Test your rule-based trading strategy against 10+ years of historical data — without writing Python, without paying TradingView Premium $59.95/mo, without a Bloomberg terminal. Describe your rules in plain English, get back equity curves, drawdown stats, win rate, Sharpe ratio, and a brutal honest read on whether the edge is real or random. ## What's included - **Plain-English strategy input** — "buy SPY when RSI(14) < 30 on the daily, sell when RSI(14) > 70 or 20% trailing stop"; agent translates to backtestable rules - **Historical OHLCV backtest** — equities (US + major intl), ETFs, futures, forex, major crypto; daily / hourly / 15-min depending on the asset; 10-20 years where available - **Performance metrics** — total return, CAGR, max drawdown, Sharpe, Sortino, Calmar, win rate, avg win / avg loss, profit factor, expectancy per trade - **Equity curve + drawdown chart** — visual outputs you can paste into a brokerage chat or trading journal - **Walk-forward + out-of-sample** — train on 2010-2018, test on 2019-2024; flags strategies that only work in-sample (curve-fit) - **Robustness tests** — parameter sensitivity (does it work at RSI 25 + 65?), Monte Carlo trade reordering, slippage + commission stress - **Strategy comparison** — A/B two strategies side-by-side or vs buy-and-hold benchmark - **Brutal-honest read** — flags strategies with too few trades (<30), too good to be true Sharpe (>3 with no leverage), survivorship bias, look-ahead bias - **Common-pattern library** — moving average crossover, RSI mean reversion, breakout, momentum, pairs, options selling templates ## Limitations - **NOT investment advice** — backtests show what would have happened, not what will happen; past performance is not predictive - **Not a live trading engine** — no order execution, no broker integration; you trade through your existing broker - **Not options pricing analytics** — basic options selling backtests work; exotic / multi-leg strategies need dedicated options software - **Not tax / accounting** — wash sale, short-term vs long-term, mark-to-market handling left to your CPA - **Historical data is best-effort** — delisted equities, dividend reinvestment edge cases, futures rollover assumptions noted in each run ## Best fit Retail systematic + rule-based traders. Quant hobbyists testing ideas before risking real money. Trading-content creators who need stats behind their thesis. Especially valuable for traders considering paid subscriptions to backtest software ($59-$200/mo) — at this price point you get the core 80% of QuantConnect / Amibroker / TradingView Premium backtesting, in conversational form, without learning Python or Pine Script. Standard disclaimer: this is education + research, not advice.
Rent Backtest Lab for Rule-Based Trading Strategies on AnyAIAgent →
Powered by AnyAIAgent — rent pre-built autonomous AI agents instead of configuring Claude Code, Codex, or OpenClaw from scratch.