LAB

FILE / THE-LAB BACKTESTER · LIVE ENGINE SUPPORTER ACCESS

TEST THE REAL
ENGINE SURFACE.

Natural language in. Real backtest engine out. Default runs now use the validated 12-year baseline window unless you specify a shorter range.

SCORECARD COMBOS

—

loading live scorecard metadata…

CATALOG

—

loading instrument coverage…

STRATEGIES

—

loading engine support…

/ 01 · DATA SOURCES

12 AVAILABLE

Stack your edge.

Surface audit only. These rows describe the research inputs we want the Lab to support, but the engine truth today comes from the backtest catalog, stored baselines, and supported strategy registry below.

MARKET MICROSTRUCTURE

Historical bars + baseline artifactsvalidated 12-year window

Exchange-direct curve + OIfutures roll + open interest

Implied vol + options flowvolatility surface + P/C ratio

EVENT + ECONOMIC

Federal macro + calendarCPI · NFP · Fed minutes

Financial news aggregatorreal-time headline stream

Earnings + event wirecorporate calendar

SENTIMENT

Professional sentiment indexinstitutional positioning

Real-time social signalstreaming sentiment vector

Retail sentiment indexcrowd positioning · flow

◆ PROPRIETARY · OUR MOAT

Strategy scorecardlive-ready · secondary · experimental · shelved

Execution modelsimulated fills · real engine summary

Forge pathwaybaseline → paper → operator approval

/ 02 · STRATEGY INPUT

NATURAL LANGUAGE

Tell the Lab what you want to test.

Example: "test asian sweep on NQ with the default settings over the last 12 years"

◆ asian_sweep

◆ orb

◆ london_breakout

◆ scalp_momentum_ignition

full-window default = validated 12-year baseline unless your query sets a different range

/ 03 · PARSED PARAMETERS

Waiting for query parse. The Lab will show the detected strategy, symbol, date range, and parameter overrides before the run completes.

/ 04 · BACKTEST RESULTS

waiting 0 trades

Run a strategy to see real backtest results.

—

PROFIT FACTOR

—

WIN RATE

—

AVG WIN SIZE

—

MAX DRAWDOWN

—

TOTAL TRADES

—

NET P&L (SIM)

—

AVG BARS HELD

—

AVG MAE

EQUITY CURVE · waiting for run

—

/ 05 · LIVE NQ SCORECARD

What the engine actually rates today.

Loading live scorecard…

◆ PROMOTION PATH

Baseline scorecard decides what is live_ready, secondary, experimental, or shelved. Promotion still requires paper validation and operator approval before anything reaches the live squad.

/ 06 · VARIANT SWEEP

Your idea · run 6 ways in parallel.

One strategy, multiple stop-placement philosophies tested simultaneously. Pick the variant with the best risk profile, then compare it against the validated baseline and the live scorecard before promotion.

TIGHT STOP

10-point stop

PF 1.54

WR 71%

AvgR 0.9

DD -5.2%

n 1,847

Sharpe 1.41

P&L +$102k

Highest win-rate but smallest R. Good for high-frequency scalping.

WIDE STOP

30-point stop

PF 1.71

WR 54%

AvgR 1.9

DD -11.4%

n 1,422

Sharpe 1.67

P&L +$168k

Absorbs noise. Lower WR, bigger winners. Higher drawdown tolerance needed.

◆ BEST VARIANT

Stop above range high

PF 2.08

WR 61%

AvgR 2.1

DD -7.8%

n 1,591

Sharpe 2.14

P&L +$224k

Structural stop at range high. Best PF, best Sharpe. Context-aware.

STRUCTURAL

Stop below range low

PF 1.82

WR 58%

AvgR 1.7

DD -9.1%

n 1,512

Sharpe 1.78

P&L +$189k

Protects against trap moves. Balanced risk/reward on range plays.

VOLATILITY

1.5× ATR stop

PF 1.93

WR 59%

AvgR 1.8

DD -8.6%

n 1,488

Sharpe 1.92

P&L +$201k

Adapts to vol regime. Stays out of trouble in low-vol, gives room in high-vol.

DYNAMIC

Trail after 1R

PF 1.67

WR 48%

AvgR 2.4

DD -6.3%

n 1,422

Sharpe 1.88

P&L +$177k

Runners let free. Lower WR but biggest R-multiples — rides the fat tail.

◆ SWEEP VERDICT

Variant 3 (stop above range high) is the clear winner — PF 2.08 · Sharpe 2.14 · outperforms all 5 siblings on risk-adjusted return. The Lab auto-flagged it as the submission candidate for Forge.

/ 07 · EXPORT · PINE SCRIPT

Take it to TradingView.

One click generates Pine Script v5 from your winning variant. Paste into TradingView for visual confirmation or live alerting. Also exports to NinjaScript, ThinkScript, and raw JSON.

