Consolidation and Competition in Emerging Market: An Empirical Trial for Malayan Banking Industry

Abstraction

This paper investigates the consequence of bank consolidation on market construction and competition in Malayan banking industry during the old ages of 1998 to 2005. The survey evaluates the grade of competition utilizing H-statistic proposed by Panzar and Rosse ( 1987 ) . The estimated H-statistics are positive runing from 0.53 to 0.81 and the Wald trial for the market construction for monopoly or perfect competition is rejected. The consequences imply that the fiscal establishments in Malaysia earned their gross in the status of monopolistic competition with the traditional interest-based market is significantly less competitory than the overall market. The grounds is nevertheless deficient to demo that there is an addition in competition due to a alteration in the market construction. Therefore, the findings suggest that extra competition policy is needed to guarantee competition in the Malayan banking market in position of farther consolidation in the banking sector.

JEL Classification: G21 ; D24, L1

Keywords: Competition ; Panzar-Rosse theoretical account ; Malaysia ; Market construction

1 Introduction

Academicians and policy shapers seem to accept the position that fiscal establishments play a important function in the effectual operation of modern economic systems. One would anticipate that the higher the grade of competition, the higher its efficiency in footings of apportioning financess and in general operating as an intermediary between depositors and borrowers. There has been considerable concern about how on-going consolidation in fiscal systems around the universe will impact competition. Indeed, much of the recent public argument seems to presume that perfect competition in banking is ideal. For much of the last century, this has non ever been the instance where policy-makers focused on stableness. Therefore, banking competition by and large can be characterized as monopolistic competition that Bankss do non offer wholly homogenous merchandises ( Allen and Gale, 2001 ) .

It is observed that the banking industry in developing economic systems has been sing a cardinal alteration in its market construction since the Asian fiscal crisis. One of the chief implicit in factors behind such a moving ridge of fiscal consolidation may be the ferocious competition among Bankss triggered by fiscal deregulating and fiscal globalisation, ensuing in a crisp addition in fiscal hazard direction. Significant advancement in information engineering necessitates that Bankss seek scale economic systems and hazard variegation to derive comparative advantage over rival Bankss. While the fiscal consolidation in industrial states has been evolved by market forces, the fiscal consolidation of emerging economic systems has been forced by government-led structural accommodation policy for the banking industry in order to get the better of future banking crisis.

This survey is motivated by several factors. First, figure of surveies have examined the effects of consolidation on competition and market construction of banking industry ; nevertheless, these surveies have concentrated on developed markets ( Molyneux et al. , 1994 ; Hempell, 2002 ; Bikker and Groeneveld, 2000 ) . Our survey contributes to the literature by comparing the competitory market construction of Malayan banking industry pre- and post-consolidation periods. Second, a famine of bank consolidation and competition surveies utilizes the non-structural theoretical account attack suggested by Panzar and Rosse ( 1987 ) , and most of them have been concentrated on the developed states with really few surveies on emerging economic systems. Therefore, we extend the literature by using a Panzar and Rosse ( 1987 ) theoretical account to through empirical observation gauge the competitory behavior in Malaysia, one of the quickly turning banking sectors in emerging economic systems. Third, the overall consequence of consolidation and greater concentration can be equivocal. It can either be that the greater bank size can take to an addition in monopolistic power that causes a rise in loaning rates or it can besides be that the greater bank size can take to greater economic systems of graduated table ( Bankss are more efficient ) and causes a lessening in loaning rates. Hence, this is an of import empirical issue for policy shapers, regulators and economic experts.

The paper is structured as follows: Section 2 provides the construction and tendencies of market concentration in the Malayan banking industry in 1997-2005. Section 3 elaborates on theoretical theoretical account of Panzar-Rosse ( 1987 ) and empirical groundss generated from application of their theoretical account. Section 4 presents the informations and theoretical accounts used in the empirical analysis. Section 5 provides the appraisal and consequences of the H-statistic for the pre- and post-consolidation periods. Section 6 summarizes the paper with reasoning comments.

