Assessment · Cross-border credit risk

Your model prices distance at zero. The data says otherwise.

Cross-border private credit defaults at 6.46% against 2.56% domestically. We price that premium on your book, corridor by corridor, from our own peer-reviewed research on 13,317 transactions worth $11.8 trillion.

Request an assessment Read the research

The evidence

Four findings that change how a corridor is priced

Default premium

3.91pp

6.46% cross-border versus 2.56% domestic, across 13,317 transactions.

Hazard ratio

2.86×

Cross-border exposures fail nearly three times faster at any point in their life.

Distance elasticity

−2.41

Comparable to physical goods trade, in a purely financial market.

Corridor concentration

68%

Of cross-border volume flows through just ten bilateral corridors.

The problem

A single PD curve across a mixed book is mispricing both halves

Most credit models treat geography as a categorical control: a country dummy, a sovereign ceiling, a rating notch. That captures level differences between countries but misses the continuous cost of separation.

Apply one global PD curve to a book holding both domestic and cross-border exposures and you systematically understate risk on the cross-border portion while overstating it on the domestic. The errors do not cancel; they concentrate exactly where exposure is largest.

Why the premium persists

Information asymmetry. Monitoring is harder at distance; warning signs surface later.
Legal complexity. Enforcement runs through an unfamiliar system with less predictable restructuring outcomes.
Monitoring cost. All of the above is fixable, at a price. That price is the premium.

None of these are eliminated by faster data or better derivatives. They are structural.

The assessment

What you receive

Cross-border premium per exposure, in basis points
Hazard multiple and adjusted PD by corridor and sector
Corridor concentration and the exposures driving it
Incremental expected loss versus domestic-equivalent treatment
Written report plus an audit trail citing the research and its limitations
Results as Excel and CSV, yours to keep and reuse

Delivery: 3 working days · enquire for pricing

Request an assessment

Who this is for

Private credit funds lending across borders, development finance institutions and export credit agencies, banks with cross-border corporate books, and valuation firms that must justify a country risk premium in a fairness opinion.

Methodology

What we do, and what we do not claim

We apply published, peer-reviewed coefficients to your deal characteristics. The hazard ratio scales your domestic PD; sector friction factors adjust it for how monitorable the asset is at distance; corridor concentration is measured against the market structure documented in the research.

Stated in every report

Coefficients are applied, not re-estimated on your data.
The distance elasticity is estimated on bilateral flow volumes; we use it as a corridor-thinness signal, not as a PD driver.
Sector friction factors are judgement-based scalings of the published effect, not separately estimated.
Where you do not supply a domestic PD, the sample average is used and will not reflect your underwriting.

Source: Cross-Border Shock Transmission in Private Credit Markets: Evidence from Global Deal-Level Data (2025), presented at the Bank of England Agenda for Research (BEAR) Conference 2026.

FAQ

Common questions

How much riskier is cross-border lending?

In our research on 13,317 private credit transactions from 2004 to 2025, cross-border deals defaulted at 6.46% against 2.56% for domestic deals, a premium of 3.91 percentage points. Cox proportional hazards estimation puts the hazard ratio at 2.86, meaning cross-border exposures are nearly three times more likely to default at any point in their life.

Why does distance still affect credit risk?

Distance proxies for monitoring cost, information asymmetry, legal complexity and relationship density. Our gravity model estimates a distance elasticity of −2.41 for bilateral private credit flows, comparable to frictions in physical goods trade, despite these being purely financial transactions.

What data do you need to run the assessment?

At minimum: borrower country, sector, exposure at default and tenor. Better results with lender country, seniority, LGD, your own domestic PD, and distance. A single deal is enough to start, using the deal entry form.

Do you re-estimate the model on our data?

No, and every report states this plainly. We apply published, peer-reviewed coefficients to your deal characteristics. Re-estimation on a client portfolio is a separate bespoke engagement. See model development.

How long does it take?

Three working days from receiving your data. You receive a written report, an audit trail citing the research, and results in Excel and CSV.

The research behind it

Read the evidence first