The commitment problem of secured lending

The paper investigates optimal financial contracts when investment in pledgeable assets is endogenous and not observable to financiers. In a setting with uncertainty, two inputs with different collateral value and investment unobservability, we show that a firm-bank secured credit contract is time-inconsistent: Once credit has been granted, the entrepreneur has an ex-post incentive to alter the input combination towards the input with low collateral value and higher productivity, thus jeopardizing total bank revenues. Anticipating the entrepreneur's opportunism, the bank offers a non-collateralized credit contract, thereby reducing the surplus of the venture. One way for the firm to commit to the contract terms is to purchase inputs on credit and pledge them to the supplier in case of default. Observing the input investment and having a stake in the bad state, the supplier acts as a guarantor that the input combination specied in the bank contract will be actually purchased and that the entrepreneur will stick to the contract terms. The paper concludes that: (1) Buying inputs on account facilitates the access to collateralized bank financing; (2) Firms using both trade credit and collateralized bank nance invest more in pledgeable assets than firms only using uncollateralized bank credit. Our results are robust to the possibility of collusion between entrepreneur and supplier.