Important Notice Recession probability forecasts are inherently uncertain and do not constitute professional financial, investment, or legal advice. Past performance does not guarantee future results. Consult a qualified professional before making any financial decisions.
Probability. Not panic.

AI Analytics Engine for Yield Curve Recession Probability Modeling

Most recession signals arrive after the move. Yield curve inversions make headlines. Positioning shifts. You are already behind. Our platform runs ensemble models on live curve data so you see probability shifts before they become consensus. We combine structural, statistical, and machine learning approaches to produce robust recession probability estimates that institutional teams can build positions around.

Hero Image

Specialized Tools for Macro Positioning

From term spread analysis to stress-testing scenarios, each module is designed to surface signals you can act on.

Term Spread Analyzers

Monitor key yield curve spreads with automated anomaly detection and historical context overlays.

Dynamic Nelson-Siegel Modeling

Fit and forecast the entire yield curve surface with regime-aware parameter estimation.

Machine Learning Ensemble Methods

Combine multiple model architectures to reduce false positives and improve signal reliability.

Beyond Traditional Econometrics

See how our AI methodology compares to conventional approaches and competitor offerings in recession forecasting.

Fast recession signal detection

Faster Detection

Traditional models rely on quarterly data releases and lagging indicators. Our system processes daily yield curve updates and flags probability shifts in real time, giving your team more runway to evaluate positioning options.

  • Daily data ingestion vs quarterly releases
  • Real-time probability updates
  • Automated anomaly flagging
Reduced false positive signals

Lower False Positives

Single-indicator models generate frequent false alarms that erode trust and trigger unnecessary repositioning. Our ensemble approach requires multiple model types to agree before flagging a signal, dramatically reducing noise-driven alerts.

  • Multi-model consensus requirement
  • Historical backtest validation
  • Regime-aware threshold adjustment
Explainable AI model outputs

Explainable Outputs

Black-box AI undermines institutional confidence. Every probability estimate includes model attribution, showing exactly which factors are driving the shift and how each model in the ensemble is contributing to the final signal.

  • Model-level attribution breakdowns
  • Factor contribution visibility
  • Audit-ready documentation

Connect Your Research Stack

Enterprise

Bloomberg Terminal

Pull live yield curve data and probability forecasts directly into your Bloomberg workflow. Our API integrates with standard Terminal functions so your team can overlay recession signals onto existing screens and models without switching contexts.
Live data feeds
  • Real-time probability overlays
  • Custom alert functions
  • Historical backtest access
Learn more
Developer

Python and R APIs

Export probability forecasts, regime labels, and alert triggers directly into your quantitative research environment. Our REST API supports both Python and R workflows with comprehensive documentation and example code for rapid integration.

Quantitative research
  • REST API with JSON output
  • Example notebooks included
  • Webhook alert support
Learn more

From Data to Decision

A clear workflow from ingestion through to actionable forecast and alert generation.

Ingest Live Data

Connect to G7 yield curve feeds through Bloomberg Terminal, Refinitiv Eikon, or direct market data providers. The platform automatically handles data cleaning, gap filling, and regime classification as new observations arrive throughout the trading day.

Configure Models

Select from pre-built ensemble configurations or define custom model weights, lookback periods, and probability thresholds tailored to your mandate. Save multiple configurations for different portfolios or research contexts.

Visualize Probabilities

Explore interactive heat maps showing recession probability across time horizons and geographies. Drill down into individual model contributions, factor loadings, and historical analogs for any given forecast date.

Trigger Alerts

Set custom thresholds for probability shifts, curve steepness changes, or regime transitions. Receive notifications via email, webhook, or direct API push when conditions match your configured criteria.

Generate Reports

Export formatted research briefings with probability charts, model attribution tables, and narrative summaries. Schedule automated report generation for regular distribution to stakeholders or compliance teams.
01

Model Accuracy

Our ensemble models are backtested across multiple recession cycles and validated against historical yield curve regimes. Individual results vary based on timing, data inputs, and external economic factors. Past performance does not guarantee future accuracy.
How accurate are the recession probability forecasts?

02

Data Sources

The platform ingests daily G7 yield curve data from major market data providers. Users can also connect proprietary data feeds through our API. All data is processed, cleaned, and validated before entering the model pipeline.
What data sources power the analytics engine?

03

Customization

Yes. The platform supports custom model configurations, alternative data integration, and bespoke probability thresholds. Our analytics team can assist with setup and validation for specialized research requirements.
Can we build custom models for specific mandates?

Ready to See the Platform?

1

Request Access

Submit a brief inquiry through our contact form. Our analytics team will review your requirements and schedule a tailored demonstration of the platform capabilities relevant to your research workflow.

2

Technical Review

Receive comprehensive API documentation, data dictionaries, and integration guides. Our team is available to walk your developers through the setup process and answer technical questions about model architecture.

3

Live Demonstration

Join a guided walkthrough of the platform with live probability models, historical backtests, and alert configuration. See how the ensemble methodology surfaces signals before they become consensus.

4

Pilot Integration

Begin a structured pilot with your existing data feeds and research infrastructure. Evaluate signal quality, integration ease, and team workflow fit before committing to a broader deployment.