Cryptowatch Macro — cryptowatch.id

Smart Alert Cards surface significant intelligence shifts with full context: base rate (historical hit rate), historical analog (when this pattern occurred before), confidence level, and suggested action.

Alerts trigger on: NGS extremes, regime flips, cycle phase shifts, consensus breaks, truth engine flips, and entity regime shifts.

Surfacing daily moves you'd otherwise miss. For each curated metric we compute the day-over-day delta and divide by the 30-day rolling standard deviation of that metric's deltas — that's a z-score. |z|>2 is a 95th-percentile move for that metric in normal regime. Top-N rows by |z| are surfaced here.

How to read it. Bigger |z| = more unusual move for this specific metric (so a 0.5% drop on something that normally moves 0.05%/day ranks higher than a 5% drop on something that swings 5%/day). Use it as the "first scroll" — anything above |z|=2 deserves a look at its panel.

Which ingredients moved the composite. For each component of the composite score we compute the day-over-day delta and divide by its 30-day rolling standard deviation — a z-score. |z|>2 is a 95th-percentile move for that component.

How to read it. This tells you why the composite changed, not just that it changed. A big |z| on "AVIV Ratio" means valuation shifted; a big |z| on "Funding" means leverage sentiment shifted. Use it to diagnose composite moves without re-reading every panel.

The decision-support layer. For every historical day going back to 2011 we compute the bottom-proximity composite — the same vote+weight logic that powers the Cycle Bottom Radar (Puell + Pi-Cycle + 2Y MA + Mayer + AHR999 + Reserve Risk + NVT + MVRV-Z + LTH-MVRV + NUPL + ETF flows). Score is 0–100 and measures how close we are to a historical cycle-bottom signature. Days are binned into six cycle-phase tiers; the chart then shows BTC's actual 7d / 30d / 90d / 180d forward returns for every prior visit to each tier.

How to read it. The "you are here" pin marks the tier your current score falls into. The big number on each card is the median 90d forward return after a score in that tier; the small range below is the p25–p75 interquartile band. Tighter range = more historical agreement; wider range = the tier is regime-mixed and forward returns vary.

How to use it. A score of 80 in DEEP ACCUM with median +35% / p25 +5% / p75 +80% over 90d is a different bet than the same 80 reading with median +10% / p25 −20% / p75 +50% — same tier, very different conditional distribution. Treat the median as a base rate, not a forecast.

Honest scope: this is the bottom-proximity composite, NOT the live ±24 SIGNAL verdict. The live SIGNAL also folds in L1 macro and L4 derivatives layers that don't have multi-year cached history yet — so they can't be back-tested honestly. When that history is built, this panel will swap to the full SIGNAL.

This panel is a CONTEXT layer. The 3-AI debate engine reads the same data you see on this dashboard and synthesises a daily narrative. The synthesis is useful for "what's the room thinking today" but its short-horizon directional accuracy has not crossed trade-grade (60%) on most fields. Do not size positions off the verdicts here. The cycle radar, Fair-Value Cluster, and Desk NO-TRADE panels are the trade-grade surfaces.

3-AI debate engine. Every morning at 08:10 WIB, three specialist AIs independently analyze the same data, then the Orchestrator weighs their arguments and produces a final verdict.

The debaters: On-Chain Analyst (on-chain framework specialist — reads AVIV, LTH SOPR, Tier 1, cycle position), Macro Analyst (macro analyst with live web search — CPI, FOMC, DXY, breaking news), Sentiment Analyst (sentiment specialist — social attention feed mood, whale alerts, narrative shifts).

The orchestrator: Deep Analysis Engine identifies where they agree (high conviction) vs disagree (requires judgment), weights each AI by their domain strength per timeframe, and resolves conflicts with transparent reasoning.

BTC Price Direction: Weighted synthesis of composite score, momentum (F&G vs price), IV regime, and on-chain stress. Not a price target — just directional bias for today.

Crypto Market Cap: Follows BTC direction, modulated by rotation unlock status and alt momentum.

Overall Sentiment: The human-readable regime label. Combines cycle position, fear level, LTH behavior, and macro backdrop into a single word.

