversifier — not a bond proxy like XLU.

XLC — Communication Services (~9%)

The Frankenstein sector created in the 2018 GICS reshuffle. GOOGL and META are half the sector. Then telecom (T, VZ, TMUS), entertainment/media (NFLX, DIS, WBD), and interactive media/gaming (EA, TTWD). The reality: XLC trades much more like Tech than like the old "Telecom" sector because of GOOGL/META dominance. Drivers are ad spending (GOOGL, META), subscriber trends (NFLX, DIS), and dividend yield for the telecom cohort. Wrinkle: if you're long META, you're really long XLK dressed up in XLC clothing — the correlation to QQQ is much higher than to VZ.

XLY — Consumer Discretionary (~10%)

AMZN and TSLA alone are about half the sector — the biggest concentration issue in GICS. Then homebuilders (DHI, LEN, NVR), retailers (HD, LOW, TJX, NKE), restaurants (MCD, SBUX, CMG), and autos beyond TSLA (GM, F). The economic story is pure cyclical: consumer confidence, real wages, credit availability. Leads in early-cycle recoveries. XLY is the sector most punished by rising mortgage rates because homebuilders are heavily weighted here. Wrinkle: AMZN has AWS, which is really Tech infrastructure. When retail traders panic that "consumer is dying" they short XLY, but a third of AMZN's operating income is cloud, which decouples it from the consumer story.

XLI — Industrials (~8%)

Aerospace/defense (BA, RTX, LMT, GD), machinery (CAT, DE, ETN), transports (UPS, UNP, CSX), and engineering (ROP, EMR, HON). The economic pulse of the market — if XLI is leading, the real economy is expanding. Drivers are ISM Manufacturing PMI (>50 = expansion), capex, global growth, and defense budgets. Defense stocks decouple during geopolitical events (2022 Ukraine invasion sent LMT and RTX up sharply while the rest of XLI stalled). XLI leads early-to-mid cycle and rolls over first when PMI dips below 50. Watch DJT (transports) and COPX (copper miners) as leading indicators for XLI itself.

XLP — Consumer Staples (~6%)

PG, KO, PEP, WMT, COST, PM, MO. Food, beverages, household products, tobacco, and yes — Walmart and Costco (which are staples-retailers, not discretionary). The story is defensive: people buy toilet paper and toothpaste in every economy. Drivers are input costs (grains, sugar, packaging), FX (many are multinationals — a strong dollar hurts KO and PG's reported earnings), and pricing power. XLP outperforms in late cycle and recession. Beta to SPX is ~0.6. Correlates positively with XLU and XLV as the classic "defensive rotation" cohort.

XLE — Energy (~4%)

XOM and CVX are half the sector. Then E&P (COP, EOG, OXY), refiners (MPC, VLO, PSX), and oilfield services (SLB, HAL, BKR). XLE is a levered play on WTI crude — the rolling 60-day correlation is typically ~0.75. When crude rips (Feb-June 2022: WTI from $88 to $122), XLE was up 60%; when crude craters (2020 pandemic), XLE was the worst sector in the S&P. Drivers are OPEC+ decisions, US shale rig counts, geopolitics (Middle East, Russia), and global growth. XLE leads late-cycle when inflation is running hot and is the best inflation hedge inside equities. Doesn't correlate much with interest rates directly — commodity price is the whole game.

XLU — Utilities (~2.5%)

NEE, DUK, SO, AEP, CEG, SRE. Regulated electric, gas, and water utilities plus a growing renewables/independent-power cohort (CEG has been a monster on the AI-datacenter power thesis). The economic story: utilities are essentially bond proxies with a modest growth kicker from rate-base expansion. They pay 3-4% dividends and their allowed returns are regulated. Because their cash flows are long-duration and stable, they trade inversely to long rates — the rolling correlation with 10Y yields is around −0.60. XLU leads in late-cycle and recession when investors bid up defensive yield. In 2022, when TLT dropped 30%+, XLU dropped in sympathy despite being a "defensive" sector — because it's really a rate trade.

XLRE — Real Estate (~2.5%)

Split out from Financials in the 2016 GICS reshuffle. Data-center REITs (EQIX, DLR), industrial (PLD), residential (AVB, EQR), telecom towers (AMT, CCI), retail malls, healthcare facilities. XLRE is even more rate-sensitive than XLU because cap rates on real estate move directly with the 10Y. When TLT rips, XLRE rips; when yields spike, XLRE crumbles (2022 was brutal — down 26%). Drivers are long rates, occupancy trends, and sub-industry rotation (data centers vs. malls tell completely different stories). Leads in late cycle and early recovery when rates start falling.

XLB — Materials (~2%)

The smallest sector. LIN (industrial gases) is nearly a quarter of it. Then SHW (paints), FCX (copper), NEM (gold miner), APD (industrial gases), and DOW/DD (chemicals). The story is commodity-driven cyclical: XLB trades on China demand, the dollar, and global capex. When copper and iron ore rip, FCX and its peers lead XLB higher. Leads twice in the cycle — in early recovery when commodities bottom, and again in late cycle with inflation. Watch COPX (copper miners ETF) as the leading indicator for XLB.

⚖️ Cyclical vs Defensive — the trader-relevant split

Forget the textbook labels for a second. Here's the split that actually matters when you're deciding what to hold:

• Cyclicals (XLK, XLF, XLI, XLY, XLB, XLE, XLC). Run with the economy. Outperform in expansion. Get hammered in recession. If PMI is above 55 and rising, these are where you want to be long.
• Defensives (XLP, XLV, XLU, XLRE). Steady demand regardless of the cycle. Outperform when growth is slowing or the market is uncertain. If PMI is rolling from 55 to 48, rotate here.
• Rate-sensitives (XLU, XLRE, XLF). This is the sneaky third category. XLU and XLRE trade like bond proxies — long rates go up, they go down. XLF is the opposite — banks LOVE a steep curve. So "rate-sensitive" splits: two are anti-rate, one is pro-rate. Get this wrong and you'll be long XLU and XLF at the same time thinking you're diversified, when you're actually taking two opposite bets on the curve.

🖥️ How to use this on the platform

• Broad Market → Sector Rotation panel shows the current relative strength across all 11 sectors versus SPY on multiple timeframes (5d, 20d, 60d). Green = outperforming, red = lagging.
• Stovall Radar reads the current cycle stage directly from which sectors are leading — early cycle looks like XLF+XLY+XLK on top, late cycle looks like XLE+XLP+XLU on top.
• CAN SLIM Scanner lets you filter for stocks whose parent sector is in the top 3 by RS — a mechanical way to enforce sector strength as a requirement.
• Best practice: never go long a stock whose sector is in the bottom 3 by RS unless you have a very specific single-name catalyst (earnings beat, activist, buyout rumor). Sector gravity will fight you.

🔄 The retail sector-rotation trade

Here's the practical application most retail traders miss. Every week:

1. Rank all 11 sector ETFs by 20-day relative strength versus SPY.
2. Identify the 2-3 sectors making fresh 20-day highs against SPY while broad market is flat or drifting.
3. Those sectors are likely leading the next 2-4 week leg.
4. Do NOT buy the sector ETF. Buy the strongest individual name INSIDE that sector — better risk/reward, cleaner setups.

Example: if XLE is breaking to new 20-day highs versus SPY while crude is above its 50-day MA, don't buy XLE — buy the strongest name inside XLE (usually one of XOM, FANG, or one of the refiners on any given cycle). You get the sector tailwind plus the individual-name alpha.

📅 Sector leadership by economic stage

💱 Intermarket — What Bonds, the Dollar, and Commodities Say About Stocks FRAMEWORK

🎯 The core belief

Stocks don't move in a vacuum. Rates, credit, the dollar, and commodities are all pricing the same economic future from four different angles. When they disagree with the stock tape, one of them is wrong — and stocks are historically the last to update.

The Dollar (DXY)

The dollar index weighs the USD against a basket of major currencies (~58% EUR, 14% JPY, 12% GBP, plus CAD, SEK, CHF). Rises when foreign capital flows into US assets or Fed policy is tighter than peers'.

• Rising DXY: headwind for US multinational earnings (roughly half of S&P 500 revenue is foreign). Small-caps outperform large-caps in strong-dollar regimes.
• Rising DXY + falling commodities: disinflationary regime, good for growth stocks.
• Falling DXY: tailwind for large-cap multinationals, emerging markets, and hard assets.
• DXY divergence with SPX: rare. When both rise together, foreign capital is chasing US stocks — the late-cycle "everything in one place" bid.

Credit Spreads (HYG vs LQD, HY-OAS)

The premium risky companies pay to borrow above what safe companies pay. The single most sensitive real-time gauge of financial stress that exists.

• HYG rising vs LQD (ratio up): credit is welcoming risk. Historically precedes equity strength. Rising with SPX = confirmation.
• HYG falling vs LQD (ratio down): credit is pulling back. Historically leads equity weakness by 2–6 weeks (Gilchrist & Zakrajšek, 2012).
• Widening spreads while SPX is at highs: the classic late-cycle warning. Bonds are pricing something equities aren't.

The BnB Broad Market intermarket panel plots the HYG/LQD ratio alongside HYG/IEF (junk vs treasuries) — three lines that together tell you whether risk appetite is expanding or contracting.

The Yield Curve (10Y − 2Y, TLT/IEF proxy)

The most reliable recession leading indicator in the US toolkit. Inverted before every recession since 1955 with a median 12-month lead.

• Steep + rising rates: early-cycle expansion. Growth stocks lead.
• Flat: mid-cycle. All sectors roughly equal.
• Inverted (2Y > 10Y): late cycle. Historically 8–24 months before recession begins.
• Re-steepening from inversion (curve un-inverting while short rates fall): the recession-arrival signature. Inversion warns; the un-inversion is the trigger.

Commodities — Oil, Copper, Gold

• Copper: "Dr. Copper" — the metal with a PhD in economics. Rising copper = global demand expanding. Falling = contracting. Leads industrials by 6–12 weeks.
• Oil: dual role — energy costs for consumers, revenue for XLE. Rising oil is inflation on the way up (rate-hike pressure) but earnings tailwind for energy names.
• Gold: real-rate barometer and dollar hedge. Rising gold + falling dollar = classic risk-off pattern. Rising gold + rising dollar = geopolitical stress bid.
• Copper/Gold ratio: risk-on/off proxy. Rising = expansion. Falling = contraction. Often leads the 10-year yield.

How to act

• Before any swing long in industrials, check copper. Before any growth stock trade, check the dollar. Before assuming a rally is safe, check credit spreads.
• The four intermarket signals aligning risk-off (falling copper, rising dollar, widening spreads, steepening curve from inversion) is the highest-conviction defensive signal in the entire framework.

🔗 Chapter 8. Correlations That Matter — The Pairs Every Retail Trader Must Know MARKET STRUCTURE

🎯 The core belief

The market is a web, not a list of independent stocks. Every asset is priced against every other asset — stocks against bonds, bonds against currencies, currencies against commodities, commodities against equities. If you trade stocks in isolation you're reading one page of a book that has fifteen open at once. The good news: you don't need a PhD. If you learn 15 key intermarket pairs cold and check the two most relevant ones before every trade, you'll catch signals that pure chartists miss and dodge losses that "the chart looked great" would have handed you. This chapter is those 15 pairs, plus how to actually use them.

📐 How correlation actually works (and why textbooks lie)

The correlation coefficient runs from −1 to +1. A +1 means two assets move in perfect lockstep; a −1 means they move in perfect opposition; a 0 means they're unrelated. In practice you'll almost never see values above +0.9 or below −0.9 outside of mathematical pairs like TLT and the 10Y yield.

Here's the honest limit that no textbook stresses enough: correlations are not stable. They shift with market regime. The classic bond-equity inverse correlation held for ~40 years from the early 1980s through 2020, then flipped positive in 2022 when both dropped together as the Fed hiked aggressively. Gold and the dollar are usually inversely correlated, but during 2011's debt-ceiling crisis and 2020's pandemic panic they both spiked. Whenever someone quotes you a correlation as a fixed number, ask "over what window and what regime?" The numbers below are typical ranges — expect them to break at exactly the worst moment (i.e., during crises, which is when you actually need them).

➖ The 8 antagonistic pairs (they move in opposite directions)

1. Gold (GLD) vs US Dollar (DXY) — typical corr −0.60 to −0.75

Gold is priced in dollars globally. When the dollar strengthens, gold gets more expensive for anyone holding euros, yen, or yuan — which suppresses demand. When the dollar weakens, gold gets cheaper globally and demand rises. This is a mechanical, almost tautological relationship. Breaks: during crisis periods when investors flee to BOTH — 2011 (US debt downgrade), March 2020 (pandemic), parts of 2022 (Ukraine invasion + US rate hikes). During those windows, both DXY and gold ripped as everything else sold off. Trading application: if DXY is in a clear uptrend, do not go long GLD or GDX on a technical breakout — the currency headwind will fight you. If DXY is rolling over, gold longs get a tailwind almost for free.

