According to the syndicated loan pricing puzzle (Carey and Nini, Journal of Finance, 2007) interest rates charged to corporate borrowers are lower in Europe than in the U.S. Our investigation suggests that controlling for region-specific credit ratings makes the Europe-U.S. gap insignificant, and solves the puzzle. We speculate that the puzzle originates from the lack of uniformity of accounting standards.Keywords: Pricing puzzle, syndicated loan, rating, Europe, U.S. indicate that syndicated lending attracts around 50% of corporate debt issuance in the U.S., and some 20% in Europe. Fig. 1 shows the evolution of the market between 1992 and 2014. The so-called SL pricing puzzle emerged from the work of Carey and Nini (2007), henceforth C&N, showing that, all else equal, credit in the European SL market is significantly cheaper than in its U.S. counterpart. Subsequent articles provided partial clues to the puzzle by emphasizing that ratings alone are insufficient to account for the riskiness of SL contracts. Gaul and Uysal (2013) blamed unobservable volatility differences between U.S. and European firms. That explanation solved the puzzle, but only for listed borrowers. Berg et al. (2016) underscored that loan category matters, and showed that the puzzle vanishes for syndicated credit lines, which account for 70% of their full sample.Stressing that ratings are based on accounting ratios (Campbell and Taksler, 2003), which depend on standards that vary across jurisdictions, this paper shows that using region-specific ratings solves the SL puzzle, not only over the period covered by C&N, but also for a larger sample of SLs.
Data and MethodsAnalyzing the SL puzzle requires first comparing the all-in spreads-the spread to LIBOR, plus fees-of deals arranged in the U.S. with those observed on the European market. C&N capture the Europe-U.S. difference by introducing a dummy variable for European loans in the OLS regression explaining the all-in-spreads. They control for a number of variables, including the borrower's credit rating. This econometric design relies on the assumption that, for lenders, the informational content of credit ratings is uniform across jurisdictions. We lift this assumption by introducing