2 Malayan banking construction and concentration

At the oncoming of the crisis, the Malaysian banking system is characterized by its high market concentration consisted chiefly of three types of establishments: commercial Bankss ( domestic and foreign Bankss ) , finance companies, and merchant Bankss. Table 1 shows the development of the Malayan banking construction in 1997-2005. Domestic commercial Bankss had the largest portion of the market commanding 71.5 per centum of the market portion in 2005. Foreign commercial Bankss controlled merely 20.8 per centum of banking assets. Foreign Bankss do non do any advancement in commanding market portion because of a calculated authorities policy to develop domestic fiscal sector, under which foreign Bankss have been prohibited to open new subdivisions and no new licence was granted to foreign establishment Bankss.

[ Insert Table 1 ]

Following the deepening of the fiscal crisis that struck the part in 1997/98, the authorities took stronger measurers to advance meeting of banking establishments. A amalgamation plan was initiated in 1998 covering merely the finance companies ab initio. The amalgamation plan was later extended to include all domestic banking establishments in July 1999 and all domestic banking establishments would be restructured with six banking groups would be formed. However, as uncertainness remains on the impact of the amalgamations on the plus quality of the freshly formed group, peculiarly on the ground tackle Bankss, therefore the program was eventually halted at the terminal of September 1999. A new amalgamation program was announced, where all domestic banking establishments must organize their ain amalgamation groups and take their ain leader in each group by the terminal of January 200l. The Malayan banking industry, after its first phase of consolidation that reduced the figure of Bankss to ten anchor Bankss, is poised to come in a 2nd stage of consolidation. Table 2 shows the banking groups in Malaysia after the completion of the 2nd stage of the consolidation procedure.

[ Insert Table 2 ]

The bank consolidation in Malaysia is characterized by the structural accommodation of the banking industry since the currency crisis of 1997 and has caused important alteration in its market construction. The first stage of structural accommodation during the period of 1998 and 1999 has been carried out to brush out non-performing Bankss. Unlike the first stage, the 2nd stage of structural accommodation has been initiated by major large-sized Bankss to obtain competition power ( Bank Negara Malaysia, 2004 ) . As such, each stage of the bank consolidation has been carried out with a contrasting motivation. The bank consolidation in Malaysia has been unusually characterized by horizontal amalgamations among Bankss with overlapping market cleavage. As the market construction in Malaysia has been reorganized chiefly by several taking Bankss or fiscal keeping companies, it is observed that the size distribution of Bankss has been widened. This has led to the aggressively reduced figure of Bankss and increased in the market concentration. Sing that extra bank amalgamations are presently in advancement, it is expected that market concentration in the Malaysia banking industry will be higher in the close hereafter than at present.

The most often applied steps of concentrations are k-bank concentration ratio ( CRk ) and Herfindahl-Herschman Index ( HHI )[ 1 ]. Market concentration ratio is used as a step of market ruling power within an industry or among companies. This paper employ a widely use bank concentration index of the highest two ( CR2 ) , three ( CR3 ) and five ( CR5 ) Bankss ‘ entire assets, entire sedimentations and entire loans as initial steps. CRk is computed as the amount of top kth-tier houses ‘ market portions and summing merely the market portions of the K largest Bankss in the market, it takes the signifier:

( 1 )

CRk is a comparatively strong step because it clearly catches the market construction through market portions of a few dominating houses. This index is based on the thought that the behaviour of a market is dominated by a little figure of big Bankss. The CRk index is really utile to analyze the market influence of a few dominating houses in the market, but it is non so utile in hold oning the general characteristics of market construction On the other manus, the Herfindahl-Herschman Index ( HHI ) is defined as the amount of the squared market portions of all Bankss in the market. HHI takes market portions as weights, and emphasize the importance of big Bankss by delegating them a greater weight than smaller Bankss. The HHI can be computed as follows:

( 2 )

where MS is the market portion of the ith house and N is figure of houses in the market. HHI has the advantage of including information of the distribution of market portion every bit good as the figure of houses which take portion in the industry. Having their ain virtues, both CRk and HHI are widely used as major steps of market concentration. As the market concentration index indicates the possibility of houses ‘ anti-competitive behaviour in an industry or a market, many states ‘ regulative governments utilize the index as one of the blessing standard of corporate consolidation.

