Transcript Benavides and Huidobro

Comments on:
Public Initiative to Support Entrepreneurs:
Credit Guarantees versus Co-Funding
by Arping, Loranth and Morrison
and
Are Loan Guarantees Effective? The Case
of Mexican Government Banks
by Benavides and Huidobro
Stijn Claessens
Assistant Director, Financial Studies Division
Research Department, International Monetary Fund
Conference on Partial Credit Guarantees
World Bank, March 13/14, 2008
General Comments



Topic little explored, theoretical or empirical
Nice papers, new theory and data usage
General comments


Do papers consider the (unique) role of government?
 In papers little special about government
 Much could be done by private sector, if given a subsidy?
What is optimal design of public support in general?
 Like to know what is optimal, first best design of financial
and other instruments given objectives, constraints, etc.
 How do partial guarantees compare to other instruments?
 Papers address some, but not all elements
Arpin, Loranth and Morrison paper:
Some corporate finance questions

Liability structure is a corporate finance question



Because of moral hazard and adverse selection corporations
have specific debt-equity structures, short, long-term, banks
versus bonds, covenants, etc
Also similar to optimal insurance contract design: deductibility,
risk-sharing, premiums
Also means partial guarantees is one of many tools


Optimal financial structure for any firm can include partial
guarantees, but also co-financing/-funding, Or for insurance:
Parallels suggest rich toolkit from which to draw optimal
contracts for (partial) guarantees: extend work if possible
Arpin, Loranth and Morrison paper:
Broaden set of tools

Can one fit partial guarantee in overall financing




Adjust financial structures/guarantees to the issues



Does this means a pecking order of tools? Are there tradeoffs?
Is there an optimal, joint combination? Probably, since that is
what we observe in most real corporate finance settings
Somewhat addressed, but can one model/quantify more general?
Nature of second-best should affect financial structure
E.g., sector, size, information, specific risks, inputs
Bring in government versus private sector

Is this a “standard” corporate finance? If not why does the
presence of government make it special?
Benavides and Huidobro: Framework

Framework not so clear



Does framework tell about efficacy of guarantee?



Why are banks risk-averse? Can one model the source:
capital adequacy requirements, managerial, are diversification
markets incomplete?
Is this (just) a matter of supply curve moving? If so, are
there other ways, e.g., subordinated loan to the bank to lower
risk?
Is this just a subsidy, but no leverage, or more?
Is the subsidy best given to the bank or to the borrower?
Minor: is break-even per unit or per program?
Benavides and Huidobro:
Data and Empirics

Many programs in Mexico




Data good, but with problems, not to be ignored
Aggregate analysis is mostly suggestive



Why are only government banks in this business?
Are the programs coordinated in some way? Seems not
Very hard to estimate credit demand functions, especially
with aggregate data. Few degrees of freedom, 67
observations. Many other reforms and changes over period
There are other, better tests of presence of credit constraints
In any case, bank or firm level analysis better


Would help provide answers to: targeting, additionality, etc.
If can not, could still consider overall crowding out of fiscal
How to model the best approach in
general for these problems?

Define objective



Is government unique relative to private sector?




Economic, social or political (economy)?
Static or dynamic?
If truly public goods in light of market failures (which?)
If better at enforcement, information (?)
If better at building institutional environment, dynamically
aided through guarantees
If not unique, or facing institutional and political
economy constraints


How can guarantees best be done by private sector/contract?
Is it (just) a question of how to optimally use a given amount
of public funds for a certain objective?
Can one model the best designs?

What is then the best design given objectives,
government and private sector constraints?
Auction of guarantees
 Support for financial intermediaries
 Subsidy to borrower
 Co-financing or co-funding
 Insurance type products
 Risk-sharing
 Guarantees, partial, full
 Etc. etc.