BTC Trade Bias: Should you be looking at longs, shorts, or neither? Based on composite direction, funding regime, IV positioning, and cycle phase. NEUTRAL = no trade zone — the expected value of taking a position is too close to zero.

Confidence levels: >65% = high conviction (multiple layers agree). 50-65% = moderate (some disagreement). <50% = low (conflicting signals, treat as coin-flip).

Calibration window: the framework's directional accuracy is graded 7 days after each call, on a rolling sample. Trade-grade threshold is 60%; below that the field is shown as context only. See the per-field scorecard at the bottom of the panel.

The first filter the legacy on-chain toolkit was missing. Per the next-gen indicator framework: any indicator that ignores macro is structurally incomplete post-2020. BTC moves in M2's direction 83% of the time on a 12-month basis (Lyn Alden, Sep 2024). Fed rate hikes in 2022 directly drained liquidity from BTC and broke the 200W MA support that had held in every prior cycle.

M2 YoY (US): rolling year-over-year change in US M2 money supply. >3% rising = RISK-ON, 0–3% = neutral, <0% = RISK-OFF, <−2% = STRONG RISK-OFF. Input class: public macro liquidity series.

DXY 3m: US Dollar Index 3-month % change. Rising DXY = dollar scarcity = risk-off. Falling = liquidity expansion = risk-on. Note: post-ETF the DXY relationship has weakened (r² 0.70 → 0.45) — never use alone, always cross-reference with M2.

BTC-SPX 30d correlation: regime detector. >0.6 = BTC behaves as risk asset, macro dominates. 0.3–0.6 = mixed. <0.3 = BTC decoupled, on-chain dominates. <0 = safe haven / gold-like. This card doesn't push the gauge directionally — it tells you which other layers should dominate your read.

How to use: if all 3 cards align RISK-ON, the macro environment is BTC-friendly and you can lean into Tier 1/2 confluence signals. If macro is RISK-OFF, even a strong on-chain confluence may struggle against tightening liquidity (see 2022 cycle).

Stablecoin velocity measures how fast USDT and USDC are moving on-chain. Computed as transfer_volume / market_cap for each stablecoin.

High velocity = tokens circulating fast (speculation, arbitrage, payments) — risk-on signal.

Low velocity = tokens sitting idle in wallets or DeFi (hoarding, risk-off) — liquidity is parked.

Divergence signal: when USDT velocity rises while USDC velocity falls, it often means retail/speculative money is moving while institutional/DeFi money is parked — or vice versa.

Hong Kong spot BTC ETF daily flows. Three HK-listed funds: ChinaAMC, Harvest, and Bosera & HashKey. These are the only spot BTC ETFs available to mainland Chinese investors via the Stock Connect program.

Net flow >0 = creations (new BTC entering custody) — bullish for Asia demand. Net flow <0 = redemptions — distribution pressure.

How to use: when HK flows are strongly positive while US flows are neutral/negative, it signals Asia-led demand rotation. Persistent outflows suggest mainland appetite is cooling.

Public and private company BTC treasury holdings from bitcointreasuries.net. Tracks Strategy (MSTR), miners, ETFs, governments, and private firms with material BTC on their balance sheet.

Total BTC held by all tracked entities is a structural supply-absorption metric. When public companies are accumulating, it removes float from circulation. When they distribute, it adds supply pressure.

How to use: watch the top holders list for concentration risk. If one entity controls >5% of circulating supply, their liquidation risk becomes a macro factor. Miner holdings are particularly sensitive to hash-price regime shifts.

Net on-chain flow by segment. Aggregated via labeled-entity intel feed. Tracks Exchange (CEX), DEX, and Smart Money flows for BTC, ETH, and SOL.

Signal: Positive exchange flow = coins moving to CEX (sell pressure). Negative exchange flow = coins leaving CEX (accumulation). Smart money direction serves as a leading indicator for structural trends.

One question, three lenses: is the current BTC bid structural or just leverage?

• Leverage (derivatives analytics, 25-exchange OI 24h change) — are traders stacking risk or unwinding? Big 24h drops = forced deleveraging = contrarian bull. Big 24h builds = crowded, fragile.
• Supply (derivatives analytics, 20-exchange CEX reserve 30d netflow) — are BTC coins leaving exchanges (outflow = cold storage = supply crunch = bullish) or piling onto exchanges (inflow = pre-sell signal)?
• Demand (derivatives analytics, US spot BTC ETF concentration) — is institutional flow broadening (healthy) or dangerously concentrated in one fund (fragile)?

