How trading signals work

A trading signal is a data-driven recommendation to buy or sell a financial instrument at a specific price. SignalsTrades generates signals by running 13 independent technical rules against live market data simultaneously. The majority direction with the highest combined weight becomes the signal.

How the 13-rule engine works

Each rule is an independent check of a different aspect of market conditions. Rules are grouped into three categories: technical indicators, market structure, and external context.

Technical indicators (price-based)

• RSI (Relative Strength Index): a 14-period daily RSI below 30 signals oversold conditions (BUY). Above 65 signals overbought (SELL). Data source: Yahoo Finance daily closes.
• Price Momentum: the 5-day simple moving average compared to the 20-day. Short-term average above long-term confirms upward momentum (BUY).
• Moving Average Stack: SMA20, SMA50, and SMA200 alignment. All three stacked upward with price above = golden cross (BUY). All stacked downward = death cross (SELL).
• Triangle Breakout: the daily range narrows across several sessions (compression), then expands. The expansion direction fires BUY or SELL. Identifies the algorithmic equivalent of a drawn triangle pattern.

Market structure (price levels)

• Round Level: price within 15 pips of a round number (e.g. 1.0900 or 1.1000). Institutions place orders at round levels, creating predictable reactions.
• Historical Support/Resistance: price near a 90-day swing high or low (within 0.15%). Touching support adds a BUY bias. Touching resistance adds a SELL bias.
• Win Streak: if the last 3 closed signals on the same pair went in the same direction with winning outcomes, trend confirmation adds one more vote.

External context (cross-market and sentiment)

• Fear and Greed Index: market-wide sentiment from alternative.me. Extreme fear reads as a buying opportunity (contrarian BUY). Extreme greed reads as caution (SELL).
• Session Open: the first 90 minutes of the London or New York session carry the highest institutional liquidity. Signals generated in these windows carry additional weight.
• USD Broad Trend: the direction of the US Dollar Index (DXY) on weekly closes. A falling dollar is bullish for EUR/USD and GBP/USD (BUY). A rising dollar is bearish.
• News Sentiment: the latest headlines for the pair are scored positive or negative by an AI model (Claude Haiku via Anthropic). Fresh negative news on USD adds a bearish USD bias.
• Calendar Event: high-impact events (CPI, NFP, FOMC) within plus or minus 4 hours add a directional bias based on the expected market reaction.
• Gold Correlation: XAU/USD (gold) moves inversely to USD strength. A rising gold price adds a bearish USD bias across USD-denominated pairs.

How confidence is calculated

Each rule carries a power weight: high, medium, or low. The confidence score reflects how strongly the winning side outweighs the opposing side, expressed as a percentage. It is clamped between 55% and 95%. A signal where all strong rules agree reads near 95%. A signal where rules are split reads near 55%.

Confidence is a measure of alignment, not a promise of profit. A 95% confidence signal does not win 95% of the time. It means the algorithm found unusually strong agreement across independent indicators.

Entry price, take profit, and stop loss

The entry price is the live market price at the moment the signal is generated. Take profit (TP) and stop loss (SL) are fixed pip distances:

• Forex majors (EUR/USD, GBP/USD, etc.): TP = 40 pips, SL = 25 pips (1.6:1 R/R)
• JPY pairs (USD/JPY, GBP/JPY, etc.): TP = 60 pips, SL = 35 pips
• Gold (XAU/USD), crypto, indices: TP = 300 points, SL = 150 points

What the engine does not do

• It does not trade automatically. Signals are for information only.
• It does not predict news events. The Calendar rule reacts to the scheduled event calendar, not breaking news.
• It does not guarantee outcomes. All signals carry market risk.