Research rules
Methodology
Every result follows an explicit timeline, cost model, and set of limits.
01
Decision timeline
A strategy reads data through a completed daily close. Its target shifts forward one bar before execution, so Monday's decision can first trade at Tuesday's open.
This delay prevents a strategy from trading at a price that was already known only after the decision was made.
02
Execution assumptions
The engine models commission, fixed fees, and adverse slippage. Orders fill completely at the adjusted next-open price.
- Long-only positions
- No borrowed cash
- Sells execute before buys
- Daily closing valuation
03
Advanced research
Position limits and optional exits use prices available at the current open. Portfolio estimates come from historical returns and covariance.
Monte Carlo paths reuse those historical estimates. They show how the assumptions behave, not where prices will go.
- Chronological 70/30 parameter study
- Long-only sampled frontier
- Seeded equal-weight simulation
- Equal-weight buy-and-hold benchmark
04
Known limits
The simulator does not model taxes, bid-ask spreads, volume limits, latency, partial fills, or market impact. A chosen symbol list can also create survivorship bias.
Strategy examples are engineering demonstrations, not evidence of a profitable trading edge.