The synthesis: when 3 of 3 are bullish, the rally has structural legs and leverage is NOT the driver. When 2 of 3 are bullish, watch the odd one out carefully. When ≤1 is bullish, any price action is suspect — it's likely leverage-led and vulnerable to a flush.

Short-horizon overlay (weeks, not months/years). While the Cycle Bottom Radar above tracks long-cycle valuation, this panel tracks tactical positioning from widely cited on-chain and derivatives indicators.

STH Cost Basis is the single most-watched line — price above this level = bull regime, below = bear regime. Currently the universal "where is the bull/bear divide" anchor across institutional research.

Funding Regime at sustained record-negative is a contrarian-bullish setup — bears paying carry, short squeeze fuel building.

Unified verdict from ~30 data sources. Most dashboards show each signal in isolation. The Truth Engine ingests all of them, normalizes to a single scale, detects where sources disagree, applies domain-weighted resolution rules, and tells you: what's the consensus, where do sources conflict, and what's the strongest counter-case?

Four primitives unified: (1) Cohort Confluence — LTH/ETF/Smart Money alignment; (2) Trap Divergence — price-vs-signals divergence flags; (3) 3-AI Debate Engine — cross-model consensus with steelman; (4) Composite Gauge — 4-layer weighted vote.

Conflict types: pairwise — two sources disagree; cross-layer — Macro says one thing, On-Chain says another; temporal — short-term (funding/sentiment) vs long-term (halving/Tier 1).

Standard cycle indicators tell you WHERE you are. This panel tells you WHO is driving it. In the ETF era, "LTH supply dropping" can mean either bearish distribution (2021 top) or bullish institutional handoff (2024 ETF lock-up). Entity resolution is the difference.

10 regime types: Institutional Handoff, Retail Distribution, Miner Capitulation, Miner OTC Absorption, Double Exit, Structural Bid, Government Dump, Smart Money Rotation, Stablecoin Buildup, and Regime Confusion.

Zero new API credits. Reads existing entity caches and cycle parquets. No additional private-source calls.

6 independent cycle indicators running in parallel to our on-chain-analytics-derived cycle_phase output. Use these for cross-validation — if all 6 agree with our composite, conviction is high. If they diverge, something in one of the data pipelines is off.

• Puell Multiple — miner revenue vs 365d MA. <0.5 = capitulation (bottom), >4.0 = euphoria (top).
• Pi Cycle Top — 111d MA vs 2× 350d MA ratio. When ratio hits 1.0, Pi Cycle signals the cycle top (fired within days of 2013/2017/2021 tops).
• 2Y MA Multiplier — price vs 730d MA. Below MA = historical buy zone, above 5× MA = historical sell zone.
• AHR999 DCA — DCA efficiency indicator popular among Chinese traders. <0.45 = DCA zone, 0.45-1.2 = moderate, >1.2 = overheating.
• Mayer Multiple — price / 200D SMA. <0.8 = bottom-quintile of cycle history (DCA), >2.4 = top-decile (sell). Trace Mayer's original 2013 ratio.
• Hash Ribbons — 30D vs 60D MA of BTC hash rate. Charles Edwards' miner-capitulation indicator. 30D below 60D = miner stress; cross back above after capitulation = historical buy signal (fires +35-92d after the actual bottom; lagging confirmation).

Three jobs this panel does: (1) risk management — tells you if you're doubled up on the same trade; (2) regime detection — auto-classifies current market from VIX + BTC/SPX; (3) one actionable signal — DXY 3-month trend, the only correlation-based forward-return predictor that survived backtest (86% WR when DXY falls >3% over 3 months, n=442).

The tiers:

• Tier 1 Macro — DXY, VIX, 10Y, M2. Tells you the environment BTC is operating in.
• Tier 2 Equity / Commodity — SPX, NDX, Gold, Oil. "Same desks" regime check.
• Tier 3 Crypto Internal — ETH, SOL, XRP, HYPE. Zero intra-crypto diversification currently (r>0.93 for the majors).

