Free module · Put it all together

Putting It All Together: Build a Complete Trading Research System

Follow one idea from a written hypothesis to a reconciled paper-trading decision.

End-to-end workflow · data contracts · signals · models · portfolio · orders · ledgers · monitoring

Lessons in this module

1. 1 / Define the job and a small research contract
2. 2 / Make a point-in-time data contract
3. 3 / Turn an idea into a causal feature and a baseline
4. 4 / Replay decisions into fills and net returns
5. 5 / Add ML, regime models or RL only at a defined interface
6. 6 / Convert forecasts into constrained portfolio positions
7. 7 / Turn targets into idempotent orders and handle partial fills
8. 8 / Reconcile fills, costs, cash and performance
9. 9 / Run the miniature system and define the promotion decision

Open the interactive module

Practice and apply

• Research hypothesis — p=.56, gain .012, loss .008, round-trip cost .001. Find expected net return per exposed dollar.
• Available training labels — Labels become available at 8, 10 and 12. Fit cutoff is 10, with equality permitted. How many labels may enter?
• Trailing feature — Closes [98,99,100], lookback 3. Find latest price relative to its trailing mean minus one.
• Net account change — $10,000 account, exposure .25, gross asset return .01 and round-trip cost .001 of notional. Find dollar profit.
• Paired improvement — Candidate [.004,−.002,.003], baseline [.003,−.001,.001]. Find mean paired difference.
• Target holdings — $10,000 equity, $100 reference, target vol .005, asset vol .02, max weight .25. Find whole-share target.
• Uncommitted order quantity — Target 25 shares, held 10, working buy 5. Find additional buy quantity.
• Reconcile ending cash — Cash starts at 10,000. Buy 25 at 100 with fee .25; sell 25 at 100.92 with fee .25. Find ending cash.
• Separate external capital — Equity rises from 10,000 to 10,522.50, including a $500 deposit. Find investment P&L.

Work through the practice exercises · Quant development tools