2. Long Bonds (TLT) vs 10Y Yield (TNX) — corr −0.95

This is a mathematical inverse. Bond prices and yields are two sides of the same coin — when yields rise, prices fall, period. The relationship is close to −1 because TLT is basically a synthetic 20-year Treasury and TNX is the 10-year yield (close in duration). TLT has an effective duration of ~17 years, meaning a 1% move in yields translates to roughly a 17% move in TLT price. That's why 2022 was so brutal for TLT holders: yields went from 1.5% to 4.3% and TLT dropped from ~$150 to under $95. Trading application: never look at TLT and TNX as separate signals — they're one signal. Watch TNX for direction, trade TLT if you want the vehicle. And know: any bond-proxy sector (XLU, XLRE, staples with high dividends) will follow TLT's lead.

3. Utilities (XLU) vs Long Rates (TNX) — corr −0.55 to −0.65

Utilities compete with bonds for yield-seeking capital. When the 10Y pays 4.5%, why would you own a utility stock yielding 3.5% with equity risk? You wouldn't — you'd sell it. This is why XLU behaves like a bond proxy. The correlation isn't as tight as TLT-TNX (utilities also have some earnings sensitivity, especially the ones with renewables exposure like NEE and CEG), but it's strong enough to be actionable. Trading application: before adding a utility long, check the 10Y trend. If TNX is breaking to new highs, do not touch XLU — you're fighting a rising tide against you.

4. Homebuilders (XHB, ITB) vs Mortgage Rates (MORTGAGE30US)

Rolling correlation typically −0.65 to −0.75. This is the most direct affordability channel in the entire market. When 30-year mortgage rates go from 3% to 7% (as they did from early 2022 through late 2023), monthly payments on a $400K house nearly double, and demand for new homes craters — which crushes DHI, LEN, NVR, PHM, and their peers. Yet 2023 was strange: rates stayed high but homebuilders ripped anyway, because existing-home inventory was so constrained that new-build was the only game in town. So the correlation is real but not always sufficient — supply matters too. Trading application: if you're long a homebuilder, watch the 10Y and the mortgage rate. Any sustained downshift in yields is a green light; any spike is a red flag.

5. Airlines (JETS) vs Oil (WTI) — corr −0.55 to −0.70

Jet fuel is 20-30% of an airline's operating costs — the single biggest variable expense outside labor. When WTI crude rips from $70 to $110 (like early 2022), airline margins compress hard, and JETS, DAL, UAL, AAL all get sold. When crude drops, the tailwind is immediate. Breaks: when demand is the dominant driver (pandemic recovery, holiday-season bookings), airlines can rally even with rising oil because revenue growth outpaces the fuel hit. Trading application: never go long airlines during a crude breakout without a hedge. If WTI is basing, JETS setups get a green light.

6. Emerging Markets (EEM) vs US Dollar (DXY) — corr −0.60 to −0.70

Emerging market corporates and governments have trillions in dollar-denominated debt. When the dollar strengthens, servicing that debt gets more expensive in local currency terms — which tightens financial conditions across all EM economies. Add in that EM exports (commodities, manufactured goods) get priced in dollars and you have a double whammy. Every major EM sell-off (1997 Asian crisis, 2013 taper tantrum, 2015 China devaluation, 2022 rate cycle) coincided with a strong dollar. Trading application: if DXY is breaking above prior resistance, cut EEM, EMB, and any single-name EM longs. It's not worth the fight.

7. Growth Stocks (QQQ) vs 2Y Yield — corr −0.50 to −0.70 (2022-2023 sample)

Growth companies get most of their intrinsic value from cash flows years or decades in the future. Discounting those future cash flows back to today uses interest rates — and the 2Y yield is the market's best proxy for near-term Fed policy expectations. When 2Y yields spike (the market is pricing hikes), long-duration cash flows get slaughtered. 2022 was the textbook: 2Y went from 0.7% to 4.7%, QQQ dropped 33% peak-to-trough. The correlation weakens in cycles where earnings growth is the dominant story (2020's tech surge despite falling rates was really about earnings acceleration, not just rate mechanics). Trading application: before sizing up a QQQ or single-name growth long, glance at the 2Y. If it's breaking higher, size down.

8. Bitcoin (BTC) vs US Dollar (DXY) — corr −0.40 to −0.60, but noisy

BTC is priced globally in dollars; when the dollar strengthens, the same BTC costs more in foreign currency and demand from ex-US buyers softens. But BTC also has periods where it correlates POSITIVELY with QQQ as a risk-on proxy (2020-2021 was that regime — BTC and QQQ moved together, both against DXY). The correlation to DXY is real but weaker than gold-DXY because BTC is a much smaller, less liquid market driven by flows we don't fully see (institutional buying, ETF creation/redemption, on-chain leverage cycles). Trading application: use DXY as one input for BTC direction but don't rely on it alone. Add QQQ trend and BTC-specific factors (hash rate, ETF flows).

➕ The 7 synergistic pairs (they move together)

1. Software (IGV) vs Hardware/Semis (SOXX)

Both live inside Tech but they play different roles in the cycle. Semis lead early cycle when capex is expanding (data centers, phones, autos all need more chips). Software leads late cycle when growth slows and investors pay a premium for recurring-revenue, high-margin business models that don't require capex. In 2023-2024, NVDA-driven SOXX ripped on AI capex while IGV was the laggard. In 2020, IGV led as SaaS multiples exploded. Trading application: watching the SOXX/IGV ratio tells you where in the tech cycle you are.

2. Copper (COPX) vs Global Industrials (XLI) — "Dr. Copper" leads industrial demand

Copper is used in everything — construction, electrification, EVs, grid infrastructure, data centers. When copper prices are rising, it's a real-time signal that industrial demand is expanding globally. XLI historically follows COPX with a small lag. The 2020-2021 copper surge preceded the industrial rally; the 2022 copper roll preceded the XLI weakness later that year. Trading application: if COPX is basing and breaking out, add exposure to XLI. If COPX is rolling over, tighten stops on industrial longs.

3. Financials (XLF) vs 10Y-3M Yield Curve — corr ~+0.55

Banks make money on the spread between short-term deposits (they pay ~0-2%) and long-term loans (they charge 6-8%). A steep curve = wide net interest margin = fat bank profits. A flat or inverted curve (like 2022-2024) squeezes NIM and hurts bank earnings. The correlation isn't tight in real-time but the medium-term relationship is powerful. Trading application: if the 10Y-3M curve is steepening (either short end falling or long end rising), XLF gets a fundamental tailwind. This was the setup that drove the massive 2016-2018 XLF rally.

4. Gold Miners (GDX) vs Gold (GLD) — GDX is a leveraged bet on GLD

Gold miners have operational leverage — their costs (labor, fuel, energy) are relatively fixed, so a change in the gold price flows disproportionately to earnings. When GLD moves 5%, GDX often moves 12-15%. This makes GDX (and its junior sibling GDXJ) the high-beta play for anyone who wants gold exposure with more punch. Trading application: if you're bullish gold, GDX gets you 2.5-3x the move. If you're bearish gold, be careful — GDX gets slaughtered in gold downtrends and its holders don't usually panic until they've already lost 30%+.

5. Small Caps (IWM) vs High-Yield Credit (HYG) — both are risk-on proxies

Small-cap companies are more credit-sensitive than large caps — they borrow at higher rates, have thinner margins, and can't ride out downturns as easily. HYG measures the market's willingness to lend to junk-rated corporates. When HYG is making new highs, credit conditions are loose and small caps benefit. When HYG rolls over, IWM almost always follows within days to weeks. The 2022 credit tightening was signaled by HYG topping in Aug 2021, months before the IWM peak. Trading application: HYG is a leading indicator for IWM. Watch HYG for the first crack.

6. REITs (XLRE) vs Long Bonds (TLT) — both duration plays

Both XLRE and TLT are long-duration cash flow assets. When long rates fall, TLT rises and cap rates on real estate compress, making REITs more valuable. When long rates rise, both get crushed. The correlation is +0.65 to +0.75 in most regimes. Trading application: if TLT is breaking out, XLRE almost always follows — often with more punch because REITs also have earnings sensitivity to the economy on top of the rate mechanic.

7. Semis (SMH) vs Growth (QQQ) — semis are the canary

Semiconductors sit at the base of the tech stack — they're the physical layer that everything else depends on. When semi demand accelerates (data centers, AI, phones, autos), it foreshadows broader tech growth. When semi demand rolls (inventory correction, capex pullback), it's usually the first warning that broader tech is about to weaken. SMH often leads QQQ by days to weeks at inflection points. The 2022 SMH top preceded the QQQ top; the late-2022 SMH bottom preceded the QQQ bottom. Trading application: if you're trading QQQ, keep SMH open on the same screen. Divergences are meaningful.

🎯 The six key trading applications

1. Trade confirmation

Before taking a sector or single-name long, check whether the underlying macro driver is with you. Long XOM on a breakout? Confirm WTI is above its 50-day. Long DHI? Confirm the 10Y isn't ripping. Long GDX? Confirm DXY is rolling. When the technical setup AND the intermarket driver align, that's a high-conviction trade. When they conflict, size down or skip it.

2. Divergence warnings

Divergences between related assets are one of the most reliable warning signs in the market. SMH cratering while QQQ still grinds higher? Risk is building — the leader is failing while the follower hasn't figured it out yet. IWM breaking down while SPY holds up? The risk-on signal is weakening under the surface. Tighten stops when these divergences show up.

3. Regime detection

If XLU and XLP have been outperforming XLK for weeks, you're in a defensive rotation regardless of what SPY looks like. Size down growth exposure and get selective. If XLE and XLB are leading, you're in a late-cycle inflation regime — energy and materials get the benefit of the doubt on breakouts. The relative strength between defensives and cyclicals is a real-time regime signal.

4. Currency check

Before trading anything gold-related or EM-related, check DXY. A ripping dollar is a headwind for gold, EM equities, EM debt, and commodities priced in dollars. Save yourself the pain of fighting the currency tape. Same for BTC — check DXY as one of several inputs.

5. Rate proxy check

Before adding utilities, REITs, or long-duration bond-proxy trades, glance at TLT and TNX. If yields are breaking higher, these trades face a mechanical headwind and technical setups are much lower probability. If yields are rolling, they get a tailwind for free.

6. Sector spread trades (advanced but real)

The pro version: instead of picking direction, pick relative direction. Long the strongest sector ETF, short an equally-weighted position in the weakest. Example: long XLE, short XLU in a rising-inflation, steepening-curve regime. This dampens SPX-beta and isolates the sector story. Not for beginners — but real capital gets deployed this way.

⚠️ When correlations BREAK — the honest limits

• Crisis periods. In March 2020, October 2008, and briefly in 2011 (US debt downgrade), everything correlated to +1. Stocks, corporate bonds, gold, emerging markets, oil — all sold off together as investors liquidated anything liquid to raise cash. Diversification failed at exactly the moment you needed it. Plan for this — hold real cash, not "cash-like" positions.
• Regime shifts. The 40-year inverse bond-stock correlation held from 1980 through 2020, then inverted in 2022. Both dropped as the Fed hiked into inflation. "Balanced portfolios" (60/40) had their worst year in decades because the diversification math broke. Regime shifts don't announce themselves — you notice them 6 months in.
• Idiosyncratic news dominates. If NVDA reports blowout earnings, it can rip 15% while SMH sells off because everyone else in the sector had a bad quarter. Correlations describe the average; single-name news can dominate for the individual name for days or weeks.

🏋️ The practical exercise

This week, before every trade you consider, run this 30-second check. Identify the two most relevant correlations for that name.

• Long XLE or XOM? Check WTI and DXY.
• Short XLF? Check the 10Y-3M curve and HYG.
• Long a homebuilder? Check the 10Y and existing home sales trend.
• Long GDX? Check GLD and DXY.
• Long QQQ or NVDA? Check the 2Y and SMH.

Do this for 30 straight trading days. It'll feel slow at first — an extra 30 seconds per trade. By day 15 it'll be automatic. By day 30 you'll wonder how you ever traded without it. That's how professionals see the market — not as isolated charts but as one big interconnected web where every asset whispers about every other one.

🌊 Market Breadth — Is the Average Stock Participating? FRAMEWORK

🎯 The core belief

An index is a weighted number. It can rise while the average stock is falling — and history says that's exactly what happens at major tops. Breadth answers one question: is the crowd going up with the index, or are only the giants dragging it along?

The Advance-Decline (A-D) Line

Every day, every stock either advances or declines. The A-D line is a running total: cumulative (advancers − decliners). It doesn't care about market cap — a $200B mega-cap and a $500M small-cap each cast one vote.

• Rising A-D + rising index = healthy uptrend, participation is broad.
• Falling A-D + rising index = divergence. Only a few big names dragging the tape. Historically leads major tops by 4–8 weeks (Zweig).
• Rising A-D + falling index = washout done; internals are healing before price confirms.

The BnB Broad Market chart splits the A-D line by market-cap tier (mega / large / mid / small). When only mega-caps are advancing and everything below is falling, that's the same divergence Zweig warned about, made visible.

New 52-Week Highs vs New 52-Week Lows

How many stocks are hitting yearly highs today vs yearly lows. A healthy uptrend has highs dominating. When the count of new lows starts creeping up while the index is still up, cracks are forming under the surface.

• Highs ≥ 5% of issues, lows < 1% = broad participation. Chase risk.
• Both above 2.8% on the same day = fragmentation. This is one of the four conditions that triggers the Hindenburg Omen.

