This guide provides general research criteria. It is not personal financial, legal or security advice.
Separate backtests from live results
Backtests depend on historical data, assumptions and execution models. They may omit slippage, latency, liquidity constraints and fees.
Ask for the denominator
A percentage result is incomplete without the measurement period, starting capital, cash flows, leverage and benchmark.
Inspect drawdown and loss periods
Return alone does not show the path taken to achieve it. Maximum drawdown, volatility and time under water are essential context.
Watch for selection effects
A provider may show its strongest strategy, account or time window. Evidence is stronger when all accounts and discontinued strategies are included under a consistent methodology.
Use this in a real evaluation
Document each answer and its source. Mark unsupported statements as provider claims, set a review date, and treat material unknowns as unresolved risk. Our methodology explains the full evidence hierarchy.