DriftState uses the Adaptive Drift Estimator (ADE) — a KAMA-variant that scales its smoothing constant to market efficiency. The For-Loop Score measures how many lookback periods the ADE is trending upward (max 60). Scores above 30 trigger LONG. Below 10 triggers CASH. The result is a noise-filtered, regime-persistent signal that avoids whipsaws.

Loading DriftState...

The Universal Valuation aggregates 12 independent momentum and valuation signals — RSI, Rate of Change, Price Momentum, Bollinger Band%, Sharpe, Sortino, Omega Ratio, Money Flow Index, CCI, Crosby Ratio, Ehlers Filter, and Chande Momentum — into a single Z-score composite. Each signal is individually Z-scored over a 115-bar rolling window before averaging, ensuring no single indicator dominates. Works on any asset and any timeframe. Above +2: historically overbought/distribution zone. Below −2: historically oversold/accumulation zone.

Computing composite...

Volatility Pulse is a VynthraQuant original composite volatility model. It combines 10 independent mathematical volatility estimators — each capturing a different statistical property of price behavior (close-to-close variance, high-low range, open-close dynamics, GARCH persistence) — into a single normalized reading from 0–100%. Readings above 70% indicate elevated risk. Below 30% signals compression, historically preceding explosive directional moves.

Analysing volatility structure...

A WMA(20) baseline with standard deviation bands at 0.5σ, 1σ, 1.5σ, 2σ, and 3σ. Shows precisely where price sits within its statistical distribution. Outer bands (±2σ, ±3σ) indicate statistically extreme extensions historically prone to mean reversion. Inner bands define normal trending structure.

Loading signal interpretation...

BMD normalizes price within adaptive SMA/SD bands (0–100%). Fast EMA(5) = Long line. Slow EMA(13) = Cash line. Enters LONG when fast > 83, exits to CASH when slow < 60. Candles colored cyan (LONG) or magenta (CASH).

Computing...

Tracks what percentage of the top 40 altcoins have positive momentum over 30 days (medium-term) and 14 days (short-term). Each row shows both readings per coin. When both gauges are high (>70%), broad market participation is healthy. A divergence — where 14D drops sharply while 30D stays elevated — signals that short-term momentum is fading before it shows up in the medium-term number. This is one of the earliest warnings of a weakening market structure.

Loading signal interpretation...

The Valuation Composite averages eight independent valuation signals into a single Z-score: MVRV Z-Score, Macro RSI (normalized), Normalized MA Composite, Power Law percentile, NUPL, Macro CCI, Sharpe Ratio, and HP Filter (rescaled). Each signal is first standardized to a comparable Z-score scale before averaging, so no single indicator dominates. The composite oscillates around zero: above +2 has historically been a danger zone (cycle top vicinity). Below −2 has historically been a maximum opportunity zone. The scoreboard shows the current Z reading of each component.

Computing composite...

Market Value to Realized Value, normalized via Z-score over 365 days. The Z-score shows how far current valuation is from its historical mean in standard deviations. Values above +3 have historically marked cycle tops. Values below −1 have marked cycle bottoms — providing high-conviction long-term entry signals.

Loading signal interpretation...

Net Unrealized Profit/Loss measures the aggregate profit or loss of all coins relative to the price they last moved on-chain. Approximated here using price vs a long-term realized price proxy (EMA-200 as realized cap basis). NUPL above zero means the average holder is in profit — euphoric territory. Below zero means the average holder is underwater — capitulation territory. The Z-score normalizes the signal over history to produce consistent cycle comparisons.

Computing NUPL...

Proper quantile regression of Bitcoin price in log-log space (log price vs log days since genesis Jan 3, 2009). Unlike simple power law + offset, each quantile band (0.1%, 5%, 50%, 95%, 99%) has its own independently fitted slope and intercept. This captures the true statistical distribution of where BTC has traded over its full history, providing probability-based price bands rather than arbitrary offsets.

Calculating quantile position...

Shows where BTC price sits within the quantile regression bands as a 0–100% score. A reading of 95% means price is at the Q0.95 band — historically near cycle tops. A reading near 5% means price is at the Q0.05 band — near historical cycle bottoms. The score is derived from the same quantile regression model as the Quantile Horizon chart above, making both models fully consistent.

