What We Do

Systematic Asset-Based Finance

We apply the data infrastructure and analytical rigour of systematic strategies to asset-based markets.

Strategy Coverage

Investment Universe

MinVaris invests across European asset-based finance, spanning public structured credit, privately negotiated transactions and special situations. A common loan-level underwriting framework enables us to evaluate risk and relative value consistently across each strategy.

Strategy	Representative Investments
Public Structured Credit	RMBS, Consumer ABS, Auto & Leasing ABS
Private Asset-Based Finance	Bilateral financings, forward-flow agreements, portfolio acquisitions
Special Situations	Non-Performing Loans (NPLs), Unlikely-to-Pay (UTP) exposures, Significant Risk Transfer (SRT) transactions, bespoke asset-backed opportunities

Investment Philosophy

Managing Dispersion & Payoff Asymmetry

We believe that expected return alone provides an incomplete basis for investment decisions. Long-term portfolio outcomes depend not only on expected cash flows but also on the distribution of potential outcomes, the asymmetry of credit losses, and the interaction of security structure with collateral performance.

Our research focuses on estimating loan-level distributions rather than relying solely on portfolio averages.

Individual loans Pool average

Illustrative — schematic, not derived from live loan-level data.

A pool’s weighted-average loan describes a point in the input space the actual collateral does not occupy. No individual loan carries the average credit bureau score, DTI, and LTV simultaneously—the average is a statistical construct, not a credit that exists in the pool.

In plain terms: there is no such thing as the average loan. Valuing the pool as if one exists prices a credit that nobody actually holds.

Volatility Drag

g ≈ μ − ½σ²Realised compound growth vs. expected return and variance

Realised long-run compound growth (g) is constrained by variance (σ²) around expected return (μ). Where models fail to measure loan-level dispersion, the investor implicitly absorbs a compounding penalty.

In plain terms: unmeasured swings in performance quietly erode long-run return, even when the average estimate was right.

Payoff Concavity

E[P(X)] ≤ P(E[X])Jensen’s inequality applied to capped credit payoffs

Credit instruments possess capped upside and uncapped downside. Under Jensen’s inequality, evaluating a portfolio using summary averages overstates true value. Evaluating the full empirical distribution of outcomes seeks to reduce the performance drag associated with mismeasured dispersion and asymmetric payoff structures.

In plain terms: credit can only pay back so much, but it can lose far more — pricing off the average loan overstates what the pool is actually worth.

Research Architecture

Four-Stage Research Architecture

Our research platform transforms regulatory loan-level disclosures into systematic investment insights through four integrated stages.

Data Foundation

We ingest and standardise monthly loan-level disclosures from European securitisation repositories, creating a consistent analytical dataset across issuers, jurisdictions and reporting standards.

Contextual Intelligence

Loan-level information is enriched with external datasets—including housing market dynamics, legal enforcement regimes and macroeconomic indicators—to provide the context required for robust underwriting.

Quantitative Underwriting

Statistical and machine learning models estimate borrower behaviour, prepayment risk, default probability and recovery dynamics, transforming millions of observations into actionable investment signals.

Continuous Learning

Every monthly reporting cycle provides new realised outcomes. These observations are used to recalibrate our models, continuously improving underwriting accuracy as additional evidence becomes available.