The non-monotonic impact of bank size on their default swap spreads: Cross-country evidence
研究了115家跨国银行的数据，发现银行规模对信用违约互换利差有非单调影响：小银行扩大规模降低风险，大银行扩大规模则增加风险，并找到了最优规模水平。
This paper studies the drivers of bank's credit default swap (CDS) spread, taken as a measure of credit risk, by considering the impact of housing market along with a number of bank level determinants, such as regulatory capital, leverage, size, liquidity, asset quality and operations income ratio. We build upon a unique dataset consisting of 115 banks (during pre- and post-crisis periods) headquartered in 30 countries from both developed and emerging countries. Results suggest that CDS spread is driven by asset quality, liquidity and operations income ratio, while bank size is found to have a non-monotonic impact on CDS spread. If the bank is small, an increase in size reduces the average credit risk. If the bank is large enough, an increase in size raises the latter. From our results we derive the level of bank size that minimizes the CDS spreads. Financial institutions growing beyond this threshold are subject to higher credit risk, implying that smaller and medium sized banks are safer than large banks. When considering the estimates in the periods before and after the 2007 crisis, we further find a different extreme point of bank size in the former (approximately 1642 billion Euros) relative to a significantly lower level of optimal bank size (around 70 billion) in the post-crisis period, implying too-big-to-fail and too-big-to-save in the pre-crisis regime.