Column legend: 30d / 90d rolling Pearson correlation on daily returns. Δ arrow = week-over-week corr trend. Values grayed out when p > 0.05 (statistically insignificant — ignore them).

Backtest findings embedded:

• Gold leading BTC by 3-7 months is a myth on daily returns (max r=+0.02 at 90d lag).
• 10Y yield is statistically insignificant (p=0.227 full period).
• ETH is BTC with leverage: r=0.93. Holding both = 2x same risk.
• DXY 3m trend is the only signal that passed the forward-return test: 86% WR.

How to use: start with the hero — if DXY is triggered, size up BTC. Check the regime detector — if VIX <18, you're in risk-on mode. Scan the crypto tier — if you hold BTC + ETH + SOL, recognize they're the same position. Ignore grayed-out (noise) rows entirely.

Head-to-head comparison of the 3 orthogonal smart-money cohorts. Most dashboards show each cohort in isolation. This panel answers the one question they don't: do the 3 lenses agree right now?

The 3 cohorts:

• LTH (Long-Term Holders) — on-chain cohort >155 days held. ~14.7M BTC. Contrarian by design: accumulates during fear, distributes during euphoria. The old "follow LTH" heuristic.
• ETF / TradFi Institutional — US spot BTC ETF creations/redemptions. ~$130B+ of BTC in custody. Behaviorally trend-following in aggregate (flows correlate with 30d returns) but with a sticky long-only component (RIAs, pensions) that adds structural bid.
• Smart-Money Labels — wallets labeled by historical PnL track record. Contrarian like LTH but smaller pool, faster rotation. The "stable rotation" metric shows when they deploy captive stables into risk.

Conviction levels:

• 🟢 HIGHEST · 3/3 aligned — all three lenses pointing the same direction. Rarest + highest-signal setup. Historically fires at major cycle inflections.
• 🟡 MEDIUM · 2/3 aligned — two cohorts agree, one disagrees. Check the odd-one-out: it's either early (the leader) or wrong (the laggard). Usually the contrarian cohorts (LTH + SM) lead, and the trend-follower (ETF) lags by days to weeks.
• 🔴 LOW · divergent — all three split. No directional signal, but the divergence itself is information. Usually happens in mid-cycle chop or regime-transition moments.

How to use this: when conviction is HIGHEST and direction is bull → highest-probability accumulation setup. When HIGHEST and direction is bear → highest-probability distribution/top-risk setup. When MEDIUM, treat the odd-cohort as the key variable to monitor for flip. When LOW, wait for resolution — the panel is literally telling you "no clear direction yet."

What this panel doesn't do: predict price direction. It tells you what the three smart-money cohorts are doing RIGHT NOW. Cohorts can be wrong. But when all three agree, they're much less wrong than any single lens alone.

Top 15 labeled on-chain transfers in the last 24 hours, ≥$10M (auto-lowers to $5M on quiet days). BTC + ETH + SOL. Requires at least one labeled side via labeled-entity intel.

Adaptive regime: when CEX flows dominate, the badge shows sell-pressure / accumulation. When DeFi protocols or funds dominate the labeled activity, the badge shifts to "DeFi rotation" or "Fund activity" — telling you to discount the CEX-centric framing.

Flow type colors: rose = CEX deposit (sell pressure), aquamarine = CEX withdrawal (accumulation), amber = MM rail, purple = DeFi, sky blue = issuer, grey = entity-to-entity. Miner rows get an extra accent.

Unified view showing bottom/top triggers (Tier 1 + Tier 2) sorted by proximity, with BTC-USD price translation for each.

Regime-Conditional Probability Distributions. Instead of pointing to a single price target, this panel projects the statistical distribution of BTC returns 10, 30, 60, and 90 days out—specifically filtered for periods where the macro regime and on-chain state match today.

Canonical Crypto Fear & Greed Index. 0–100 scale aggregating volatility, market momentum, social media, surveys, dominance, and Google trends into one number.

How to read: 0–24 Extreme Fear · 25–49 Fear · 50 Neutral · 51–74 Greed · 75–100 Extreme Greed. Historically a contrarian indicator — readings below 25 have marked cycle bottoms (March 2020, June 2022, November 2022 FTX), readings above 85 have marked local tops.

