Ten Corridors Carry 68% of Cross-Border Private Credit
Private credit is routinely described as a global market. Measured by the number of countries involved, it is. Measured by where the money actually goes, it is not.
Network analysis across 13,317 transactions worth $11.8 trillion finds that the top 10 bilateral corridors account for 68% of cross-border volume. Two-thirds of the market flows through ten lender-country → borrower-country pairs.
Why an average conceals this
Concentration is invisible in aggregates. A market spanning dozens of countries produces reassuring summary statistics — wide geographic coverage, many active jurisdictions, no single dominant destination. The distribution tells a different story: a small number of dense corridors, and a long tail of pairs with almost nothing moving through them.
This is the same phenomenon we see inside portfolios. The average corridor in our data sees roughly one deal a year, but the standard deviation is 4.7 times the mean. Averages are where concentration risk goes to hide.
The systemic consequence
Concentration matters because it determines what happens when something breaks.
- Common exposure. When many lenders route through the same corridors, they hold correlated positions without having chosen to. A shock to one corridor is not idiosyncratic to one lender.
- Contagion pathways. Networks with dense clustering propagate stress efficiently. Lenders who share exposure to one borrower tend to share exposure to others.
- Illusory diversification. A portfolio spread across ten borrowers in three countries may sit entirely within two corridors — diversified on paper, concentrated in practice.
The market looks robust because it is broad. It may be fragile because it is deep in only a few places.
What to do with this
The practical test is not "how many countries am I in?" but "how many corridors am I actually in, and how correlated are they?"
- Measure concentration at corridor level, not country level. Lender-country → borrower-country is the unit that carries the risk.
- Compute effective N, not raw count. A Herfindahl index and its effective-number equivalent will tell you how many corridors you are genuinely diversified across. It is usually fewer than the count suggests.
- Stress the corridor, not the country. If 68% of market volume moves through ten pairs, a corridor-level shock is a systemic event, not a name-specific one.
You can run the first two on your own book in seconds with our free Portfolio Diagnostics tool — it computes HHI by sector and geography, effective N, and Euler risk contributions, entirely in your browser.
The limit of this finding
Concentration is a description of structure, not a prediction of failure. Dense corridors exist because they are efficient: shared language, legal familiarity, established relationships, lower monitoring cost. The same features that concentrate the market also make those corridors safer, deal by deal.
The risk is not that concentration is irrational. It is that it is invisible in the metrics most institutions actually report.
Findings from "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.
Need this applied to your portfolio?
We turn research into models and advice you can act on — economic capital, climate ICAAP, and model governance.