To analyze the tendencies of market concentration in the Malayan banking industry we compute the market concentration index based on CRk and HHI over the period 1998-2005, which includes periods of bank consolidation, restructuring every bit good as structural reform of the whole banking system. Table 3 shows the HHI, CR2, CR3, and CR5 of entire assets, entire sedimentations and entire loans, as the indexs of market concentration in Malaysia. The consequences show that for the period 1998-2005, Malayan banking industry is considered a reasonably concentrated market with HHI estimation is in between 1,231-1,644.

[ Insert Table 3 ]

The market concentration ratio in Malayan banking industry shows an increasing tendency in between 1998-2000 after the first stage of consolidation. The HHI estimation based on entire loans increases to 1,644 ( 1,346 in 1998 ) , CR2 estimation additions to 0.43 ( 0.36 in 1998 ) , CR3 estimation additions to 0.58 ( 0.49 in 1998 ) and CR5 estimation additions to 0.71 ( 0.70 in 1998 ) . This increasing tendency is due to Government actions to reconstitute the banking system by including all domestic banking establishments so that six banking groups would be formed. Even though the first stage of consolidation has led to increase in concentration ratios, it was non sufficient to bring on the banking system to go anti-competitive. The restructuring of the banking system was chiefly to take the unhealthy fiscal establishments and to retain sound fiscal establishments in the system.

The 2nd stage of consolidation is between 2001-2004 where HHI, CR2, CR3, and CR5 based on entire loans show a diminishing tendency of 1,231 ( 1,427 in 2001 ) , 0.36 ( 0.40 in 2001 ) , 0.46 ( 0.52 in 2001 ) and 0.62 ( 0.65 in 2001 ) severally. The market concentration ratios based on entire sedimentation besides reveals a similar tendency. These tendencies indicate the failure of the first program of consolidation to unify the domestic commercial Bankss into six anchor Bankss. The concentration ratios, nevertheless, increase aggressively in 2005. The crisp rise in the market concentration ratios reflects the alterations in distribution of market portions originating from the completion of amalgamation and acquisition exercising. This event has led to the enlargement of market portion of the staying fiscal establishments. Hence, the program to originate farther amalgamation and acquisition without opening the market to further competition will increase the market concentration and lead to the fiscal establishments acting in anti-competitive mode.

3 The panzar-rosse attack: theory and grounds

The Panzar-Rosse ( P-R ) attack for proving market power relies on the premiss that Bankss will use different pricing schemes in response to alter in input costs depending on the market construction in which they operate. The advantage of non-structural attack is that it is rooted in theory and in rule distinguishes between market power and efficiency as a beginning of concentration and profitableness. It is really utile because it depends on single bank ‘s informations, for which there are sufficient observations. Therefore, whether a bank operates in a competitory market or exercises some monopoly power can be inferred from the analysis of that bank ‘s entire gross as it responds to altering input monetary values.

The trial is derived from a general banking market theoretical account, which determines equilibrium end product and the equilibrium figure of Bankss by maximising net incomes at both the bank degree and the industry degree. Two critical deductions exist for this equilibrium theoretical account. First, at the bank degree, net income is maximized where fringy gross is equal to fringy cost:

( 3 )

is the fringy gross map, is fringy cost map, is the end product of bank I, is the figure of Bankss, and consists of exogenic variables that shift the bank ‘s gross and cost maps, severally and is a vector of bank I ‘s factor input monetary values.

The H-statistic can be used to place the three major market constructions, viz. , monopoly/perfect collusion, monopolistic competition and perfect competition/contestable market. Decisions about the type of market construction are made based on the size and mark of the H-statistic. That is, both the size and mark are used to distinguish between the different market constructions. The intuition behind the H-statistic remainders entirely on microeconomic theory, which outlines how income or grosss react to alterations in input monetary values for the different market constructions. A sum-up of the testable hypotheses of the different market constructions is presented in Table 4.

[ Insert Table 4 ]

There are several surveies in the literature that have used Panzar-Rosse method to mensurate the degree of competition in the banking sector. A sum-up of old P-R surveies on banking sector is presented in Table 5 for assorted states. Most of the old empirical appraisals of P-R theoretical account is applied to developed states and late, several surveies have employed this methodological analysis to quantitatively measure the grade of competition and market construction of banking industry in developing and passage states. In general, all of these surveies find that banking market is best described as monopolistic competition. However, none of the surveies in banking literature investigated the consolidation, competition and market construction in the Malayan banking sector.

