Market entry, privatisation and bank performance in transition

Market entry, privatisation and bank
performance in transition
Steven Fries, Damien Neven,
Paul Seabright and Anita Taci
EIASM Workshop on Financial Development
Prague, 26 May 2006
Introduction

Many policies used to change socialist banking
systems into market-oriented ones, including
– Restructuring and privatisation of state banks
– Entry of new private banks, including foreign

Gauge impact of reforms on banking development
by examining revenue and costs
– Foreign banks are low marginal cost entrants
– Privatised banks attract greater demand
– Differences among private banks diminish over time
– State banks under perform on both demand and costs
Approach of the paper

Develop unique model of monopolistic competition
in bank lending and deposit taking to analyse
revenues
– Look for effects of market entry and ownership on
deposit and loan margins

Use standard trans-log specification for costs
– Estimate marginal costs by ownership type

Compare margins and marginal costs for mark-ups
– Indicates ability to attract demand for loans and deposits

Allow parameters of revenues and costs to vary
over time
Existing literature

Banking market structures in transition economies
(TEs)
– Gelos and Roldós (2002), Yildirim and Philippatos
(2002a) and Drakos and Konstantinou (2003)
– Evidence of monopolistically competitive markets
– Increasing competition over time

Cost / profit efficiency of banks in TEs
– Grigorian and Manole (2002), Yildirim and Philippatos
(2002b), Bonin, Hasan and Wachtel (2005) and Fries
and Taci (2005)
– Greater efficiency of foreign banks
Existing literature (cont’d)

Determinants of bank net interest margins in
TEs
– Drakos (2003)
– State-owned banks have relatively low margins

Theoretical models of imperfect competition
and institutional development in TEs
– Hainz (2003a, 2003b)
– Studies show that the market power of banks
decreases as the quality of institutions improves
The starting point: bank profit function

Accounting identity
i  rilLi - ridDi + rinNi - R(Li + Ni - Ei - Di) - C(Di,Ni,Li,Wi)
Loan
margin
Deposit
margin
Non-loan assets
margin
Operating costs
i  (ril - R)Li + (R - rid)Di + (rin - R)Ni + Rei - C(Di,Ni,Li,Wi)
Interbank rate

Equilibrium determination of loan and deposit
margins in monopolistically competitive markets
The empirical specification

Revenue function is therefore
REVijt = Mijtl(•)Likt + Mijtd(•)Dikt + ρNijt + σEijt + ijt


Margins are a function of number of
competitors, market shares and bank
ownership
Combine loans and deposits into a single
scale variable because of their co-linearity
REVijt = Mijtl+d(•)(Likt + Dikt) + ρNijt + σEijt + εijt
Empirical specification (cont’d)

Trans-log specification for costs
lnCijt = αo + Σst βs lnQs,ijt + Σmn Xm lnWm,ijt

Higher order terms omitted because they
were insignificant in preliminary estimations

Allow estimated constants to vary by bank
ownership and over time

Identifies differences in average costs across
bank types
Dataset

Bankscope database for accounting data
– 477 banks in 15 transition economies
– Sample period covers 1995 – 2004
– Unbalanced panel, 115 banks for entire sample
period

Unique coverage of bank ownership
– Time varying measures of ownership based on
EBRD staff research
– Newly established, privatised and state-owned
– Domestic and foreign
Basic data on bank revenues

Revenues to total assets in per cent
9.0
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
New domestic
1995-98
New foreign
Privatised
domestic
1999-2001
Privatised
foreign
State owned
2002-04
Basic data on bank costs

Operating costs to total assets in per cent
8.0
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
New domestic
1995-98
New foreign
Privatised
domestic
1999-2001
Privatised
foreign
State owned
2002-04
Results – Revenue equation
(revenue adjusted for inflation)
1995 - 1998
1999 - 2001
Loans
0.0320***
0.0194
-0.0350***
Deposits
0.0007
0.0609***
0.0475***
Newly established, domestic
-0.0096
0.0095***
0.0105***
Newly established, foreign
-0.0184***
-0.0094***
0.0081***
Privatised, domestic
0.0199***
0.0275***
0.0282***
Privatised, foreign
0.0052**
0.0051**
0.0038***
Number of banks per million of population
-0.0005
-0.0012**
0.0002
Share of loan and deposit market
0.0015***
0.0289***
0.0063*
IAS dummy variable
-0.0101
-0.0289***
-0.0089**
Non-Loan financial assets
0.0125
0.0391***
0.0194***
Equity
0.2408***
0.2749***
0.2906***
2002 - 2004
Explanatory variables
Interacted with loans plus deposits
Results – Cost equation
1995 - 1998
1999 - 2001
2002 - 2004
Ln (loans + deposits)
0.7995***
0.6101***
0.6332***
Ln (loans + deposits ^ 2
0.0021
0.0360***
0.0410**
Ln (operating costs / total assets)
0.3681***
1.3556***
0.9657***
Ln (operating costs / total assets) ^ 2
0.1715***
0.2036***
0.0370*
Ln (non-loan financial assets)
0.0331
0.1075***
0.0529***
Equity to total assets ratio
-0.0019
-0.1170
0.3072**
Newly established, domestic
-0.2649***
-0.0833*
-0.0298
Newly established, foreign
-0.3767***
-0.0051
0.0303
Privatised, domestic
-0.2767***
-0.0340
0.0083
Privatised, foreign
-0.0209
0.0263
-0.0377
Marginal costs and mark-ups

Foreign bank marginal costs below that of state bank
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
-2.5
New domestic
1995-98
New foreign
Privatised, domestic Privatised, foreign
1999-2001
2002-04
Marginal costs and mark-ups

Mark-ups above state bank benchmark
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
New domestic
1995-98
New foreign
Privatised, domestic
1999-2001
Privatised
2002-04
Conclusion


Foreign banks are low marginal cost entrants
Privatised banks attract greater demand
– Evidence that newly established banks attract more demand
with a lag

New domestic banks had weaker performance than
other private banks
– While differences diminish over time, this casts doubts on
liberal entry policies for domestic banks


State banks are the weakest performers throughout
Policy implications
– Openness to foreign entry (costs) and privatisation of state
banks to both domestic and foreign owners (demand)