How to use: combine with direction (rising from extreme fear = early accumulation signal; falling from extreme greed = early distribution signal). The 7d and 30d deltas shown here capture that velocity. A single low reading isn't enough — look for sustained multi-week extreme readings for cycle-turn conviction.

Why a separate panel from Sentiment: the Sentiment panel next to this one shows social attention mood from the AI synthesis layer. F&G is the aggregated canonical index. When they agree, conviction is high. When they diverge (e.g., extreme fear here but neutral social mood), something in the composite is driving it that social isn't catching — usually volatility or the dominance component.

Live BTC price chart with the full cycle-level stack overlaid: Realized Price (cyan), LTH Realized Price (green floor), STH Cost Basis (rose ceiling), True Market Mean (white dashed), 200-Week MA (purple). All levels are computed live — RP from on-chain analytics, 200WMA from HL daily candles, LTH/STH/TMM from the on-chain pipeline.

DCA zone shading: the 5 accumulation zones (A→E) are drawn as semi-transparent horizontal bands anchored to live RP × the backtested multipliers (0.70 / 0.88 / 1.00 / 1.11 / 1.25 / 1.33). They auto-rescale as RP moves.

How to read: when price enters a shaded zone, that's the trigger to deploy the DCA allocation for that zone. When price closes below RP for > 14 days, the Tier 2 "Price vs RP" signal fires.

Input class: BTC candle feed + on-chain analytics pipeline (overlay levels).

Plain English: we logged the Bottom Proximity score every day for ~1.5 years. For each historical day, we know what BTC's price was — and what it was 7d / 30d / 90d later. This panel groups those days by score range (0-20, 20-40, …) and shows what the typical "what happened next" looked like for the range your current score falls into.

How to read: the headline tells you the median return at the most decision-relevant horizon (30d). The short-horizon row shows the 7d / 30d / 90d median and p25→p75 range. The full bin comparison is collapsed below because it is reference detail, not the main read.

Caveats: sample sizes are small (~1.5 yr of daily samples, sometimes <20 per bin at long horizons). Past forward returns ≠ future. If the macro regime has shifted (ETF flows, Fed policy), apply judgement.

The AI debate engine's qualitative directional read on each major alt. Macro's headline BTC Composite Signal IS the BTC verdict — this panel is the alt-companion. ETH, SOL, and HYPE often diverge from BTC and from each other; this panel exposes those divergences so you can size each book independently rather than copying the BTC stance across the board.

Why no BTC card here. BTC's bias is the BTC Composite Signal verdict above. Showing it twice would be duplicated framing — the gauge there is the canonical BTC read.

How to read. Qualitative bias = the AI's directional read after weighing macro / on-chain / sentiment / cycle for that specific asset. It is not yet an asset-specific cycle model. Conviction = how much the three analyst voices agreed. Size = recommended notional exposure as a % of the slot you've allocated to that asset. Rationale = the one-sentence "why" that drove the bias.

How to use. If BTC Composite Signal reads "MILD BUY +6" and SOL reads "Neutral 30-40%", the SOL slot should run smaller and tighter regardless of how good the BTC setup looks. Divergence is itself a signal — when ETH/SOL conviction is meaningfully higher than BTC, alts may be driving the next leg.

Phase 1 — surfaces what the existing AI debate already produces. Phase 2 will add per-asset cycle phase math (MVRV, NUPL, Puell) once on-chain collectors include ETH + SOL — at which point this panel may merge back into BTC Composite Signal as an asset toggle.

Curated regime-conditional forward returns. Each preset asks one fixed historical question using stored BTC and macro/cross-asset series. This is intentionally not a free-form query builder yet; missing source history returns a visible unavailable state.

Two-axis regime classifier. Macro liquidity (Expanding / Neutral / Contracting) intersected with crypto risk state (Risk-On / Compression / Risk-Off). Each of the 9 cells carries historical forward-return stats and an altcoin beta recommendation.

How to read the grid. The highlighted cell is today's classification. Forward returns show what BTC did over 7d / 30d / 90d after prior visits to that cell. Altcoin beta suggests how much alt exposure to carry: 0.0 = all BTC, 1.5 = high-beta alts.

