Abstract
This thesis reviews the theory and evidence on bank lending to
companies and uses an event study to test the hypothesis that banks
obtain inside information about borrowers. The argument is that a
stock market response to announcements concerning bank loans
indicates that banks do obtain inside information, which is signalled
by the announcements. US event study evidence supports the
information hypothesis for smaller quoted companies but not larger
ones. This study is the first to use UK data and the results show
less response to loan announcements than in the USA, which is
consistent with other evidence that banks in the UK do not, as a
rule, obtain inside information about large borrowers.
It is important to test the information hypothesis because the
established view of the rationale for bank lending assumes that it is
true. There are several reasons why more information would give
banks an advantage and enable both the average cost of their loans to
be lower and the cost of each loan to reflect the risk of the
borrower more accurately. But as the evidence indicates that banks
do not have an information advantage for large borrowers, it is
suggested that their lending in these cases is better explained by
the service and commitment they can offer and by their capacity to
negotiate with companies in difficulties. Bond investors are as well
informed as banks but are not organised to cope with bad debts, which
explains why the eurobond market is restricted to very safe issuers.
This is believed to be the main factor limiting the development of
markets in corporate debt.
The thesis includes a brief history of the financing of
business in the UK which establishes that securities markets have
been an important source of funds for quoted companies since the
1920s and that banks have not been involved in corporate management
or ownership since the mid nineteenth century. There is also a
comprehensive review of event study methods which demonstrates their
surprising diversity and justifies the choice of method for this
study, which is believed to be the first in the UK to use daily data
throughout.