[ Insert Table 5 ]

4 Data and theoretical account preparation

For the empirical analysis, 22 fiscal establishments comprise of domestic commercial Bankss, finance companies and merchant Bankss from 1998-2005 are used which accounted for about 121 observations. Malayan Bankss runing in Islamic banking environment will be excluded in the sample of the analysis due to different regulative model. Unconsolidated bank-specific one-year informations are taken from published balance sheet of one-year studies of each single fiscal establishment. Table 6 nowadayss drumhead statistics of the variables used in the panel theoretical accounts, measured in 1000000s of Malayan Ringgit ( RM ) .

[ Insert Table 6 ]

In the empirical analysis, a methodological pick demands to be made on how to suitably specify a bank ‘s production procedure. The attack to input/output definition used in this survey follows the intermediation attack, which was originally developed by Sealey and Lindley ( 1977 ) and posits that entire loans and securities are end products, whereas sedimentations along with labor and capital are input to the production procedure of Bankss. Specifically, the input variables used in this survey are the mean cost of labor, sedimentations and capital. Following Shaffer ( 1982 ) , Nathan and Neave ( 1989 ) , Molyneux et Al. ( 1994 ) , Perera et Al. ( 2006 ) and Al-Muharrami et Al. ( 2006 ) , the reduced-form entire gross ( TREV ) and entire involvement gross, ( TINT ) equation can be written as:

( 4 )

for t = 1, aˆ¦.T, where T is the figure of periods observed and I = 1, aˆ¦.I, where I is the entire figure of Bankss. Subscripts one and t refer to bank I and at clip t. In this survey, we use two steps as dependent variables. TINT is the ratio of entire involvement gross to entire assets and TREV is the ratio of entire gross to entire assets. The dependent variable is divided by entire assets in order to account for size differences. The first specification in which the dependant variable is merely the involvement portion of entire gross is consistent with the attack that fiscal intermediation constitutes the nucleus concern of fiscal establishments. Although involvement grosss still constitute the chief beginning of Bankss ‘ earning, recent surveies on banking activities report a dramatic increased of other income from fee-based merchandises and off-balance sheet activities in recent twelvemonth given the increased degree of competition in fiscal markets ( Nathan and Neave, 1989 ; De Bandt and Davies, 2000 ) . This can besides be explained partially by the desire of fiscal services houses to spread out their gross bring forthing beginnings without changing their hazard and therefore their capital construction. For this ground, it is appropriate to include entire grosss in the theoretical account in add-on to involvement grosss.

This survey follows old surveies ( Molyneux et al.,1994 ; Perera et al. , 2006 ; Al-Muharrami et al. , 2006 ) presuming that all financess are input in Bankss ‘ production map. Under this intermediation attack, Bankss use three inputs viz. labour, sedimentation and capital. As we do non hold exact informations on input monetary value of labor ( PL ) , we approximate this variable with a placeholder of the ratio of forces disbursals to the entire assets. A placeholder for cost of labor ( PK ) is the ratio of other runing outgo to fixed assets and a placeholder for cost of capital and placeholder for monetary value of financess ( PF ) , is the involvement disbursals over entire sedimentation. All dependent and independent variables are taken in natural logarithm.

The input monetary values are followed by a set of bank-specific factors ( BSF ) that are relevant to the modern banking concern. The BSF includes the ratio of entire equity to entire assets ( EQASST ) to command for differences in the capital construction. The expected coefficient can be positively related to entire gross of well-capitalized Bankss that are involved in riskier operations and portfolios and in the procedure tend to keep more equity, voluntarily or involuntarily. However, harmonizing to Molyneux et Al. ( 1994 ) the coefficient can be expected to be negatively related to the entire gross since lower capital ratios should take to higher bank gross. In add-on, the ratio of loan proviso to entire plus ( RSKASST ) is included to account for a different step of bank-specific hazard ( Al-Muharrami et al. , 2006 ) . The RSKASST is expected to hold a positive impact on gross.