Transition matrix. Empirical probability of moving from cell X to cell Y based on logged history. More observations = sharper transitions.

Narrative Gravity Score (NGS) measures the gap between social narrative intensity and on-chain fundamental reality.

Positive NGS (red) = narrative is overhyped relative to fundamentals — gravity will pull it down. Negative NGS (cyan) = fundamentals are stronger than the narrative — upside asymmetry as attention catches up. Near zero (blue) = roughly aligned.

Scale: -5 (deep value) to +5 (extreme overhang). Built entirely from existing CryptoWatch data — no Kaito, no LunarCrush.

Post-ETF replacement layer for legacy MVRV-Z. Built on the published Cointime Economics framework (August 2023). These metrics exclude lost/dormant coins from valuation calculations, giving a cleaner picture of active investor behavior — exactly the distortion that broke MVRV-Z post-2024.

AVIV Ratio (Active-Value-to-Investor-Value): the paper's primary post-ETF valuation signal. Long-term mean is consistently ~1.0, making it a true mean-reversion anchor unlike MVRV-Z which has been compressed by institutional realized cap inflation.
Bands: <0.55 extreme undervaluation · 0.55–0.75 deep accumulation · 0.75–1.00 accumulation (BUY) · 1.00–1.50 mid bull · 1.50–2.50 late bull · >2.50 cycle top.

Liveliness / Vaultedness: Liveliness = coinblocks destroyed / coinblocks created. Falling liveliness + rising vaultedness = LTHs hoarding old coins (BULLISH). Rising = LTHs distributing aged supply (BEARISH).

Realized Loss: total $ value of BTC sold at a loss. Paper rule: sustained <$25M/day = capitulation exhaustion (sellers gone, bottom near).

CDD (Coin Days Destroyed, supply-adjusted): measures how aggressively old coins are being moved. High = LTH distribution. Useful for detecting tops.

Honest backtest. For each major cycle inflection since 2015, this table shows what the composite framework would have scored using the exact same logic that's live today, applied retroactively to on-chain analytics data as of that date. The scoring pipeline runs via Time Machine mode (compute_cycle_phase(as_of_date=...)), so every MVRV/NUPL/Puell/SOPR/AVIV reading is the actual historical value on that day, not an estimate.

How to read the accuracy label:

• exact — framework nailed it. Position <5 + Tier1 ≥3 for bottoms, or position >80 + Tier1 0 for tops.
• good — framework got it mostly right (position <10 with 2+ Tier 1, or >60 for tops).
• partial — framework was in the right zone but not at extreme. Still actionable with lower conviction.
• miss — framework did not flag it. These are the honest failures.

What this table is not: it is NOT a future-prediction promise. Past performance ≠ future results. The framework has been tightened post-2022 (e.g., NUPL threshold lowered from <0 to <−0.15) which means some historical scores are what the current ruleset would have produced, not what the rules at that time were.

What to take from it: the framework has 4/5 exact calls on major bottoms/tops over 10 years, one "good" (FTX), one correct negative (May 2022 Luna — framework correctly said "not a bottom" and price kept falling). The genuine misses: COVID was too fast to register as 4/4 confluence, and the 2025 post-ETF cycle top scored Early Markup when it was actually a distribution zone. Know what the framework catches and what it doesn't.

Does the composite signal actually make money? This panel runs the live scoring logic retroactively over 3 years of BTC history and shows the equity curve of a simple strategy vs buy-and-hold + DCA.

Strategies:

• Conviction — long 100% at composite ≥ +6, flat at ≤ −6. Few trades, deep-value entries only.
• Mild — long at ≥ +3, flat at ≤ −3. More trades, less extreme thresholds.
• Buy & Hold — baseline: bought BTC on day 0, never sold.
• DCA — baseline: $1/day equal allocation regardless of price.

Honest scope: backtest uses only 7 Tier-1 L2+L3 signals (LTH MVRV, STH MVRV, LTH NUPL, Puell, MVRV Z, LTH SOPR, AVIV) because those are the metrics with multi-year parquet history. Live composite uses 21 signals including L1 Macro + L4 Execution which don't have long enough cached history to backtest honestly. Expect live results to differ. 10bps taker fee modeled per trade.

