Transcript The IMF Role in Bank and Bond Markets By Eichengreen, Kletzer

Monitoring International
Borrowers: The IMF Role in Bank
and Bond Markets
By Eichengreen, Kletzer, and Mody
Discussant
Athanasios Vamvakidis
Empirical question
• Impact of IMF programs on market access
and cost of funds
• 6,700 loan transactions between emerging
market borrowers and banks, 3,500 new
bond issues
• 1991-2002
• Banks act as delegated monitors on behalf of
depositors
• Dispersed bondholders lack the capacity or/and
the incentives to incur the large costs of
monitoring
• Therefore, the IMF’s public monitoring role
should be more valuable to bondholders
• The IMF could signal valuable information to the
private sector
• The IMF program can be a commitment
mechanism for good policy plans/intensions
Main Results
• IMF programs reduce spreads more in
markets dominated by bonds than in those
dominated by bank loans
• This result is stronger for countries with
intermediate levels of external debt/GDP
ratios
• This result is stronger for programs that
turn precautionary
Discussion
• The theoretical framework discussed in
the paper helps put the empirical tests in
perspective
• The empirical results are consistent with
the theoretical predictions
• Global variables:
US industrial growth (results are consistent
with Arora and Vamvakidis (2004a), log of
swap rate, EMBI volatility
How about:
Growth in trading partners (Arora and
Vamvakidis (2004b),
Regional growth
Sector shocks
• Increased IMF transparency after 1997
Most program information/country reports, on
policies, projections, and risks, are now public
information
Emphasis on transparency by the IMF for its
members (Fiscal and data ROSC, FSAP and
other similar facilities)
Has the ability of an IMF program to provide a
positive signal increased as a result of more
transparency?
Results for before and after 1997
Results for programs that publish the LOI and staff
reports versus programs that don’t
• IMF programs (positive signals):
All IMF programs
Precautionary
Programs turned precautionary
Negative program related signals?
Program negotiations that failed, indicating the need for a
program, but the absence of good policy plans, of IMF
monitoring, and of policy commitment (can use staff
reports to get the data)
Precautionary arrangements that turn non-precautionary
Programs that collapsed: IMF pulling the plug
• Causality
The authors claim that the use of data on
transactions reduces causality concerns.
However, the paper needs to do more to
explain this claim and address remaining
causality concerns
• The paper tests if the results differ by
issuer (bank lending versus bond issues),
by the level of external debt/GDP ratio,
and by the size of IMF lending
Other possible dimensions:
Region (both for lenders and borrowers)
Level of GDP per capita
History of crises
Openness of capital markets
Level of development of the financial sector