Realised Return

Realised return is the bedrock metric for EA evaluation because it incorporates every friction that backtests miss: spread costs at actual broker rates, slippage at actual market conditions, swap costs at actual rollover rates, broker-side execution delays, and the inevitable mistakes the strategy makes during edge cases the backtest didn't simulate.

The distinction between realised return and other return concepts:

• **Realised return**: actually happened on a real account under real broker conditions during the measurement window • **Backtest return**: theoretical result of replaying historical data through the strategy logic; subject to over-fitting, look-ahead bias, slippage modelling errors, and tick-data quality issues • **Marketed return**: vendor's published claim; may match realised return (if vendor is honest), may approximate backtest return (if vendor uses backtest as primary marketing), may be aspirational without specific source (if vendor is opaque about methodology) • **Annualised return**: any of the above scaled to a 12-month equivalent for comparability; computation involves geometric compounding so monthly returns don't simply add

The editorial expectation: vendors publishing realised returns on verified accounts is mandatory for serious products. Vendors who publish only backtest returns or marketed returns without verification fall below the diligence floor regardless of the headline numbers' attractiveness.

Common realised-return patterns by tier:

• **Safety tier** (Smart Robot AI, Trendopedia, Fortuna): 2-5% monthly realised return, sustained over 12+ months • **Most-profitable tier** (Scalperology, GoldStrike, Breakopedia): 5-10% monthly realised return, sustained over 12+ months • **Aggressive tier** (excluded from editorial endorsement): 10-25% monthly marketed returns that rarely match realised live performance; usually grid or martingale systems that haven't yet experienced their inevitable blow-up

For EA evaluation, the buyer-side workflow:

1. Confirm the vendor publishes realised returns on a verified account (not just backtest or marketed claims) 2. Compare realised returns over the past 6-12 months against vendor marketing claims; significant divergence (>30%) is a vendor-honesty signal 3. Decompose realised returns into per-month observations to assess consistency; aggregate annual returns can mask quarterly variance 4. Identify the worst observed month and worst observed quarter; realised drawdown during these periods represents the buyer's likely worst-case experience 5. Cross-reference realised returns with documented broker conditions (spread, commission, slippage); buyer's broker conditions should approximate verification broker for realised returns to translate

Vendor-marketed returns that significantly exceed realised returns on verified accounts indicate either (a) the verification account doesn't reflect the strategy's typical deployment (e.g. vendor uses smaller capital or different risk settings on the verified account) or (b) the vendor's marketing is aspirational without specific basis. Both reduce the buyer's confidence in expected outcomes.