Does Financial Liberalization Spur Growth? Geert Bekaert Columbia University and NBER Campbell R....

Does Financial Liberalization Does Financial Liberalization Spur Growth?Spur Growth?

Geert BekaertColumbia University and NBER

Campbell R. HarveyDuke University and NBER

Christian T. LundbladIndiana University

Emerging Markets Corporate FinanceFebruary 2003

2

Real GDP Growth Five Years Before and After Financial Liberalizations

-0.02

-0.01

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

Pre-Liberalization Post-Liberalization

3

Financial Development

Growth

Financial Liberalization

Relaxing FinConstraints

Investment

GrowthOpportunities

Efficiency ofInvestment

Cost of Capital

4

Controversial exercise• Liberalization implies consumption booms and

inefficient investment (crisis literature)

• Liberalization may lead to reduced savings (endogenous growth literature)

• Liberalization may lead to “hot speculative capital” and induce capital flight (Stiglitz & others)

Financial Liberalization and Growth

5


What we already know (too many references to list!):

• Financial/banking development associated with higher growth

• Cost of capital decreases

• Investment increases

6

Financial Liberalization and GrowthOutline:1. Did liberalization spur growth?

– Large panel of data– Cross-sectional growth regression with temporal dimension

2. How did liberalization spur growth?3. Accounting for the liberalization effect

– Is is macro-economic reforms?– Is it financial development?– Other simultaneity biases?

4. Conclusions

7


Caveats:Not much guidance from theory.

• As a result, it is important to conduct extensive robustness experiments

8


Econometric Framework:

ktititiikkti LibXQy ,,,1980,,,

where yi,t+k,k is real per capita GDP growth between t and t+kQi,1980 is initial GDP,Xi,t represents control variablesLibi,t is a Liberalization indicator variable

9



)](Z)SX[()](Z)SX[(ˆ

as written becan estimator The

],,[

],,[

Let

111

,,1980,,

YZXZ

LibXQx

TT

titiiti

10



NN

kktitii

X

X

X

Z

X

X

X

yYxX

00

00

00

,

][ and ][given where,

2

11

,,,

ST is the variance covariance matrix of the sample orthogonalityconditions

11


Key issues:• Temporal dimension• Different weighting matrices• Liberalization variable• Choice of “k”• Endogeneity of the liberalization decision

12


Liberalization dates:

• Use Bekaert and Harvey (JF 2000) “official liberalization” dates

• These dates are based on a detailed chronology of important regulatory events

• Augmented with IFC frontier markets and three developed markets, Spain, New Zealand and Japan

13


Data: Four samples determined by availability of data

• Sample I: 95 countries

• Sample II: 75 countries

[macroeconomic and demographic data]

14


Data: Four samples determined by availability of data

• Sample III: 50 countries• Sample IV: 28 countries[add financial development indicators]

As data requirements become more stringent, the variance of GDP levels across countries in the sample decreases.

15


Liberalization dates:• Use Bekaert and Harvey (JF 2000) “official

liberalization” dates• These dates are based on a detailed

chronology of important regulatory events • Augmented with IFC frontier markets and

three developed markets, Spain, New Zealand and Japan

16


Liberalization dates:• Robustness of our results checked by examining

Bekaert and Harvey (2000)’s “First Sign” dates• These dates based on the earliest date of {official

liberalization, first ADR and first closed-end fund}

• Example: Thailand – “Official” 1987:09– “First Sign” 1985:07

18


Liberalization dates: • Capturing “intensity” or “comprehensiveness”

of the liberalization– Ratio of IFC investable market cap to global stocks

(Bekaert (1995) and Edison and Warnock (2002))– U.S. holdings of domestic market capitalization

• Is it just a proxy for capital account openness? [See Rodrik-Edwards debate]

19


Are the dates exogenous?

