Building up close main-bank relationships can be beneficial to small and medium sized firms for several reasons. For mature firms, relationship lending eases access to bank finance and positively affects firm innovation activity. As a further advantage, a relationship bank is able to "lean against the wind" if a firm is in financial distress. Relationship banking therefore provides an insurance for liquidity and is especially attractive for high risk firms. But firms face transaction costs in providing information. Cost sensitive firms with low risk might not want to bear these costs and ask for more transaction oriented banking. A bank's ability to offer relationship banking depends on its portfolio characteristics, hierarchy, investment in specialization, and strategy. For banks that have higher refinancing costs than others, it is hard to invest in specialization to offer the advantages of relationship banking and simultaneously offer competitive interest rates. Such banks have incentives to compete in debt repayments. In other words, those banks offer cheap loans and services to cost sensitive firms with low default risk and which do not demand liquidity insurance.
Do firms select the main bank relationship according to their risk or preferences for being helped in difficult times? I empirically test this for newly established German firms. High market shares of public and cooperative banks make the German banking system particularly interesting with regard to this question. Both Sparkassen and cooperative banks have a mission statement and implicit or explicit guarantees, reducing refinancing costs. But private banks are not restricted to a certain lending or liquidation policy and have incentives to offer transaction oriented banking, thus attracting low risk firms. I analyse a firm's main bank choice related to its demand for liquidity insurance, its sensitivity to costs, and its ex ante default risk.
The KfW/ZEW Start-up Panel, a representative sample of young firms in Germany, paints a rich picture of firm and entrepreneurial characteristics. It includes detailed information on firms' criteria for choosing a main bank relationship, the selected bank, and previous private relationships with that bank. I use nearly 1,900 observations on firms established between 2009 to 2011. Alternative banks and their characteristics are identified using the ZEW Bank Panel. I test for risk considerations in a firm's initial main bank choice in two steps. In the first step, I employ a logit model to estimate the probability of choosing a certain bank type. In the second step, I employ a nested logit model to estimate the probability of choosing a bank out of a set of alternative banks.
I find that firms for which "expected bank support in financial distress" is of the utmost importance choose a relationship oriented bank. Entrepreneurs who consider their personal bank relationship to be valuable to their firm are also more likely choose a relationship oriented bank. However, I do not find that firms select their main bank according to ex ante risk measured by predicted default probability. Cost sensitive firms are more likely to choose a private bank. Furthermore, I find that a bank's local competence is an important selection factor. Banks with a high regional market share and those that are regionally concentrated are more likely to be chosen. The distance between firm location and bank branch is not economically significant.