Asymmetric Information and Corporate Lending: Evidence from SME Bond Markets
Alessandra Iannamorelli, Stefano Nobili, Antonio Scalia, Luana Zaccaria
Review of Finance, Volume 28, Issue 1, January 2024, Pages 163–201, https://doi.org/10.1093/rof/rfad024
Different existing theories offer opposite predictions regarding the unobservable credit quality of firms that access bond financing in lieu of bank loans. On the one hand, non-bank lending may be prone to adverse selection and market breakdowns, due to the comparative advantage that distinguishes banks from other intermediaries in acquiring and processing information about borrowers (Leland and Pyle, 1977; Diamond, 1984; Fama, 1985). On the other hand, banks can be more flexible than public markets when firms experience liquidity shocks, allowing for renegotiation of loan terms instead of (inefficient) liquidation (see for example Chemmanur and Fulghieri, 1994). Therefore, entrepreneurs who privately observe a high probability of financial distress may find it optimal to use bank loans, while financially solid firms can issue publicly traded debt.
A novel measure of asymmetric information in credit markets
Testing these different hypotheses is challenging, since private information on firms’ credit quality is, by definition, unobservable to outsiders (including researchers). To overcome this limitation, we propose a novel measure labelled ‘Risk Spread’, which we use to examine the empirical relationship between private information on risk and firms’ funding choices. Specifically, Risk Spread is computed as the difference between credit quality as perceived by outside investors, i.e. based solely on accounting data, and that observed by insiders, which combines information on both financial statements and credit history (including usage of credit lines and payments), as recorded in the national Credit Register. Importantly, while financial statements are easily accessible to non-bank lenders and investors, either directly or through third-party data providers, firm-level information from the Credit Register is only known to firm managers and (to a lesser extent) banks. Therefore, the difference between the two risk scores (i.e. public–private) represents the contribution of private information to the total estimated credit risk. Said differently, holding ‘public’ risk score constant, a positive (negative) Risk Spread value indicates that firm insiders observe lower (higher) credit risk than that perceived by firm outsiders.
Evidence of favorable selection among bond issuers
Using data on both issuer and non-issuer Italian SMEs, we estimate the firm-level probability of issuing bonds for the first time in public markets. Our coefficient estimates for the Risk Spread variable is positive, significant, and robust to different specifications, suggesting that access to capital markets is more common among firms that are more creditworthy than an analysis of their public balance sheet data would suggest. In other words, we document favorable selection of bond issuers among SMEs.
We additionally explore differences in funding costs between loans and bonds, and, surprisingly, we find that the yields of bonds at issuance are on average 100 basis points larger than rates on new loans (inclusive of non-contingent fees) obtained by the same firm in the same quarter.
Why do better firms voluntarily choose to pay higher rates on public debt markets? As a possible mechanism that reconciles these seemingly contradictory findings, we propose that firms may use public debt as a signalling tool.