Trade Ideas Track Record
Review resolved model-cohort outcomes in public
What the public track record shows
The Trade Ideas Track Record is a public transparency page for AI Trade Ideas. It reports rolling and lifetime outcomes by model cohort using resolved ideas and is updated nightly.
Open it at optionomics.ai/trade_ideas/track_record. No login is required.
The page includes:
- Total resolved ideas included in the current report
- Aggregate lifetime win rate
- Number of model cohorts represented
- Open and awaiting-resolution counts
- Latest resolved winners, each linking to its public, timestamped receipt
- A rolling win rates by model table—30-day, 90-day, and lifetime win rates per cohort, each with its resolved count, plus the average favorable peak and open/awaiting counts
- The measurement rules used by the public report: how directional ideas are counted as wins (a minimum favorable move within the monitoring window, with a higher bar for option contracts), how policy-specific idea types are checked, and the “no hidden denominator” rule—only resolved ideas enter a win rate, and open ideas are listed separately, never dropped
Why results are grouped by cohort
Different idea families can have different horizons, structures, and market contexts. Cohort rows keep those results separate so one aggregate number does not hide every difference.
Use the cohort table to compare a model’s rolling result with its longer history. A short window may react more quickly to a market-regime change, while a lifetime result contains more observations but can blend very different regimes.
Resolved, open, and awaiting
- Resolved ideas have enough outcome information to be counted under the published measurement rules.
- Open ideas remain active inside their monitoring window.
- Awaiting resolution ideas are past their window but not yet included as final wins or losses.
A cohort below the minimum sample size shows “Not enough data” with its count rather than a percentage, so a small sample cannot masquerade as a track record.
The public page explains how resolved outcomes are counted. Read those rules before comparing the displayed rate with a brokerage account, because the report is not a model portfolio and does not represent your fills, sizing, fees, or taxes.
Use the track record responsibly
- Check the number of resolved observations, not only the percentage.
- Compare rolling and lifetime windows.
- Review the idea detail and risk context, not only a cohort label.
- Expect results to vary as market regimes change.
- Do not treat historical win rate as the probability that the next idea will win.
For the live feed, levels, current assessment, option-leg performance where available, and research links, use AI Trade Ideas.
