Overview

PSRC Aurum AI is a systematic momentum and breakout framework developed for forward testing on XAUUSD using the five-minute chart.

The indicator identifies high-quality bearish displacement candles and issues unified alerts when momentum, higher-timeframe structure, daily directional bias, execution conditions, and risk controls align.

The current beta model is short-only. You may test full functionality of the cbot by accessing the Tradingview code which has the same functionality but no auto trading.

The Tradingview version is free on our website.

Core Logic

1. Momentum Candle Detection

The indicator searches for confirmed M5 candles displaying institutional-style displacement.

A qualifying candle must demonstrate:

• Range expansion relative to ATR.
• Strong bearish body dominance.
• A close near the candle low.
• Limited opposing wick.
• A break below recent short-term structure.
• Acceptable volume and volatility conditions.

Signals are evaluated only after the candle closes. Intrabar signals are not used.

2. Higher-Timeframe Alignment

The M5 setup must align with the confirmed H1 trend framework.

The default model considers:

• Price relative to the H1 EMA20.
• H1 EMA20 relative to EMA200.
• Optional H1 EMA slope confirmation.
• Confirmed higher-timeframe data only.

This filter is intended to prevent entries against established directional structure.

3. Daily Bias Confirmation

The primary beta model requires an exact PSRC Daily Bias score of:

-3

This represents maximum bearish agreement within the daily-bias model.

The score may include:

• Prior-day value-area positioning.
• Liquidity sweep behavior.
• VWAP alignment.
• Cumulative volume-delta direction.
• Intraday structural confirmation.

Historical research showed that weaker bearish scores of -1 and -2 did not provide the same continuation quality.

4. Markov Regime Monitoring

The Markov Regime Confidence Filter estimates whether the current daily market state has shown sufficient historical persistence.

It evaluates:

• Kaufman Efficiency Ratio.
• BULL, BEAR, or CHOP regime classification.
• Rolling state-transition probabilities.
• Minimum sample adequacy.
• Regime persistence probability.

The primary beta model records Markov information diagnostically but does not require it as a hard entry condition.

The challenger model requires Markov persistence to pass.

5. Entry Process

A short signal is generated only when all required conditions are satisfied at the confirmed M5 candle close.

The indicator then calculates:

• Entry reference price.
• Protective stop.
• Partial-profit target.
• Final target.
• Position size.
• Broker-lot equivalent.
• Campaign and event identifiers.

The default trade-management model uses:

• Partial take profit at 1R.
• Final target at 2R.
• Break-even protection after the partial target.
• Hard maximum holding period.
• Session force-flat control.

Unified Alerts

The indicator uses one unified alert stream for operational events.

Alert types include:

• Short entry command.
• Partial-profit event.
• Final-target event.
• Protective-stop event.
• Market-exit command.
• Setup-rejection diagnostic.
• Connection heartbeat.

Each alert may contain:

• Event ID.
• Campaign ID.
• Strategy model.
• Broker route.
• Account alias.
• Symbol.
• Side.
• Quantity in units.
• Quantity in lots.
• Entry price.
• Stop price.
• TP1 and TP2 prices.
• Signal timestamp.
• Maximum permitted signal age.

For TradingView deployment, create one alert using:

Any alert() function call

The external bridge must handle order acknowledgement, broker-side stop placement, take-profit placement, OCO behavior, and break-even modification.

Recommended Asset and Timeframes

Validated Asset

XAUUSD — Gold versus U.S. Dollar

The beta model was developed specifically around the volatility, liquidity, and session behavior of spot gold.

It should not be assumed to transfer directly to:

• Silver.
• Forex pairs.
• Equity indices.
• Cryptocurrencies.
• Futures.
• Individual equities.

Each additional market requires independent research and validation.

Required Chart Timeframe

M5

The entry logic, candle thresholds, session behavior, stop distances, and historical research were designed for the five-minute chart.

Supporting Timeframes

• M5: signal and execution timeframe.
• H1: trend and structure confirmation.
• Daily: directional bias and Markov regime analysis.

Preferred Trading Conditions

The model is designed for liquid intraday periods, particularly around the New York trading session.

Best conditions generally include:

• Strong directional expansion.
• Elevated but controlled volatility.
• Clear bearish daily alignment.
• Adequate distance from opposing liquidity.
• Normal executable spread.
• Stable broker connectivity.
• No imminent high-impact economic release.