PINE SCRIPT v5 · STOP-ABOVE-RANGE VARIANT · AUTO-GENERATED

// @version=5
// Generated by ChartHustle · The Lab
strategy("NY Open Fade · Stop Above Range", overlay=true,
          initial_capital=25000, default_qty_type=strategy.fixed, default_qty_value=2)

// Inputs sourced from backtest variant 3
gapThreshold = input.float(0.5, "Gap % Threshold", step=0.1)
vixMax       = input.float(18.0, "VIX Ceiling")
rangeMins    = input.int(5, "Opening Range (min)")

// Opening range (first 5 min of NY cash)
isFirstBars  = (hour == 9 and minute < 35)
rangeHigh    = ta.highest(high, rangeMins)
rangeLow     = ta.lowest(low, rangeMins)

// Entry: gap > 0.5%, VIX < 18, fade after first bar closes below range low
gapPct       = ((open - close[1]) / close[1]) * 100
canEnter     = gapPct > gapThreshold and not isFirstBars and close < rangeLow

// Stop: above range high · Target: 2R below entry
stopLevel    = rangeHigh + syminfo.mintick
riskPts      = stopLevel - close
targetLevel  = close - (riskPts * 2)

if canEnter and strategy.position_size == 0
    strategy.entry("Short", strategy.short)
    strategy.exit("Exit", "Short", stop=stopLevel, limit=targetLevel)

◆ LINE-BY-LINE AUDIT

Every input matches your backtest parameter. No hidden logic, no magic numbers.

◆ ALSO GENERATES

NinjaScript (.cs), ThinkScript (tos), JSON schema for API consumers, Python (backtrader).

◆ CAVEAT

TradingView caps bar history. The Lab runs on the server-side engine against the validated baseline window, so treat TradingView as a viewer, not the source of record.

/ 08 · THE CAP DIFFERENCE · NOT TRADINGVIEW

TradingView caps the bars.
We don't.

TRADINGVIEW

◆ Free tier: 500 bars of history
◆ Pro tier: 5,000 bars on 1m charts
◆ Premium: 20,000 bars on 1m
◆ That's ~14 trading days of 1-minute NQ
◆ Your backtest is a rounding error

CHARTHUSTLE LAB

◆ Validated 12-year baseline for NQ
◆ Real strategy catalog from the live engine
◆ Simulated execution model on the server
◆ Baseline scorecard plus promotion workflow
◆ The lab is the source of record

THE VALUE · TradingView is the viewer. We're the lab. Build + test here, visualize there — that's the workflow.

/ 09 · WHY THIS BACKTESTER · NOT THE USUAL

Every other backtester lies about fills.
This one uses ours.

REAL EXECUTION

Broker-adjusted fills

Other backtesters assume perfect fills at the signal price. Ours replays through the server-side execution model with slippage, latency, and commission assumptions kept in one place.

PROPRIETARY DATA

Specialist Archive fused in

Every trade our specialists have taken — 48,192 fills across 6 archetypes — is available as a data layer. Test your strategy against regimes where they won or lost. Nobody else has this.

MULTI-SOURCE FUSION

12 feeds, not just price

Price action alone is 90's backtesting. Ours layers VIX regimes, FOMC windows, news sentiment vectors, social flow, options OI — and lets you toggle which feeds the strategy considers.

COMBINATORIAL

1.2 billion and counting

Our grid search has already tested 1.2 billion parameter combinations. Every supporter submission adds to the shared knowledge base — your backtests compound the platform's edge, not just yours.

FORGE PIPELINE

Winning strategies graduate

Pass the three gates (PF ≥ 1.3, n ≥ 100, survived 2+ regime samples) → submit to Forge. If it holds through 30 days of paper, it joins the live squad as a named specialist with your attribution.

/ 10 · COMMUNITY FORGE · LIVE LEADERBOARD

Strategies in the pipeline.

Member submissions at each stage of graduation. Clear view of what's been proposed, what's being tested, what's live.

01

Opening Range Fade · NQ · Gap > 0.7%

submitted by kai.trade · 14 days paper

PF 2.41

n 143

◆ GRADUATING · 16d REMAIN

02

FVG Reclaim · RSI Divergence

submitted by jamal_ict · 3 days paper

PF 1.68

n 87

◆ EXPERIMENTAL

03

Liquidity Sweep After FOMC Minutes

submitted by drew_nq · queued

PF 3.02

n 44

◆ IN BACKTEST QUEUE

04

VWAP Stretch · High-Vol Regimes Only

submitted by trader_matt · queued

PF 1.89

n 112

◆ IN BACKTEST QUEUE

05

Fade The Fed Presser

submitted by anon_47 · shelved

PF 0.82

n 61

◆ SHELVED · PF <1.3

VIEW FULL FORGE QUEUE →

/ 06 · AVAILABLE STRATEGIES

Engine support, not wishlist copy.

This list is pulled from the live catalog and scorecard. If it is not in the engine or baseline artifacts, it should not be marketed as ready.

Loading catalog…

/ 07 · BASELINE WINDOW

Validated baseline before promotion.

The Lab now defaults to the validated 12-year baseline window unless the query explicitly asks for a shorter period. That keeps the front-end promise aligned with the stored artifacts.

CURRENT RULE

Natural-language runs default to 2014-01-02 → 2026-03-31. Promotion requires baseline scorecard strength, then paper validation, then operator approval.

Loading baseline metadata…

/ THE LAB · SUPPORTER TOOLS

Your ideas · tested against the validated baseline · scored against the engine.

Run a supported strategy honestly, compare it against the live scorecard, and only move it toward paper when the baseline and operator review agree.

Supporter access · $79/mo · lab, replay, archive, autopsies, and builder tools around the public room

◉ WATCH WAR ROOM ◆ ARCHITECTURE ← INDEX

TEST THE REALENGINE SURFACE.