slide1 l.
Skip this Video
Loading SlideShow in 5 Seconds..
Financial System Efficiency PowerPoint Presentation
Download Presentation
Financial System Efficiency

Loading in 2 Seconds...

play fullscreen
1 / 31

Financial System Efficiency - PowerPoint PPT Presentation

  • Uploaded on

Bank Control, Capital Allocation, and Economic Performance Randall Morck University of Alberta M. Deniz Yavuz Oblin, Washington University in St Louis Bernard Yeung NUS Is the control structure of the banking system related to capital allocation efficiency and economic outcomes?

I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
Download Presentation

PowerPoint Slideshow about 'Financial System Efficiency' - elina

An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.

- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript

Bank Control, Capital Allocation, and Economic PerformanceRandall MorckUniversity of AlbertaM. Deniz YavuzOblin, Washington University in St LouisBernard YeungNUS

financial system efficiency
Is the control structure of the banking system related to capital allocation efficiency and economic outcomes?

family, state, and independent control of banks


Fraction of banks control by business families is correlated with

less efficient allocation of capital,

more non-performing loans,

more bank crisis,

more macro volatility, and

slower growth

Opposite is true for independent bank control

State control is correlated only with less efficient allocation of capital and more NPL

Financial System Efficiency
motivation type of wealth and growth
Motivation: Type of wealth and growth

The Cross-Country Relationship Between Growth in Real Per Capita GDP and Capital Ownership, Controlling for Current per Capita Income, Capital Investment Rate, and Level of Education (from Morck, Stangeland, and Yeung, 2000)

H1 includes only the wealth of billionaires known positively to be heirs, politicians or politicians= relations. H2 also includes the wealth of billionaires who are probably heirs. H3 includes H1 plus fortunes jointly controlled by a founder and his heirs. H4 includes all the above. H5 through H8 are analogous to H, H2, H3 and H4 but do not include politician billionaires and their relations.

why such results
Around the world firms have “controlling owners,” LLSV 1999

a large fraction of these dominant owners are wealthy families.

Via pyramids, etc., these wealthy families gain control of a large fraction of corporations and corporate assets

+ more micro efficiency among family controlled set of firms

Overcome inefficiency in using the market – due to information asymmetry and regulatory/bureaucratic burdens

– what is good for a set of firms can be bad for the economy

Intrinsic results due to biased resources allocations

Economic entrench – rent seeking

Lobby for entry barriers (Fogel 2006)

Capture the capital market

Why such results?
why capture the financial market
Developed financial markets encourage “creative destruction” and promote competition. (Schumpeter, 1942)

Not always in the best interest of elite/powerful-families to promote efficient financial systems.

Capture the institutional development – save capitalism from capitalist

Morck, Stangeland, Yeung, 2000, 2002, 2005; Rajan and Zingales 2003, 2004;….Perotti and Vorage 2008 etc

Morck et al 2002, Rajan and Zingales 2003

An economy may have no stock markets, but, cannot do without banks

Capturing the banking system

Family bank ownership was first looked at Caprio et al JEI 2007

Why capture the financial market?
let us not presume
The positive and negative consequences of family control of banks

Efficiency view (mostly applies at the micro level):

Interest alignment and mitigating information asymmetry (Diamond 1984; Hoshi et al. 1991; Khanna et al. 2000; Fisman & Khanna 2004; Almeida & Wolfenzon 2006, Caprio et al 2007, etc)

Entrenchment view (has macro level implications):

Family banks may favor related firms (see, e.g., La Porta et al. 2003), limit capital to potential competitors,

ATM – Engage in risk shifting, government bails out (see, e.g., La Porta et al. 2003, Haber 2002).

Add state ownership

LLSV, “Government Ownership of Banks,” JFE, 2002

Let us not presume
empirical investigation of
Relationship between bank ownership structure and

Efficiency in capital allocation

NPL, Bank Crisis

Macro Volatility


Efficiency vs. economic entrenchment?