What the Sharpe and Max DD tell you: Sharpe = risk-adjusted return (higher is better; >1 is good, >2 is excellent). Max DD = worst peak-to-trough loss during the period. A high-Sharpe strategy with small DD is more valuable than a high-return strategy with brutal drawdowns, because humans can't actually hold through -80%.

Rolling pairwise correlation between the 4-layer framework's representative indicators over the last 90 days using daily log-returns. Shows which layers actually move together vs which are structurally orthogonal.

Why this matters: the framework's edge depends on orthogonal layers. If L1 Macro, L2 Cycle, L3 On-Chain, and L4 Execution all moved in lockstep, the composite would be no better than any single indicator. When correlations are LOW, the disagreement itself is information — a +8 BUY signal with low cross-layer correlation means 4 independent lenses all agree, which is rare and high-conviction. A +8 BUY with everything correlated 0.95 means you're really reading one signal four times.

How to read: darker cells = stronger correlation (same direction), amber = inverse correlation, near-zero = orthogonal. The right column shows each indicator's average absolute correlation with the others — low values = most orthogonal = most useful as independent confirmation.

Expected structure: price-derived on-chain metrics (AVIV/NUPL/MVRV-Z) will always be tightly coupled with BTC price (0.9+) — they're derivatives of price. The real orthogonality test is between macro (DXY, SPX) and market-structure signals (Funding) versus the on-chain stack. Funding near 0 correlation across the board = market structure is doing its job as an independent lens.

Prediction markets price real-money probabilities. Every contract pays $1 if YES resolves true, $0 if NO. Prices are probabilities in real dollars — participants have skin in the game. This panel doesn't dump the raw watchlist. Instead it distills three things that actually matter to our framework: next Fed decision, policy regime for the year, and BTC implied range.

Fed · next meeting — aggregates the April 2026 FOMC contracts (hold / cut 25 / cut 50+ / hike). The dominant probability tells you whether a rate surprise is pricing in. Near-100% HOLD means no Fed catalyst this month.

Policy regime · 2026 — "zero cuts all year" probability + Fed Chair contracts. This is the direction of the year, not one meeting. High "zero cuts" + hawk Chair contract = hawkish regime pricing; low + dove Chair = dovish.

BTC implied range — takes the BTC price threshold contracts (upside at $150K / $250K, downside at $15K) and extracts tail probabilities. Narrow tails = consolidation regime. Fat upside = market pricing parabolic move. Fat downside = capitulation risk priced in.

Dashboard tie-in — the cyan bar compares market-implied pricing against our own LTH cohort read. When prediction markets price a fat downside tail but our LTH is in accumulation signature (or vice versa), that's a trade-worthy divergence — one side is wrong.

How to use: prediction markets are a secondary confirmation layer, not a primary signal. They answer "what does money think?" — useful as a sanity check on our on-chain read. The raw watchlist is still available in the collapsible section below if you want to see individual contracts.

The mirror of Tier 1. Tier 1 detects bottoms with zero false positives. This panel detects cycle tops using 6 distribution indicators backtested against the 2013, 2017, 2021, and 2025 tops. Require 4/6 triggered for STRONG SELL conviction.

The 6 indicators:

AVIV > 2.0 — cointime valuation extreme (replaced MVRV-Z for post-ETF era)
NUPL > 0.55 — Net Unrealized Profit/Loss in euphoria zone
Puell Multiple > 4.0 — miner revenue overextended
MA Stack 4/4 — full bull technical regime (all MAs aligned)
Funding > +54% — extreme leveraged speculation (annualized)
Halving regime = late_markup or distribution_danger — inside the historical top window

How to use: 0-1/6 = no top risk. 2-3/6 = early warning, start trailing stops. 4+/6 = STRONG SELL — execute the Distribution Playbook below. Cross-check with composite gauge: if composite < -8 AND dist confluence >= 4, highest conviction sell signal the framework can produce.

Public track record — every Daily Intel call is timestamped with 30/90/180d forward returns.

Probability-weighted catalyst overlay — tests multi-layer propagation scenarios.