The McClellan Oscillator & Summation Index

Same raw material (advancers minus decliners), smoothed. Two EMAs — 19-day and 39-day — subtracted:

• Oscillator > +100: strong up-thrust. Often the "breakaway" moment out of a base.
• Oscillator < −100: capitulation. Contrarian buying zone (with confirmation).
• Summation Index rising through zero: intermediate uptrend confirmed. The McClellans' original signal.
• Summation making lower highs while price makes higher highs: the classic breadth divergence Zweig also identified — but this one has a slow-moving, high-signal-to-noise smoothed form.

How to act

• Confirm swing entries against the A-D line: if you're going long into a breakout, make sure breadth is expanding, not contracting.
• Treat any 2+ Hindenburg triggers inside 30 sessions as a regime warning, not a short signal — cut new-long size, not existing winners.
• Watch the McClellan Summation cross of zero as your intermediate-trend switch. Fewer whipsaws than daily A-D reads.

🧭 Market Internals — The Tape Reader's Compass MACRO · ALL PAGES

🎯 Picture this

You're about to enter a trade. Before clicking BUY, you take ONE breath, glance at four numbers, and KNOW — in 2 seconds — whether the market is helping or fighting you. Four numbers. Two seconds. One go/no-go answer.

The four numbers

The 3 habits

1. The Veto — Before EVERY entry, glance at verdict pill. Long needs ≥ BALANCED. Short needs ≤ BALANCED. Eliminates ~30% of losing scalps without changing your setups.
2. The Size Dial — Verdict is your volume knob. STRONG_RISK_ON + long = 1.0×. BALANCED = 0.5×. RISK_OFF against you = 0.25× or skip.
3. The Reversal Tell — SPY rising + TICK falling + TRIN rising for 15+ min = distribution → fade coming. Inverse = absorption → bounce coming.

Anchor this: The strip is a veto first, a size dial second, a reversal alert third. Make ONE habit automatic before all others: glance at the verdict BEFORE every click.

😨 Sentiment — Where Is the Crowd, and Are They Right? FRAMEWORK

🎯 The core belief

At extremes, the crowd is reliably wrong. In the middle, they're mostly right. Sentiment data is only useful when it's at extremes — a moderately bullish or bearish reading tells you nothing you didn't already know from the tape.

The Sentiment Triad

Three feeds cover the surveyed opinion, the options market's positioning, and the pricing of fear itself. Any one alone is noise. Two aligned is a caution. Three aligned at an extreme is the contrarian signal.

1 · VIX — The Fear Gauge

Implied volatility of the SPX 30-day option strip. Measures what options traders are paying for insurance right now.

• VIX < 12: complacency extreme. Historically the environment where corrections start — not tomorrow, but with elevated risk over the next 1–3 months.
• VIX 12–18: normal. No signal.
• VIX 18–25: elevated. Caution, but common in mild pullbacks.
• VIX > 30: fear extreme. Historically buying zones — but never buy the first spike, always the second lower high.

The percentile in the trailing 252 sessions matters more than the absolute number. VIX 15 during a low-vol regime is complacency; VIX 15 coming off a spike is fear that's already released.

2 · AAII Bull/Bear Survey

Weekly survey of the American Association of Individual Investors. Simple: what do retail investors say they are — bullish, bearish, or neutral for the next 6 months?

• Bull-Bear spread > +25: retail is euphoric. Historically precedes mean reversion within 4–8 weeks.
• Bull-Bear spread < −25: retail is capitulated. Historically precedes rallies within 4–8 weeks.
• Neutrals > 40%: nobody knows anything. Range-bound market signal.

Retail is a lagging indicator — that's the whole reason it's contrarian. When your barber, cab driver, or Twitter timeline is bullish, retail has already bought.

3 · Put/Call Ratio

Total put volume ÷ total call volume, usually smoothed by a 10-day moving average. Measures what options traders are doing, not saying.

• P/C > 1.20 (10dMA): heavy protection buying. Historically bottoms within 2–3 weeks.
• P/C < 0.60 (10dMA): complacency, everyone chasing calls. Historically tops within 2–4 weeks.
• Equity-only P/C is cleaner than total P/C — index P/C is polluted by portfolio-hedging flows that aren't really "sentiment."

Alignment: the money signal

When all three fire at once — VIX in its extreme percentile, AAII spread past its 2σ band, put/call at the outer decile — you have sentiment alignment. Historically these occur 3–5 times per year and have edge over the following 20–60 sessions.

Alignment does not mean trade tomorrow. It means size up when your setup fires in the direction the crowd isn't looking.

⚠ The Hindenburg Omen — A Warning, Not a Clock REGIME FILTER

🎯 The core belief

Major market tops don't happen when everything is going up. They happen when the market is fragmenting — some stocks hitting new highs while others hit new lows, on the same day, while the index looks fine on the surface. The Omen tries to catch that specific texture.

The Four Simultaneous Conditions

All four must occur on the same session:

1. New 52-week highs ≥ 2.8% of advancers+decliners — enough stocks are still rocketing to certify an uptrend surface.
2. New 52-week lows ≥ 2.8% of advancers+decliners — but at the same time, at least as many are cratering.
3. New highs ≤ 2× new lows — the ratio can't be too skewed (breakout thrusts naturally have highs >> lows and don't count).
4. NYSE Composite > its close 50 sessions ago — the index itself is still in an uptrend (this filters out obvious downtrends where the pattern is meaningless).
5. Plus the modern refinement: McClellan Oscillator negative on the trigger day (McHugh).

Cluster Detection — Why It Matters

A single Omen is noise. Two or more Omens inside 30 sessions is the configuration that carries statistical edge. Miekka's original research and every subsequent replication finds that clusters — not lone triggers — precede the meaningful drawdowns.

• 0 triggers: no signal.
• 1 trigger in 30 sessions: watch, don't act.
• 2+ triggers in 30 sessions: regime warning. Historical median lead time to a 5%+ drawdown: 6–12 weeks.
• 3+ triggers in 30 sessions: rare. When it happens, take it seriously.

What NOT to Do

• Do not short on the trigger. The signal has weeks-to-months of lead time. The market often makes higher highs first.
• Do not sell existing winners. Trend-following rules override regime filters on positions that are still working.
• Do not size new longs at max. This is the actionable part: cut size on new entries, demand tighter stops, require higher-conviction setups.

How to act

• On a fresh 2+ cluster: reduce new-long risk by ~50% (half normal position size) until 30 sessions pass without another trigger.
• Pair the Omen with the A-D line — if breadth is diverging on top of the cluster, treat it as a stronger signal.
• Pair the Omen with sentiment extremes — a cluster during retail euphoria (AAII spread +25, put/call < 0.6) is the highest-signal configuration.

🌐 Intermarket Regime MACRO · TOP OF PAGE

What it is

The 4-asset canonical chain (Stocks / Bonds / Commodities / Dollar) mapped to one of 16 named macro regimes. Each regime has a specific trade thesis with citations.

How it computes

5-day % returns of SPY / TLT / DBC / UUP are signed (U = up, D = down) and concatenated into a 4-letter chain (e.g. "UDUD"). The chain is looked up against the 16-regime table.

Persistence

A regime is shown as ACTIVE only if it has been the same chain for 3+ consecutive sessions. Otherwise TRANSITIONING with the prior active regime preserved.

The 16 regimes

How to act

🔄 Sector Rotation & Cycle — Broad Market SECTOR · BROAD MARKET

What it is (and why it was restructured 2026-08-23)

The sector area on Broad Market used to stack three unrelated things under one "Sector Valuations" header, and rendered the same leadership series twice. It is now organized as three questions, three surfaces, zero duplicated data:

How to read it (workflow)

• Start at the cycle verdict. It sets expectations: a MARKET_TOP read (energy/materials leading) means new longs deserve late-cycle skepticism even when the index looks fine.
• Confirm with the RS chart. Is the leadership persistent (months) or a rotation flicker (weeks)? Use the range buttons.
• Drill into the flow widget for which names inside the leading sector are actually being accumulated.
• Check valuation last — leadership plus a cheap PEG is a different bet than leadership at 2.5× PEG.

Housekeeping note

The 2026-08-23 tape/events audit also removed a dead loader that had been fetching sector-rotation data every 15 minutes into a container deleted on 2026-08-04, and clarified the two regime verdicts: 🌡️ Today's Market Read = intraday tactical score from market internals; 🧭 Regime Transmission = multi-day macro regime from release surprises. They answer different horizons and are labeled as companions.

🌊 Sector Rotation SECTOR · TOP OF PAGE

What it is

A composite 0-100 score per sector ETF combining 4 proven methodologies. Strict mode (default) requires all three filters to pass for ROTATING IN: Weinstein Stage 2 + O'Neil Trend Template 8/8 + Wyckoff Phase D or E.

The 4 methodologies

Buckets

• Growth — XLK, SOXX, XBI, XLY, IWM
• Cyclical — XLI, XLB, XLE, URA
• Financial — XLF, KRE
• Defensive — XLV, XLP, XLU, XLRE
• Safe Haven — TLT, GLD, UUP, TIP
• Speculative — IBIT (crypto)

Rotation Pairs

The system identifies the dominant out→in bucket flow when the spread between top and bottom buckets is ≥ 25 points. Each pair has a named interpretation (e.g. "RISK-OFF / FLIGHT TO QUALITY -- growth rotating OUT into bonds/gold").

How to act

🎯 Stocks in Rotating Sectors SECTOR

What it is

For every sector with rotation score ≥ 60, the system pulls that sector's members from a curated universe (~280 tickers) and applies the same O'Neil Trend Template at the individual stock level. Only stocks passing 6/8 or 8/8 criteria are surfaced.

What you see

• Ticker with template score (8/8 = green, 6-7/8 = yellow)
• RS rating — stock's 12m relative strength vs SPY, scaled 0-100
• IN PLAY badge if currently in today's SIP composite
• Catalyst bias (BULL / BEAR) if a fresh catalyst is present

🎯 The Regime Stack — Five Models, One Reality MASTER FRAMEWORK

🎯 The core belief

Every asset moves within a regime — a persistent state of the world where certain setups work and others don't. What kills traders is not lack of setups; it's trading a setup outside its regime. A bull-flag breakout has edge in an uptrend and loses money in a downtrend. A defensive-sector rotation trade has edge at cycle tops and no edge in early expansions. The platform runs five regime models simultaneously because no single one is right about everything — but read together, they tell you exactly which setups have edge right now.

📐 The five models — what each one actually measures

The critical insight: these models are not competing versions of the same signal — they operate at different horizons. Keller can flip from confirmed uptrend to warning in 3 days. Stovall's stage read changes over weeks. The yield curve's message evolves over months to years. When they agree, you have a coherent story. When they disagree, one is early and one is late.

1 · Keller Regime — the tactical uptrend switch

The Institutional Uptrend Timing Model chapter covers this in depth. Summary in regime terms:

• CONFIRMED_UPTREND: full risk-on. Follow-through day fired, distribution days < 5 in last 25 sessions. Growth breakouts have edge.
• UPTREND_UNDER_PRESSURE: distribution accumulating. Reduce new-long size, tighten stops on existing.
• RALLY_ATTEMPT: market bottoming. No new positions until Day 4+ with follow-through day.
• DOWNTREND: risk-off. No long breakouts. Cash or short-only.
• CIRCUIT_BREAKER: hard stop — a severe internal breakdown has fired the regime kill switch.

Why it matters at the top of the stack: Keller answers "can I take a long right now?" faster than any other model. It's the go/no-go tactical filter — everything else on this list refines what to trade once Keller says green.

2 · Stovall Cycle — the sector-rotation stage

The Stovall Radar maps today's 63-session sector leadership to one of four canonical business-cycle stages:

• MARKET_BOTTOM (full recession) — expected leaders: XLF, XLY. Rate cuts + early consumer bid.
• BULL_MARKET (early recovery) — expected leaders: XLK, XLI, XLB. Growth + capex + materials cycle.
• MARKET_TOP (full recovery) — expected leaders: XLE, XLB. Late-cycle inflation, commodities peak.
• BEAR_MARKET (early recession) — expected leaders: XLP, XLV, XLU. Defensives bid.

Historical validation (23 years, 5,500 daily reads): only BULL_MARKET has real forward edge (+0.51% median 63d, 62% hit) — the other three are descriptive-only. Day 1 of a fresh transition is where the BULL edge lives; by day 20+ of MARKET_TOP it decays to −1.44% median.

Why it matters: Stovall answers "which sectors have historical edge right now?" It doesn't tell you when the cycle will turn — it tells you where in the cycle you probably are based on what leadership is doing.

3 · Yield Curve — the macro anchor

The 10Y-2Y spread (proxied by TLT/IEF ratio) is the single best-documented recession leading indicator in the U.S. — inverted before every recession since 1955 with a median 12-month lead (Estrella & Mishkin, 1996). Four states:

• EXPANSION (curve steep, rising rates) — economic anchor says Bull expected.
• LATE_CYCLE (curve inverted or inverting) — Top expected.
• RECESSION_ARRIVING (curve re-steepening from inversion) — the recession-arrival signature. Bear expected.
• RECOVERY (curve steep + still steepening) — Bottom expected.

Why it matters: The yield curve is the slowest, deepest signal in the stack. When it disagrees with Keller or Stovall, ask which one is late — usually the yield curve is right and the faster signals are early.