Computing cycle position...

A 400-day RSI applied to BTC price strips out all short-term noise and reveals the macro momentum cycle. Unlike the standard 14-day RSI used for trading, this version operates on a timescale of years. Above 80: historically aligns with cycle tops and late-bull euphoria. Below 30: historically aligns with cycle bottoms and maximum fear. The 50 level acts as a bull/bear regime separator — sustained readings above 50 are characteristic of bull markets.

Computing macro RSI...

CVDD divides the Realized Cap by total coin days destroyed — a measure of long-term holder conviction. It produces a price floor that has historically caught every major BTC cycle bottom. When BTC touches the CVDD floor it is one of the highest-conviction accumulation signals in on-chain analysis.

Loading on-chain data...

Measures how far BTC price has deviated from its long-term average across five different MA types: SMA(350), EMA(350), WMA(350), and two exponential variants. Each deviation is normalized to price (percentage above/below), then averaged and Z-scored over the full history. The result shows whether BTC is extended above or compressed below its structural trend. Positive Z = price above trend (expensive). Negative Z = price below trend (cheap on a structural basis).

Computing MA composite...

The Commodity Channel Index at a 500-day length strips out all short-to-medium term noise and operates on the same timescale as crypto market cycles. The standard CCI measures how far price is from its 500-day average in units of its own mean absolute deviation. Applied to BTC daily closes, it produces a cycle oscillator: extreme positive values correspond to late bull tops, extreme negative values correspond to bear market bottoms and accumulation zones.

Computing CCI...

Rolling 400-day Sharpe Ratio measures the risk-adjusted return of BTC over a ~1.1 year lookback. Sharpe = (mean daily return × √400) / (std dev of daily returns). High positive Sharpe means BTC has been delivering strong returns per unit of risk — typical of healthy bull markets. Sharpe turning negative marks bear market conditions where returns do not compensate for risk. Extreme positive readings historically precede mean-reversion. The signal is then Z-scored over history for cycle comparison.

Computing Sharpe...

The Hodrick-Prescott Filter decomposes price into a long-run structural trend component and a cyclical deviation component. This indicator shows the normalized deviation of BTC price from its HP trend — essentially measuring how far price has strayed above or below its smoothed long-run trajectory. Values near zero mean price is tracking its structural trend closely. Large positive deviations have historically preceded tops; large negative deviations have preceded bottoms.

Running HP filter...

Volume Profile maps where the most trading activity occurred over the past 200 days. The Point of Control (POC) is the price level with the highest volume. The Value Area (VAH/VAL) contains 70% of total volume. Price tends to gravitate toward high-volume nodes and accelerate through low-volume gaps.

Calculating volume structure...

For each historical candle, this model estimates where leveraged long and short positions opened near that price would be liquidated at 5×, 10×, 20×, 50×, and 100× leverage. These levels persist horizontally until price sweeps through them. The result is a 2D heatmap showing where the highest concentrations of liquidatable leverage sit at any price level — the "liquidity pools" that large players target. Yellow/green = extreme concentration, dark blue = low.

Scanning for liquidity zones...

Cumulative Volume Delta tracks the running difference between aggressive buy volume (taker buys) and aggressive sell volume (taker sells) using Binance 4H kline data. Each bar represents net buying (green) or selling (red) pressure. The anchored view resets at each UTC day, showing per-session orderflow clearly. Rising CVD confirms bullish order flow. Falling CVD while price rises is a divergence warning. The CVD Profile shows delta accumulation per price level — the Point of Control (POC) is the price where the most aggressive buying or selling occurred.

Analysing order flow...

The DriftState Allocator is a Winner-Takes-Most rotation system across 8 assets using the Adaptive Drift Estimator (ADE). Each asset gets a Loop Score (1–60 lookback). Assets above the LONG threshold (30) qualify. The single strongest asset gets 100% allocation. Below CASH threshold (10) = exit. Fees: 0.06% on each side of turnover. When no asset qualifies, the portfolio goes to full cash. Equity curve shows net-of-fees performance vs BTC buy-and-hold.

