Estimation of the costs of changing corporate banks

Companies avoid changing when it is costly, so it is difficult to observe the costs of change in data. In this blog, I examine the forced changes brought about by bank closures to provide a new estimate of the lower bound of bank switching costs for businesses. Lithuania provides an ideal setting for identification due to the unexpected simultaneous closings of two banks, which I call “troubled bank” and “healthy bank”, and due to an extensive credit register, which shows rates interest rate and other characteristics of all outstanding loans in Lithuania. When a struggling bank was shut down and its customers were forced to switch, their borrowing costs immediately fell and converged on the rest of the market (see Figure 1). This suggests that businesses were stranded and overcharged, and the fact that they paid the overcharge instead of changing suggests that the costs of switching were even greater than the overhead. Therefore, the overhead provides the estimate of the lower limit of switching costs.

What explains the drop in borrowing costs after the bank closes? The theory classifies the costs of change into the costs of information and “shoe-leather” change. Information costs arise from interbank information asymmetries, whereby a company’s internal bank (a bank with which a company has a lending relationship) knows the company better than external banks. If the business tries to change, outside banks may suspect that the company’s most knowledgeable in-house bank has refused to lend because the business is bad or risky. External banks would thus offer high interest rates, which creates change costs for companies. Internal banks can exploit this by overcharging their customers. If an internal bank closes, interbank information asymmetries disappear and external banks can offer interest rates without stigmatizing the money changers. This would lead to lower borrowing costs for businesses.

Another explanation for the decline is that companies could always have borrowed at lower interest rates, but, due to the “cost of shoe leather”, chose not to do so until they did. are forced to do so. Shoe-to-leather conversion costs may include transaction costs (eg research costs), artificial costs (eg loss of benefits provided by an internal bank, such as lower collateral requirements. , larger loans, longer maturities or expected support during crises), psychological costs (breaking the habit of borrowing from the same bank) and learning costs (setting up and getting used to systems of a new bank). These costs are intangible and difficult to quantify, but they allow inside banks to extract rents to the point where companies are indifferent between changing and staying. An estimate of these rents would reveal the lower bound of the total costs of change.

The costs of changing information are more relevant to better and more opaque companies, i.e. younger and smaller companies, while the costs of shoe leather are expected to affect all companies. I find that the better, younger, and smaller companies have been overcharged a lot more, suggesting that the costs of change come mostly from information asymmetries. I also find that the “healthy bank” has not overcharged its borrowers, suggesting that troubled banks may be less concerned about their reputation and, therefore, are more likely to exploit the costs of change.

I contribute to the literature by showing that credit relationships with troubled banks can be harmful, and by showing, for the first time, how heist costs disappear when a bank is closed. In this way, I propose a new identification of the hold-up problem and one of the first empirical estimates of the costs of changing firms in the loan market. I would expect to find comparable results in Norway and other similar lending markets, characterized by relatively small, opaque, bank dependent businesses and a high concentration of banks.

My results have political implications. First, although bank closures are generally expensive, I offer one benefit that regulators should consider when resolving failing banks: A bank closure can improve credit allocation, because better quality businesses. , which are particularly important for productivity and employment, could borrow more cheaply. Second, regulators could monitor bank lending rates for early signs of bank distress. Third, information asymmetries are a major source of change costs, so regulators could be urged to aim to reduce these asymmetries further. For example, preventing loan renewals could improve the reliability of businesses’ credit history.

Bankplassen er en fagblogg av ansatte i Norges Bank. Synspunktene som uttrykkes its representative forfatternes syn og kan ikke nødvendigvis tillegges Norges Bank. Har du spørsmål eller innspill, kontakt oss gjerne på bankplassen@norges-bank.no.