The strategy should avoid:

• Thin rollover conditions.
• Abnormally wide spreads.
• Uncontrolled news spikes.
• Illiquid holiday sessions.
• Late-stage price exhaustion.
• Repeated entries after consecutive losses.

Risk and Protection Framework

The indicator includes institutional-style safety controls such as:

• Maximum daily loss.
• Maximum consecutive losses.
• Maximum trades per day.
• Session restrictions.
• Spread filter.
• ATR volatility filter.
• Drawdown lock.
• Cooldown period.
• Margin-aware sizing.
• Maximum notional exposure.
• Emergency kill switch.
• Signal-freshness validation.
• Manual execution arm.
• Broker-bridge readiness gate.

Risk per campaign should remain conservative during beta testing.

Best Practices

Use Demo Accounts First

The cbot should remain on demo or paper execution until a statistically meaningful forward sample has been collected.

A minimum of 60 independent campaigns is recommended. A larger sample of 100 campaigns is preferred.

Keep Parameters Frozen

Do not change entry thresholds, bias rules, sessions, stop distances, or targets during the forward-validation period.

Changing settings after losses invalidates the test and creates curve-fitting risk.

Use Realistic Costs

Forward and historical evaluation should include:

• Actual broker commission.
• Realistic spread.
• Slippage.
• Quantity rounding.
• Lot-step limitations.
• Broker execution delay.

Monitor Campaigns, Not Exit Legs

Partial exits and final exits belong to one trade campaign.

Performance should be evaluated on a flat-to-flat campaign basis rather than treating each exit leg as a separate independent trade.

Review Monthly Concentration

Profitability should be distributed across multiple months and market conditions.

The model should not be promoted if most profits are generated by one isolated volatility episode.

Maintain Broker-Side Protection

Stops and take profits should be placed at the broker immediately after entry confirmation.

The external bridge should not rely solely on TradingView remaining connected.

Current Limitations

Historical results are promising at 1.6 to 2.2 PF across different regimes but not statistically conclusive.

The strongest historical performance was concentrated in a limited 2026 gold regime, with a substantial portion of profit generated during one month.

Short-Only Architecture

The current approved model does not issue long entries.

Historical testing showed that the long-side implementation reduced overall performance.

Asset Specificity

The system has not been validated on assets other than XAUUSD.

Applying it to another symbol without recalibration may materially change signal frequency, risk, and expectancy.

Timeframe Specificity

The strategy is designed for M5 execution.

Its ATR thresholds, structure logic, session controls, and risk model may not remain valid on other chart timeframes.

Markov Model Lag

The Markov filter is a confirmation tool, not a leading indicator.

It may react slowly during rapid regime changes and should not be treated as an independent entry signal.

Daily Bias Lag

The Daily Bias model can confirm direction after a meaningful move has already developed.

This may cause missed early entries or entries during mature expansions.

Broker-Feed Dependence

Spot-gold price, spread, tick volume, session boundaries, and candle construction vary between brokers.

Results may differ across TradingView data feeds and broker platforms.

Execution Dependence

TradingView alerts do not confirm broker fills.

The external execution bridge must independently manage:

• Fill acknowledgement.
• Stop placement.
• Take-profit placement.
• OCO cancellation.
• Break-even modification.
• Duplicate-message protection.
• Disconnection handling.

News Risk

The indicator cannot guarantee safe execution during major economic releases.

Gold may experience severe spread expansion, slippage, and price gaps during high-impact events.

The indicator is a research and execution-support tool.

Past performance, historical simulations, and demo results do not guarantee future profitability.

Recommended Beta Configuration

• Symbol: XAUUSD.
• Chart: M5.
• Model: Primary — Exact Daily Bias -3.
• Direction: Short only.
• Risk: 0.25% to 0.35%.
• Execution: Demo account.
• Protective orders: Broker-side.
• Alert type: Any alert() function call.
• Parameter changes: Prohibited during validation.
• Review cycle: Weekly operational review and monthly performance review.

Status

Development stage: Forward-validation beta

Approved use: TradingView alerts and demo-account execution

Live-capital status: on cBot

Project name: PSRC Aurum Pulse BetaInternal reference: PSRC-APB-01