4 · PPO Stack — the momentum alignment

Multi-timeframe PPO on any ticker (or the index): LT (21,34), MT (12,26 daily / 5,13 weekly), ST (1,5). Five alignment states:

• FULL_ALIGN (LT+MT+ST all positive) — full risk-on trend continuation.
• PULLBACK (LT+MT positive, ST negative) — buyable dip inside uptrend.
• DETERIORATING (only LT positive) — swing trend has rolled; tighten.
• DOWNTREND (all three negative) — capital preservation.
• EARLY_TURN (LT negative, MT+ST positive) — unproven bottom or bear-market rally.

Historical validation (Stovall × PPO interaction test, 5,500 rows): forward SPY returns are consistently better when the daily PPO is depressed at the time of the stage read, worst at FULL_ALIGN. Buying weakness beat chasing alignment in every stage. Plus the sharp warning: MARKET_TOP + DETERIORATING = expected leaders (XLE/XLB) underperform SPY by −3.21% median over 63 sessions.

Why it matters: PPO is the ticker-level momentum lens that refines index-level regime reads. Keller might say uptrend, but if a specific stock's PPO is DETERIORATING, that stock is not participating.

5 · Weinstein Stages — the structural position

Stan Weinstein's four-stage cycle applies to any ticker or index over months-to-years:

• Stage 1 (basing) — sideways after downtrend, 30-week MA flat. No trend, low returns, tight ranges. Do not chase.
• Stage 2 (advancing) — breakout above the base, 30-week MA rising. The only stage worth being long.
• Stage 3 (topping) — sideways after uptrend, 30-week MA flattening. Distribution phase. Sell into strength.
• Stage 4 (declining) — breakdown below the top, 30-week MA falling. Short-only or cash.

Why it matters: Weinstein's rule — "never take a long unless the weekly chart is Stage 2" — is the single strictest position filter in the entire framework. It overrides pattern signals, breadth reads, everything. If a stock isn't Stage 2 on the weekly, you don't buy it.

🎯 The Alignment Matrix — when they all agree

The five models rarely all point the same way. When they do, size up. Read the matrix top-to-bottom:

⚠ The Divergence Playbook — when models disagree, one is early

Divergences between fast and slow models are the highest-information moments the platform generates. Nine specific configurations to watch:

1. Keller UPTREND + Yield Curve LATE_CYCLE

Meaning: The tactical read says risk-on, the deep macro read says late-cycle warning. Historically: the yield curve is warning about a top that hasn't formed yet — Keller will catch up in weeks to months.

Play: Long, but with reduced size and tighter stops. Watch for Keller to flip to UNDER_PRESSURE — that's the alignment moment. This is the current configuration as of most of 2025-2026.

2. Stovall MARKET_TOP + Yield Curve EXPANSION

Meaning: Sector leadership has already rotated to late-cycle (energy/materials), but the yield curve is still showing normal expansion. Historically: sectors are the tell — they front-run the curve by a quarter.

Play: Trust sectors. Reduce growth exposure, tilt toward the actual leaders (XLE, XLB), don't add tech into strength.

3. Stovall MARKET_TOP + PPO DETERIORATING (the sharp one)

Meaning: Sector rotation says late-cycle AND daily momentum is rolling over. The specific interaction with the strongest historical evidence.

Play: Historically expected leaders (XLE/XLB) underperform SPY by −3.21% median in this specific combination. Do not chase energy or materials into this configuration. This is the "late-cycle leader exhaustion" signature.

4. Keller CONFIRMED + PPO FULL_ALIGN (the trap of comfort)

Meaning: Everything looks great. The tape says up, the momentum says up. But this specific combination is where the historical edge is weakest.

Play: Historical median forward 63d excess with all bull signals lit: +2.9% to +3.6% (lowest of all combinations). Full alignment is when the crowd is already positioned — reduce chase behavior, take profits into strength.

5. Keller DOWNTREND + PPO DETERIORATING but PPO ST rising

Meaning: The tape is broken but the immediate week is bouncing. Bear-market rally in progress.

Play: This is not a bottom. Historical: bear-market rallies inside PPO DOWNTREND are median −1.5% forward 63d. Fade the rally with tight stops, or stay cash.

6. Yield Curve RECOVERY + Stovall BEAR_MARKET

Meaning: The curve says the recovery has started, but sectors still show defensive leadership. Historically: this is the classic bottoming pattern — curve leads, sectors lag, tape follows sectors.

Play: Bottom-fishing zone. Watch for Stovall to flip to MARKET_BOTTOM (XLF/XLY leading) — that's the entry signal for early-cycle names.

7. Weinstein Stage 2 (weekly) + Keller UNDER_PRESSURE (daily)

Meaning: The structural uptrend is intact but the daily tape has short-term stress.

Play: The Weinstein rule wins — Stage 2 is Stage 2. This is a buyable dip. Reduce size on breakouts but do not exit longs. Standard Bellafiore/O'Neil "trend intact, breathe" configuration.

8. Weinstein Stage 3 (weekly) + Keller CONFIRMED (daily)

Meaning: The stock (or index) has topped structurally but the daily tape still shows strength — a bounce inside a topping pattern.

Play: Weinstein again wins — do not add longs to a Stage 3 stock. This is what O'Neil calls "late-stage base" and Minervini calls "the third base failure zone." Sell into strength.

9. All five bullish EXCEPT the yield curve

Meaning: Everything short-term says up, but the bond market is warning. Common configuration in 2024-2026.

Play: Trade tactically bullish (Keller/Stovall/PPO/Weinstein all say go) but keep the position size below what full alignment would warrant. Cap open risk at 4-6% total portfolio heat. The yield curve is telling you not to be maximally aggressive even when it feels safe.

📋 The Hierarchy Rule — which model wins when they fight

When models genuinely conflict (not just "different horizons"), a decision hierarchy is needed. In order of override authority:

1. Weinstein weekly stage (structural) — never violate. Stage 4 on the weekly = no long, period. Stage 2 = the necessary condition for any long.
2. Yield curve (macro anchor) — never fight the recession-arrival signature. When the curve re-steepens from inversion, cap risk at 50% of normal until Keller confirms the downtrend.
3. Keller regime (tactical switch) — the go/no-go for new positions. CIRCUIT_BREAKER overrides everything below it.
4. Stovall stage (sector selection) — refines what to trade, not whether. Reads that persist > 21 days start to decay historically.
5. PPO stack (entry timing) — refines when. Enter on weakness (PULLBACK / DETERIORATING) not on FULL_ALIGN.

⏰ The Daily Regime-Check Workflow

Five minutes before the open, in order:

1. Weinstein SPX weekly stage — glance at the Wave Analyzer. Stage 2 = longs allowed. Stage 3 = trim, no adds. Stage 4 = no longs.
2. Yield curve state — Stovall Radar Ideal-vs-Actual anchor. Note which macro regime the curve is showing.
3. Keller regime — CAN SLIM Scanner header. Note the current state and distribution-day count.
4. Stovall stage — Stovall Radar current stage + persistence chip. Note who's supposed to lead.
5. PPO SPY stack — Broad Market TREND ALIGNMENT panel. Note the alignment state.

Write the five readings down. If they align (all bull or all bear), size accordingly. If they diverge, identify which of the nine divergence configurations you're in and act per that playbook.

🎯 The Regime Discipline (Bellafiore's rule, applied)

Mike Bellafiore's rule: trade only the setups your firm has proven work in the current regime. Applied to this framework:

• CAN SLIM breakouts — edge only when Keller is CONFIRMED_UPTREND and Weinstein is Stage 2. Historically zero edge in Stage 3 or Stage 4.
• Bull-flag continuation trades — edge in Stovall BULL_MARKET or MARKET_BOTTOM with PPO PULLBACK. Poor edge in MARKET_TOP or DETERIORATING.
• Defensive rotation trades (XLP/XLV/XLU longs) — edge in Stovall MARKET_TOP → BEAR_MARKET transitions with yield curve LATE_CYCLE. No edge otherwise.
• Mean-reversion / oversold bounces — edge only when Keller is not DOWNTREND and Weinstein is Stage 2 or Stage 1 (basing). Never buy oversold in Stage 4.
• Turtle Soup / false-breakout fades — edge in range-bound regimes (Stovall stage flipping, Keller UNDER_PRESSURE). Bad edge in trending regimes.

🧠 What NOT to do

• Don't chase alignment. By the time all five say bull, the crowd has bought. Full alignment is lower expected forward returns than mixed with PPO in pullback — the data says so.
• Don't over-diversify the regime read. If you're reading 15 macro indicators to make a call, you're not filtering — you're rationalizing. Five models, five readings, in order.
• Don't override the hierarchy. When the yield curve is inverted and re-steepening, no amount of "but the tape looks fine" changes what history says happens next.
• Don't confuse regime with prediction. Regime doesn't say what happens tomorrow. It says which setups have edge right now. Trade the setup, not the forecast.
• Don't skip Weinstein. The single most-common expensive mistake on this platform: buying a Stage 3 or Stage 4 stock because Keller and Stovall are green. The weekly chart is Stage 2 or you don't buy. This one rule prevents most portfolio disasters.

🏛️ Dow Theory — The Original Trend Framework FRAMEWORK

The Six Tenets

1. The averages discount everything: all known information is already priced in. News confirms, doesn't create, trends.
2. The market has three trends: Primary (major, 1+ years), Secondary (weeks-months), Minor (days-weeks).
3. Primary trends have three phases: Accumulation, Public Participation, Distribution (parallels Wyckoff exactly — Dow described this in 1900, Wyckoff formalized it in 1931).
4. Averages must confirm each other: In Dow's era, Industrial average (blue-chip stocks) and Rail average (transports) had to agree for a valid trend signal. Modern version: SPY must confirm QQQ, or NYA must confirm SPY.
5. Volume confirms the trend: volume expands in the direction of the primary trend, contracts on counter-trend moves.
6. Trends persist until definitive reversal: a trend is in force until proven otherwise. Don't fight the tape.

The Three Trends in Detail

Dow's analogy: primary trends are like ocean tides — powerful, slow, dominant. Secondary reactions are the waves within the tide. Minor movements are ripples on the waves.

The Three Phases of a Primary Trend

• Phase 1 (Accumulation): smart money buys when everyone else is fearful and pessimistic. Sentiment terrible. Fundamentals still look bad.
• Phase 2 (Public Participation): trend becomes obvious, retail piles in. Longest and biggest phase — the "juicy middle."
• Phase 3 (Distribution / Excess): smart money sells to euphoric retail. News is universally bullish. Sentiment maxed. Divergences appear.

Confirmation Requirements

• Higher highs AND higher lows on closing prices (not intraday) = valid uptrend
• Lower highs AND lower lows on closing prices = valid downtrend
• Multiple averages must agree — SPY breakout with QQQ NOT confirming = suspect
• Volume must confirm — breakout on low volume = suspect (Dow considered this bearish for the move)

How Dow Theory Signals a Trend Change

1. A rally fails to make a new high (LH — first warning)
2. Prior low is broken on strong volume (LL — confirmation)
3. Secondary averages must confirm the same pattern
4. Trend is now downtrend until proven otherwise

🌊 Wyckoff Method — The Complete Framework FRAMEWORK

Cross-reference: The Wyckoff Phase D/E chip in the Sector Rotation scorer uses this framework at the sector-ETF level. This section teaches the underlying method so you can read every ticker the Wave Analyzer scans, not just the chip verdicts.

Wyckoff's Three Laws

1. Supply and Demand determine price: When demand exceeds supply, price rises. When supply exceeds demand, price falls. Nothing else matters.
2. Cause and Effect: The amount of accumulation or distribution (the CAUSE, measured on a P&F chart) determines the size of the resulting move (the EFFECT). Big base = big move.
3. Effort vs Result: When volume (effort) doesn't produce price movement (result), something is off. Divergence = imminent reversal.

The Composite Operator

Wyckoff's key mental model: imagine every stock is run by a single very smart operator. This "Composite Operator" (institutions, in reality) accumulates in secret at low prices, marks the stock up by creating public excitement, distributes to the retail crowd at high prices, then marks it back down. Your job is to read the Composite Operator's intentions from price and volume.

The Complete Market Cycle

Canon rule (drives the Wave Analyzer): every stock, at every moment, sits in one of these 5 states + a phase A-E within that state. There is no "no schematic" — only insufficient data to identify one.

Accumulation Phase Cycle A → E

• Phase A — Stopping Action: PS (Preliminary Support) → SC (Selling Climax on huge volume) → AR (Automatic Rally) → ST (Secondary Test, retest of low on lower volume). Downtrend halted.
• Phase B — Building the Cause: sideways trading range. Multiple STs. Composite Operator absorbs supply from weak hands. Can last weeks to months.
• Phase C — Spring / Shakeout: brief violent break BELOW range low that reverses. Traps late shorts. The highest-conviction long entry point in the entire cycle. Grade No.3 Spring (deep test that fully recovers) is best.
• Phase D — Signs of Strength (SOS): higher lows within range, wide up bars on volume expansion, LPS (Last Point of Support) pullback = second-chance entry after breakout.
• Phase E — Markup: sustained uptrend confirmed. Trend follow. Buy pullbacks toward moving averages.

