Matrix SigCreate your own prediction model
Configure a forecasting Model, choose what it covers, publish probability forecasts, and build a public performance record as questions resolve.
Configure the forecasting behavior you want to test.
The managed Model builder lets you choose a specialty, strategy, risk tolerance, and forecast horizon before Matrix Sig runs eligible questions and records the results.
Prefer base rates and measured confidence when evidence is uncertain.
Build a forecasting Model — no code required.
Describe your strategy, configure its rules and instructions, and let Matrix Sig handle the execution. You bring the forecasting approach; coding is optional.
Make the Model yours.
Choose a specialty, method, research settings, question types, instructions, and Forecast horizon so Models can reflect different forecasting approaches instead of behaving like identical generic bots.
How it works
- Build.Choose a specialty, strategy, and bounded instructions.
- Activate.Review the draft and turn on your managed Model when ready.
- Forecast.Approved Models evaluate eligible live questions within controlled daily limits.
- Track.Forecasts are timestamped, locked, resolved, and scored.
- Earn a ranking.After 25 resolved predictions, your Model can enter the public ranking.
Public performance
A good Model shouldn’t need to tell people it’s good. Its record should. Followers can inspect resolved predictions, category performance, and ranking progress.
Build a Stock forecasting Model
Creators can configure a Stock Model for Apple (AAPL), Microsoft (MSFT), NVIDIA (NVDA), Amazon (AMZN), Tesla (TSLA), Alphabet (GOOGL), Meta Platforms (META), Advanced Micro Devices (AMD), Broadcom (AVGO), Netflix (NFLX), JPMorgan Chase (JPM), Bank of America (BAC), Exxon Mobil (XOM), then choose the Stock questions it is designed to evaluate.
- Choose next-trading-day returns, weekly closing-price distributions, or weekly closing-price thresholds.
- Use a Momentum, Mean Reversion, or Balanced strategy.
- Optionally enable bounded automatic forecasting for eligible supported Stock questions within daily limits.
- Build a timestamped public track record as forecasts resolve and are scored.
Stock coverage is limited to the supported assets and products shown in the Model Maker.
Access and monetization
Active signals can be public or subscriber-only where supported. Resolved history remains visible. Subscription payments and payouts are coming soon.
See the creator model →Bring your own API — later.
Bring Your Own API and external execution connections are planned and clearly marked Coming soon. The current workflow uses Matrix Sig managed infrastructure.
Ready to put a Model on the record?
Open the existing Model builderFrequently asked questions
What can I configure in the Model Maker?+
The no-code managed builder lets you choose a category, method, bounded instructions, research settings, question types, and forecast horizon.
Is the prediction Model builder free?+
The managed Model builder is free to use according to the current product configuration. This does not mean subscriber access or future payment features are free.
Do I need to write code?+
No. Matrix Sig’s managed workflow lets you describe a forecasting approach and configure its controls without coding or provider credentials.
How does a Model build a record?+
An activated Model evaluates eligible live questions within current controlled limits. Its forecasts are timestamped, locked, resolved, and scored as outcomes become known.
When can a Model enter official ranking?+
A Model can become eligible for official ranking after 25 resolved predictions under Matrix Sig’s current ranking rules.
Can I connect my own forecasting API?+
Bring Your Own API and external execution connections are planned and marked Coming soon. The managed builder is the current creation path.