Docs / Reading the Card

Reading the Card

Each symbol in your watchlist gets its own card. The card gives you a complete regime picture — from structural state to tail risk to intraday patterns. This guide explains every number and what to do with it.

⚡ Attention Score always visible

A 0–100 score that answers one question: how much does this symbol deserve your attention right now? It aggregates multiple signals into a single number so you can scan quickly across your watchlist.

0–25LOW
26–50MODERATE
51–75ELEVATED
76–100HIGH
FieldDescriptionHow to use it
Score (0–100) Composite score from regime instability, vol spikes, tail risk, and structural signals Use as a triage filter. HIGH (75+) means something structural is happening — dig into the details.
Reasons Top 2 drivers of the attention score Tells you why the score is elevated. "Regime shift detected" is more significant than "Fat tails increasing".
💡
Beginner tip Start your morning scan by looking at Attention scores. Any symbol above 50 is worth reading in detail. Below 25 — the market is quiet, your current setup probably still applies.
💊 Stat Pills (Micro & Macro) always visible

Four statistical properties of price returns, measured at two timeframes: Micro (short-term, ~4h) and Macro (longer-term, ~7d). Each pill shows a value and a confidence score.

Micro summaryCHAOTIC · STABLE
Pred0.27 c0.61
Drift0.08 c0.61
Tails1.00 c0.77
Miss0.00 c0.77
PillWhat it measuresHow to interpret
Pred Predictability — how much structure exists in recent returns High (0.7+) = returns have exploitable patterns. Low (below 0.3) = random noise. Most markets are between 0.1–0.4.
Drift Directional bias — is price drifting up or down? High positive = strong upward drift. Near zero = no directional tendency. Useful for trend filters.
Tails Fat-tail intensity — how heavy the return distribution tails are High (0.8+) means extreme moves are more likely than normal. This is when position sizing matters most.
Miss Model miss rate — how often the model's predictions were wrong High miss rate suggests the market is in a structural break or the model is struggling. Reduce reliance on signals.
c (confidence) Statistical confidence in the pill value, 0–1 Below 0.5 means the value is based on limited data — treat with caution. Above 0.8 = reliable estimate.
ℹ️
Micro vs Macro Micro captures what's happening right now. Macro captures the broader structural context. If Micro shows HIGH_VOL but Macro shows LOW_VOL, you're in a short-term spike within a calm environment — likely to revert.
🎯 Risk Score always visible

The large number in the top-right of the card. A 1–10 score of current market risk level, derived from volatility forecasts. Think of it as the "temperature" of the market for this symbol.

1–3LOW 🟢
4–5MEDIUM 🟡
6–7ELEVATED 🟠
8–9HIGH 🔴
10EXTREME 🧊
⚠️
Score 8+ = reduce size A score of 8 or higher means the model sees elevated volatility risk. This is not a signal to buy or sell — it's a signal to adjust your position sizing. High risk score + high tail asymmetry is the most dangerous combination.
📋 Metrics Grid always visible

Six core metrics displayed in a 3×2 grid below the header.

MetricDescriptionHow to interpret
Price Last known price Reference point. For crypto and forex, updates every 30 seconds. For stocks, updates during market hours.
Return 24h Price change over the last 24 hours Context for the current move. Compare with σ 24h — a 2% move when σ=5% is small; the same move when σ=0.4% is large.
σ 24h Expected daily volatility (1 standard deviation move) The typical size of a daily move for this instrument. Use for position sizing. The arrow (↑/↓) shows if current vol is above or below its historical average. ~$X shows the dollar equivalent at current price.
Regime Current market structure label with confidence See the Regime section below for full breakdown. Confidence score in parentheses — above 0.7 is reliable.
Uncertainty Model uncertainty about the current regime HIGH uncertainty means the model sees conflicting signals. Treat regime label with caution. LOW = model is confident.
Attention Attention score repeated here for quick reference Same as the ⚡ score at the top of the card.

Regime Types

TREND
Price is moving consistently in one direction with momentum. Trend-following strategies work best. Mean-reversion strategies fail here.
RANGE
Price oscillates within a defined zone. Mean-reversion and range-trading setups perform well. Breakout strategies get faked out.
TRANSITION
The market is shifting between regimes. This is the most dangerous state — strategies that worked recently may suddenly stop working. Reduce size.
HIGH VOL
Elevated volatility with erratic price action. Both trend and range strategies struggle. Risk management becomes the priority.
UNKNOWN
Insufficient data or conflicting signals prevent a clear classification. Treat as TRANSITION — proceed with caution.
💡
Practical rule Before entering a trade, check that the regime matches your strategy's assumptions. A momentum entry during RANGE regime, or a mean-reversion entry during TREND regime, is fighting the environment.
🎲 Tail Asymmetry always visible

The probability that the price will move more than ±3% in the next 24 hours — in each direction separately. This reveals whether the market has asymmetric downside or upside risk.