Distribution Phase Cycle (mirror image)

• Phase A: PSY (Preliminary Supply) → BC (Buying Climax) → AR → ST. Uptrend halted at the top.
• Phase B: sideways range. Multiple UTs (Upthrusts = false breakouts above range high). Institutions distribute.
• Phase C — UTAD: false breakout above range high that fails. Traps late longs. Highest-conviction short entry.
• Phase D — Signs of Weakness (SOW): lower highs, wide down bars on volume, LPSY (Last Point of Supply) rally = second-chance short.
• Phase E — Markdown: sustained downtrend confirmed. Trend follow. Short rallies.

The 9 Buying Tests (Pruden's Framework)

Before entering LONG in a suspected accumulation, check how many of these tests pass. 5+ = high conviction:

1. Downside price objective accomplished (target reached)
2. Preliminary Support, Selling Climax, Secondary Test present
3. Activity bullish (volume increasing on rallies, decreasing on reactions)
4. Downtrend line broken
5. Higher lows forming (transition from LL to HL)
6. Higher highs forming (transition from LH to HH)
7. Stock stronger than the market (positive relative strength)
8. Base forming (horizontal accumulation on P&F)
9. Estimated upside profit potential ≥ 3× downside risk to stop

The 9 Selling Tests are the exact mirror for shorting distributions.

P&F Cause & Effect (Wyckoff's Signature Method)

• Horizontal Count: measure the width of the trading range on a P&F chart in columns. Multiply columns × box size = accumulated cause. Add to range low (accum) or subtract from range high (distr) for target.
• Vertical Count: measure the height (in boxes) of the first strong thrust column after climax. Multiply by 3 (Dorsey multiplier) = projected target from the anchor point.
• Convergence: when horizontal and vertical counts point to the same target (≤5% apart), the target is HIGH-CONVICTION.

Weis Wave Volume Method

David Weis's contribution: chart cumulative volume by "wave" (unbroken directional move). Compare up-wave volumes to down-wave volumes.

• Up-wave volume EXPANDING while down-wave volume CONTRACTING: bullish sponsorship building. Institutions accumulating.
• Up-wave volume CONTRACTING while down-wave volume EXPANDING: bearish sponsorship draining. Institutions distributing.
• Weis is a CONFIRMATION tool for schematic — not a standalone signal.

LPS vs LPSY — The Two Highest-Probability Entries

Both LPS and LPSY are the second-chance entries that follow a completed terminal event (Spring or UTAD) and a Sign of Strength / Sign of Weakness. Per Bogomazov/Pruden, these are the highest-probability discretionary entries in the entire Wyckoff method on daily/weekly timeframes.

Ideal Volume Sequences — What to Actually Look For

These are the two 5-step volume footprints that separate valid Wyckoff setups from every other range-bound pattern. When you can trace all five steps on your chart, the setup has textbook institutional footprints.

Effort vs Result — The Fundamental Diagnostic Matrix

The single most useful real-time diagnostic tool per Wyckoff's Third Law. Every significant bar can be classified in this 2×2 matrix. On daily/weekly charts these readings are cleaner and more reliable than on intraday because institutional volume dominates the totals.

Modern HFT/AI Era — What Actually Changed

Wyckoff's underlying logic has not been invalidated. Large operators still must accumulate before advancing and distribute before declining. What HFT and AI-driven execution changed is the speed, depth, and visual appearance of the process. The classical principles still hold, but the events themselves look different on modern charts.

What Has NOT Changed

• The need for a proper Cause (trading range) before a major Effect
• The requirement that terminal events (Spring / UTAD) occur after meaningful absorption
• The superiority of LPS / LPSY entries over aggressive entries at the extreme
• The diagnostic power of Effort versus Result on higher timeframes
• Weekly structure still dominates Daily structure

Preferred Entry Framework — The Practical Rules

Per section 7.2 of the Wyckoff canon (Pruden/Bogomazov systematization). Every wave-analyzer trade plan on this platform follows this framework.

Non-Negotiable Requirements

1. A clear, well-defined trading range must exist before labeling any Spring or UTAD
2. Volume must confirm the event (climactic on the extreme, declining on the test)
3. Prefer Terminal Springs and UTADs over early Phase B undercuts or overshoots
4. Weekly context overrides Daily. Do not take Daily Springs against a clear Weekly distribution
5. Define the invalidation level (Spring low or UTAD high) BEFORE entry

Daily / Weekly Analysis Checklist

Run this sequence every time you analyze a chart:

1. Identify the larger cycle position (Accumulation, Markup, Distribution, Markdown) on the Weekly chart first
2. Locate any clear trading range and mark its support and resistance boundaries
3. Confirm Phase A stopping action has already occurred (SC/AR/ST or BC/AR/ST)
4. Look for evidence of absorption (Effort vs Result mismatches) inside the range
5. Determine whether a Terminal Spring or UTAD has occurred or is forming
6. Check that the subsequent LPS or LPSY shows the correct volume behavior (declining)
7. Establish the logical invalidation level and calculate risk before considering entry
8. Confirm the potential reward justifies the risk (range height measured-move is a common starting point)

Common Mistakes on Higher Timeframes

• Labeling every undercut as a Spring without a proper prior range and volume confirmation
• Ignoring Weekly structure while trading Daily signals
• Entering too early in Phase B before absorption is complete
• Using stops that are too tight for Daily volatility (give the trade room under the Spring low or above the UTAD high)
• Treating modern multi-leg Springs/UTADs as failures simply because they are not textbook single-bar events
• Over-weighting lower-timeframe noise when the Daily/Weekly structure is clear

🌀 Elliott Wave Theory — The Fractal Structure of Markets FRAMEWORK

The Core Insight

Elliott discovered that markets move in repeating fractal patterns of 5 waves in the direction of the primary trend (impulse), followed by 3 waves against it (correction). This 5-3 pattern repeats at every timeframe — from minute charts to century charts.

The Impulse Pattern (5 waves)

Three Unbreakable Rules

1. Wave 2 cannot retrace more than 100% of Wave 1. If it does, the count is wrong.
2. Wave 3 is never the shortest of Waves 1, 3, and 5. Usually the longest.
3. Wave 4 does not overlap Wave 1's territory (in equities/futures). If it does, the count is wrong.

The Correction Pattern (3 waves = ABC)

• Wave A: first move against the prior trend
• Wave B: bounce (often traps traders who think the trend has resumed)
• Wave C: final move in the correction direction, usually deeper

Correction Types

• Zigzag: sharp A, small B, sharp C (5-3-5 structure). Most common in strong trends.
• Flat: A and B roughly equal, C makes new low (3-3-5). Common in sideways markets.
• Triangle: 5 waves (ABCDE), each smaller. Continuation pattern — expect breakout.
• Complex/Combination: WXY or WXYXZ. Two or three simple corrections joined.

Fibonacci Retracements & Extensions (mandatory tool)

• Wave 2 typically retraces 50-61.8% of Wave 1
• Wave 3 typically extends 161.8% of Wave 1 (measured from Wave 2 low)
• Wave 4 typically retraces 38.2% of Wave 3
• Wave 5 typically equals Wave 1 (equality) OR extends 61.8% of Wave 1-3 range
• Wave A/C equality: Wave C often equals Wave A in a zigzag correction

Degrees (the fractal levels)

Every wave is composed of smaller waves. Standard degree hierarchy (largest → smallest):

1. Grand Supercycle (century+)
2. Supercycle (decades)
3. Cycle (years)
4. Primary (months to a year)
5. Intermediate (weeks to months)
6. Minor (days to weeks)
7. Minute (hours to days)
8. Minuette (minutes to hours)
9. Sub-Minuette (minutes)

When we say we're "in Wave 3," we always mean at a specific degree. A stock can be in Wave 3 at Primary degree while being in Wave 2 at Minor degree simultaneously.

Elliott + Wyckoff Confluence (Pruden's Discovery)

Pruden showed that Elliott's 5-wave impulse maps neatly onto Wyckoff's schematics:

• Wyckoff Accumulation = origin of Elliott Wave 1
• Wyckoff Phase C Spring = Elliott Wave 2 low
• Wyckoff Phase D breakout = Elliott Wave 3 start
• Wyckoff Phase E markup = Elliott Waves 3, 4, 5
• Wyckoff Distribution BC = Elliott Wave 5 top

💰 Chapter 18. Fundamental Analysis for Traders — What Charts Can't Tell You FUNDAMENTAL

🎯 The core belief

Charts tell you if and when a stock is moving. Fundamentals tell you why — and whether that move has legs. A trader who ignores fundamentals is trading squiggles; a fundamentalist who ignores charts is a very smart, very early bag-holder. This chapter is the retail trader's minimum-effective-dose of fundamental analysis: enough to filter out garbage before you draw a single trendline, enough to know when a chart pattern has a real catalyst behind it, and enough to recognize the situations where fundamentals get temporarily overwhelmed by flow, sentiment, or macro.

Why a trader should care — even a 3-day swing trader

You don't need to model discounted cash flows to trade well. But you do need to answer three questions before risking capital on any name:

• Is there a reason it could move? An earnings beat, a product launch, a sector tailwind, a short squeeze setup. Charts show energy; fundamentals show the fuel.
• Is the company solvent enough to survive my holding period? Even a 5-day swing trade can blow up on a going-concern warning or a covenant breach.
• Am I on the right side of the sector? Trading NVDA-style semiconductor names in an inventory-glut cycle is fighting the current, no matter how clean the base looks.

Even pure technicians like William O'Neil built their systems on top of fundamentals — the "C-A-N" in CAN SLIM stands for Current earnings, Annual earnings, New products/management. The "SLIM" is chart and flow. O'Neil didn't pick one side; he demanded both.

The two schools — and why you must pick one

Every stock-picking framework in the last century boils down to one of two philosophies:

Neither is wrong. Both compound wealth over full cycles. But they require opposite temperaments and opposite risk rules. The single biggest failure mode in retail investing is trading a growth stock with value patience — holding a broken momentum name through a 40% drawdown because "the story is still intact." Growth stocks that break their trend rarely come back. Value stocks that break their trend often stay broken for a decade. Pick one lens. Get good at it. See Chapter 21 for the full growth-vs-value breakdown.

The retail trader's fundamentals workflow — 5 steps, 5 minutes

You don't have time to read a 10-K per idea. You need a triage. Here is the exact sequence to run before you commit real money to any ticker:

1. Sector check. Pull up the Sector Rotation panel. Is this ticker's sector in the top-3 by 3-month relative strength, or the bottom-3? Being right on the sector is roughly half the battle. If you're buying a name in a bottom-3 sector, you'd better have a very specific catalyst reason (turnaround, spinoff, activist stake).
2. Earnings trajectory. Look at the last 4 quarters of revenue and EPS. You want acceleration, not just growth. Revenue growing 10 → 12 → 15 → 19% YoY is a very different animal from 19 → 15 → 12 → 10%. Deep Analysis surfaces this as a bar chart on every ticker page.
3. Margin trend. Is gross margin expanding, flat, or compressing? Margin expansion is the single most powerful positive fundamental signal — it means pricing power. Margin compression on flat revenue is the biggest silent killer of growth stocks.
4. Valuation reality check. P/E vs sector median. For growth names, PEG. You're not looking for "cheap" — you're looking for "not insane." A 40x P/E on a 40% grower is a PEG of 1, which is defensible. A 40x P/E on a 5% grower is a PEG of 8, which is a slot machine.
5. Flow of smart money. Any 13F inflows from top funds in the last quarter? Any recent insider selling clusters (Form 4 filings)? Insiders can sell for a hundred reasons but they only buy for one — they think the stock is going up. A cluster of insider buying is one of the highest-quality signals in all of investing.

When fundamentals fail — be honest about it

Fundamental analysis has real, well-documented failure modes. Pretending otherwise gets people wiped out. There are three regimes where fundamentals go on vacation:

• Meme stocks and short squeezes. GME in January 2021, AMC through 2021, BBBY in 2022 — these were pure order-flow phenomena. Short interest, gamma exposure, and social-media coordination overrode every fundamental metric. Trying to "value" GME at $300 was a category error; it wasn't a company at that moment, it was a war between hedge funds and Reddit. Trade the flow or stay out; don't bring a DCF to a gamma fight.
• Distressed and turnaround plays. Names in bankruptcy, near-bankruptcy, or under active litigation trade on news headlines and legal outcomes. Ratios don't work because the denominators are unstable (equity might be zero, EBITDA might be negative, debt might be renegotiated overnight). If you must trade these, treat them as event-driven, not fundamental.
• Rapid regime changes. March 2020 (COVID crash), October 2008 (Lehman), December 2018 (Fed pivot). When correlations spike to 1.0, every stock trades like the S&P index. Individual fundamentals are drowned out by macro. In these windows, cash and index-level risk management beat stock picking.

Recognize these regimes and step aside. Nothing in your fundamental toolkit will save you when the market is in a liquidity panic. The trader who doesn't trade during those windows preserves capital to deploy when fundamentals matter again.

Fundamentals as a filter, not a fortune-teller

The best mental model for retail traders is this: fundamentals filter your universe; technicals time your entry. Before you sit down to scan charts, you should have a shortlist of maybe 30–60 names that pass your fundamental screen. Then and only then do you look for setups. This inverts what most retail traders do (chart first, "check the fundamentals" as an afterthought), and it's why most retail traders underperform.

