Transcript Financial Deepening and Bank Productivity in Latin American

Financial Deepening and Bank
Productivity in Latin America
Georgios Chortareas
University of Athens
Claudia Girardone
University of Essex
Jesus G. Garza Garcia
University of the West of England
Cass Business School
Emerging Scholars in Banking and Finance
December 9th, 2009
Introduction
Financial liberalisation has been a recent
important trend in the financial sectors in Latin
America.
 Main idea was to generate a more competitive
banking sector.
 Aizenman (2005) suggests that financial
liberalisation gained from increased savings in the
economy (mainly through foreign capital).
 This increase in savings in the economy was
expected to increment the level of financial
deepening in the economy (particularly the credit
to the private sector).

Financial Deepening
Figure 1
Bank Credit to the Private Sector in terms of GDP,
Average Percentage
1999-2006
160
140
120
100
80
60
40
20
UK
EURO Area***
South East
Asia**
Latin America*
Venezuela
Uruguay
Peru
Paraguay
Costa Rica
Colombia
Chile
Brazil
Argentina
0
Data obtained from IFS (International Financial Statistics) from the IMF (line 32D.ZF).*The Latin American average is calculated using the following countries:
Argentina, Brazil, Chile, Colombia, Costa Rica, Paraguay, Peru, Uruguay and Venezuela. For Venezuela the average is calculated from 1999-2004 due to data
availability. **The South East Asia average is calculated using the following countries: Brunei, Cambodia, Malaysia, Laos, Myanmar, Singapore, Thailand, Hong
Kong, South Korea, Philippines and Vietnam. The average for Brunei and Vietnam was computed for the years 1999-2005 and for Myanmar from 1999-2004
due to data availability. ***The data for the EURO Area was elaborated including all the countries in the European Union.
Figure 2
Financial Deepening in Latin
America
Source: IFS (International Financial Statistics from the IMF)
*The Latin American average is calculated using the following countries: Argentina, Brazil, Chile, Colombia, Costa Rica,
Paraguay, Peru, Uruguay and Venezuela; except for M2/GDP in which Peru and Venezuela were excluded due to data
availability.
Literature Review

Economists have largely disagreed on the role of
financial development and economic growth.

Increased numbers of financial institutions and
financial instruments help reduce information costs in
the economy (Rioja and Valev, 2004).

Many other studies find a positive relationship
between finance and growth (King and Levine (1993a,
b, c), Roubini and Sala-i-Martin (1992), Pagano (1993),
Jayaratne and Strahan (1996), Levine (1997a, 1998),
Arestis and Demetriades (1997), Rajan and Zingales
(1998), Lindh (2000), Levine et al. (2000) among
others.
Literature Review
“Thus, if finance is to explain economic
growth, we need theories that describe
how financial development influences
resource allocation decisions in ways that
foster productivity growth...”
 Levine, 2004: p. 6

Literature Review
Recent studies address how financial development
affects economic growth through greater
productivity (economic).
 Particularly, Levine et al. (2000), Arestis et al.
(2002) and Arestis et al. (2006) argue that
financial development affects economic growth
through productivity growth.
 There are various studies which suggest that
financial development may enhance greater
economic productivity and contribute to a more
efficient allocation of capital.

Motivation
Latin American financial systems are highly bankbased.
 Extensive literature on the relationship between
financial deepening and economic growth.
 No literature focusing on the microeconomic
relationship: financial deepening and banking
productivity and/or vice-versa.
 Financial deepening forms the broad environment
in which banks operate.

Data
Data obtained from Bankscope and the
IFS from the IMF.
 Study includes 9 Latin American countries:
Argentina, Brazil, Chile, Colombia, Costa
Rica, Paraguay, Peru, Uruguay and
Venezuela.
 Observations: 973
 Years: 2000 - 2006

Methodology
GMM Panel Data: endogenous variables (TFP, PCR) exogenous variables: macro
variables
TFPi ,t   i ,t  1TFPi ,t 1   2 PCRi ,t  3 PCRi ,t 1   4 PCRi ,t 2   5CPIi ,t   6TRADEi ,t t
  7GOVi ,t  8 XRATEi ,t  8GDPpercapitai ,t  i   i ,t
TFP = TOTAL FACTOR PRODUCTIVITY
PCR= CREDIT TO THE PRIVATE SECTOR /GDP
CPI = CONSUMER PRICE INDEX
TRADE = SUM OF EXPORTS + IMPORTS / GDP
GOV = TOTAL GOVERNMENT EXPENDITURE / GDP
XRATE = AVERAGE ANNUAL EXCHANGE RATE
GDP per capita = GROSS DOMESTIC PRODUCT per capita
TFP or the Malmquist Index



