Finance & Insurance
Portfolio risk, default probability, policy pricing. Every decision is about uncertainty over future outcomes. Bayesian methods make that uncertainty explicit.
Finance and insurance decisions are about uncertainty over future outcomes. Portfolio risk, default probability, policy pricing, reserve adequacy. Bayesian methods make uncertainty first-class, not an afterthought, with full posterior distributions that regulators accept and stakeholders understand.
How can we figure out the unknown from the things we know? The answer lies with Bayesian Statistics.
Work in This Space
Everysk
Bayesian PE Index from Capital Cash Flows
Estimated private equity returns from capital cash flow data. Bayesian model development with upgraded samplers. Index aligned with Cambridge Associates VC Index benchmarks.
VisualVest
CLV for Robo-Investment Platform
Modified Shifted Beta-Geometric survival model for variable AUM-based payments. Hierarchical individual-level churn parameters with Streamlit dashboard.
Trusted By
- Everysk
- VisualVest
- Nürnberger Versicherung
- Charles Schwab
What We Solve
Private Equity Index Modeling
Estimate returns from capital cash flows without transaction-based prices. Time-varying value-added factor estimation aligned with Cambridge Associates benchmarks.
Customer Lifetime Value
Contractual CLV for robo-advisors, wealth management, and subscription banking. Shifted Beta-Geometric survival model for churn with variable payment amounts.
Insurance Risk Modeling
Hierarchical Bayesian models for insurance risk pricing with proper uncertainty bounds. Biweekly SLA coaching with client team upskilling on Bayesian actuarial methods.
Causal Inference for Financial Analytics
Interrupted time series and causal impact for policy and product changes. Quasi-experimental methods including Regression Discontinuity and Synthetic Control.
Why Bayesian
Why Bayesian for Finance & Insurance?
- Full posterior over risk — Not just a probability of default but the full distribution, enabling expected loss and tail risk in one model.
- Interpretable to regulators — Regulators require model explainability. Black-box ML fails regulatory validation; Bayesian models pass with transparent assumptions and credible intervals.
- Works with sparse data — Private equity, structured credit, and specialty insurance have sparse historical data. Bayesian priors let domain expertise fill gaps without overfitting.
“They went a step further after the first final model. It’s not common in consulting to keep challenging and improving something that’s already delivered. Transparent, honest, and always looking for the best result.”