📄 Chapter 19. Reading Financial Statements — Income, Balance Sheet, Cash Flow FUNDAMENTAL

🎯 The core belief

Every public company must file three financial statements: the income statement, the balance sheet, and the cash flow statement. Each answers exactly one question. The income statement asks: is the company making money? The balance sheet asks: how healthy is it right now, this instant? The cash flow statement asks: is the cash real, or is it an accounting illusion? The three together are a lie detector. Any two can be manipulated with aggressive accounting; getting all three to agree is much harder. That's why Buffett famously reads the cash flow first, the balance sheet second, and the income statement last.

1. The Income Statement — "did they make money?"

The income statement (also called the P&L, for profit and loss) runs top-to-bottom. Each line subtracts a category of costs from revenue. Here's a stripped-down illustrative example — call this company Acme Widgets, a $1B revenue business:

Revenue is the total dollar value of goods and services sold. This is the top line and the least manipulable number on the statement. COGS is the direct cost of producing what was sold — raw materials, direct labor, freight. Gross profit is what's left, and gross margin (gross profit ÷ revenue) is arguably the single most important number in the entire filing. It tells you whether the company has pricing power. A software company with 80% gross margin has structural advantage; a grocer with 25% gross margin is playing a volume game where every basis point matters.

Operating expenses are the costs of running the business that aren't tied directly to production: R&D (product development), SG&A (salespeople, executives, HR, office rent). Subtract these from gross profit and you get Operating Income — sometimes called EBIT (Earnings Before Interest and Taxes). Operating margin is the second-most-important number on the page. It tells you how efficient management is.

Below the operating line come interest expense (cost of debt), taxes, and finally net income — the bottom line, what belongs to shareholders. Divide by shares outstanding and you get EPS. EPS is what analysts forecast and what the "beat/miss" headlines reference.

What to actually watch on the income statement

• Revenue growth YoY — quarter-over-quarter is noisy; year-over-year is the honest comparison. Growth companies should show >15% YoY consistently.
• Gross margin trend — 8 quarters in a row. Expanding = pricing power. Compressing = commoditization or cost inflation.
• Operating margin trend — same 8-quarter view. Rising operating margin on flat revenue = operational leverage kicking in (very bullish).
• EPS growth vs. revenue growth — if EPS is growing faster, buybacks and margin expansion are helping. If EPS is growing slower, dilution and cost creep are hurting.

2. The Balance Sheet — "how healthy are they right now?"

The balance sheet is a snapshot in time. It follows one iron rule: Assets = Liabilities + Shareholders' Equity. Everything the company owns is financed either by money it owes (liabilities) or money the owners have put in and retained (equity). Both sides must balance to the penny.

What to actually watch on the balance sheet

• Cash position. How many quarters of operating expenses does the cash pile cover? For an unprofitable growth company, this is life expectancy. Below 4 quarters of runway = danger zone.
• Debt-to-Equity ratio. Total debt ÷ shareholders' equity. Rule of thumb: <1 for tech/software, <2 for industrials, higher is fine for regulated utilities and banks.
• Current ratio. Current assets ÷ current liabilities. Above 1.5 = comfortable. Below 1 = watch for a liquidity squeeze.
• Receivables growing faster than revenue. Huge red flag. It means the company is booking sales but not collecting cash — often because it's stuffing the channel or extending generous credit to weak customers.
• Inventory growing faster than revenue. Same idea from the other side — product is piling up unsold, which usually forces future markdowns and gross-margin compression.
• Goodwill as a % of equity. If goodwill is more than half of shareholders' equity, the company has grown mostly by acquisition and is one bad quarter away from a big impairment writedown.

3. The Cash Flow Statement — "is the cash real?"

This is the statement Buffett reads first, because it's the hardest to fake. Net income is an accounting construct with dozens of judgment calls (when to recognize revenue, how fast to depreciate assets, what counts as a one-time charge). Cash is cash. The cash flow statement reconciles net income back to actual cash movement, in three sections:

• Cash from Operations (CFO). The real one. Starts with net income, then adds back non-cash charges (depreciation, stock-based compensation) and adjusts for changes in working capital (receivables, inventory, payables). This is the cash the actual business generated.
• Cash from Investing (CFI). Money spent on capital expenditures (capex — new factories, equipment, software), acquisitions, and investments in securities. Almost always negative for a growing company.
• Cash from Financing (CFF). Money raised or returned via debt issuance/repayment, stock issuance, share buybacks, and dividends. A mature company usually shows negative CFF (returning capital); a growth company often shows positive CFF (raising capital).

The single most important derived number: Free Cash Flow

Free Cash Flow (FCF) = Cash from Operations − Capital Expenditures. This is the money the business genuinely throws off after keeping the lights on and investing to maintain itself. FCF is what pays dividends, funds buybacks, retires debt, and — for the owner — compounds real wealth. A company that grows revenue but never generates FCF is a treadmill; a company that grows FCF per share at double digits for a decade is a compounding machine.

What to actually watch on the cash flow statement

• CFO > Net Income. Over a full year, operating cash flow should exceed net income (because of depreciation add-back). If net income is growing but CFO is flat or declining, earnings quality is deteriorating — a classic pre-blowup pattern.
• FCF trend. Is FCF per share rising over 3–5 years? Buybacks that shrink share count while FCF grows are the cleanest form of shareholder return.
• Stock-based comp (SBC) as % of revenue. Tech companies often show great "adjusted" earnings by treating SBC as a non-cash item. But SBC dilutes you as an owner. If SBC is >10% of revenue, adjust reported earnings down.

The Six Red Flags Every Retail Trader Should Watch

1. Declining gross margins for 2+ consecutive quarters. Pricing power is eroding.
2. Receivables growing faster than revenue. Channel stuffing or non-paying customers.
3. Inventory growing faster than revenue. Unsold product piling up; future markdowns coming.
4. Rising debt with flat or declining EBIT. The interest coverage ratio (EBIT ÷ interest expense) is falling. Below 3x is a warning; below 1.5x is a crisis.
5. Operating cash flow lagging net income for 2+ consecutive quarters. Earnings quality is deteriorating.
6. Frequent "one-time" charges. If restructuring, impairments, or goodwill writedowns appear every 3–4 quarters, they aren't one-time — they're the business.

How the three statements connect — the tie-in that matters

The three statements are not independent — they're three views of the same underlying reality, and they must tie together. Net income from the income statement flows into retained earnings on the balance sheet. Depreciation on the income statement gets added back on the cash flow statement (because it's a non-cash charge) but reduces PP&E on the balance sheet. Capex on the cash flow statement increases PP&E on the balance sheet.

This inter-connection is why fraud is hard. To fake earnings, you have to fake one statement. To fake the whole picture, you have to fake all three consistently — and consistent fakery leaves telltale patterns (like the receivables blowout that always accompanies revenue fabrication, or the CFO-vs-net-income divergence that always accompanies aggressive revenue recognition). When Enron and WorldCom blew up, the tells were visible in the cash flow statement for years before the news broke. Reading cash flow first isn't just a Buffett quirk; it's a fraud detector.

Segment reporting — under-appreciated by retail

Buried in every 10-K and 10-Q is a segment reporting section that breaks revenue, operating income, and sometimes assets by business unit or geography. This is where the real story often lives. A company reporting flat consolidated revenue may be masking a fast-growing cloud segment offsetting a shrinking legacy hardware segment. That mix shift can be a huge investment thesis by itself — Microsoft's re-rating in 2015–2020 was fundamentally a story of Azure growing from 5% to 40% of revenue while Windows became a smaller share. If you don't read segments, you miss the composition change.

When you look at segments, watch for: which segment is growing fastest, which segment has the highest margin, and whether management is investing (capex, R&D) in the fastest-growing segment or milking it to fund the declining one. Milking growth to fund decline is a classic value-destruction pattern (Kodak, Sears, Nokia).

The 5-Minute 10-K Skim — the retail workflow

You will never read a 10-K cover-to-cover. Nor should you. Here's the ruthless triage:

1. MD&A (Management Discussion & Analysis) — the first 10 pages. This is management's own narrative about what happened and why. Read it, but with skepticism — every management team spins. Look for what they don't mention as much as what they emphasize.
2. Risk Factors — skim, looking specifically for new risks not in the prior year's 10-K. New risk factors are legally required disclosures; they always signal something. A comparison tool (many available online) will diff this year's Risk Factors against last year's in 30 seconds.
3. Segment reporting — 60 seconds. Which segment is growing, which is dying? Where is capital going?
4. Income statement 3-year trend — revenue, gross margin, operating margin, net income. 30 seconds.
5. Cash flow statement — is FCF growing? Is CFO tracking net income?
6. Legal proceedings section — surprises hide here. New lawsuits, SEC investigations, regulatory actions.
7. Related-party transactions footnote — buried near the back. Executives lending money to their own companies, or family members on the payroll. Red flag for governance issues.

Total time: 15–25 minutes for a competent skim. This is 10x more work than most retail traders do, and it will keep you out of most disasters.

Quarterly vs. annual — what changes

The 10-K is the annual report — audited, comprehensive, roughly 100–300 pages. The 10-Q is the quarterly report — reviewed (not fully audited), shorter (30–80 pages), and less detailed. Auditors sign off on 10-Ks but not 10-Qs, which is why fraud tends to show up in the quarters and then get "discovered" in the year-end audit. As a retail trader, you'll read 10-Qs 3 times a year and a 10-K once. Read each 10-K in full skim mode; treat 10-Qs as delta reports — what changed since the last quarter?

📊 Chapter 20. Key Ratios & Metrics — P/E, PEG, P/B, ROE, ROIC, and How to Actually Use Them FUNDAMENTAL

🎯 The core belief

Ratios exist so you can compare apples to apples. A $500 stock isn't expensive; a $10 stock isn't cheap. What matters is what you're getting per dollar of price: how much earnings, how much book value, how much cash flow. Every ratio in this chapter is a way to normalize price. But ratios only work in context — a P/E of 5 on an airline is a warning, not an opportunity; a P/E of 40 on a genuine 40% grower is defensible, not insane. This chapter teaches the ten ratios that matter for retail traders, when each works, and — critically — when each misleads.

P/E — Price to Earnings

Formula: Price per share ÷ EPS. Or equivalently, market cap ÷ net income.

Meaning: How many dollars you're paying for each dollar of annual earnings. A P/E of 20 means you're paying $20 for $1 of yearly profit — a 5% earnings yield.

Typical ranges:

• S&P 500 historical average: ~16–17x
• High-growth tech: 25–40x (sometimes 50+ for hypergrowth)
• Utilities and staples: 15–20x
• Banks and financials: 8–14x
• Cyclicals (autos, chemicals, airlines) at peak: 5–8x — this is a warning, not a bargain

When it misleads: Cyclicals at peak earnings look absurdly cheap. Airlines in 2007 traded at P/E of 5 — right before a 90% collapse. The trick: at the top of a cycle, "E" is temporarily inflated. When earnings mean-revert, P/E balloons and price crashes. Rule: be suspicious of a low P/E in a cyclical business near a macro top. Also useless for unprofitable growth companies (P/E is negative or undefined).

PEG — P/E to Growth

Formula: P/E ÷ expected annual earnings growth rate (in %).

Meaning: Peter Lynch's favorite. Adjusts P/E for growth. A P/E of 30 on a 30% grower has PEG of 1.0 — reasonable. A P/E of 30 on a 5% grower has PEG of 6.0 — insane.

Rules of thumb: PEG < 1.0 = potentially undervalued growth. 1.0–1.5 = fairly priced. >2.0 = paying for hope. Best for genuine growth stocks (revenue growth >15% YoY); breaks down for value or cyclical names.

When it misleads: Depends entirely on the "G" — expected growth. Analyst estimates for growth are systematically optimistic, especially for beloved names. Always sanity-check PEG using trailing growth, not forward estimates.

P/B — Price to Book

Formula: Price per share ÷ book value per share (equity ÷ shares outstanding).

Meaning: How many dollars you're paying for each dollar of accounting net worth. Ben Graham's classic value screen was P/B < 1.5.

Where it works: Asset-heavy businesses — banks, insurers, REITs, refiners, industrials with real factories. For a regional bank, P/B of 1.0 vs. 1.5 is a meaningful difference in what you're paying.

Where it's useless: Asset-light businesses — software, consulting, media, luxury brands. Microsoft's "book value" is almost meaningless; its real value is its installed base and network effects, none of which appear on the balance sheet. Applying P/B to software makes everything look permanently "expensive."

P/S — Price to Sales

Formula: Market cap ÷ trailing 12-month revenue.

Meaning: The fallback when a company isn't profitable yet and P/E is meaningless. Useful because revenue is the least-manipulable line on the P&L.

Rough typical ranges by industry:

• Software / SaaS: 5–15x (10x is common for high-growth)
• Consumer staples: 1–3x
• Retail: 0.5–1.5x
• Commodities and refining: 0.5–2x
• Biotech (pre-revenue): meaningless

When it misleads: Ignores profitability entirely. Two SaaS companies at 8x P/S can be radically different if one has 80% gross margin and the other 40%. Always pair P/S with gross margin.

EV/EBITDA — Enterprise Value to EBITDA

Formula: (Market cap + total debt − cash) ÷ EBITDA (Earnings Before Interest, Taxes, Depreciation, Amortization).