Entire assets ( ASST ) controls for the size of the bank and can be considered as a placeholder for economic systems of graduated table ( De Bandt and Davies, 2000 ; Shaffer, 2002 ) . However, the consequence of this variable is undetermined on the evidences that any positive influence on gross may be offset by larger Bankss capable of diversifying their concern and distributing the hazard of concern. Entire funding to entire assets ( LOANASST ) , a placeholder for grade of intermediation, is expected to be positively related to gross, as higher proportion of plus on the bank ‘s book is expected to bring forth higher gross, since higher commissariats are associated with higher hazard and higher expected return.

An of import characteristic of the H-statistic is that the trials must be undertaken on observations that are in long-term equilibrium. As suggested in the old surveies ( Molyneux et al. , 1994 ; De Bandt and Davies, 2000 ; Bikker and Haaf, 2002 ) , the trial is based on the proposition that in competitory capital markets risk-adjusted rates of return will be equalized across Bankss. The equilibrium trial can be performed by recalculating the Panzar and Rosse ‘s H-statistics replacing the dependent variable of entire gross over entire assets with the natural log of return on assets ( ROA ) . This confirmation is of import for the instances of perfect competition ( H=1 ) and monopolistic competition ( H & gt ; 0 ) , while H & lt ; 0 is a long-term status for monopoly. Therefore, if the sample is non in the long-term equilibrium, H & lt ; 0 no longer establishes monopolistic market conditions, but remains true that H & gt ; 0 disproves monopoly or divinatory fluctuation short-term oligopoly ( Shaffer, 1985 ) . The long-term equilibrium can be estimated in the undermentioned signifier:

( 5 )

It should be noted that following Claessens and Leaven ( 2004 ) , the step of ROA included in equation ( 5 ) is equal to ln ( 1+ROA ) and therefore adjusted for little negative values due to Bankss ‘ losingss in any twelvemonth. The long-term equilibrium trial measures the amount of the snap of return on assets with regard to input monetary values. If the E-statistic, this implies that the banking market is in long-term equilibrium. If rejected the market is assumed non to be in equilibrium ( Claessens and Leaven, 2004 ) . It should be noted nevertheless that the equilibrium does non intend competitory conditions are non allowed to alter during the sample period. It merely implies that alterations in banking are taken bit by bit.

In contrast to most literature, we do non trust on a simple cross sectional appraisal, but carry out panel appraisal with fixed effects, leting the coefficients on the unit input monetary values to alter over clip and to command for unseen heterogeneousness. Throughout the survey White ‘s ( 1980 ) heteroscedasticity consistent statistics is used and preliminary probes of informations reveal that there is no terrible multicollinearity which hinders efficient appraisal of the theoretical accounts. The Wald trial which follows an F-distribution is used in the competition theoretical accounts to prove whether or non the deliberate H-statistics are statistically different from zero and integrity.

The empirical trials are implemented by spliting the period 1998-2005 into two sub-periods and interacting the input monetary value variables with a silent person variable, DumCon for full period that takes the value of one in the 2nd sub-period to capture for structural interruption. If the interaction term outputs important estimations, they indicate structural interruption in the statistical relationship between input monetary values and grosss and we can ascertained in which way the amount of the snap changed. We choose twelvemonth 2001 as the twelvemonth taging the structural interruption due to existent consolidation procedure significantly took topographic point in the banking sector even though the initial proclamation of consolidation was made in 1999.

5 Empirical consequences

In order to prove competition and equilibrium of banking industry in the long-term, equation ( 4 ) is estimated for full sample period and two sub-periods ( 1998-2001 and 2001-2005 ) . The appraisal consequences for the competitory place utilizing the entire gross ( TREV ) as the dependant variable are reported in Table 7. The consequences for the full sample indicate a structural interruption for the consolidation period as predicted. The dummy variables for the period are important at the conventional degrees. Therefore, we re-estimate and discuss in more inside informations for the two sub-periods instead than the full sample period.