Tail ±3% (24h) 16.80% ↓  |  8.30% ↑
ElementMeaningAction
p_down (↓) Probability of a −3%+ move in next 24h High downside probability = hedging or reduced longs make sense.
p_up (↑) Probability of a +3%+ move in next 24h High upside probability = market is pricing in potential squeeze or breakout.
Color Red if down/up ratio >2.5×, yellow if >1.5×, green if up > down Red means significantly more downside risk than upside — be careful with long positions.
⚠️
Not shown for low-volatility instruments EURUSD and similar forex pairs rarely move ±3% in a day, so tail values are 0 and the row is hidden. This is correct — it means tail risk at that threshold is negligible for that instrument.
🚨 Alert Chips always visible

Small badges in the card footer that fire when specific structural events are detected.

ChipMeaningWhat to do
Shift: X → Y A regime transition was detected with high confidence The market just changed character. Re-evaluate any open positions that were predicated on the previous regime.
Vol INVERTED Short-term volatility is higher than long-term — inverted term structure Usually signals stress or a spike. Similar to VIX backwardation in equity options — often precedes a move.
Confidence ↓ Data quality has degraded vs recent history Treat current signals with more caution. May indicate unusual market conditions or data gaps.
🔮 Regime Forecast (HMM) details panel

A Hidden Markov Model trained on realized volatility forecasts the probability of regime change at three horizons. This tells you how stable the current regime is likely to be.

Current HMM stateHIGH_VOL
Confidence0.46
1h P(change)45.1%
4h P(change)60.6%
24h P(change)61.5%
FieldMeaningInterpretation
Current state LOW_VOL / MEDIUM_VOL / HIGH_VOL — the HMM's classification of current volatility regime Note: this is the volatility regime from the HMM, separate from the structural Regime (TREND/RANGE/etc) in the main card.
Confidence Model confidence, 0–1 Below 0.3 = model is uncertain. Above 0.7 = high confidence in the current state classification.
P(change) 1h/4h/24h Probability that the volatility regime will be different at that horizon P(change) >60% at 1h = regime is unstable, likely to shift soon. Low P(change) = current regime expected to persist.
ℹ️
How to read this together with the Regime field If Regime = RANGE and HMM P(change 24h) = 65%, it means: the market is ranging now but the model expects this to change within a day. Don't build range-trading positions expecting them to work for long.
📈 Volatility Surface details panel

Volatility estimates across four time horizons (1h, 4h, 24h, 7d), comparing the model's prediction vs empirical (realized) vol. The ratio tells you whether short-term vol is elevated or suppressed relative to longer-term.

Term structureNORMAL (0.25)
1h model / emp / ratio0.40% / 0.40% / ≈1.000
24h model / emp / ratio3.57% / 2.39% / ▲▲1.495
FieldMeaningInterpretation
Term structure NORMAL / INVERTED / FLAT NORMAL = short vol < long vol (healthy). INVERTED = short vol > long vol (stress signal — similar to yield curve inversion).
Model vol GARCH-based forecast for each horizon What the model predicts volatility should be at that horizon.
Empirical vol Realized (historical) volatility at that horizon What actually happened. Compare with model — large divergence signals unusual conditions.
Ratio ≈ / ▲ / ▲▲ Model/empirical ratio. ≈ = within 10%, ▲ = 10–25% above, ▲▲ = >25% above ▲▲ at 24h means the model expects much higher vol than recent history — the market may be underpricing risk.
🕐 Intraday Heatmap details panel

Shows which hours of the day (UTC) are historically most and least volatile for this instrument. Updated from 30 days of data.

FieldMeaningPractical use
Cell brightness Relative activity vs daily average. 1.0x = average, 2.0x = twice as active Bright cells = high-volume hours with wider moves. Enter during active hours if you want liquidity. Avoid if you want quiet fills.
NOW marker The current UTC hour, highlighted with a white border Tells you where you are in the daily activity cycle right now.
Now label NORMAL / HIGHER / MUCH_HIGHER / LOWER / MUCH_LOWER — current hour vs its own baseline MUCH_HIGHER means this specific hour is seeing more activity than usual for this time of day. Often precedes a move.
Peak / Quiet The most and least active hours historically For EURUSD the peak is 13–15 UTC (London/NY overlap). For BTCUSD it's relatively flat. Plan your trading sessions accordingly.
⚙️ Calibration details panel

How well-calibrated the model's probability estimates are. A well-calibrated model that says "50% probability" should be right 50% of the time. This section shows the actual vs target coverage.

GradeGOOD (0.002)
Band 50%49.9% → 50% ✓
Band 80%80.0% → 80% ✓
Band 95%95.0% → 95% ✓
FieldMeaningInterpretation
Grade GOOD / FAIR / POOR with a calibration error score POOR calibration means the model's confidence levels are unreliable. Weight tail and scenario probabilities less.
Band 50/80/95% Actual coverage vs target. Error shows the gap. ✓ = perfect. A −5% error on Band 80% means the 80% confidence interval only captured 75% of actual outcomes — the model is slightly overconfident.
ℹ️
Why this matters If calibration is POOR, the model's volatility forecasts may be systematically too high or too low. A well-calibrated model (GOOD) means you can trust the σ 24h number as a true 1-standard-deviation estimate.
₿ Crypto Block crypto only

Additional metrics sourced from Bybit. Only shown for crypto symbols. Provides market microstructure context beyond price.

