Early Access · Now Open

Do Fair Value Gaps
Actually Work?
Find Out.

EdgeLab runs institutional-grade statistics on your trading ideas — FVGs, session bias, opening range breakouts, HP filter trends — and tells you if the edge is real before you risk a dollar.

No spam. Notified when EdgeLab opens to the public.

✓ You're on the list. We'll be in touch.
15+Pre-screen tests
4yrHistorical depth
OOSValidated results
0Lines of code needed
pre-screen result — XAUUSD FVG
Edge Detected
fvg_fill_rate_binomial · XAUUSD
P-VALUE
0.0000
FILL RATE
89.5%
SAMPLE
3,515
Path Dependency Analysis
Class
High
Avg MAE
11.2 pips
Clean
31.5%
MAE Distribution — how far against you before fill?
< 1x (clean)
31.5%
1–2x gap
22.9%
2–3x gap
12.5%
3–4x gap
7.1%
> 4x (brutal)
26.0%
Median
1.8x
75th pct
4.2x
90th pct
9.6x
PROCEED TO QUANT RESEARCH →
THE AI WILL INVESTIGATE THIS EDGE IN DEPTH

Research backed by academic literature

151 Trading Strategies · Kakushadze & Serur ORB Edge · Zarattini, Barbon & Aziz 2024 FX Microstructure · Lyons 2001 HP Filter Trend · Harris & Yilmaz 2009 Carry Trade Anomaly · Lustig & Verdelhan Intraday Momentum · SSRN 2024

How It Works

From Hypothesis
to Deployed Strategy

EdgeLab follows the same workflow institutional quant desks use — validate the idea statistically before building the strategy.

1

State Your Hypothesis

Tell EdgeLab what you think works — "XAUUSD fair value gaps fill reliably in the London session" or "AUDJPY has afternoon bullish bias after ADR exhaustion." No jargon required.

pre-screen →
2

Statistical Pre-Screen

A fast statistical test (1–5 seconds, no AI involved) runs against years of price data. Binomial tests, t-tests, Mann-Whitney U. Real statistics, not vibes. If the edge is there, you proceed.

p < 0.05 →
3

AI Builds & Backtests

The quant AI designs a strategy around your confirmed edge, runs in-sample and out-of-sample backtests with OOS validation, and reports whether the result holds up to scrutiny.

oos pass →

What You Can Test

16 Pre-Screen Types.
All Statistically Validated.

Not indicator templates. Rigorous statistical tests grounded in market microstructure and quantitative finance research.

Fair Value Gap

Tests FVG fill rates, path dependency (MAE distribution), and whether entering at adverse excursion levels creates a measurable edge across 1h and 15m timeframes.

fill rateMAE distributionpath dependency

Opening Range Breakout

Tests all four standard OR windows simultaneously (5m, 15m, 30m, 60m). Breakout must align with the opening candle direction — the key filter documented in Zarattini et al. 2024.

5m / 15m / 30m / 60mdirection filter

Session Bias

Tests whether a market moves directionally during a specific trading session — London open momentum, NY session reversals, Tokyo range behavior — across hundreds of sessions.

londonnew yorktokyo

Momentum vs Mean Reversion

The most fundamental question in strategy design. Tests autocorrelation structure to determine whether an instrument trends or mean-reverts at a given timeframe — and how strongly.

Ljung-Box testHurst exponent

ADR Exhaustion

When the morning session consumes ≥70% of the 20-day ADR before the afternoon opens, tests whether a statistically significant directional bias emerges.

session rangebinomial test

HP Filter Trend

Applies the Hodrick-Prescott filter (λ=100×n²) to extract the trend component from noise. Tests whether MA crossovers on the clean trend signal have statistically significant directional accuracy.

151 strategiesharris & yilmaz 2009

Conditional Probability

Tests whether one market period predicts the next. If the first hour of NY open is bullish, does the second hour continue? Reveals sequential market structure that can be traded systematically.

chi-square testsession sequences

Time of Day

Finds which specific hours carry the strongest directional edge. Markets are not uniformly random throughout the day — liquidity patterns and institutional participation create repeatable hourly tendencies.

