Essays on Connected Lending, Misallocation, and Aggregate Productivity

Connected lending is prevalent in many countries and might play an important role in generating the idiosyncratic investment distortions faced by firms. Connected lending occurs when financial intermediaries extend loans to firms based on special connections rather than firm characteristics. Controlling owners of commercial banks might use their banks to channel funds to their non-bank businesses. Connected lending can therefore cause a misallocation of resources, which in turn reduces aggregate productivity. My three papers focus on connected lending, resource misallocation, and aggregate productivity.;My first paper, "Misallocation and Manufacturing Total Factor Productivity (TFP) in Thailand," uses census data on Thai manufacturing plants to quantitatively study resource misallocation and its effect on manufacturing TFP in Thailand. I find that there is more resource misallocation across plants in Thailand than in China, India, and the United States. When misallocation of capital and labor were hypothetically reduced to the extent observed in the United States, manufacturing TFP would increase by about 70 percent. I also find that plants which have government ownership or are located in the northern or the northeastern regions have lower plant productivity and face lower distortions than other plants. In addition, plants with medium size face larger distortions than small and large plants.;My second paper, "Connected Lending and Aggregate Productivity," uses cross-country data to document a negative association between connected lending and aggregate productivity. I develop a model incorporating entrepreneurship, financial frictions, and connected lending to quantitatively study the effects of connected lending on aggregate productivity. The quantitative results show that connected lending has a moderately negative effect on aggregate productivity. Along with the stylized facts documented in this paper, the results also indicate that connected lending is better explained by the crony view than the information view. In other words, special connections between firms and banks generally do not reduce the asymmetric information between them.;My third paper, "Connected Lending and Concentrated Control," uses data from more than 2,600 firms across 25 countries. I study whether the control rights of the controlling bank owners are associated with the extent to which firms need special connections to obtain loans. I find that the control rights of the controlling owners increase the need for special connections with banks. I also find that protection of shareholder rights decreases the need for special connections. The opposite holds true for bank supervisory power. Different from its significant effect on the corruption of bank officials, I find that private monitoring of banks does not significantly decrease the need for special connections. No evidence is found that bank officials become less corrupt as the control rights of the controlling owners increase. The results indicate that an increase in the control rights of the controlling owners of banks might reduce the integrity of bank lending.