Topic 7 – Competitive Issues in Banking. Competitive Issues in Banking Outline Output Measurement Productivity Measurement Economies of Scale and.

Topic 7 – Competitive Issues in Banking

Competitive Issues in Banking

OutlineOutput MeasurementProductivity MeasurementEconomies of Scale and ScopeS-C-P vs Efficient Market Hypothesis

Measurement of Bank Output

Output produced by financial institutions is difficult to determine due to problems identifying quantities of measurement.

Difficult to account for quality in a banking service. i.e. an ATM may improve the quality of payment services

as well reduce a customer’s transaction cost – difficult to measure.

Two common approaches: The production approach The intermediation approach

The Production Approach

Banks are treated as firms for measuring output. Banks use capital and labour to produce deposits

and loan accounts and output is measured as the number of accounts/number of transactions per account.

Problems:How to weight each of the bank’s services in the

computation of output?Unlikely to find comparable quality data from

different banks – comparison is difficult.

The Intermediation Approach

This approach recognises intermediation as the core activity.Output is measured by the value of loans

and investments.Cost is measured as operating cost plus

deposits.

Most bank productivity studies used intermediation approach.Fewer data problemsUniversal Banking?

Productivity Measures

Two types of productivity measures are used.PartialTotal

Partial measures are based on financial ratios (coursework).

Total measures take into account the multiple nature of outputs and inputs in banking. One such measure is total factor productivity (TFP).

Total Factor Productivity

TFP employs a single productivity ratio by using multiple inputs and outputs.

The widely used TFP measure is data envelopment analysis (DEA).DEA compares the observable outputs (yjp) and inputs

(xip) of several banks.

It then identifies the relatively most efficient bank (and therefore the relatively less efficient banks.)

To do this, it maximises the following:

p = ipijpj Xv/Yu subject to p 1 for all p and weights vi,uj >0, p

represents several banks

Total Factor Productivity

The model is run repetitively with each bank appearing in the objective function once to derive individual efficiency rating.

The decision about efficiency or inefficiency is based on the following:E=1 ‘efficient’E<1 relatively inefficient

However, ‘efficient’ does not mean efficient in absolute terms but efficient compared to other banks in the data set.

Economies of Scale

Economies of scale : Long-run concept where all the factor inputs can be varied.

A bank exhibits scale economies if an increase in factor inputs yields a greater than proportionate increase in output the bank is operating on the falling portion

of their average cost curve

Economies of Scale

Consider a bank with three factors of inputs capital (deposits), labour (bank employees) and property (branch network). The bank produces one output (loans).

Economies of scale are said to exist if, as a result of doubling each of three inputs, the bank is able to more than double its loan portfolio.

In practice, even this simple example is problematic, b/c all factor inputs are not variable to the same extent (e.g. it is difficult to double deposits).

Economies of Scale

If inputs are doubled and loans are doubled, then risk portfolio is bound to change (shareholder wealth).

All this implies that it is difficult to apply the concept of economies of scale in financial sector.

Hence it calls into question the underlying concept of mergers and acquisition – the hope for economies of scale and scope.

Economies of Scope

Economies of scope exist if the joint production cost of producing two or more outputs is lower than if the products are produced separately.

For example a bank offers three services to customers (deposits, loans and payment services).

If the bank can supply these services more cheaply through a joint production process than producing and supplying them independently, it is said to be enjoying economies of scope.

Strategic Inference

Evidence of economies of scale will mean large banks have a cost advantage over small banks.

If synergy is present, universal banks will be more profitable than traditional banks.

Structure-Conduct-Performance

Model states that a change in the market structure or concentration of the banking sector affects the way banks behave and perform.

Market structure is determined by the interaction of cost (supply) and demand.

Structure-Conduct-Performance

Conduct (decisions on prices, levels of advertising etc.) is a function of number of sellers and buyers, barriers to entry and cost structure.

The outcome is market performance (profitability).

In simple words less number of firms (structure: high concentration), higher prices (conduct) and higher profit (performance)

Efficient Market Hypothesis

Some banks earn supernormal profit b/c they are more efficient than others.

The relative efficiency model predicts the same positive profit-concentration relationship as the SCP model.

However, the positive relationship is not explained by collusive and abusive behaviour (as with SCP), but greater efficiency (and hence concentration).

Strategic Inference

According to SCP, concentration is exogenous, resulting in higher prices for consumers and higher bank profitability.

In the relative efficiency model, exogenous bank specific efficiencies results in more concentrated markets.

If the relative efficiency model is found to hold, it would suggest markets are best left alone.