Meaning: Better than P/E for comparing companies with different debt levels, because EV captures the full cost of buying the whole business (equity plus debt, minus cash you'd inherit).

Typical ranges: 10–15x is "normal" for stable businesses. 6–9x = potentially cheap. >20x = premium growth or bubble.

Why pros prefer it: A company loaded with debt can look "cheap" on P/E while being genuinely expensive on EV/EBITDA. Private-equity buyers price everything in EV/EBITDA terms.

Limitation: EBITDA excludes real costs (depreciation is real — machinery wears out). Charlie Munger famously said "every time you see the word EBITDA, substitute 'bulls*** earnings.'" Take his point seriously, especially for capex-heavy businesses.

ROE — Return on Equity

Formula: Net income ÷ shareholders' equity.

Meaning: Buffett's favorite quality metric. How many cents of profit the company generates per dollar of owner capital. Consistently >15% for 5+ years = high-quality business.

Rule of thumb: Elite compounders show 20–30%+ ROE year after year. Mediocre businesses show 5–10%. Banks target 12–15% ROE as a hurdle.

Watch out: ROE can be artificially inflated by leverage. If a company doubles its debt and does the same operating profit, ROE rises even though the business hasn't improved. Always cross-check ROE with debt-to-equity. Which is why the next ratio matters more.

ROIC — Return on Invested Capital

Formula: NOPAT (net operating profit after tax) ÷ (equity + debt).

Meaning: The truer quality measure. ROIC tells you the return on all the capital in the business — equity plus debt. Leverage can't juice it.

Rule of thumb: Elite companies compound at >20% ROIC over cycles (think MSFT, VISA, MA, MCO). Above 15% consistently = high-quality. Below the company's cost of capital (typically 8–10%) = the company is destroying value even if it appears profitable.

Damodaran's rule: Long-term stock returns approximate ROIC. A business that compounds capital at 20% will, over decades, produce roughly 20% shareholder returns — provided it can keep reinvesting at that rate.

Dividend Yield

Formula: Annual dividend per share ÷ price per share.

Meaning: The cash yield you get for holding the stock. Sweet spot: 2–4% with a track record of annual increases.

Danger zone: Above 5–6% often signals distress — the market is pricing in a dividend cut. Kinder Morgan (KMI) yielded 10%+ in late 2015 right before slashing its dividend 75%. AT&T (T) yielded 7%+ in 2021 before a major cut in 2022. If the yield looks too good, the market usually knows something.

What matters more than yield: Dividend growth. A stock yielding 2% and growing the dividend 10% a year will out-yield a 5% stock with a flat dividend inside a decade.

Debt-to-Equity

Formula: Total debt ÷ shareholders' equity.

Meaning: How leveraged the company is. Interpretation is entirely industry-dependent.

• Tech / software: <0.5 is healthy
• Industrials: 0.5–1.5 is normal
• Utilities: 1.0–2.0 is normal (stable regulated cash flows can support debt)
• Banks: not comparable — use different metrics (Tier 1 capital ratio)

Red flag: D/E rising quarter after quarter while EBIT is flat or falling. That's a company borrowing to plug operational holes.

Current Ratio

Formula: Current assets ÷ current liabilities.

Meaning: Can the company pay its bills over the next 12 months? Above 1.5 = comfortable. Between 1.0 and 1.5 = tight. Below 1.0 = liquidity crunch risk.

Note: A very high current ratio (>4) can indicate poor capital allocation — the company is sitting on too much cash and receivables rather than reinvesting or returning capital.

Using ratios together — the 30-second retail screen

Individual ratios are noise. Combinations tell a story. Here are three battle-tested profiles:

Notice that "growth" and "value" filters barely overlap. That's the point — they're describing different animals. The "quality compounder" filter is the intersection Buffett and Munger hunt for: businesses that grow steadily, generate real cash, and compound at high rates of return with minimal debt. There are maybe 100 of these on the whole planet at any given time.

The Piotroski F-Score — nine yes/no questions

Joseph Piotroski (University of Chicago, 2000) proposed a 9-point checklist to separate genuine value stocks from value traps. Each check scores 1 if yes, 0 if no. A stock scoring 8 or 9 out of 9 has historically outperformed low-score stocks by a meaningful margin over subsequent 12 months.

1. Positive net income
2. Positive operating cash flow
3. Operating cash flow greater than net income (earnings quality)
4. ROA higher than the prior year
5. Long-term debt lower than the prior year
6. Current ratio higher than the prior year
7. No new share issuance in the past year
8. Gross margin higher than the prior year
9. Asset turnover higher than the prior year (using assets more efficiently)

The F-Score is especially useful in the low-P/B universe, where roughly half the "cheap" stocks are junk. Applying the F-Score filter turns a mediocre value screen into a genuinely useful one.

Ratios by industry — the context cheat sheet

Ratios only make sense in context. A P/E of 30 is expensive for a bank and cheap for a software company. Use this table as a sanity check when a screen produces something that looks unusual for its industry.

⚖️ Chapter 21. Growth vs Value — The Two Disciplines and How to Know Which You Are FUNDAMENTAL

🎯 The core belief

There are exactly two ways to make money in stocks. Value says: the market misprices things; buy $1 of assets for 50 cents and wait for the discount to close. Growth says: the market underestimates compounding; buy $1 of earnings that will become $5 of earnings and let time do the work. Both philosophies have generated billionaires. Both have gone through decade-long droughts where they underperformed the other by 300+ percentage points. They require different data, different patience, different risk rules, and — most importantly — different temperaments. Trying to switch between them opportunistically is the fastest way to be bad at both.

The value trader's signatures

A value trader hunts for cheap, boring, and unloved. The mental frame is: "The market is a manic-depressive; my job is to buy when it's depressed and sell when it's manic." Concrete signatures on a screen:

• P/E below sector median (often 30–50% below)
• P/B < 3 (for asset-heavy names, sometimes < 1.5)
• Dividend yield above sector median, with a multi-year record of payment
• Insider buying > insider selling in the last 6 months (Form 4 cluster analysis)
• Multi-year record of consistent (not spiking) free cash flow
• Boring, sometimes ugly, business — the kind you'd never brag about at a dinner party

Canonical practitioners: Benjamin Graham, Warren Buffett, Seth Klarman, Joel Greenblatt, Howard Marks. Their books are your syllabus.

Sample tickers today (illustrative, not recommendations): mature financials (JPM, WFC, BAC), consumer staples with pricing power (KO, PG, PEP), integrated energy (XOM, CVX), select industrials trading below trend multiples.

Value traps to avoid: The dangerous cousin of a value stock is a melting ice cube — a business whose earnings are permanently declining, so it always looks "cheap" but keeps getting cheaper. Print newspapers post-2005, video rental post-2008, mall REITs post-2015. The tell: revenue declining year over year for 3+ consecutive years, not just one bad year. A true value stock is temporarily depressed; a melting ice cube is structurally declining. Learn the difference or the value discount will eat you alive.

The growth trader's signatures

A growth trader hunts for accelerating, expanding, and dominant. The mental frame is: "The market underestimates how big the winners can get; my job is to identify the compounders early and hold through the ride." Concrete signatures on a screen:

• Revenue growing 20%+ YoY for 4+ consecutive quarters
• Expanding gross and operating margins (not compressing)
• Insider selling absent or minimal — steady holding, not distribution
• Relative Strength Rating 80+ (O'Neil's RS scale — top 20% of market by price performance)
• Stock making new 52-week highs while its sector is also strengthening
• Clear "new" story — new product, new management, new market, new regulation

Canonical practitioners: Philip Fisher, T. Rowe Price (the original), Peter Lynch, William O'Neil, Mark Minervini, David Ryan.

Sample tickers today (illustrative, not recommendations): high-quality large-cap tech leaders, best-in-class scientific instruments and biotech, dominant platform businesses expanding into adjacencies.

Growth traps to avoid: The dangerous cousin of a growth stock is the parabolic story stock — a name that runs 5x in a few months on narrative alone, with no earnings backing. When the story cracks (and it always cracks), these fall 80–95% and rarely recover. The tell: revenue growth without corresponding earnings growth, cash-burn accelerating rather than decelerating, and a chart that has moved 100%+ above its 200-day moving average. Peter Lynch's warning: "Long shots almost always miss the mark."

When each style outperforms — the historical record

Growth and value trade off in long cycles. If you don't understand the cycle, you'll think whichever style is currently working is the only one that ever works.

The pattern: growth wins when rates are falling and money is cheap; value wins when rates are rising and cash today matters more than promises. This isn't a law of physics, but it's the strongest empirical regularity in the style-rotation literature. See Chapter 17 (Regime Stack) for how to read the current regime.

Which one YOU should be — a retail-honest section

Most retail traders never make this choice consciously, which is why most retail traders lose. Read these two profiles and be honest about which one you actually are:

Pick one style. Learn it deeply. Trade it consistently. Judge yourself against the appropriate benchmark (Russell 1000 Value for value, Russell 1000 Growth for growth) and against the appropriate time horizon (3–5 years for value, 6–18 months for growth). Do not try to be both, do not switch styles mid-drawdown, and do not judge a value strategy by 6-month growth-market returns or vice versa.

How our platform serves both disciplines

• Growth toolkit: The CAN SLIM Scanner filters for O'Neil-style acceleration (Current + Annual earnings, New products, RS 80+). The Wave Analyzer and VCP scan identify momentum setups and Volatility Contraction Patterns for entry timing.
• Value toolkit: The Deep Analysis page surfaces P/E vs sector median, P/B, dividend yield, FCF trend, and insider buying/selling — the raw material for value screens. The Piotroski F-Score (referenced in Chapter 20) filters value candidates by 9 fundamental quality checks, separating true bargains from melting ice cubes.
• Regime-aware for both: The Regime Stack (Chapter 17) tells you which style is currently favored by the macro environment — so you know whether your discipline is in tailwinds or headwinds.

📅 Chapter 22. Earnings — Season Mechanics, Guidance, Surprise, and Post-Earnings Drift FUNDAMENTAL

🎯 The core belief

Earnings are the closest thing to a scheduled catalyst that public markets offer. Four times a year, every U.S.-listed company reports quarterly financials, and for a few days around each report the stock trades on that information alone. This is when short-term price discovery is at its most concentrated — and when the biggest single-day gains and losses in a trader's career are typically made or avoided. Earnings can also be the single fastest way to blow up an account, because the moves are binary and gap-through-stops. This chapter is about how earnings season actually works, what data points matter, and the four coherent strategies a retail trader can run around prints.

Earnings season basics

There are four earnings seasons per year, each roughly 6 weeks long, kicked off by the big banks and closed out by retailers and small caps:

Between seasons is "no-man's land" — a 2–4 week window when major catalysts are rare and price action is dominated by macro. Peak volume in any given quarter is usually week 3, when the biggest names (AAPL, MSFT, GOOGL, META, AMZN, NVDA) report clustered within 5–7 trading days. That week alone can move the entire S&P.

The 4 key data points on any earnings day

When a company reports, four numbers determine the immediate stock reaction:

1. EPS actual vs consensus. The classic "beat" or "miss." Analyst consensus is the average of published estimates. Beating by pennies is meaningless; beating by 5–10%+ is a real surprise.
2. Revenue actual vs consensus. Same idea for the top line. In modern markets, revenue misses are often punished more than EPS misses — because EPS can be engineered via buybacks and cost-cutting, but revenue can't easily be faked.
3. Forward guidance. The single most price-sensitive data point. Management tells the market what to expect next quarter or next year. Options: raise, maintain, lower, or withdraw. Withdrawing guidance ("suspending guidance due to uncertainty") is almost always taken as bearish — even in a nominally strong print.
4. Special items. Restructuring charges, spinoffs, dividend changes, buyback authorizations, executive departures. A big new buyback announcement can turn a mediocre print into a green day; a surprise CFO resignation can wreck a great print.

The three post-earnings scenarios

The market's reaction is almost never just about the beat/miss — it's about the combination of the print and the guidance. Three patterns dominate:

Post-Earnings Announcement Drift (PEAD) — the academic edge

In 1989, accounting professors Victor Bernard and Jacob Thomas published a landmark paper documenting one of the most robust anomalies in market history. Their finding:

• Stocks that beat consensus by more than 5% tend to keep drifting upward for roughly 60 trading days after the print.
• Stocks that miss consensus by more than 5% tend to keep drifting downward for roughly 60 trading days after the print.
• The magnitude of the drift is roughly proportional to the size of the surprise.

Why does this happen? Analysts are slow to revise their estimates. When a company delivers a big beat, analysts don't immediately raise their forward estimates to reflect the new run-rate — they revise gradually over the next 4–8 weeks. Each upward revision triggers a fresh wave of buying. The market, in effect, digests the news too slowly.

Why hasn't PEAD been arbitraged away? It has been partially compressed since 1989, especially in mega-caps. But it persists in mid-caps, small-caps, and less-covered names, where fewer analysts create slower estimate revisions. It also persists during periods of market inattention (mid-August, late December).

How to trade PEAD: On the day after a big beat (>5% surprise) with raised guidance and a positive price reaction, enter on the next day's pullback. Hold for 20–60 trading days. Cut on a break of the earnings-day low.

