Backtesting built by traders, from first test to paper review.
CuteMarkets is built for developers who want to test stock and options strategies with the same discipline they use in production systems: explicit data contracts, causal replay, realistic execution assumptions, WebSocket-aware monitoring, and paper-trading runs that can be audited.
Backtesting
Rebuild the stock and options market state a strategy could actually see before it entered a trade.
Backtesting starts with point-in-time contracts, listed expirations, stock bars, quotes, trades, and aggregate bars. The goal is not a prettier equity curve; it is a replay loop that prevents stale contract leakage, same-bar fills, and last-price assumptions from deciding the result.
Research
Compare strategy families, failed branches, and portfolio candidates without hiding weak evidence.
Research workflows need reproducible artifacts: manifests, candidate grids, holdouts, PBO/DSR diagnostics, and negative-result notes. CuteMarkets fits the data layer under that process, from stock context and chain state to quote history and endpoint-level scripts a developer can version and audit.
Live Paper Trading
Move a frozen research object into a paper-trading run while preserving parity and reject reasons.
Paper trading should be a validation surface, not a silent rewrite of the backtest. The run checks live or delayed data, streams market updates where needed, selects executable contracts, records quote and order decisions, and compares the route against the exact assumptions that promoted the strategy.
Workflow stack
The same data spine supports each use case.
A developer new to backtesting should start with a simple rule, then improve the data assumptions before expanding the strategy search. The same stocks, options, quote, bar, and stream objects that make a backtest honest also make research reproducible and paper trading debuggable.
Data reconstruction
Fetch the stock context, listed contracts, expirations, quotes, trades, bars, IV, Greeks, and open interest that were observable at the decision time.
Realism gates
Reject stale quotes, wide spreads, impossible same-bar entries, unsupported DTE windows, and strategy variants that only work under a permissive fill model.
Research artifacts
Keep manifests, summary tables, selected trades, daily PnL, diagnostics, and no-go reports together so the next run starts from evidence.
Paper readiness
Freeze the selected profile, verify parity, dry-run the launch contract, and only then route paper orders with explicit risk controls and stream checks.
Summary