Research & Publications

Institutional research on orbital credit risk.

Analysis bridging space systems engineering, Bayesian inference, and structured finance. Produced for lenders, insurers, and institutional investors deploying capital into orbital infrastructure.

Research Briefings · Subscribe
Quarterly briefings. No promotional content. Unsubscribe at any time.
Publications
Insight

Space Operators Are Paying an Equity Premium for Debt-Compatible Assets

In 2023, space insurers collected $557 million in premiums and paid out $995 million in claims. Those numbers contain a credit signal that the lending market has not yet read.

March 23, 2026Read →
Insight

The Space Sector Is Financeable. Lenders Cannot Prove It Yet.

Private investment peaked at $18 billion in 2021, fell to $5.9 billion in 2024, and rebounded to $12.4 billion in 2025. The problem is analytical, not structural.

March 16, 2026Read →
Portfolio Analysis

Concentration Risk in Orbital Portfolios

Why diversification across operators fails under debris stress, and how altitude-correlated default modelling exposes hidden portfolio concentration.

Forthcoming
Methodology

From Orbital Mechanics to Default Probability

Why engineering risk metrics fail to translate into credit decisions, and the standardised pathway from conjunction density to loss given default.

Forthcoming
Methodology

Three-Layer Risk Architecture for Orbital Assets

Unified data architecture, multidimensional risk quantification, and portfolio-level analytics as the framework for institutional space underwriting.

Forthcoming
Series · Article IV

A Proposed Orbital Credit Framework for Institutional Lenders

A structured framework for integrating orbital risk metrics into standard credit underwriting. Addresses Basel III risk weight assignment, Solvency II SCR calculation, and IFRS 9 expected credit loss modelling.

Forthcoming
Quantitative Methods

Bayesian Uncertainty in Orbital Risk

Prior calibration and posterior inference for satellite default forecasting.

Forthcoming
Quantitative Methods

Ensemble Methods for Non-Linear Default Detection in Mega-Constellation Operators

Gradient-boosted and mixture-model approaches to credit risk in high-asset-count LEO portfolios.

Forthcoming
Market Analysis

SPV Structures in Space Finance: How Programmatic Vehicles Create Institutional Demand

Programmatic SPVs resolve three structural barriers simultaneously: repossession, valuation, and regulatory capital treatment.

Forthcoming
Regulatory Analysis

Berlin Space Protocol: Why Zero Ratifications After 14 Years Matters to Lenders

The absence of a ratified space debris liability framework is the single largest source of unpriced tail risk in satellite-backed lending.

Forthcoming
Valuation Analysis

The Musk Premium: SpaceX Pre-IPO Valuation Across Three Frameworks

DCF, comparable transactions, and sum-of-the-parts applied to SpaceX's reported FY2025 financials. The key driver is not Starlink revenue.

April 9, 2026Read →
Structured Finance

Covenant Design for Satellite-Secured Lending: DSCR Triggers, ICR Floors, and Orbital Risk Maintenance Tests

How lenders can structure maintenance covenants for orbital assets, translating SIGMA scores and PD trajectories into enforceable financial controls.

Forthcoming
Data Studies
ML Challenger Validation

Gradient Boosting Challenger Model: AUC Analysis on 991 Operator-Quarter Samples

Validation of the gradient-boosting ML challenger component on 991 operator-quarter samples from institutional financial databases (22 publicly traded space operators, 2010–2025). The AUC reflects discriminatory power on operators with complete financial data. System-level validation is reported separately via the 41-case SR 11-7 operational backtest.

0.87
Challenger AUC
991
Operator-Quarters
+34pp
vs. Base Rate
Default History Calibration

Bayesian Prior Calibration Using Moody's Annual Default Study

How Moody's infrastructure default series is used to construct Beta-Binomial priors for orbital assets with zero or sparse default histories. Includes hyperparameter selection methodology and sensitivity analysis.

881
Labelled Events
5-layer
Bayesian Stack
Model Families
Research topics
Probability of defaultBayesian inferenceStructured financeSovereign riskIFRS 9Basel III/IVSPV structuresLEO constellationsModel validationBerlin Space ProtocolECA financeMarket commentary