The period 2002-2005 shows a comparatively higher grade of competition than in 1998-2001 which indicate that single bank have some ability to post higher net income due to their alone characteristics and market scheme. The estimated arrested development equations for panel theoretical account and fixed consequence theoretical account explain in the scope of 79 % – 96 % of the variableness in the entire gross ( TREV ) equation. The consequences for the period 1998-2001 better explained the fluctuation of the entire gross compared to the appraisal consequences for the period 2002-2005. All diagnostic trials confirm the good tantrum of the theoretical accounts.

[ Insert Table 7 ]

The estimation of H-statistics in both pooled OLS and fixed consequence equations for full sample are positive runing between 0.591-0.724 for TREV equation. The consequences of H-statistics for the period 2001-2005 varies from 0.728-0.813 while the consequences for the H-statistics for the period 1998-2001 varies from 0.599-0.654 which are significantly lower than the recent periods. This is consistent with surveies on other developing states ( Al-Muharrami et al. , 2006 ; Perera et al. , 2006 ) that find H-statistics between zero and one ( i.e. monopolistic competition ) . Therefore, the consequences suggest that the Malayan banking sector earned their gross in market status of monopolistic competition and any signifier of divinatory fluctuation oligopoly and monopoly can be clearly rejected during the sample periods.

In the appraisal consequences where TREV is used as the dependant variable, all the banking cost elements such as unit monetary value of labor ( PL ) and unit monetary value of financess ( PF ) have the positive marks, connoting the increased factor costs taking to the higher gross while the unit monetary value of capital ( PK ) has a negative mark in both pooled OLS and fixed consequence theoretical account. However, merely two of the variables, unit monetary value of labor ( PL ) and unit monetary value of financess ( PF ) are statistically important at the conventional degree. The major part to the H-statistic chiefly comes from unit monetary value of financess ( PF ) and followed by unit monetary value of labor ( PL ) which is barely surprising given the fact that support and labor cost are the chief factor in the production map of Bankss. The positive mark of unit monetary value of labor ( PL ) suggests that force costs are every bit of import as operating expense costs which are comparatively high in the banking industry. Nevertheless, in both specifications, the consequence of the monetary value of capital ( PK ) on the overall snap appears to be minimum and statistically undistinguished compared to other input monetary values. Our consequences are consistent with other surveies that find that the mark of the coefficient on the unit monetary value of capital varies and, in most instances, its impact is negligible on the factor monetary value snap.

The coefficient of the EQASST is positive in the period 1998-2001 and negative in the period 2002-2005 but statistically undistinguished. The coefficient of the variable depiction hazard leaning ( RSKASST ) has a positive consequence on income and statistically important. The coefficient of the ASST variable is negative and statistically important in both OLS and fixed consequence theoretical account except for OLS theoretical account in the period 2001-2005. The consequences suggests that size-induced differences between Bankss may take to lower entire gross per unit of assets and that larger Bankss seem to be less efficient compared to smaller Bankss. This may connote that as a whole the banking market in Malaysia faces diseconomies of graduated table. Finally, the consequences show that the ratio of loans to entire assets ( LOANASST ) ever has the expected positive mark and is important in both specifications. The positive coefficient reflects that the higher fraction of loans on the entire assets ‘ composing envisages greater involvement income and entire gross.

The appraisal consequences for the competitory place utilizing the entire involvement gross ( TINT ) as the dependant for full sample and two sub-periods are reported in Table 8. Similar to the appraisal for TREV, the full sample period appraisal indicates a structural interruption in 2001. All trials confirm the good tantrum of the theoretical accounts. The estimated arrested development equations for panel theoretical account and fixed consequence theoretical account explain in the scope of 90 % – 97 % of the variableness in the entire involvement gross ( TINT ) equation. The consequences for the period 1998-2001 are better explained the fluctuation of the entire gross compared to the appraisal consequences from the period 2002-2005. H-statistics in both pooled OLS and fixed consequence equations for full sample are positive runing between 0.532 -0.583. The consequences of H-statistics for the period 2002-2005 runing from 0.569-0.631 while the H-statistics for the period 1998-2001 runing from 0.531-0.576. With respect to the grade of competition, the H-statistics when utilizing TINT as the dependant variable by and large produces lower estimation than TREV for both OLS and fixed consequence. This indicates that traditional interest-based banking is less competitory than the overall market for both periods under probe. This may suggests that the turning involvement by bank in the fee- and commission-based merchandise market section.

