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.
Yield curve intelligence

Research methodology and economic foundations

Understand how our ensemble models translate yield curve dynamics into recession probability estimates you can build positions around.
Forecast horizon $ 240,000.00
months ahead
  1. Signal detection $ 120,000.00
  2. Position review $ 180,000.00
  3. Recession window $ 240,000.00
Now 20Y
  • Transparent methodology
  • Peer-reviewed foundations
  • Open-source components
  • Continuous validation

Why the yield curve predicts recessions

The yield curve captures the market's collective expectations about future interest rates, inflation, and economic growth. When short-term rates exceed long-term rates — an inversion — it typically signals that investors expect future rate cuts in response to economic weakness. Our models decompose the yield curve into term premium and expectations components, isolating the signals that have historically preceded recessions with meaningful lead times. The term premium reflects the extra compensation investors demand for holding longer-dated bonds, while the expectations component embeds forward-looking economic assessments. By separating these elements, our analytics distinguish between genuine recession signals and distortions caused by central bank policy or safe-haven flows. This decomposition is central to our methodology and explains why our ensemble approach outperforms single-indicator models that treat all inversions as equivalent.

AI techniques explained

Gradient boosting

Sequential decision trees correct errors from previous iterations, capturing complex non-linear relationships between yield curve segments and subsequent economic outcomes.

Neural networks

Deep learning architectures identify latent patterns across multiple time horizons, processing high-dimensional curve data that traditional econometric methods struggle to model.

Neural network data visualization
Quantitative research team working

NLP of communications

Natural language processing extracts sentiment and policy signals from central bank statements, enriching yield curve data with qualitative forward guidance.

Ensemble methods

Multiple model outputs are combined through weighted aggregation, requiring consensus before flagging recession signals and reducing individual model weaknesses.

Historical validation

How our signals performed across major recession periods.

2001

Dot-com recession

Our ensemble configuration detected elevated recession probability twelve months before the official NBER peak, with term spread compression and corporate credit deterioration providing confirming signals.

2008

Global financial crisis

The model flagged sustained inversion across multiple tenors eighteen months ahead of the Lehman collapse, with regime classification shifting to contraction well before market stress became acute.

2020

Pandemic recession

While the trigger was exogenous, the model had already identified elevated fragility from yield curve dynamics and credit spread behavior in late 2019, positioning the signal as context rather than prediction.

2023

Yield curve inversion

The most prolonged inversion on record generated sustained elevated probability readings, demonstrating the model's ability to maintain signal discipline through periods of policy distortion and delayed transmission.

Independent recognition

Our forecasting methodology has been referenced in academic literature, presented at central bank research conferences, and validated through independent backtesting studies.

Published methodology
Conference presentations Peer review Independent validation
Open data standards
Academic citations
Reproducible research
Continuous audit Transparency awards Research partnerships Policy engagement
Model registry

Research leadership

Our team combines academic rigor with practical market experience in macroeconomic forecasting and quantitative modeling.

Dr. Eleanor Whitfield

Dr. Eleanor Whitfield

Head of Research

Yield curve econometrics and recession forecasting

Education

Doctor of Philosophy in Financial Economics, London School of Economics

Eleanor leads the development of our ensemble forecasting framework and oversees model validation protocols. She previously directed macro research at a sovereign wealth fund and has published extensively on yield curve dynamics.

Experience

Sovereign Wealth Research - Director of Macro Research
Central Bank Analytics - Senior Economist

Core skills

Forecasting Econometrics Team leadership

Methodologies

Dynamic Nelson-Siegel modeling Bayesian vector autoregression

Certifications

Chartered Financial Analyst designation
Certificate in Quantitative Finance
Dr. Marcus Chen-Okonkwo

Dr. Marcus Chen-Okonkwo

Lead Quantitative Analyst

Machine learning for macroeconomic prediction

Education

Doctor of Philosophy in Theoretical Physics, Massachusetts Institute of Technology

Marcus architects the machine learning pipeline and develops the neural network components of our ensemble models. His background spans particle physics and quantitative finance, bringing rigorous statistical discipline to macro forecasting.

Experience

Global Macro Hedge Fund - Quantitative Strategist
Academic Research Institute - Postdoctoral Fellow

Core skills

Model architecture Python development Statistical inference

Methodologies

Gradient boosting machines Recurrent neural networks

Certifications

Certificate in Machine Learning from Stanford Online
Financial Risk Manager certification
Dr. Priya Sharma-Rasmussen

Dr. Priya Sharma-Rasmussen

Chief Economist

Monetary policy transmission and business cycle analysis

Education

Doctor of Philosophy in Economics, University of Chicago

Priya directs the economic interpretation of model outputs and ensures alignment between statistical signals and macroeconomic theory. She advises institutional clients on integrating probability forecasts into positioning frameworks.

Experience

Investment Bank Research - Managing Director of Economics
International Monetary Fund - Senior Economist

Core skills

Policy analysis Client advisory Economic writing

Methodologies

Structural vector autoregression Regime-switching models

Certifications

Chartered Financial Analyst designation
Certificate in Advanced Econometrics
Dr. James O'Brien

Dr. James O'Brien

Head of Engineering

Distributed systems for financial data processing

Education

Doctor of Philosophy in Computer Science, Carnegie Mellon University

James oversees the data infrastructure, API development, and platform reliability. He ensures that research models translate into production-grade analytics that institutional clients can depend on for real-time decision support.

Experience

Market Data Provider - Principal Engineer
Technology Startup - Co-founder and CTO

Core skills

System design Data engineering Platform operations

Methodologies

Event-driven architecture Real-time stream processing

Certifications

AWS Solutions Architect Professional
Certificate in Financial Technology