Abstract: We study the dynamics of credit relationships between small businesses and specialized banks and analyze the real effects of specialization on this important yet understudied segment of the credit market. Using micro-level data on the universe of corporate credit in Belgium, we show that banks leverage their industry specialization to build and retain relationships with small businesses. In the relationship-building phase, banks charge lower rates in their industries of specialization. In the relationship-retaining phase, lenders subsequently raise rates faster in specialized industries, until they charge similar rates regardless of their level of specialization. Specialized banks internalize the intertemporal value of credit relationships, combining both industry knowledge and market power to extract value from their relationships. Small businesses benefit from bank specialization in the long run through higher growth in investment, profitability, productivity, and equity value. The real effects of bank specialization inform policies that could inhibit banks’ incentives to specialize, such as open banking policies.

Abstract: We establish that banks internalize firm production networks when making lending decisions. We formulate a parsimonious model of bank lending with production networks and find that banks cross-subsidize suppliers at the expense of their customers by offering lower and higher rates respectively. This cross-subsidization benefits customers on net by keeping the supplier's marginal cost down and thus the intermediate input price that the customer purchases from its suppliers. We merge credit registry data and firm-level data on the near-universe of business-to-business transactions in Belgium to get an exhaustive map between a firm's credit relationships and its supply chain linkages. We find that lenders charge a lower interest rate to suppliers and a higher average interest rate across customers, consistent with our model.

Abstract: We study the effect of past price movements on selling decisions in the housing market to draw implications about the presence of reference dependence on this market. Using data on the universe of private housing transactions in Singapore, we test a commonly used assumption that homeowners use their unit's purchase price as a reference point when making selling decisions. We decompose a unit's cumulative price change since purchase into a series of annual potential gains and estimate how the timing of these gains affects the homeowner's decision to sell. We rule out the null of the purchase price being a reference point and find that gains occurring early in a homeowner's tenure are the most predictive of selling decisions. Homeowners prefer to hold on to their properties after recent gains, consistent with the presence of momentum on housing markets.