Counter examples

• Spain in the EU

• Some countries cannot liberalize their financial markets

20


Findings so far:

• We document a liberalization effect on growth with certain "standard control variables"

21

Table 2Classic Growth Regression and the Impact of LiberalizationAnnual Average Real GDP Growth Rate (Five-Year Horizon)

A: Classic Growth RegressionSample I II III IV

Constant -0.2369 -0.2581 -0.1559 -0.3438 Std. error 0.0181 0.0216 0.0295 0.0538

Log(GDP) -0.0071 -0.0069 -0.0093 -0.0156 Std. error 0.0006 0.0006 0.0007 0.0011

Govt/GDP -0.0044 -0.0175 -0.0212 -0.0158 Std. error 0.0085 0.0096 0.0105 0.0164

Enrollment 0.0277 0.0059 0.0162 0.0624 Std. error 0.0079 0.0098 0.0119 0.0174

Population Growth -0.5978 -0.6037 -0.9389 -1.2056 Std. error 0.0577 0.0642 0.0791 0.1121

Log(Life Expectancy) 0.0746 0.0806 0.0624 0.1182 Std. error 0.0049 0.0057 0.0079 0.0140

22

Table 2Classic Growth Regression and the Impact of Liberalization

B: Classic Growth Regression with Liberalization IndicatorConstant -0.2281 -0.2374 -0.1493 -0.2018 Std. error 0.0179 0.0214 0.0286 0.0658

Log(GDP) -0.0094 -0.0088 -0.0115 -0.0158 Std. error 0.0007 0.0007 0.0008 0.0011

Govt/GDP -0.0039 -0.0178 -0.0187 -0.0301 Std. error 0.0087 0.0098 0.0105 0.0165

Enrollment 0.0305 0.0112 0.0243 0.0566 Std. error 0.0077 0.0097 0.0116 0.0171

Population Growth -0.5594 -0.5731 -0.8159 -1.1013 Std. error 0.0621 0.0691 0.0835 0.1151

Log(Life Expectancy) 0.0755 0.0781 0.0627 0.0838 Std. error 0.0049 0.0056 0.0076 0.0167

Official Liberalization Indicator 0.0095 0.0083 0.0113 0.0130 Std. error 0.0016 0.0017 0.0020 0.0036

23


Findings so far:

• The liberalization effect is robust to– different definitions of liberalization dates– to business cycle or interest rate controls– allowing for intensity of liberalization

...and independent of capital account liberalization

24

Table 3Robustness of the Liberalization Effect

A: Sensitivity to Regional InfluencesSample I II III IV

Official Liberalization Indicator (Latin) 0.0089 0.0075 0.0068 0.0138 Std. error 0.0034 0.0034 0.0035 0.0071

Official Liberalization Indicator (Not-Latin) 0.0108 0.0099 0.0136 0.0133 Std. error 0.0016 0.0017 0.0019 0.0035

B: Sensitivity to Contemporaneous World Growth and Real Interest RatesSample I II III IV

OECD GDP growth (contemporaneous) 0.1469 0.1177 0.1435 0.1695 Std. error 0.0301 0.0328 0.0315 0.0394

World real interest rate (contemporaneous) -0.0625 -0.0484 -0.0132 -0.0182 Std. error 0.0282 0.0309 0.0300 0.0372


25

Table 5Alternative measurements of liberalization

Sample I II III IV

Official Liberalization Indicator (Table 2) 0.0095 0.0083 0.0113 0.0130 Std. error 0.0016 0.0017 0.0020 0.0036

First Sign Liberalization Indicator 0.0102 0.0091 0.0105 0.0094 Std. error 0.0015 0.0016 0.0018 0.0033

Intensity Indicator A 0.0101 0.0089 0.0131 0.0132 Std. error 0.0015 0.0018 0.0016 0.0038

Intensity Indicator B 0.0120 0.0110 0.0159 0.0155 Std. error 0.0015 0.0018 0.0016 0.0036

Official Lib. Indicator*(U.S. Holdings/MCAP) 0.0352 0.0325 0.0291 0.0173 Std. error 0.0067 0.0067 0.0067 0.0073

IMF Capital Account Lib. Indicator 0.0022 0.0014 0.0061 0.0043 Std. error 0.0007 0.0008 0.0008 0.0010

IMF Capital Account Lib. Indicator 0.0004 0.0001 0.0044 0.0033 Std. error 0.0008 0.0009 0.0007 0.0009


26


Channels of increased growth:

• Both: » increased investment, partially through a cost of capital

effect and » increased productivity (which is different from the

financial development literature)

27


On the mechanism ...