ANOVAhourly breakdown

Day of Week Bias

Tests whether an instrument has a statistically significant directional tendency on specific days — Monday reversals, Wednesday trend days, Friday liquidation patterns.

t-testweekly patterns

Level Bounce

Tests whether price statistically respects key historical levels — prior day high/low/close, prior week levels. Measures bounce rate, break rate, and average forward move at each level type.

prior day levelsbounce rate

Correlated Pair

Tests the statistical relationship between two instruments — AUD and gold, CAD and oil. Measures Pearson correlation, cointegration, and Ornstein-Uhlenbeck half-life to determine if the relationship is fast enough to trade.

cointegrationOU half-life

Volatility Regime

Splits historical data into volatility regimes using ATR percentile rank. Tests whether returns and Sharpe ratios are regime-dependent — telling you whether a volatility filter should be part of every strategy you build.

ATR percentileMann-Whitney U

Rejection Candle

Identifies pin bars, hammers, and shooting stars at 20-bar swing highs and lows. Tests whether follow-through in the rejection direction is statistically reliable.

swing levelsbinomial test

Carry Bias

Tests whether a currency pair exhibits persistent directional drift consistent with the carry premium — the UIRP violation empirically documented since the 1980s. Tests across 5, 20, and 60-bar horizons.

AUDJPY · NZDJPYmulti-horizon

NR7 Compression

Crabel's Narrowest Range in 7 bars. Tests whether NR7 compression predicts directional expansion, and whether the NR7 bar's own direction predicts which way the expansion goes.

crabel 1990range expansion

Vol Term Structure

Tests whether the ratio of 5-day to 20-day realized volatility creates statistically distinguishable behavioral regimes with predictable directional bias.

RV5 / RV20regime detection

Research Library

The AI Reads the
Academic Literature.
So You Don't Have To.

When the AI designs a strategy around your edge, it doesn't guess — it searches a curated library of quantitative finance papers and cites the relevant research. Every strategy recommendation is grounded in published, peer-reviewed work.

The same papers institutional quant desks use. Accessible to any trader with a hypothesis.

151 Trading Strategies
Kakushadze & Serur · FX, futures, volatility, carry, momentum strategies with full mathematical specifications
Opening Range Breakout Edge
Zarattini, Barbon & Aziz 2024 · Sharpe 2.81 on 5-minute ORB with direction filter — the paper that validates the strategy
Intraday Momentum & Gap Events
SSRN 2024 · Intraday price imbalances, gap fill behavior, and momentum persistence in futures markets
Observable & Latent Regimes
SSRN 2024 · Volatility regime classification and its impact on strategy performance across asset classes
FX Microstructure · Stat Arb · Mean Reversion HFT
Additional papers on order flow, liquidity sweeps, carry trade dynamics, and high-frequency mean reversion

The difference

Built Like
Institutional
Quant Research

  • Out-of-sample validationEvery backtest reserves 30% of data as unseen. Results must hold forward — not just fit history.
  • Statistical significance requiredPre-screens use binomial tests, t-tests, and Mann-Whitney U. No edge, no proceed. p < 0.05 is the gate.
  • Academic research backingAll pre-screen tests are grounded in peer-reviewed literature. The AI references the papers when explaining results.
  • Plain language resultsEvery result comes with a plain-English explanation. No PhD required to understand what the AI found.
  • Live deployment via MetaAPIStrategies that pass OOS validation can be deployed directly to your MT4 or MT5 broker account.
AUDJPY ADROOS PASS
+0.111R
OOS expectancy · 18 trades
XAUUSD FVGOOS PASS
84.2%
OOS win rate · Sharpe 8.71
NQ HP FilterEDGE CONFIRMED
60.2%
Directional accuracy · 954 signals · p=0.0000
YM FVG Layered MAEOOS FAIL
−0.035R
Correctly identified as curve-fit · 292 OOS trades

Real results from live research sessions

Limited Early Access

Stop Wondering.
Start Testing.

Join the waitlist. Be first when EdgeLab opens to the public. We'll send you early access and research walkthroughs from the YouTube channel.

No spam. Unsubscribe anytime.

✓ You're on the list. Early access coming soon.