Do Banks Affect the Level and Composition of Industrial Volatility

In theory, better access to bank credit can reduce or increase output volatility depending on whether firms are more financially constrained during contractions or expansions. This paper finds that the volatility of industrial output is lower in coun-tries with more bank credit. Most of the reduction in volatility is idiosyncratic, which follows from the ability of banks to pool and diversify shocks. Systematic volatility is reduced less strongly. Volatility dampening is achieved via countercyclical borrowing: At the firm level, short-term borrowing is less (or more negatively) correlated with sales and inventories in countries with high levels of bank credit. THERE IS A DEBATE ABOUT THE EFFECT OF FINANCIAL INTERMEDIARIES—or financial development more generally—on the volatility of output. In theory, the effect on volatility of having more financial intermediation is ambiguous, depending on the stage of the country’s development (Aghion, Bacchetta, and Banerjee (2004)) or the type of shocks that affect the economy (e.g., real vs. monetary shocks as in Bacchetta and Caminal (2000), or credit demand versus credit supply shocks as in Morgan, Rime, and Strahan (2004)). The existing empirical evidence is also inconclusive. For instance, while Acemoglu et al. (2003) show that macroeconomic aggregates are less volatile in more developed countries, Beck, Lundberg, and Majnoni (2004) find no robust relation between financial intermediation and output volatility when considering different types of macro shocks. Understanding output volatility is important because volatility has a negative effect on growth (Ramey and Ramey (1995), Aghion et al. (2005)), in other words, because stability breeds growth. This paper revisits the debate on finance and volatility using industry-level and firm-level data. From the perspective of a financially constrained firm,