Deriv Signal Desk — Manual Execution Only

Live market chart

Confirmed arrows are plotted on the candle that triggered the signal.

READ-ONLY

MARKET CLASS

INSTRUMENT

TIMEFRAME

Waiting for market data M5 Awaiting candles

—

Waiting for first quote

Waiting for live price history…

The chart will draw as soon as Deriv returns candles.

Latest candles

EMA 20 EMA 50↑ BUY↓ SELL

0 candles loaded

Candle labels: Lagos time (WAT, UTC+1) · timeframe changes candle interval.

PRE-SPIKE WATCH · WARMING UP

Waiting for sufficient candles and a live price. This is an early setup watch, not a guaranteed prediction.

NO SETUP

Multi-horizon forecast engines

Separate short (3), medium (12) and long (36) candle outlooks for the selected chart timeframe.

WARMING UP

↔

WAITWaiting for enough completed candle history.

—consensus / 100 · not probability

Short trend engine—EMA 20 / 50

Momentum engine—RSI(14) · short model

MACD engine—histogram

Structure engine—20-bar short model

Candle pressure—last 3 candles

SHORT TERM · 3 BARSWARMING UP

WAIT

Confluence score —

Rolling hit-rate—

Evaluated forecasts0

Simple baseline—

Difference—

Short-horizon analysis.

MEDIUM TERM · 12 BARSWARMING UP

WAIT

Confluence score —

Rolling hit-rate—

Evaluated forecasts0

Simple baseline—

Difference—

Medium-horizon analysis.

LONG TERM · 36 BARSWARMING UP

WAIT

Confluence score —

Rolling hit-rate—

Evaluated forecasts0

Simple baseline—

Difference—

Long-horizon analysis requires at least 230 candles.

M5 / H1 / H4 / D1 confirmationLOADING HISTORY

M5 · 5 MINUTES

LOADING

Waiting for independent candles.

H1 · 1 HOUR

LOADING

Waiting for independent candles.

H4 · 4 HOURS

LOADING

Waiting for independent candles.

D1 · 1 DAY

LOADING

Waiting for independent candles.

Confirmation is valid only when all four independent timeframes are ready and agree with the selected-chart consensus. Mixed, neutral, missing or rejected history means WAIT / NO CONFIRMATION. A public quote is still not your broker quote.

The horizon is not a trade timer. A 3-bar forecast does not tell you to close or reverse after 3 bars. Forecast states refresh from new completed candles; chart arrows mark state transitions only. You decide entry, exit and holding period.

Short-horizon (3-bar) historical direction checkWaiting for matured outcomes

Directional hit-rate—not a calibrated probability

Last-candle baseline—simple direction benchmark

Difference vs baseline—percentage points

Evaluated forecasts03-candle horizon

Chronological validation labWaiting for historical windows

ASSUMED ROUND-TRIP COST (BPS)

A cost sensitivity input, not an observed broker cost. 10 bps = 0.10% of entry value. Adjust it to test assumptions.

Horizon	Training calibration	Expanding walk-forward	Validation window	Chronological holdout	Holdout net mean
Waiting for enough candle history.

Calibration is fitted only on the earliest training window using score-strength bands and shrinkage. Events are spaced by their own horizon so outcomes inside each segment do not overlap. The holdout is the latest chronological slice of the currently loaded candles, not a permanently sealed external dataset. Net mean is a directional horizon proxy after the assumed cost; it is not broker-grade P&L, a complete trading-strategy result, or evidence of future profitability. Small or unavailable samples are not promoted.

Hypothetical entry/exit simulation · not broker P&LWAITING FOR HISTORY

TRAINING WINDOW

Awaiting candle data.

Trades —

VALIDATION WINDOW

Awaiting candle data.

Trades —

CHRONOLOGICAL HOLDOUT

Awaiting candle data.

Trades —

Hypothetical rules: enter at the next candle open after a closed-candle consensus transition; stop at 1.5 × ATR; target at 2R; exit after at most 36 bars; allow only one position at a time; assume the stop is hit first when stop and target are both touched in one candle; apply the round-trip cost setting above. This remains an OHLC-only simulation: no tick-by-tick path, broker spread history, slippage model, commission schedule, swap, rejected orders, or actual fills are included. The holdout is re-cut from the currently loaded data, so it is not a permanently sealed external test set.

Confluence is an indicator-agreement score, not a probability of winning. Historical checks use only outcomes that have already matured and exclude the developing candle; results are preliminary, do not include execution costs, and are not proof of predictive edge or profitability.

Forecast signal confirmation

A chart arrow appears only when the ensemble outlook moves from neutral or opposite direction into a confirmed directional state on a closed candle.

CLOSED-CANDLE ONLY

✓Trend engine: EMA 20 / EMA 50 separation and the candle close relative to the fast EMA.

✓Momentum engine: RSI(14) momentum and MACD histogram direction contribute to the composite score.

✓Structure engine: close location within or beyond the prior 20-candle range plus the pressure of the last 3 candles.

✓Decision threshold: ensemble score of +25 or higher for UP, −25 or lower for DOWN; otherwise NEUTRAL. The score is not a probability.

✓Confirmation: arrows use completed-candle forecasts only. The rolling 3-candle historical direction check is separate from independent validation.

✓Risk reference: stop distance is 1.5 × ATR(14); suggested target is 2R from the reference entry. These are analysis levels, not broker orders.