The Malmquist Productivity Index is a
distance function which measures the degree
of productivity using a multi input and multi
output approach.
Created by the Swedish statistician
Malmquist in 1953, it was first proposed by
Caves, Christensen, and Diewert (1982).
It has since been further developed by Fare
(1998), Fare et al. (1994) among others.
TFP or the Malmquist Index
D (y , x ) D (y , x ) 
t
t 1 t
t 1
M t ,t 1 ( y , y , x , x )   t t t  t 1 t 1 t 1 
D (y , x )
 D (y , x )
t 1
t 1
t 1
t
t 1
t 1
Where M is the Malmquist Productivity Index
D is the distance function made up of inputs and outputs.
A value of M >1 implies an increase of productivity
A value of M=1 means no change in productivity and
M <1 is a reduction in productivity.
The distance function is the DEA input-oriented approach.
x = inputs (interest rate expenses, personnel expenses and other
operating expenses)
y= outputs (loans and other earning assets)
1
2
TFP or the Malmquist Index
D (y , x )  D (y , x ) D (y , x ) 
t t 1 t t 1
M t ,t 1 ( y , y , x , x )  t t t  t 1 t t  t 1 t 1 t 1 
D (y , x ) D (y , x ) D (y , x )
t 1
MI(TFP)
t 1
TEC
t 1
t
t
t
t
t 1
t 1
TC
Where MI represents the Malmquist Index (TFP- Total Factor
Productivity), TEC is the technical efficiency change and TC is the
technological change.
• The technical efficiency change relates to how close firms are
operating in relation to the best practice frontier.
• On the other hand, technical change refers to the shift in the best
practice frontier.
1
2
Empirical Results
Figure 3
TFP, TC and TEC average annual geometric average in Latin America
(%)
TFP, TC and TEC are the geometrical averages for the corresponding years in Latin America.
The Latin American average includes the countries in study: Argentina, Brazil, Chile, Colombia,
Costa Rica, Paraguay, Peru, Uruguay and Venezuela.
Empirical Results
Table 1
GMM dynamic panel data, TFP as the dependent variable
TFP lagged(t-1)
PCR
PCR lagged(t-1)
TFP
(a)
.061
TFP
(c)
-.019
1.56***
1.488***
-.845
.832*
.228
.031***
.267
-2.857*
-.095*
-.039
1.128
-1.58
(0.113)
-0.19
(0.853)
28.31
(0.166)
513
Yes
.043***
-.082
-2.647**
-.058*
-.184
2.511**
-1.87
(0.062)
-0.03
(0.977)
19.48
(0.427)
973
Yes
1.133***
PCR lagged(t-2)
CPI
TRADE
GOV
XRATE
GDP per capita
Cte.
AR(1)
p-value
AR(2)
p-value
Hansen J test
p-value
Observations
Year dummies
TFP
(b)
-.059
.032***
-.34
-2.618***
-.036***
-.059
1.481**
-2.04
(0.042)
1.22
(0.224)
25.89
(0.359)
973
Yes
Empirical Results
Table 2
GMM dynamic panel data, PCR as the dependent variable
PCR
PCR
PCR
(d)
(e)
(f)
PCR lagged(t-1)
.463**
.986***
.835***
TFP
.016***
.008*
TFP lagged(t-1)
-.011
-.015
TFP lagged(t-2)
-.024**
-.005
CPI
-.005***
-.004***
-.006***
TRADE
.286***
.22***
.223***
GOV
.683***
.21
.301
XRATE
-.001
-.009
-.006
GDP per capita
.095***
.11***
.117***
Cte.
-.8***
-.902***
-.951***
AR(1)
-2.6
-1.44
-1.05
p-value
(0.009)
(0.150)
(0.293)
AR(2)
-2.02
-1.15
-1.02
p-value
(0.043)
(0.251)
(0.310)
Hansen J test
44.96
36.91
22.57
p-value
(0.006)
(0.024)
(0.257)
Observations
973
513
513
Year dummies
Yes
Yes
Yes
Results
The main findings indicate there is a strong
positive relationship between TFP and PCR and
vice-versa (model f).
 CPI is positive and significant when explaining
banking sector productivity but negative and
significant when explaining financial deepening.
 TRADE, is positive and significant when explaining
financial deepening.
 Government expenditure is negative and
significant with TFP, implying that greater
government expenditure decreases banking
productivity.

Other results
XRATE variable is negatively and
significantly related TFP.
 GDP per capita is positively and
significantly related with PCR.

Conclusions 1/2
The above results are in general supportive of
a positive relationship between financial
deepening and banking productivity.
 This finding suggests that a channel through
which the beneficial effects of financial
deepening find their way through the
economy is the banking system.

Conclusions 2/2
Evidence of reverse causality between
financial deepening and banking productivity.
 The finding of reverse causality may indicate
the existence of a virtuous cycle between
financial deepening and banking productivity
where one facilitates the other.

Policy Implications

Policy oriented measures in the region should
take in consideration the positive causality
between financial deepening and banking
productivity and try to increase the level of
credit to the private sector as a stimulant of
economic growth.