Distance Still Prices Credit: A -2.41 Elasticity in a Market With No Shipping
Gravity models were built for trade. Goods move between two economies in proportion to their economic mass and in inverse proportion to the distance between them — because distance means freight, insurance, spoilage, and time.
Private credit has none of those. A loan is a contract and a wire transfer. Moving it 8,000km costs essentially nothing. So the distance coefficient should be somewhere near zero.
It is -2.41.
What the estimate says
Fitting a gravity specification to 2,841 bilateral country-pair observations, our research finds that bilateral private credit flows fall sharply with distance — an elasticity comparable to frictions observed in physical goods trade. A purely financial transaction, with no physical constraint whatsoever, behaves as if it were being shipped.
That is not a rounding error. It is the central puzzle.
Distance is a proxy, not a cause
Kilometres do not cause defaults. Distance is standing in for things that are real but hard to measure directly:
- Monitoring cost. Site visits, management access, local intelligence — all scale with distance.
- Information decay. The further away the borrower, the later you learn what you needed to know.
- Legal and institutional distance. Different enforcement regimes, different restructuring norms, different creditor hierarchies. Geographic distance correlates with all of them.
- Relationship density. Lending follows networks, and networks are geographically clustered.
The gravity coefficient bundles these together into one number. That is its weakness as explanation — and its strength as a pricing input.
Why this matters if you are allocating capital
Most credit models treat geography as a categorical control: a country dummy, a sovereign ceiling, a rating notch. That approach captures level differences between countries but misses the continuous cost of separation — the fact that a UK lender's Spanish exposure and its Indonesian exposure differ by more than their sovereign ratings imply.
Three consequences:
- Country risk premiums built on sovereign spreads alone understate the friction. The sovereign tells you about the state's creditworthiness, not about your ability to monitor a mid-market borrower 11,000km away.
- Corridor matters, not just destination. The same borrower country looks different depending on where the lender sits. Flow frictions are bilateral.
- Diversification is less than it appears. If flows cluster by distance, then "geographic diversification" often means adding exposures that share the same corridor characteristics.
The honest caveat
A gravity elasticity is an association across corridors, not a causal claim about any single deal. It tells you the market behaves as though distance is expensive. It does not tell you which of the underlying mechanisms dominates for your book — and the mix almost certainly differs by sector, deal size, and sponsor.
That is precisely why it belongs in a model as a documented, challengeable input rather than a plugged number.
Related
The flow friction has a credit-outcome counterpart: cross-border deals default at 6.46% versus 2.56% domestically. See the cross-border default premium.
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.