Whisper numbers — the unofficial consensus

The published analyst consensus is the "official" number. But there's often an unofficial consensus circulating among institutional traders, tracked by services like WhisperNumber.com and increasingly reflected in options positioning. This "whisper" number is what buy-side desks actually expect, and it can differ materially from the published number.

The critical pattern: a stock beats the published consensus but misses the whisper. To retail traders scanning news feeds, this looks like a beat — "why is it down 8%?" The whisper was 5 cents above consensus; the beat was only 3 cents above. The stock sells off because the buy-side hurdle was higher than the sell-side hurdle.

Rule of thumb: If a stock has run 10%+ in the two weeks before earnings, assume the whisper is higher than consensus. A published beat may not be enough. Conversely, if a stock has been weak into earnings, published consensus is probably the real hurdle — and a small beat can produce a big rally.

Estimate revisions — the leading signal

You don't have to wait for the print to guess what's coming. Analyst estimate revisions in the 2–4 weeks before earnings are one of the strongest leading signals in finance:

• Cluster of estimate cuts in the 30 days before a print → typically a miss coming, or at best a low-quality beat.
• Cluster of estimate raises (especially within 5 days of the print) → typically a beat coming. Analysts are getting whispers from IR desks and adjusting.
• Wide dispersion of estimates (analysts far apart) → high uncertainty, big move likely in either direction. Options prices will already reflect this.

The retail trader's earnings playbook — 4 options

Broadly, there are four coherent things a retail trader can do around any given earnings print:

1. Skip earnings entirely. The safest option. Close positions the day before the print, re-enter after the dust settles (day 2 or 3). You avoid the binary event. You give up the possible gap-up, but you also avoid the gap-down that gaps through your stop. For anyone still learning position sizing, this is the default recommendation.
2. Half-position through earnings. Trim to half your normal size the day before the print. If it gaps up, you make half. If it gaps down, you lose half. Bridge between paths 1 and 4.
3. Trade the drift (PEAD). Ignore the print itself. On day 2 after a big beat + raise, when the initial reaction has settled and the setup is clean, enter on a pullback. Hold 20–60 days. This is a systematic edge, not a gamble.
4. Trade the reaction (advanced). Options structures (long straddles, iron condors) or gap-fill plays on day 1. Requires understanding of implied volatility, options pricing, and gap statistics for the specific name. Not for beginners; expected value is often negative for retail without an edge.

Special earnings-adjacent events to know about

• Pre-announcements. Some companies pre-announce results 1–3 weeks before the official print, usually because the number will be far above or below expectations and legal counsel wants to avoid selective disclosure. A positive pre-announcement is bullish; a negative one is very bearish.
• Guidance-only updates. Between earnings, some companies issue mid-quarter guidance updates (usually cuts). These are treated as effectively an earnings event and can move the stock 15%+.
• Analyst days / investor days. Scheduled events where management presents longer-term guidance. Often coincide with new product cycles and multi-year targets. Can produce sustained multi-week rallies.

🕵️ Insider Transactions & 13F — Following the Money That Knows FUNDAMENTAL

🎯 The core belief

Two groups have information advantages over us: the executives running the company, and the institutions with billion-dollar research budgets. Both are legally required to disclose what they do — but with a delay. Reading those disclosures right is edge; reading them naively is a trap.

Insider Transactions (SEC Form 4)

Officers, directors, and 10%+ shareholders must file a Form 4 within 2 business days of any transaction. It's the fastest legal window we get into what people running the company are actually doing with their own money.

The Buy vs Sell Asymmetry

This is the single most important rule of reading insider data:

• Insider selling has many innocent explanations: taxes on vesting, diversification, house purchase, divorce, ordinary planned sales (Rule 10b5-1). Selling is weak signal.
• Insider buying has exactly one explanation: they think the stock is going up. Buying is strong signal.
• Seyhun's landmark work: insider buying outperforms the market by ~5% annualized; insider selling has almost no predictive power.

What Counts as a Real Buy Signal

• Cluster buying: 2+ insiders buying inside 30 days. One insider might be wrong; three officers writing checks together is a message.
• Buy after a decline: insiders buying into a 15%+ drawdown. They're not chasing — they see value at those prices.
• Meaningful dollar amounts: a $50K buy by a CEO earning $2M/year is symbolic. A $500K buy is a bet.
• C-suite > directors: CEO/CFO/COO carry more information weight than outside board members.

Selling Signals Worth Watching

Most selling is noise. But two patterns matter:

• Cluster selling: 3+ insiders selling inside 30 days without any offsetting buying, when the stock is at highs. Rare, and historically precedes underperformance.
• 10b5-1 plan cancellations: an insider who set up a scheduled sales program and then cancels it was selling into strength and no longer wants to. Contrarian buy signal.

13F Institutional Ownership

Institutions with $100M+ AUM must file Form 13F within 45 days after quarter-end. Discloses long equity positions. Two big caveats:

What 13F Actually Tells You

• New positions: managers who didn't own it last quarter and do now. High-conviction initiations are worth researching.
• Position increases (adds): existing holders increasing conviction. Especially notable when a manager doubles a position.
• Position decreases (trims): partial exits. Can be tax-driven, rebalancing, or real conviction loss — need context.
• Full exits: managers who dumped completely. Read the news for what happened.
• Ownership concentration: what % of a stock is held by 13F filers? Very high (>80%) means retail plays second fiddle; very low (<30%) means retail-driven — very different price dynamics.

Following the Names Who Actually Know

Not all 13F filers are equal. Focus on managers with documented long-term outperformance, transparent process, and concentrated portfolios:

• Berkshire Hathaway — Warren Buffett's holdings are the most-copied 13F on Earth.
• Baupost, Third Point, Pershing Square, Greenlight — value/activist managers with concentrated books.
• Tiger Global, Coatue, Whale Rock — growth-oriented, tech-heavy.
• Renaissance, Two Sigma, DE Shaw — quant funds; positions turn over fast so 13F is less useful for them.

How to act

• On watchlist adds: cluster insider buying is a reason to research the name, never a reason to buy blind. Combine with the platform's CAN SLIM quality gate before entry.
• On existing positions: cluster selling by C-suite when the position is up big is a valid trim signal — take partial profits.
• 13F is a hunting ground, not a signal: use it to discover names you haven't seen, then run your own process. The 45-day lag disqualifies it as an entry trigger.

🐋 Chapter 24. Institutional Flow — How to See What Big Money Is Actually Doing FUNDAMENTAL

🎯 The core belief

Big money leaves footprints. You will never see the order book that a Citadel or a Millennium desk sees, and the trades that matter most are hidden in dark pools by design. But the residue is public: 13F filings, 13D/13G threshold alerts, unusual options activity, block prints on the tape, Form 4 insider transactions, and simple volume-vs-price behavior. Learning to read that footprint is one of the highest-leverage skills a retail trader can build — it compresses roughly two years of "how does this actually move" experience into a single, learnable discipline. This chapter shows you the six windows and how to line them up.

13F filings — the quarterly hedge-fund window

Every institutional investment manager with more than $100M in US equity assets under management must file Form 13F within 45 days of quarter-end, disclosing their long positions in reportable securities. That means a March 31 (Q1) snapshot lands in the public record by roughly May 15. A June 30 snapshot lands around August 14. The data is a photograph of what a fund owned on the last day of the quarter — not what they own today.

What to actually check. Use a free aggregator like WhaleWisdom, HedgeFollow, or Fintel and build a small watchlist of "tier-1" managers whose thesis quality is high: Berkshire Hathaway, Baupost, Third Point, Pershing Square, Appaloosa, Greenlight, Scion, Viking, Tiger Global. Scan the change columns: new positions, added >25%, and concentrated bets where one manager owns more than 5% of the free float. When three or more high-quality funds independently add to the same name in the same quarter, that is a meaningful cluster.

Honest limits. The 45-day lag means you are always late. Fast money (event-driven, macro) may have completely rotated by the time you read the filing. 13F does not disclose short positions, options exposure, non-US listings, cash, or fixed income — so a "value" investor whose 13F shows big financial-stock longs may actually be net short via puts. Treat the filing as a starting hypothesis, not a signal. If Baupost adds to a beaten-down industrial and the stock is still near the same price 45 days later, the thesis probably has room. If it already ran 40%, you missed it — write down the name and wait for a pullback.

13D and 13G — the >5% concentration alerts

The Securities Exchange Act requires anyone acquiring beneficial ownership of more than 5% of a public company's voting shares to file with the SEC within ten days (recently tightened to five business days for 13D under 2024 SEC amendments). This is a much fresher signal than 13F.

13D = activist intent. The filer is signaling they may seek to influence control — board seats, capital return, strategic review, sale of the company. 13D announcements historically produce a 5-15% pop on the day the filing hits, and the stock often trends for weeks as the market prices in an activist campaign. 13G = passive holder who crossed 5% without intent to control (index funds, pension funds, quant shops that simply grew into the position). 13G is informative — it tells you a large, patient buyer exists — but it is not a catalyst.

What to actually check. Set an alert on SEC EDGAR full-text search or a service like WhaleWisdom for 13D filings in names on your watchlist. Read the "Purpose of Transaction" section (Item 4) — that is where the activist telegraphs the plan.

Dark pool prints and block trades

Roughly 35-45% of daily US equity volume executes off-exchange, in alternative trading systems (ATSs), internalizers, and dark pools. This is not a conspiracy — it is how institutions move size without pushing the price against themselves. Michael Lewis's Flash Boys made the mechanics public: a portfolio manager who wants to buy 500,000 shares does not send that order to the NYSE; they slice it across dark venues, print in blocks, and only the final trade reports show up on the consolidated tape, often with a delay.

Block trades — prints of 10,000 shares or more — are visible on the tape but you have to look. Retail data feeds show them as spikes in the time-and-sales window, or as "unusual volume" bars on daily charts.

The rule of thumb. If today's total volume is more than 2.5x the 20-day average and the closing price barely moved (say, within 1% of yesterday's close), someone was accumulating or distributing discreetly. The direction is decoded by the following days: if price grinds up on continued elevated volume, it was accumulation; if it fades, distribution. This is the essence of Wyckoff (Level 3 · Wyckoff Method) and it works on any liquid name.

How the platform helps. The Confluence Score treats abnormal-volume days as a flow signal and combines them with price behavior, so you do not have to eyeball 200 tickers by hand.

Unusual Options Activity (UOA)

Options are leveraged, dated, and require a directional bet — which makes them the preferred vehicle for anyone with a real edge. When options volume runs more than 5x the 20-day average on a specific strike or expiry, someone is positioning. When it happens without a corresponding move in the stock, that "someone" is often ahead of information.

Bullish tells. Large call sweeps — aggressive market-buy orders that sweep across multiple strikes and exchanges — are the signature of a fund trying to get filled quickly, price be damned. Short-dated OTM calls with heavy volume ahead of a known catalyst (earnings, FDA date, investor day) are common. So are call ratios where volume dwarfs open interest, meaning the position is new.

Bearish tells. The mirror image: put sweeps, heavy near-dated OTM put buying, and put/call volume ratios that spike above 2.0 on a name that hasn't fallen yet.

Real-world patterns. Pre-M&A option runs are the classic case — Activision-Microsoft, Twitter-Musk, and dozens of smaller deals saw abnormal call volume in the sessions before announcement. Pre-earnings positioning shows up as directional skew in the weekly expiry. Biotech pre-data prints show up as violent long-vol buying (both calls and puts) as funds bet on magnitude, not direction.

Honest limit. Not every UOA print is smart money. Retail 0DTE and meme-driven flow can also spike the tape. Filter for premium paid (dollar-weighted), not just contract count. A million-dollar sweep is different from ten thousand $0.05 lottery tickets.

Insider transactions — Form 4

Corporate insiders — officers, directors, and beneficial owners of more than 10% of a company's stock — must file Form 4 within two business days of any transaction. This is the freshest primary-source flow signal you can get.

The Peter Lynch rule. Insider buying is a stronger signal than insider selling. Insiders buy for exactly one reason — they think the stock is going up. They sell for many reasons: diversification, taxes, tuition, divorce, a boat. So weight buys heavily and discount sells unless the pattern is extreme (mass C-suite exits, for example).

Cluster buys — three or more insiders buying in the same week, especially with personal cash rather than option exercises — are a strong bullish setup and have been backtested to outperform the market by several percentage points over the following six to twelve months (Larry Williams's work, later replicated by academic studies).

CFO buys outrank CEO buys. The CFO lives in the numbers; the CEO lives in the narrative. When the person who signs the 10-K puts personal money in, listen. Ignore 10b5-1 plan sales — those are pre-scheduled and largely noise. Cross-reference this with Level 4 · Insider Transactions and 13F for the full workflow.

Volume-price divergence — the free tell

You do not need a Bloomberg terminal to spot institutional footprints. The relationship between price change and volume tells you who is in control:

• Rising price on rising volume = confirmed accumulation. Trend intact.
• Rising price on declining volume = distribution. Institutions are selling into retail buying. Warning.
• Falling price on rising volume = confirmed distribution. Get out.
• Falling price on declining volume = capitulation ending. Selling is exhausted, often buyable.

This is Wyckoff's core insight — accumulation and distribution leave signatures — applied to plain daily bars. Anyone can read it in five seconds per chart.