Empirical investigation of ….
dataset and ownership variables
Starting sample is the 10 largest banks in 44 countries (Banker 2001).

Caprio et al. (2007) provide control structure of 244 public banks.

We add the ownership structure of private banks (total 318 banks).

Most of the data come from Bankscope 2001.

Controlling shareholder = largest ultimate shareholder with stake ≥ 10%.

Controlling classifications: State, family, independent.

Bank Control variables: Fraction of the banking system controlled by state, family and independent banks weighted by total loans.

Same if weighted by assets

Dataset and Ownership Variables
capital allocation efficiency measure 1
Elasticity of capital spending with respect to the value added, by industry.

Version 1: UNIDO (UN General Industrial Statistics) Data for 1993 through 2003.

Version 2: UNIDO (UN General Industrial Statistics) Data for 1963 through 2003.

Capital Allocation Efficiency Measure 1

as in Wurgler (2000)

capital allocation efficiency measure 2
Capital Allocation Efficiency Measure 2
  • Follow Rajan & Zinglaes (1988)
    • US industries’ external financing dependence
      • [(capital expenditure – cashflow from operation)/ capital expenditure], average over 80-90
        • Dummy = 1 if above median
        • Or just the raw data
measures of capital allocation efficiency
Measures of Capital Allocation Efficiency
  • Regress
    • growthi,cis annualized compound value-added growth rate from 1993 to 2003 of industry i in country c;
    • δi = 1 if industry i is more dependent on external financing than the median industry
    • the k,c are the fractions of country c’s banking system controlled by families (k = f), state-controlled (k=s), or widely held (k=w)
    • si,c is the share of industry i in country c
    • dc and di are country and industry fixed effects
data and econometric problems
Please hold your questions on

Data timing (contemporary data on ownership (2001-2003) and target dependent variables (1993 – 2003)



Will deal with them after providing a snap shot of the results

Data and econometric problems
other critical variables
Nonperforming loans /Total loans outstanding.

WDI site – World Development Indicators (World Bank)

Average between 1993 and 2003.

Do a logistical transformation

Bank crisis Dell’ Ariccia et al 2008

Deposit runs, bank holidays or nationalization, fiscal cost of bank rescues > 2% GDP, NPL > 10% bank assets

1993 – 2003

Add 2008 – IMF, direct government rescue attempt 2008

Other Critical variables
other critical variables14
Macro variables – Beck et al 2003, Beck et al 2000, Penn world Tables,

Real per capita GDP growth

1993 through 2003 and 1963 to 2003

Estimate as the time trend

TFP Growth – 1993-2003

a= 30%

using logarithms of first differences for Y, K, and L to estimate the rate of change of the parameter A = TFP growth

Capital accumulation – Beck et al (2000), 1993 - 2003

d = 7%

1964 as the starting points

Forward iteration to get capital stocks

Calculate the rate of change

Economic stability, 1993 - 2003

the standard deviation of log first differences in real per capita GDP

Other Critical variables
Family ownership of banks is associated with

Less allocation efficiency

more NPL and bank crisis,

more macro volatility

slower growth

Opposite result for independently owned banks

State ownership is associated with

Less allocation efficiency

More NPL


Economic entrenchment, not efficiency argument

Previous results on the effect of state owned bank may need refinement

So …
robustness bank control data
Robustness – Bank control data
  • Different ways of constructing bank control variables:
    • Calculate bank control variables by using total assets as weights.
    • Control assumed at 15% or 20% instead of 10%.
    • Ignoring countries with number of banks < 3.
robustness bank control data20
Contemporary data – ownership and the target variables

Should use lagged bank control variables

Cannot get them

May not be that important if control structure changes slowly.

Only 14 banks (4.4% of the total 318) switch category between 2001-2007.

Still, many bank privatizations between 1993-2001 from Megginson (2004).

