The Panzer-Rosse Test for Assessing Completion in Banking:

Up on the competition measurement, literature can be splitted into analysis that will take structural approach that in turn will take non structural-approach. The structural approach is based up on the SCP (structure-conduct-performance) model. The SCP theory considered a casual relationship going on from the market structure to the behavior of firms pricing.

It consists of two theories. They are:

  1. Influence conduct is known as structure.
  2. Conduct is supposed to influence performance.

This implies the banking power can be produced by concentrating the banking industry that consent to lower deposit rates in banks and lending rates are increased and this can gain monopolistic profits. 

The two techniques which are utilized broadly which follows non-structural approach to measure empirically the competitive behavior degree in the market, termed contestability are been developed by Lau (1982), Breshanan (1982), Rosse and Panzar and (1987). The Breshanan model uses the idea of the firms which are profitably running in stable and a general market equilibrium model. The firms in profit will choose the quantity and price such that to maintain marginal revenue to be equal to the marginal cost.

This will perfectly meets the demand price in the competition or with the marginal revenue of the industry under collusion. The technique suggested by Lau (1982) and Breshanan (1982) need to estimate the model with the simultaneous equation based up on the industry data which is an aggregate, where the degree of market power is represented by a parameter is included. In Shaffer (1993), Haaf (2002), Alexander (1988) and Bikker investigation, test is done in this technique of Empirical implementation. 

In the Panzar-Rosse (PR) the future possibilities are: in this model, this investigates the relationship among change in input price factor and the revenue acquired by a particular bank. The Panzer-Rosse model will place on the suggestion that banks make use of various pricing strategies for changing the input cost based on the market structure in where the model is functioned.  To approach this model there will be benefit of utilizing the data of specific-bank and therefore it captures the single characteristics of various banks.

This static analysis which is comparative needs the evaluation of a revenue function in a reduced form. For a particular firm, the total revenue equilibrium is given by equilibrium quantity times the equilibrium price. Together equally based upon demand, cost and conduct; consequently all the demand and cost must be included in the revenue functions, through a particular concentration towards the factor prices. In the time period “t” and for the “ith” term, the revenue equation in the reduced form is given by specification: 

                   Rit  =  f(wit, zit, yit, εt )                                                      


                   wit   →  vector of factor price,

                   zit    →   the variables that shift the cost function,

                   yit    →   the variable that shift the demand function,

                   εt     →    is the error term.

In this case, the derivative is ∂Rit/∂Witk which is of total revenue with respect to the price of kth input, the PR and H-test (statistic) in then represented as: 

H = ∑[(∂ Rit/∂ Witk) * (Witk/ Rit)]                                                             

In the reduced form of revenue, H is the sum of elasticities regarding to all the factor prices. Otherwise, in banks percentage change is measured by the statistic. In banks equilibrium revenue is produced by 1 percent change in all the input prices of banks. Consequently, factor prices and particular data from firm is essential for H statistic for computation. In advance cost information is not required, even though the introduction of variables affecting demand or cost is required. 

When related to the market study, PR test has clear-cut understanding. In the equilibrium revenue, percentage variation is represented by H and this result in increase of unit percentage in the price by the firm which uses all the factors. In an collusive environment, PR points out assuming maximum profit, when the input prices are increased then the marginal cost will increase and equilibrium output and revenues are decreased.  For monopoly, a homogeneous conjectural difference oligopoly or a perfectly colluding oligopoly, H<0.  Shaffer (1982) has proven that the H =1 for operating a natural monopolist in a market which is perfectly contestable and also in a firm where the maximum sales subject to break even constraints.

Under a competition which is perfect, marginal revenue and the marginal cost will increase with increase in input prices by the equal amount and therefore H=1. Monopolistic competition is considered when the value of H is between 1 and 0, in revenues proportional increase becomes less due to increase in input prices. In a monopolistic competitive market which is symmetric, 0<H<1. It is worth highlighting that not only H sign is important, but magnitude is also equally important. As explained in Haff and Bikker (2000), the approach includes four conditions. 

They are: 

  1. At the long run equilibrium, banks are operated.
  2. By the participant’s action of the other market, performance of the bank is influenced.
  3. The cost structure is homogeneous and
  4. Compare to the unity, price elasticity of demand is greater. 

From the standpoint of econometric, the rejection of H≤0 rules which is out of monopoly model, rejection of H≤1 excluded by the three models and rejection of H≤0 and H=1 hypothesis but should not consider the H≤1 then it mean that only competition model which is of monopolistic is consistent with the data. The relevant summary of cross-country and country specific studies is illustrated in below table: 

During the year 1996-2004, degree of competition is evaluated in the Indian banking industry utilizing data on the scheduled commercial banks and PR approach is employed. 

Author(s) Countries Years Outcome*
Developed Economies      
Shaffer (1982) New York(United


1979 MC
Nathan and Neave (1989) Canada 1982-84 1982: PC; 1983-84:


Molyneux et al. (1994) Germany, United



1986-89 Italy: M;UK,France,

Spain: MC;Germany:

MC (except 1987: PC)

Corrorese (1998) Italy 1988-96 MC (except 1992 and

1994: PC)

Bikker and Groenveld (2000) 15 EU countries 1989-96 MC (exceptBelgium

andGreece: PC)

De Bandt and Davis (2000) France,Germany,Italy,

and United States

1992-96 Large banks: MC;

small banks: M (except

Italy: MC)

Bikker and Haaf (2002) 23 countries Various ranges

(up to 1998)

MC (for overall


Emerging Economies      
Gelos and Roldos (2002) Argentina,Brazil,Chile,


Hungary,Poland, and


1994-99 MC
Philippatos and Yildrim (2002) 15 Central and Eastern

European countries

1993-2000 MC (except large

banks: PC)

Belaisch (2003) Brazil 1997-2000 MC (except foreign


Levy-Yeyati and Micco (2003) Argentina,Brazil,Chile,

Colombia,Costa Rica, El

Salvador, andPeru

1993-2002 MC

* M — monopoly; MC — monopolistic competition; PC — perfect competition

Table: Application of the Panzar-Rosse Methodology to Banking Studies