224 S. Drucker and M. Puri
Berger and Udell (1993)develop the “monitoring technology hypothesis”, which
attempts to explain loan sales as a reaction to improvements in monitoring and informa-
tion technology.^37 As information technology improves, banks can sell loans to direct
lenders because these loan buyers increase their ability to monitor loans. For high qual-
ity borrowers, the monitoring cost advantage of banks falls below the signaling costs of
intermediation, which enables the sale of loans. An important implication of the theory
is that banks keep risky, essentially illiquid loans for which their monitoring advantage
is important.Berger and Udell (1993)find empirical support.
Drucker and Puri (2006)also find evidence consistent with the monitoring technol-
ogy hypothesis. Using a sample of loans that are originated between 1999 and 2004, the
authors identify individual loans that are traded in the secondary market. They find that
banks sell the loans of more informationally transparent borrowers—larger firms who
have long-term debt credit ratings. The monitoring cost advantage of banks is presum-
ably smaller for these types of loans. Interestingly, sold loans have additional, tighter
financial covenants as compared with loans to similar firms which are not traded in the
secondary market. This is consistent with loan buyers directly monitoring borrowers
through covenants.
The “comparative advantage hypothesis” argues that loan sales arise out of exoge-
nous differences in the comparative advantages of financial intermediaries. Researchers
have explored a number of different comparative advantages that could motivate loan
sales.^38 Hess and Smith (1994)claim that banks may have a comparative advantage in
originating and servicing loans but not in funding or interest risk management.Pavel
and Phillis (1987)provide empirical support, showing that banks with origination and
servicing advantages have a higher probability of selling loans and also sell more loans.
Carlstrom and Samolyk (1995)assume that banks have an advantage in finding and
screening profitable local projects and loan sales arise because, otherwise, financially
constrained banks would have to pass up positive investments when there were many
good opportunities in the local market. Some empirical studies have supported this the-
ory, as a typical bank with a binding capital constraint is more likely to sell a higher
activities result in banks using off-balance sheet activities, such as loan sales (Pennacchi, 1988; Pavel and
Phillis, 1987), the “collateralization hypothesis”, in which loan sales provide a mechanism to shift risk from
risk-averse to risk-neutral investors (Benveniste and Berger, 1986, 1987) or to help avoid debt overhang for
banks (James, 1988), and the “moral hazard hypothesis”, which suggests that banks use loan sales to book
income immediately and increase leverage to take advantage of deposit insurance (Benveniste and Berger,
1986 ; James, 1988).
(^37) This theory is an extension ofBhattacharya and Thakor (1993), who find that intermediary monitoring
dominates direct monitoring when the benefits from scale economies in monitoring exceed the costs of sig-
naling the value of assets to investors.
(^38) In addition to the information-based comparative advantages that are discussed here,Pennacchi (1988)
discusses another comparative advantage that is based on funding differences between banks. Loan sales
provide a means by which the inexpensive funds that are raised by some banks can be used to finance the
loans at other, higher cost banks. The empirical evidence on this non-information-based view is mixed (see
e.g.Berger and Udell, 1993; Haubrich and Thomson, 1996).