Loading allocator...

Gamma Exposure (GEX) measures the total dollar gamma that market makers hold across all open option contracts. Because dealers delta-hedge continuously, their combined gamma position creates mechanical buying and selling pressure in the underlying. The formula is:

GEX = 0.01 × S² × Σ(OI_C·Γ − OI_P·Γ)

The heatmap shows GEX across the full strike × implied volatility surface, using live Deribit open interest. The white dotted lines mark the current spot price and ATM implied volatility — their intersection is where the market sits right now.

■ Green zones = dealers are long gamma here. They buy dips and sell rallies → suppresses volatility, pins price.

■ Red zones = dealers are short gamma here. They must buy when price rises and sell when it falls → amplifies moves, fuels trends.

■ Gamma Flip = the strike where total GEX crosses zero. Above the flip: pinning regime. Below: trending regime. The flip level acts as a structural support/resistance.

■ DTE filter = narrowing to shorter expirations shows gamma that expires soon — highest sensitivity and pin potential for this week/month.

Loading GEX data from Deribit...

GEX is a snapshot metric — it reflects the current open interest distribution. It updates on each refresh. Data source: Deribit Public API. Formula based on BackQuant methodology.

Note: Deribit options are on BTC and ETH only. SOL/SUI/PAXG not available.

Absolute GEX splits total gamma exposure into its two components. Call GEX (green) represents the gamma dealers hold from call options — when price rises, they sell to stay delta-neutral, creating resistance. Put GEX (red) is the gamma from put options — dealers buy as price falls, creating support. Where call GEX dominates above spot: ceiling. Where put GEX dominates below spot: floor. The 0DTE view isolates same-day expiry — highest gamma per dollar, highest pin potential.

The GEX Profile shows Call GEX minus Put GEX per strike as a filled curve. Green above zero means dealers are net long gamma at that strike — they act as shock absorbers, suppressing volatility and pinning price. Red below zero means net short gamma — dealers amplify moves in both directions. The transition from red to green is the structural support/resistance boundary. In 0DTE view: this boundary is the key intraday level — a sustained break below it shifts the market from dampened to explosive regime.

Loading GEX profile...

The Probability Surface derives the risk-neutral probability density function (PDF) from live Deribit options data using the Breeden-Litzenberger formula. For each expiry, the IV smile is interpolated across strikes, Black-Scholes call prices are computed on a fine grid, and the second derivative with respect to strike yields the PDF. Each curve shows the market's implied probability distribution of where the asset price will be at that expiry. The mode is the most likely price, EV the expected outcome. Skewness above zero means the market prices a right tail — asymmetric upside — while negative skewness implies crash risk is priced in. Excess kurtosis above zero indicates fat tails (higher chance of extreme moves than a normal distribution implies).

Loading options chain...

The VQ Neural Composite processes 5 technical features (RSI, MACD histogram, Bollinger Band width, Stochastic oscillator, ATR ratio) through z-score normalization and a tanh activation function with optimized weights. The output oscillates between −1 (maximum bearish) and +1 (maximum bullish). Dynamic standard deviation bands define overvalued and undervalued zones. When the composite crosses above the midline, trend shifts bullish. When it enters the upper SD band, the asset is in distribution territory. The methodology is inspired by single-layer neural networks but adapted for real-time market analysis.

Loading model...

Estimates the probability of BTC experiencing a 10%, 20%, or 30% drawdown within the next 30 days, based on current realized volatility and historical return distributions. Uses a 30-day rolling annualized volatility to parameterize a log-normal distribution, then calculates the cumulative probability of each drawdown threshold. This is not a prediction — it's a statistical risk gauge. High probabilities during low-volatility periods often precede regime breaks.

Loading risk model...

Calculates the z-score of each asset's price relative to its rolling EMA at three timeframes (14, 30, 90 days). A z-score above +2 means the asset is 2 standard deviations above its mean — statistically overextended. Below −2 indicates deep undervaluation. When all three timeframes align at extremes (all above +2 or all below −2), mean reversion probability is historically elevated. The signal column shows OVEREXTENDED, UNDERVALUED, or NEUTRAL based on cross-timeframe agreement.

Loading z-score model...