Financial Development

1
Financial Development Corporate Growth

• IN DIRECT INTERSECTORAL COMPARISONS

Jan Bena and tepán Jurajda
LSE and CERGE-EI
LSE November, 2006
2
MOTIVATION

• There is positive cross-country correlation
  between financial development and economic
  activity (Goldsmith, 1969 King Levine, 1993).
• Size of finance-growth effect?
• Finance-growth effect on particular firm types?
• But how can we disentangle two-way causality?
• Supply ? Firms need external finance to reap
  growth opportunities.
• Demand ? Financial development reflects future
  growth opportunities.
• Growth opportunities are unobservable.
• No episodes of exogenous changes in financial
  development.
• Identification relies on strong assumptions.

3
DEALING WITH REVERSE CAUSALITY Literature

• Cross-country studies
• Initial period indicators of country financial
  development
• (King and Levine, 1993 Levine and Zervos,
  1998)
• Instrumental variables, mostly legal origin
• (La Porta et al., 1998 Levine et al., 2000)
• Regional differences within a single country
• Controls for unobservable country-level growth
  determinants
• (Jayarante and Strahan, 1996 Bertrand et al.,
  2004)
• Experience of specific industries across
  countries
• Quantify industry need of external finance
• (Rajan and Zingales, 1998 Beck et al., 2004
  Guiso et al., 2004)

4
THE RAJAN-ZINGALES STRATEGY Assumptions

• (A1) Industry growth opportunity shocks are
  global
• the same need to expand production
• (A2) Industry technology is also constant across
  countries
• the same of external finance to expand
  production by a unit
• ? Cross-industry differences in the need for
  external finance are the same across countries.
• (A3) U.S. listed firms face a perfectly elastic
  supply of external finance.
• ? Observed industry external finance dependence
  in the U.S. (US_EFD) serves as counterfactual for
  outside financing need in other countries.

5
THE RAJAN-ZINGALES STRATEGY Implementation

• Regress industry growth on
• country and global industry fixed effects
• interaction term US_EFDINDUSTRY
  Financial_DevelopmentCOUNTRY
• to ask whether industries more dependent on
  outside finance grow faster in financially more
  developed countries.
• STRONG Reverse causality endogeneity at country
  level.
• WEAK Constant industry differences in demand
  for external finance across countries such as
  U.S., Finland, Philippines, Zimbabwe, ...
• Direct tests of assumptions are not available.
• Similarity of technological content of industries
  across development levels threatened by empirical
  trade research (Schott, 2003).

6
AN ALTERNATIVE STRATEGY Assumptions

• (A1) Industry growth opportunity shocks are
  global.
• (A4) Corporate growth in a given industry would
  be the same in absence of differences in country
  financial development.
• Apply both assumptions in a more appropriate
  context
• EU-15 single market,1995-2003 (harmonized
  product market regulation)
• Industry growth is verifiably highly synchronized
• Comparable firms Age, Size, Leverage,
  Tangibility, Quoted, Ownership, ...
• ? Do two comparable EU-15 firms in the same
  industry but facing different financial system
  grow at different rates?

7
AN ALTERNATIVE STRATEGY Cost and Benefits

• Benefits of (A4)
• Avoid quantification of industry EFD
• Obtain economically measurable estimates
• Costs of (A4)
• Heterogeneity in financial development in EU-15
  assumed orthogonal to other country-level
  determinants affecting growth.
• ? Control for initial-period GDP to capture
  convergence effects.
• Use initial-period (predetermined) indicators of
  country financial development. What if markets
  are forward looking?
• ? Control for country-level future growth
  opportunities

8
BASIC SPECIFICATION
Gijkt a ßFDi ?GDPi dtj Xk'? eijkt
9
DATA

• Firm level Amadeus TOP 250 thousand for EU-15
• Real value-added growth of manufacturing firms
• Only public and private limited liability
  companies
• Remove state-owned firms
• Best firm-level EU data source available to date
  
• Country financial development indicators
• World Bank Financial Structure and Economic
  Development Database
• Total capitalization Includes debt securities
  (Hartmann et al., 2006)
• Control premium Private benefits of control
  (Dyck Zingales, 2004)
• Industry level OECD STAN
• Industry growth rates used to identify
  synchronized industries.

10
CORPORATE DESCRIPTIVE STATISTICS Firm-Year Data
over 1995-2003
11
FINANCIAL DEVELOPMENT The EU-15 over 1990-1994
12
FINANCIAL DEVELOPMENT AND CORPORATE GROWTH Basic
Estimates
13
FINANCIAL DEVELOPMENT AND CORPORATE GROWTH Basic
Estimates
14
FOCUS ON SYNCHRONIZED INDUSTRIES

• Our strategy is based on (A1) synchronization of
  industry growth shocks,
• so it will fail where industry growth is driven
  by local regulation.
• Hence, we identify synchronized industries using
  ANOVAs of industry growth with YEAR and COUNTRY
  factors.
• Synchronization corresponds to strong YEAR
  factors.
• Differentiate Low-, Medium-, and
  High-synchronization industry groups.
• Or use continuous synchronization measure.

15
DEVELOPMENT AND GROWTH Industry Synchronization
Groups
16
FIRM-TYPE INTERACTIONS
Gijkt a ß0FDi ß1xkFDi ?GDPi dtj
Xk'? eijkt
17
AGE Interaction
18
SIZE Interaction
19
COLLATERALIZATION Interaction
20
TANGIBILITY Interaction
21
ROBUSTNESS CHECKS

• Financial development measures misleading
• if they reflect not only differences in
  available supply of finance,
• but also demand for finance driven by future
  country growth opportunities.
• ? Control for predicted future country growth
• Take industry averages of EU-15 realized growth
  over 1995-2003, and weight them by initial-period
  country-level shares of each industry.
• Robustness to removing UK and Greece
• Median Regressions

22
Robustness to AGGREGATE GROWTH OPPORTUNITIES
23
Robustness to Removing UNITED KINGDOM and GREECE
24
CONCLUSIONS

• We apply simple cross-country comparisons in an
  appropriate setting
• within EU-15,
• to synchronized industries,
• to many similar firms, both large and small.
• As a result, we obtain
• coefficients that translate to easy-to-interpret
  magnitudes,
• differences in the finance-growth effect by firm
  types.
• Findings
• Move from the least to the most developed
  financial system within the EU-15 boosts firm
  annual growth rate by 2 to 3 percentage points.
• Young firms have limited access to financial
  markets.
• Ability to provide collateral helps to get
  outside finance, especially if the firm is small
  or young.
• We do not find a size effect interaction.