Financial Sector Reform

1
Financial Sector Reform
Discussion on Bekaert, Harvey and Lundblad
Galindo and Micco

• Sergio Schmukler
• Rethinking Structural Reform Conference
• Federal Reserve Bank of Atlanta
• Inter-American Development Bank
• October 2003

2
Discussion Outline

• Reform agenda logic and scope
• Broader than these papers
• Broader than financial sector reforms
• Outcomes for Latin America mixed or
  disappointing
• Extensive to other regions
• Not as predicted by these papers
• If reforms did disappoint, why?
• Policy alternatives to these papers
  recommendations
• Liberalization papers
• Creditor rights paper

3
1. Reform Agenda Logic Generate more and cheaper
financing
GOVERNMENT
FIRMS
Supervision and regulation
More financing
BANKS (fragile and inefficient)
Increased competition
Higher returns
Cheaper financing
CAPITAL MARKETS
Wider set of instruments
INVESTORS
4
1. Reform Agenda Logic Sweeping reforms
throughout financial sector

• Depository and clearing
• Trading platforms

• Supervisory agency
• Investor protection

Lower transaction costs
Improve exchange infrastructure
Mitigate information and agency problems
More retail investment
CAPITAL MARKETS
SUPPLY
DEMAND
Long-term financing and specialization
Foreign capital available
Increase liquidity
Promote stocks dissemination
More discipline and efficiency

• Pension funds
• Mutual funds
• Insurance companies

• Financial liberalization

• Privatization
• Tax incentives

5
1. Reform Agenda Actions Financial
liberalization increasing since late-1980s
Latin America Indices of Financial
Liberalization by Sector
More liberalization
Less liberalization
Countries included are Argentina, Brazil, Chile,
Colombia, Mexico, Peru, and Venezuela. The value
1 means repression, 2 partial liberalization, and
3 full liberalization. Figures correspond to
end-of-month values. Source Kaminsky and
Schmukler 2002
6
1. Reform Agenda Actions Other reforms
implemented in the early-1990s
Percentage of Latin American Countries that
Implemented Reforms
The figure shows the percentage of countries
(from a group of 16 countries) that had
implemented reforms before 1990, in 1990-1995,
and in 1996-2001. Countries included are
Argentina, Bolivia, Brazil, Chile, Colombia,
Costa Rica, Ecuador, Guatemala, Honduras,
Jamaica, Mexico, Panama, Paraguay, Peru, Uruguay,
Venezuela. Sources Bhattacharya and Daouk 2000,
ICRG, and local data
7
1. Reform Agenda Actions Reforms occurred mostly
after liberalization
Sequencing Financial Liberalization and
Institutional Reforms
The table shows the percentage of coutries that
had implemented reforms before partial and full
liberalization. Countries included are Argentina,
Brazil, Chile, Colombia, Mexico, Peru, and
Venezuela. Sources Bhattacharya and Daouk 2000,
Kaminsky and Schmukler 2001, and ICRG
8
2. But Results Not as Expected

• (Reforms are good according to these papers)
• Equity markets
• Visible growth (market cap), but not as in OECD
• Rising concentration and delisting
• Migration to international equity markets
• Small relative to bond markets

9
2. But Results Not as Expected

• Bond markets
• Dominated by government paper
• Dollarization
• Short duration
• Currency and maturity mismatches
• Institutional investors
• Fast growth but tapped by governments

10
2. But Results Not as Expected Latin America
Equity Market Growth Is Poor
Stock Market Capitalization
Percentage of GDP
Countries included are Argentina, Brazil, Chile,
Colombia, Mexico, Peru, and Venezuela. Asian
countries Hong Kong, Indonesia, Korea, Malaysia,
Philippines, Taiwan, and Thailand. G-7countries
Canada, France, Germany, Italy, Japan, U.K., and
U.S. Figures correspond to end-of-year
values. Sources IFC's Emerging Markets Database,
World Federation of Exchanges (FIBV), and The
World Bank
11
2. But Results Not as Expected Stock Exchanges
Affected by Increasing De-listing
Domestic Stock Exchanges Number of Companies
Source IFC's Emerging Markets Database
12
2. But Results Not as Expected Large Companies
Migrate to International Markets
Latin American Companies Value Traded
USD Millions
Countries included are Argentina, Brazil, Chile,
Colombia, Mexico, Peru, and Venezuela.
Internationalized companies are defined as
companies that cross-list or raise capital in
international stock markets at some point in
time. Figures correspond to end-of-year
values. Sources The Bank of New York, Euromoney,
and IFC's Emerging Markets Database
13
2. But Results Not as Expected Equities Traded
Belong to a Handful of Large Firms
Stock Market Concentration (end-2000)
Percentage
Source IFC's Emerging Markets Database
14
2. But Results Not as Expected Public Bond
Trading Outweighs Equity Trading
Value Traded in Domestic Markets (2000)
Percentage of GDP
The figure shows the value traded through the
domestic exchanges. However, in the case of
Mexico, repo operations (conducted or not through
the exchange) are also considered. Sources Local
data, The Handbook of World Stock, Derivative
Commodity Exchanges 2001, Federacion
Iberoamericana de Bolsas de Valores (FIABV), and
The World Bank
15
3. If Reforms Did Disappoint, Why?