18

Empirical Testing of Competitive Conditions in

Banking

19

Outline

DEA Studies on Efficiency Scale and Scope Economies

USA StudiesUK StudiesEU Studies

S-C-P TestingContestability Testing

20

Efficiency (USA)

Rangan et.al. (1990) used DEA with data from 215 US banks using the intermediation approach of output. They break down the efficiency score into

technical inefficiency (wasted resources) and scale inefficiency (non-constant return to scale).

The study showed the efficiency score of 0.7 implying 30% wastage, all due to technical inefficiency.

21

Efficiency (USA)

Humphrey (1992) measured productivity and scale economies using flow and stock measures of banking output. He obtained measures of total factor productivity by using 202 US banks. Banking productivity had been flatOnly moderate increase in TFP

22

Efficiency (UK)

Field (1990) applied DEA to a cross section of 71 UK banks. The results were that 80% were found to be inefficient due to scale inefficiencies.

Drake et.al. (1991) used DEA to analyse building societies in 1988 after deregulation in 1987. 37% of these societies showed increase in their overall efficiency post-deregulation.

23

Scale and Scope Economies (USA)

Shaffer and David (1991) tested for the presence of economies of scale for very large US multinational banks.

Traditional production cost function used with and without hedonic terms (qualitative factors). In the absence of hedonic terms they found evidence of positive scale economies.

In the equation with the hedonic terms included, economies of scale were reduced from the level without hedonic terms.

24

Scale Economies (USA)

Humphrey (1992) obtained estimates of scale economies for US banks of a multitude of sizes.the intermediation measure of output was initially used.

His study results suggest diseconomies of scale for all sizes of bank.

However, when alternative measures of output (production approach) were used, Humphrey found significant economies of scale for small banks, constant costs for medium sized banks and scale diseconomies in large banks.

25

Scale Economies (USA)

This study however raised an important question.Which measure of output should be used? The author suggested that stock measure was

more accurate than flow measure of output.

The study overall results suggest there are slight economies of scale for small banks, but slight diseconomies for large US banks.

26

Scale Economies (UK)

Hardwick (1990) tested for scale and scope economies using UK building society data and a marginal cost approach.

Overall economies of scale were found - except for very large building societies, where strong diseconomies were present.

Drake (1992) found no evidence to support this.

27

Scale Economies (EU)

Altunbas and Molyneux (1993) examined the cost structure in four European countries (France, Germany, Italy and Spain). Italy – economies at all levels of output. Spain – economies only for smallest banks. France – economies at all levels of output. Germany - diseconomies of scale at all levels

of output.

28

Scope Economies (USA)

Gilligan and Smirlock (1984) study supported the hypothesis of economies of scope.

Mester (1987) concluded there was no strong evidence to either support or refute the presence of economies of scope.

Lawrence (1989) reached similar conclusions

Hunter et. al.(1990) found no evidence of cost complementaries.

29

Scope Economies (UK)

Hardwick did not find evidence either for or against economies of scope for large building societies.

However, he found significant diseconomies of scope for building societies with assets worth £1.5 billion or less.

The results of the study showed no case for diversification of most building societies into broader banking market.

Drake (1992) largely agreed, but found that building societies with assets in the range of £500m-5 billion displayed diseconomies of scope.

30

Scope Economies (EU)

Italy - diseconomies for all sizes.Spain - scope economies evident for banks with

assets of < $1.5 billion. France - middle sized banks showed economies

of scope. Germany - largest banks showed scope

economies, smaller banks showed diseconomies.

31

S-C-P vs Relative Efficiency (USA)

Berger and Hannan (1989) conducted direct tests of the SCP and relative efficiency models. The concentration variables were found to be

negative, implying more concentrated the market the lower the rate paid on deposits (consistent with SCP).

Jackson (1992) that ‘price’ is a non-linear U shape. This supports the relative efficiency model.

32

S-C-P vs Relative Efficiency (EU)

Molyneux and Forbes (1993) tested the SCP vs relative efficiency hypothesis using European data.

The authors agreed with Berger and HannanThey concluded that the SCP hypothesis is

supported by this European sample.

33

Empirical Studies on Contestable Banking Markets

Shaffer (1982) (Rosse-Panzer Statistics-RPS), US banking Banks in the sample behave neither as

monopolists nor as perfect competitors in the long run.

Nathan and Neave (1989), Canadian banking Authors derived a positive but significantly

different from both zero and unity RPS confirming the absence of monopoly power among Canadian banks and trust companies.

consistent with a banking structure exhibiting some features of monopolistic, contestable competition.

34

Empirical Studies on Contestable Banking Markets

Molyneux, Lloyd, Williams and Thornton (1994)- German, UK, French, Italian and Spanish banking marketsThe authors found the RPS for Germany, the UK,

France and Spain, to be positive and significantly different from zero and unity.

Commercial bank revenues behaved as if they were earned under monopolistic competition.

For Italy however, monopoly.