The liquidity shocks of ’08–’09 revealed that measures of liquidity risk being used in most financial institutions turned out to be woefully inadequate. The construction of long-short portfolios based on liquidity proxies introduces errors such as extraneous risk factors and hedging error. We develop a new measure for liquidity risk using exchange-traded funds (ETFs) that attempts to minimize this error. We form a theoretically-supported measure that is long ETFs and short the underlying components of that ETF, i.e., long and short a similar set of underlying securities with the same weights. Pricing discrepancies between the long and short positions are driven by liquidity differences between the ETF and its underlying components. Constructing liquidity risk factors in a number of markets, we undertake several tests to validate our new liquidity metric. The results show that our illiquidity measure is strongly related to other measures of illiquidity, explains bond index returns, and reveals a systematic illiquidity component across fixed-income markets.

We examine whether religion affects the terms of bank loans. We hypothesize that lenders value the traits of religious adherents, such as risk aversion, ethical behavior and honesty, and thus offer favorable loan terms to religious borrowers. Consistent with this hypothesis, we find that corporate borrowers located in counties with a high level of religiosity are charged lower interest rates, have larger loan amounts and fewer loan covenants. These results suggest that the corporate culture of borrowers influences the availability and cost of bank loans.