14 (out of 324) banks change control from 2001 to 2007

2 family banks  state controlled, 4 becomes widely held

4 state controlled  widely held

2 widely held  family owned, and 2 become state controlled

Beware of the “Chen” factor –

incompetent and corrupt government passes the control of state owned banks to rich families

Result = corrupt and incompetent government

Do the data both way

Move all family owned back to state owned

Or, like we did before, keep them as family owned

Or, take a weighted average for each

Identical results

So, the mis-classification not a factor

Robustness – Bank control data
Alternative Regressions:

Regression robust to outliers.

Weighted least squares regressions.

Correcting for arbitrary heteroscedasticity vs simple OLS

Expanding the sample to 2007 (GDP data)

no change

Expanding the crisis data

E.g., use both Dell’ Ariccia et al (2008) and Demirguc-Kunt et al (2006) , result the same

robustness in the interpretation using banks to entrenchment
Proxying for under development in financial markets?

Control for capital market development

In the Rajan and Zingales regression (panel B table 5), we put in cross terms between external finance dependence and capital market development,

Same result

Just due to “oligarchic” behavior?

Put in an oligarchy variable as an additional control

Or form another cross term in the R& Z regression

same result

Robustness in the interpretation:using banks to entrenchment
causality and endogeneity
Causality and endogeneity
  • Reverse causality – poor environment and therefore own banks?
    • Cannot do Granger
      • Limited sample size and data
      • Instrument?
        • done that, works, but, not really credible because there could be many hidden channels these instruments affect growth or stability
    • Still, the R&J regression results
      • Fixed industry and country effects included, results therefore not due to latent variables at the industry or country level
      • If reverse causality, families own banks to support their firms, their bank ownership should mitigate the slower growth of external financing dependent industry in poor environment.
        • Found the opposite
    • Clinical country level exists – Haber (2004)
    • Shall give more results
instrumental variable
Instrumental variable
  • To deal with endogeneity using time invariant instrumental variables:
    • legal origin (La Porta et al. 1998, 2008) – financial development
    • religion (Stulz and Williamson, 2003) – credit rights
    • latitude (Hall and Jones, 1999; Acemoglu et al. 2001) – Western influence
    • Use them to predict ownership (Tobit)
      • French legal origin = family ownership most prevalent
      • Protestant countries = independently owned banks most prevalent
      • Highest latitude = independently owned banks most prevalent
  • Instrumental variable result = table X
    • Plead guilty – legal origin and religion are exogenous, but, they may have a relationship with our target performance variables
more on robustness in the economic interpretation
Incompetent heir effect?

Step one, improve the bank ownership data

Get ownership data to identify families that do and do not own other business

“Orbis” to identify other business owned

Use also the internet

Reclassify those with no “known other business” as independent

Same result

More on robustness in the economic interpretation
Step two

Dig up alternative stories

Incompetent family?

Find that family ownership is highly related to “entry barriers

Number of procedures,

Time and costs to set up a firm

Aside – family bank ownership is related to more income inequality and a smaller middle class too

Control of the banking system is important for efficiency of capital allocation.

Family and state control over banks have similarly economically significant and negative implications for capital allocation efficiency.

However, family control is also correlated with lower economic growth, elevated macroeconomic instability, and more income inequality.


Do not let families capture the banking system!

Multiple governments (e.g., Singapore, Canada) actually stipulate against bank owning families from holding non-banking business

"Measures to separate financial and non-financial activities of banking groups“ – Speech by DPM Lee Hsien Loong At The Association Of Banks in Singapore (ABS) 21 June 2000

“with banking and non-banking activities inter-meshed within a conglomerate, there will be a strong tendency to stretch any safety net intended for the banking system also to cover non-bank operations in the group.”

Why is family ownership of banks common?

Political Economy Outcome.

Rajan and Zingales (2004); Morck, Wolfenzon, Yeung(2005); Stulz (2005); Perotti and Volpin(2007); Haber and Perotti(2008); Acemoglu, Johnson and Robinson (2008).

Families are the highest bidders because their efficiency gains are the highest.