» Liberalization does not lead to consumption binge – investment increases– trade balance decreases

28

Table 7The Channels of the Liberalization Effect Liberalization and the Components of GDP

Sample I II III IV

Investment/GDP 0.0141 0.0116 0.0079 0.0127 Std. error 0.0025 0.0026 0.0034 0.0053

Consumption/GDP 0.0082 -0.0246 -0.0085 0.0123 Std. error 0.0043 0.0046 0.0049 0.0071

Government/GDP -0.0027 0.0006 0.0053 0.0041 Std. error 0.0006 0.0009 0.0013 0.0019

Exports-Imports/GDP -0.0211 -0.0107 -0.0061 -0.0231 Std. error 0.0017 0.0013 0.0029 0.0024

29

Macroeconomic Impact of Liberalization

-0.0300

-0.0250

-0.0200

-0.0150

-0.0100

-0.0050

0.0000

0.0050

0.0100

0.0150

0.0200

I II III IV

Investment/GDP Consumption/GDP Government/GDP Exports-Imports/GDP

30



» Investment increases - but you need a minimum “country quality level” to see effect– decreased cost of capital associated with more

investment

31

Table 8Investment, the Cost of Capital and the Liberalization Effect

Sample II III IV

Official Liberalization Indicator -0.0581 -0.1090 -0.1802 Std. error 0.0166 0.0321 0.0466

Log(Credit Rating) -0.0016 -0.0068 -0.0181 Std. error 0.0016 0.0061 0.0095

Log(Credit Rating)*Lib Indicator 0.0172 0.0315 0.04900.0041 0.0081 0.0113

Minimum credit rating requiredfor liberalization to positively 29.3 31.8 39.6impact investment

Official Liberalization Indicator 0.0213 Std. error 0.0071

Dividend Yield -0.0719 Std. error 0.0403

Div Yield*Lib Indicator -0.2209 Std. error 0.0538

32



» Productivity increases– and this is not just a banking development effect

33

Table 9Liberalization, Total Factor Productivity and Capital Growth

Sample I II III IV


Private Credit/GDP 0.0036 0.0037 0.0033 -0.0023 Std. error 0.0013 0.0013 0.0014 0.0019


34


Accounting for the liberalization effect:

• We investigate whether part of the effect can be ascribed to» macroeconomic reforms» financial development» other regulatory reforms

35



Macroeconomic reforms ...

» Liberalization not spuriously reflecting macroeconomic reforms– we control for trade openness, inflation, black market

premiums, and government deficits

36

Table 10Macroeconomic Reforms and Financial Liberalization

Sample I II III IV

Trade 0.0090 0.0105 0.0089 0.0084 Std. error 0.0007 0.0007 0.0010 0.0012

Log(1+Inflation) (Latin) 0.0010 0.0004 0.0011 -0.0003 Std. error 0.0017 0.0016 0.0017 0.0024

Log(1+Inflation) (Not Latin) 0.0088 0.0057 0.0036 -0.0450 Std. error 0.0021 0.0024 0.0053 0.0149

Log(1+Black Market Premium) -0.0093 -0.0096 0.0005 0.0048 Std. error 0.0014 0.0015 0.0014 0.0071

Fiscal Deficit -0.0738 Std. error 0.0178


37



Financial development ...

» Degree of banking and equity market development is important but independent boost from liberalization – we examine the size of private credit, equity market

activity, and equity market size

38

Table 11Financial Development and Financial Liberalization

A. Banking sector development Sample I II III IV

Private Credit 0.0204 0.0234 0.0159 0.0298 Std. error 0.0030 0.0033 0.0046 0.0082

Private Credit*Lib indicator -0.0213 -0.0244 -0.0131 -0.0329 Std. error 0.0031 0.0034 0.0045 0.0082


Impact on growth of increasing private credit from median of segmented countries to median of liberalizedNonliberalizing country 0.0097 0.0101 0.0061 0.0101Liberalizing country 0.0139 0.0126 0.0145 0.0145Increment to growth by liberalizing 0.0042 0.0025 0.0084 0.0044