• Wrong sequencing
• Financial integration went too fast
• Imperfections in international markets lead to
  vulnerabilities when domestic sector is not ready
  
• Thus, follow right sequencing
• But, are there sufficient incentives to reform
  institutions without the discipline from openness
  and crisis?
• Wait for the fruits of reforms and do more
• The dividends from reforms have long gestation
  period
• Thus, accelerate pace
• But, we all know what happens in the long run

16
3. If Reforms Did Disappoint, Why?

• Wrong expectations
• Recommendations based on cross-sectional evidence
• Reforms did not tackle well some basic issues for
  EMs
• Macro policy
• Systemic risk
• Domestic and external shocks, e.g. capital flows
  volatility
• Institutional factors size, information
  asymmetries, moral hazard
• Globalization
• Thus, redesign reform to address the basic
  issues, and revise expectations
• But, many of these issues are intrinsic to
  emerging markets and solving them is a daunting
  task

17
4. Liberalization Papers

• Great papers!
• Important contributions
• Welfare implications relative to previous work
• Growth mean and volatility
• Solid work
• Almost painful to read, overwhelming force
• Incorporated most comments received in the past
  years, so difficult to discuss
• Particularly for the growth mean paper
• Volatility paper needs to address the same
  comments that spur paper already did

18
4. Liberalization Papers

• Potentially relevant issues to consider
• One of the most fundamental national policy
  decisions of the past 25 years has been the
  liberalization of equity markets across the
  world.
• Here, allowing foreigners to purchase domestic
  stocks
• But witness size of stock markets
• And witness effects of capital account
  liberalization
• Which samples/estimations make sense?
• Mix of developed and developing, including very
  poor countries
• Including countries with no change in
  liberalization
• Time series for emerging markets
• Fixed effects estimations
• Coefficient diminishes to 0.56
• But why including countries with no variation in
  liberalization here?

19
4. Liberalization Papers

• Potentially relevant issues to consider
  (continued)
• More on econometrics
• Overlapping observations
• How are common factors treated besides SUR
  specification?
• How is endogeneity addressed, specially given
  that it comes up in the growth paper?
• Removing 97-00 seems arbitrary
• Crises as outliers? Rationale?
• Why including 99 and 00?
• Perhaps, crises consequence of liberalization, as
  suggested
• Even with crises, other evidence points towards
  lower volatility

20
4. Liberalization Papers

• Caveats related to liberalization measures
• Wider set of financial liberalization measures
• Pace of liberalization
• Reversals
• Intensity of liberalization
• Capital account liberalization
• Might seem more important to many readers
• Treatment of capital account liberalization

21
4. Liberalization Papers

• Other specific points
• More on economic significance would be welcomed
• Results weakened with other controls
• For LAC countries, mean results insignificant
• Other regressors
• Other controls, like GDP per capita
• Level of inflation might matter
• Risk sharing
• Perhaps, could be another paper
• Ability versus actual risk sharing

22
5. Creditor Rights Paper

• Great paper!
• Important points of the paper
• Effectiveness of creditor rights, interacted with
• With rule of law
• With efficiency of the judiciary
• With small and medium firms
• Perhaps nicest results
• Three effects of creditor rights
• Credit/GDP
• Share of bank credit financing of small and
  medium firms
• Change in real credit (procyclicality)

23
5. Creditor Rights Paper

• Potentially relevant issues to consider
• More consistency across empirical analysis
• Creditors rights, rule of law, or broad
  institutions?
• No multivariate analysis, omitted variables
• Only done very partially for change in credit,
  when controlling both for creditor rights and
  rule of law
• Perhaps, control for GDP per capita, given that
  GNP and GDP growth not significant
• Interaction might result more from law and order
  explanatory power than from creditor rights
• Perhaps interact with GDP per capita and other
  variables
• Why is creditor rights insignificant by itself
  second part?

24
5. Creditor Rights Paper

• Potentially relevant issues to consider
  (continued)
• Why procyclicality?
• Risk can be reduced ex-ante, knowing state of
  institutions
• More explanation would be welcomed
• Policy recommendations
• Cross section evidence versus time series
  recommendation
• More reforms needed
• Foreign jurisdictions
• Systemic shocks and creditors right
• Repudiation of contracts
• Shocks to collateral
• In fact, acknowledged importance in institutional
  analysis

25
5. Creditor Rights Paper

• More on data and variables
• Macro variables not significant, except budget
  deficit
• WEBS survey
• Only for 1999
• Share of investment financed with bank credit and
  legal protection
• What about other type of financing
• Other specific point
• Econometrics
• Procyclicality results weak using IVs
• Clustering standard errors