Pressure 30LOW / NEUTRAL
Funding z/pct−0.72 · 26.5% · tail 47%
OI z / Δ24h / Δ1h−0.35 · −2.32% · +0.68%
Basis−0.0367%
Imbalance0.913 / 0.411
FieldMeaningInterpretation
Pressure 30 Net buying/selling pressure over last 30 periods HIGH = strong buying pressure. LOW = selling. NEUTRAL = balanced. Useful for short-term directional bias.
Funding z/pct Funding rate z-score and percentile vs history Extreme positive funding = longs are paying shorts — crowded long trade. Extreme negative = crowded short. Both are contrarian signals.
OI z / Δ24h / Δ1h Open Interest z-score, 24h change, and 1h change Rising OI + rising price = strong trend with new money entering. Rising OI + falling price = shorts building. Falling OI = positions closing.
Basis (Mark price − Index price) / Index price Positive basis = perpetual trading at premium. Negative = discount. Large positive basis often precedes funding spikes.
Imbalance top10/top50 Order book bid/ask imbalance at top 10 and top 50 levels Above 0.7 = strong bid pressure near price. Below 0.3 = strong ask pressure. Good for short-term entry timing.
📐 Levels & Liquidity details panel

Key price levels derived from volume profile and VWAP analysis. Shows where price is relative to important reference points, and how liquid the market currently is.

FieldMeaningHow to use it
VWAP 24h Volume-weighted average price over the last 24 hours The "fair value" anchor for intraday traders. Price above VWAP = bullish intraday bias. Below = bearish. Used as the scenario trigger.
VWAP dist Current price distance from VWAP in % Large negative = price stretched below VWAP, potential mean-reversion opportunity. Large positive = stretched above.
Range pos Position within the 24h range, 0%=low, 100%=high Below 20% = near the bottom of the daily range. Above 80% = near the top. Context for entry timing.
Zone HVN / LVN / MID — High Volume Node, Low Volume Node, or middle HVN = price near a heavy trading zone — expect support/resistance. LVN = price in a vacuum zone — moves can accelerate quickly.
POC Point of Control — the price level with most volume in the last 24h Strong magnetic level. Price tends to return to POC. Good target for mean-reversion or area to watch for reaction.
Liquidity score 0–100 score of current market liquidity Below 50 = thin liquidity, expect slippage. Above 80 = deep market, normal execution. Important for sizing larger positions.
🗺️ Scenarios details panel

Two conditional price scenarios based on VWAP position. Not predictions — conditional projections based on current structure and volatility.

Bull triggerPrice holds above VWAP 24h
Target 1 / 2227.09 / 234.93
Bear triggerPrice breaks below VWAP 24h
Target 1 / 2178.74 / 172.78
FieldMeaningHow to use it
Bull scenario If price holds above VWAP — upside projection based on volatility and range structure Use Target 1 as a realistic near-term objective, Target 2 as extension. Not guaranteed — these are volatility-based projections.
Bear scenario If price breaks below VWAP — downside projection Use as stop/target reference. If tail↓ is high, the bear scenario has higher probability than the number alone suggests.
Tail note Warning if tail asymmetry makes one scenario more likely "If tail↓ high — possible cascade" means combine with tail asymmetry. High downside tail + bear scenario = elevated risk of fast move lower.
⚠️
These are not trade recommendations Scenarios show where price could go given structure and volatility — not where it will go. Always use your own risk management and stop-loss levels.
📖 Glossary

Quick reference for terms used throughout the card.

TermDefinition
σ (sigma)Standard deviation — the statistical measure of volatility. "1σ move" means a move of one standard deviation, which happens about 68% of the time in a normal distribution.
HMMHidden Markov Model — a statistical model that infers hidden states (like market regimes) from observable data (price returns).
GARCHGeneralized AutoRegressive Conditional Heteroskedasticity — a model for forecasting volatility that accounts for volatility clustering (calm periods and stormy periods tend to cluster).
VWAPVolume-Weighted Average Price — the average price weighted by trading volume. Often used as a benchmark and intraday fair value.
POCPoint of Control — the price level with the highest traded volume in a given period. Acts as a gravitational center for price.
HVN / LVNHigh Volume Node / Low Volume Node — areas of high or low trading activity in the volume profile. HVN = support/resistance, LVN = fast-move zones.
Funding rateIn perpetual futures: the periodic payment between longs and shorts to keep the contract price close to spot. High positive = longs pay shorts (crowded long).
Open Interest (OI)Total number of outstanding futures contracts. Rising OI = new money entering. Falling OI = positions closing.
BasisDifference between the perpetual contract price and the spot index price. Positive basis = futures at premium to spot.
Tail riskThe probability of extreme outcomes — the "fat tails" of the return distribution. Markets have fatter tails than a normal distribution predicts.
CVaRConditional Value at Risk — the expected loss given that a tail event occurs. More informative than simple probability.
Z-scoreHow many standard deviations a value is from its historical mean. Z > 2 = unusually high. Z < −2 = unusually low.