Impact on growth resulting from liberalization with no change in private credit/GDPLow level of private credit/GDP 0.0143 0.0130 0.0134 0.0156High level of private credit/GDP 0.0042 0.0025 0.0084 0.0044

39

Table 11Financial Development and Financial Liberalization

B. Trading activity

Turnover 0.0170 0.0097 Std. error 0.0037 0.0071

Turnover*Lib Indicator -0.0145 -0.0078 Std. error 0.0038 0.0074

Official Liberalization Indicator 0.0135 0.0151 Std. error 0.0019 0.0040

Impact on growth of increasing turnover from median of segmented countries to median of liberalizedNonliberalizing country 0.0042 0.0009Liberalizing country 0.0132 0.0132Increment to growth by liberalizing 0.0090 0.0124

Impact on growth resulting from liberalization with no change in turnoverLow level of turnover 0.0126 0.0131High level of turnover 0.0090 0.0124

40



Other regulatory reforms ...

» The financial liberalization/growth effect is not a post-banking crisis effect

» The enforcement of law and institutions are important

41

Table 12Banking Crises and Financial Liberalization

A. Growth During Crisis PeriodsSample I II III IV

During Systemic Crisis -0.0048 -0.0066 -0.0062 -0.0095 Std. error 0.0007 0.0008 0.0009 0.0018


During Systemic and Borderline Crisis -0.0048 -0.0064 -0.0081 -0.0075 Std. error 0.0005 0.0006 0.0007 0.0011


42

Table 12Banking Crises and Financial Liberalization

B. Post Crisis Period Growth

Post Systemic Crisis 0.0062 0.0060 0.0001 0.0017 Std. error 0.0009 0.0010 0.0013 0.0042


Post Systemic and Borderline Crisis 0.0049 0.0048 0.0033 0.0050 Std. error 0.0006 0.0007 0.0007 0.0011


43

Table 13Insider Trading and the Liberalization Effect

Sample I II III IV


Insider Trading Law 0.0001 0.0004 -0.0005 -0.0004 Std. error 0.0006 0.0006 0.0008 0.0009

Official Liberalization Indicator 0.0092 0.0080 0.0116 0.0138Std. error 0.0016 0.0019 0.0017 0.0044

Insider Trading Prosecution 0.0036 0.0031 0.0037 0.0015Std. error 0.0008 0.0009 0.0010 0.0012

44


Conclusions

• Financial liberalization spurs growth by 1% per annum over the five years

• Survives a battery of robustness experiments

• We understand better the channels whereby growth impacted by financial liberalization

• Liberalization effect not spuriously accounted for by a host of other events such as macro-economic reforms

45


Conclusions

• Financial liberalization has a very important economic effect

46

Gov/GDP 75th to 50th

Enrollment 25th to 50th

Pop Growth 75th to 50th

Life Exp. 25th to 50th

Liberalization

Total Growth = 3.02%


Conclusions• Financial liberalization has a very important economic effect

• Consider economic impact of improvements plus a equity market liberalization

Liberalization

47


Future and on-going research

• Growth volatility

• Financial development

• Liquidity and asset pricing

• The sequencing of liberalizations

48

Table 1

Growth Volatility and LiberalizationGDP growth rate standard deviation

I I1980-2000 1980-1997

Constant 0.1936 0.2179 Std. error 0.0142 0.0127Initial GDP 0.0046 0.0052 Std. error 0.0003 0.0003School -0.0116 -0.0077 Std. error 0.0025 0.0032Log(Life) -0.0471 -0.0547 Std. error 0.0036 0.0035Gov/GDP -0.0106 0.0048 Std. error 0.0067 0.0074PopGR 0.1159 0.1184 Std. error 0.0989 0.1082Official Lib. -0.0057 -0.0107 Std. error 0.0007 0.0009

Does Financial Liberalization Spur Growth? Geert Bekaert Columbia University and NBER Campbell R....

Documents

Transcript of Does Financial Liberalization Spur Growth? Geert Bekaert Columbia University and NBER Campbell R....