Abstract: Bridge loan financing,as a short-term and provisional informal financing tool,can help firms rollover loans at the micro-level and transfer risks across time at the macro-level. Though bridge loan borrowing is quite pervasive in practice,we still lack a general picture of what is going on. To this end,we match  loan-level data from 19 large-scale commercial banks and financial data of firms listed on the Chinese A-share market from 2007 to 2013. With these in hand,we are able to conduct some preliminary empirics to offer several stylized facts of Chinese A-listed firms involved with bridge loan borrowing. The following results are obtained by comparing firms with bridge loan and those without bridge loans. First,those firms with bridge loan are associated with lower profitability and higher leverage ratio in the past,implying that those firms must once undergo negative business shocks and have to bear higher level of debt to be repaid before they borrow bridge loans. Second,those firms with bridge loans have higher financial costs,mainly because financing via bridge loan is more costly. Thirdly,the future return on equity and Tobin's Q for firms with bridge loan turn out to be lower. This might be because the exceptionally high financial cost of raising funds from bridge loan hinders those firms from making more productive investments and therefore worsens their long-term performance. The above empirical evidence is the first step for understanding the cause and consequence for firms to borrow bridge loans. 

Based on the stylized facts,the paper then constructs a two-period game-theoretical model with bank,loan officer and firm to examine the mechanism on how bridge loan financing affects financial stability. There are a continuum of firms in the economy. We first assume that the productivity of each firm is independently drawn from a distribution function across two periods and call this the “transitory productivity shock” case. Later we discuss the “persistent productivity shock” case where each firm's productivity is the same across two periods. Each firm has to borrow a unit of fund from bank to finance their investment in period one and two. Those firms with poor realized earnings in the first period are unable to pay their debt back in the end of the period,so they would like to borrow bridge loan to repay their debt first and then get funds for period-two investment. Those firms that fail to repay their first-period debt go bankrupt and the bank only recoup a part of their salvage value from those failing firms. However,the bank cannot observe the realized output of a firm while the loan officer can know this key information. The objective function of a loan officer has two components:one is that the bank wants him to release more loan,the other is that the bank also aims to constrain default risk from bad firms by detecting whether the loan officer continues to grant a loan to a firms involved with bridge loan and if so,the loan officer would therefore be punished. In equilibrium,the loan officer continues to grant a loan to firms whose productivity is above an endogenous cutoff and rejects to grant a loan to firms with productivity below such cutoff. The firms that survived from the first period are composed of two parts:those “healthy firms” that do not borrow bridge loans and those “sub-healthy firms” that have to borrow bridge loans. A “healthy firm” would invest the full part of loan. As  contrast,a “sub-healthy firm” needs first pay back to the bridge loan company at a high interest rate and then invest the remaining proceeds,but the bank still charge the repayment by a full loan scale,so the default risk of those “sub-healthy firms” becomes high. This is  the cost induced by bridge loan from the perspective of bank. The associated benefit,directly speaking,is that those “unhealthy firms” avoid going bankrupt in the short run and have a chance to survive to the next period. 

Our model has drawn several important conclusions. First,bridge loan is able to stabilize a firm's financing and maintain its investment and production in the short-run,especially for those firms affected by liquidity shocks,and consequently delays financial risks to the future. Second,when confronting the tradeoff between inter-temporal risk transfer and two-period profits,the regulatory authority may have different attitude towards bridge loan from commercial banks,as the former's objective is to maximize the financial stability for the whole system while the latter's is to maximize expected profit without taking systemic risk induced by bridge loan borrowing into consideration. The optimal size of bridge loan lending chosen by commercial banks is suboptimal in terms of financial stability,so the regulatory authority must step in to oversee commercial banking and constrain their negative externality on financial system. Third,as the likelihood of systematic crisis goes up,the regulatory authority should adopt heavier regulations on bridge loan financing. That is,the regulatory authority must take a more stringent measure to control the bridge loan borrowing.

What's more,our model also sheds light on the contemporary debate on whether the loan officer should be accountable for granting loans for lifelong. We characterize the loan officer's incentive under the lifelong accountable system and compare it with the benchmark case of “bridge-loan-detecting system“ where the loan officer would be only penalized if he has been detected to grant a loan to a firm with bridge loan borrowing. We find that the optimal financial stability under the two systems can be the same,but contracting under the lifelong accountable system is less robust. More precisely,when the environment changes and the policy cannot be adjusted in time,the financial risk under the lifelong accountable system turns out to be higher. Our analysis reveals that the so-called “lifelong accountable system” does not looks perfect as it sounds like. Finally,we also discuss the economy where there are two types of firms in the economy:firms subject to transient liquidity shocks and firms subject to permanent production shocks. The production shocks experienced by the former in the two dates are independent with each other while those experienced by the latter are the same. If the bank can tell one type from the other,it would rather lend to transiently-shocked firms rather than permanently-shocked firms. However,firm's type is private information and only known to itself. As the proportion of firms affected by permanent shock increases,the negative impact induced by bridge loan borrowing on financial stability in the long run grows. As a result,the optimal scale of bridge loan lending should be consequently downsized,until all such lending activities should be banned.