Ch. 5: Banks in Capital Markets 219
loans from AAA lenders significantly exceed the excess returns from lenders rated BAA
or lower. While the univariate results indicate that the announcement effects are concen-
trated among loans from high-quality lenders, as in previous studies, the authors refine
their test through regressions that control for other characteristics that could be driving
the result (such as differences in borrower characteristics). Even after controlling for
these other factors, higher quality lenders continue to be associated with significantly
higher abnormal announcement returns.
Overall, these studies find that the announcement day abnormal returns are signifi-
cantly positive for loan announcements, stronger for loan renewals and changes rather
than initiations, larger for smaller firms rather than larger firms, stronger when public in-
formation about the firm is noisy rather than clear, and larger for higher quality lenders.
It is important to note that the positive abnormal returns around loan announcements
contrasts with the strongly negative announcement effects of equity, the moderately
negative abnormal returns around convertible debt issuance, and the insignificant abnor-
mal returns surrounding straight public debt announcements (seeAsquith and Mullins
(1986), Mikkelson and Partch (1986), and Masulis and Korwar (1986)for equity is-
suances;Eckbo (1986)andMikkelson and Partch (1986)for debt issuances). These
latter results are consistent with a firm’s public security issuance decision revealing its
private information on its prospects (Myers and Majluf, 1984) and strengthen the view
that bank loan announcements convey positive private information to the capital mar-
kets.^27
To further examine the information content of loans,Dahiya, Puri, and Saunders
(2003)take a different approach and study the announcement of asaleof a borrower’s
loans by its lending bank. The termination of a banking relationship through a loan
sale may convey a negative signal to the market about a firm’s prospects. To test this
hypothesis, the authors employ the event-study methodology, using a sample of loan
sale announcements by the originating bank for sub-par loans. The data is collected by
cross-matching loan sale information in two market newsletters (Gold SheetsandBank
Letter) with CRSP.^28 Using 3-day, 5-day, and 7-day event windows, the authors find
a highly significant negative abnormal return of between−1.61% and−8.11%. These
results are consistent with a negative information effect arising from loan sales. The
authors also perform two additional tests to examine if loan sales are valid signals of a
bank’s negative private information about the firm. First, the authors find that firms that
have their loans sold are more likely to file for bankruptcy than other comparable firms
trading day. The data used in previous studies does not allow for the authors to distinguish announcements
that occur during or after the trading day, which forces the use of two-day event windows.
(^27) See however, some contrary evidence inBillett, Flannery, and Garfinkel (2003)who find in the long run,
bank loan announcements appear no different from seasoned equity or public debt issuance. Thus the positive
abnormal return is subsequently followed by a negative return.
(^28) Dahiya, Puri, and Saunders (2003)focus on the sales of seasoned sub-par loans, where the information
effects of bank sales are likely to be highest and where they have a more representative sample of loan sales.