Do cultural differences between contracting parties matter? Evidence from syndicated bank loans

This paper by Giannetti and Yafeh examines whether cultural similarities or differences between banks and their corporate customers have an impact on the decision about whether to extend a loan, and, if so, the size, interest rate, and conditions associated with the loan.

From analysis of the syndicated loan market between 1980-2005, the authors find that banks from all over the world are more likely to offer smaller loans, at higher interest rates, and require more guarantees, the more ‘culturally distant’ the borrowing company is from the lending bank, i.e. the terms of a loan between counter parties in the US and Japan will be more onerous to the borrower than the loan terms between counter parties in the US and Canada, or between Japan and South Korea. Moreover, this difference persists in susbsequent transactions suggesting that it is not related to an initial lack of familiarity between lender and borrower but for some other reason. The authors’ theory is that higher contracting costs between culturally dissimilar parties leads to this pricing difference.

Analysis

The research considers loan contracts in more than 70 countries and focuses on two key relationships where cultural differences could come into play – the interaction between the lead bank in the syndicate and the borrowing company, and the interaction between the lead bank and other participant banks in the same syndicate. The primary loan characteristic used for comparison is the interest rate charged inclusive of all fees (the ‘all-in’ spread, taken from a standard industry source, Dealogic).

Measuring cultural distance is much much more subjective and therefore difficult. Giannetti and Yafeh use a social science study, the World Values Survey (see chart on p. 10), which evaluates the position of countries along two major dimensions of cultural variation – the relative importance of religion (traditional versus secular), and the relative development of its society along a scale from prioritising economic and physical survival to a greater focus on self-expression and quality of life. Plotting countries along these two dimensions results in a map grouping culturally similar countries, and allowing mathematical measurement of the distance between any two points to be used as a proxy for cultural distance. This is supplemented with other factors such as the physical distance between countries; whether there is a common border, a similar legal framework, the same language or historic colonial ties, and any differences between the rights of creditors in each country.

Results

The findings point to the cultural distance between borrower and lender having a consistent positive and significant correlation with the loan spread. For example, a one standard deviation increase in cultural difference, which is approximately the same as between the US and Canada, adds 6.5 basis points to the overall cost of a loan. Between the lenders within the syndicate, the greater the cultural distance between the lead bank and a participant bank, the lower the share of the risk that the participant bank takes.

Other key findings include:

• Foreign and culturally distant banks appear more likely to offer cheaper loans than domestic banks, but are more expensive than foreign but culturally close banks.
• Foreign and culturally distant banks give smaller loans and are more likely to want guarantees.
• A subsidiary of the lender local to the borrower reduces but does not eradicate the effect of cultural distance (the interest rate significantly decreases but the loans are likely to be smaller with greater security required).
• Repeated loans mitigate the effect of cultural differences but only disappear when a borrower has received more than four loans from a particular lead bank and this is extremely rare in reality.
• Risk sharing between culturally different banks within the syndicate increases with repeated joint deals, but only disappears after more than thirty interactions and again this is very rare.

The results are robust when controlling for differences in loan amounts, maturity and collateral; as well as differences in trade relationships between countries, currencies, industry, and the borrower’s ownership structure. The results also hold when lead banks from the US and the UK are excluded from the sample, suggesting that the phenomenon is not limited to the largest and most well-known international banks or to the Anglo-Saxon financial sphere. Most importantly, the relationship holds when the authors control for external factors influencing borrower-lender matching, such as credit rating, i.e. it is not the case that the worst borrowers match with culturally distant banks and therefore pay higher interest rates on their loans. In fact, the financial profile of companies borrowing from distant lenders actually appears to be stronger than those borrowing from domestic or culturally close lenders.

Conclusions

Giannetti and Yafeh discuss various theories to explain the relationship between cultural differences and loan cost including whether monitoring costs are greater when lending to a culturally distant borrower. However, the persistence of the impact of cultural differences despite repeated interactions with the same lender makes the monitoring cost theory less likely. They conclude that an increase in contracting costs is the most likely explanation. Writing contracts between two culturally different parties is likely to consume more time and resources than between two parties that share common cultural norms, language, legal framework, communication style or organisational structure. Professional decision makers may therefore be inclined to offer better terms to culturally similar counter parties.