[ Insert Table 8 ]

Similar to the appraisal consequences for TREV, the unit monetary value of capital ( PK ) , unit monetary value of labor ( PL ) and unit monetary value of financess ( PF ) for the theoretical account TINT have the positive marks ( except for the unit monetary value of capital ( PK ) in the period of 2002-2005 ) , connoting that addition in factor costs taking to the higher gross. All the variables of banking cost, PL, PF and PK are statistically important at the conventional degree. The major part to the H-statistic chiefly comes from unit monetary value of financess ( PF ) and followed by unit monetary value of labor ( PL ) which is similar to findings for TREV equation given the fact that support and labor cost is the chief factor in the production map of involvement gross. The positive mark of unit monetary value of labor ( PL ) suggests that force costs are every bit of import as operating expense costs which are comparatively high in the banking industry whereas the negative mark of unit monetary value of capital ( PK ) may indicates that penchants of bank client to cover with labor intensive Bankss. However, in both specifications, the consequence of the monetary value of capital on the overall snap appears to be minimum compared to other input monetary values. The consequences are consistent with other surveies that its impact is negligible on the factor monetary value snap.

The coefficient of the ASST variable is negative and statistically important in both panel OLS and fixed consequence theoretical account in both periods which suggests that size-induced differences between Bankss may take to lower entire gross per unit of assets and that larger Bankss seem to be less efficient compared to smaller Bankss. This besides suggests that as a whole the banking market in Malaysia faces diseconomies of graduated table with regard to bring forthing interest-based gross. The consequences show that the ratio of loans to entire assets ( LOANASST ) ever has the expected positive mark and is important in both specifications. This implies that a higher fraction of loans on the entire assets ‘ composing envisages greater involvement income and entire gross. The coefficient of the variable depiction hazard leaning ( RSKASST ) proposing that Bankss with a higher degree of commissariats indicates a more hazardous loan portfolio and accordingly a higher degree of counterbalancing return and therefore has a positive consequence on income but statistically undistinguished. Finally, the coefficient of the EQASST is positive and statistically undistinguished in the period 1998-2001 whereas it is negative and statistically important for the period 2002-2005. The consequences for TREV and TINT for both periods show the hardiness of P-Z appraisal.

In order for the above trial consequences to be valid, the banking industry should be in the long tally equilibrium during these periods. The ground for this is that if the market is in equilibrium, a dependant variable will hold no correlativity with monetary values of input factor. The equilibrium place in the banking industry is assessed by gauging the equation with ROA as a dependant variable which is presented in the Table 9. The Wald trial can non reject the void hypothesis H=0 for both OLS and the fixed consequence theoretical account taking to the decision that the banking industry is in the long-term equilibrium in both periods 1998-2001 and 2002-2005. Although banking consolidation is still come oning, it is possible that the market is in equilibrium in reaction to the establishment at different times.

[ Insert Table 9 ]

6 Decisions

The survey examines the competitory status of Malayan fiscal sector for the period of 1998-2005 and with two sub-periods 1998-2001 and 2002-2005. These periods correspond to a period characterized by significant reform to restructure and consolidate the banking into a market-driven based economic system, and to farther liberalize and deregulate sufficiently the systems in order to incorporate economically with the international fiscal market. The footing for the rating of competitory conditions is the extant of oligopoly theory in the new industrial organisation literature, specifically, the competition theoretical account developed by Panzar and Rosse ( 1987 ) .

The findings of the market competition, the estimated values of H-statistics for both sample periods, are positive runing from 0.599-0.813 for TREV equation and 0.531-0.631 for TINT equation. This is consistent with other surveies on developing states that find H-statistics between zero and one ( Al-Muharrami et al. , 2006 ; Perera et al. , 2006 ) . The Wald trial rejects the hypothesis for the market construction of monopoly or perfect competition for both periods of probe. Although the consequences pointed to the monopolistic competition in the banking sector, it does non demo any indicant of the alteration in the market lead to increase in market competition in the recent old ages. Furthermore, the grounds suggests that there is less competition in the traditional interest-based market compared to the overall market which includes other runing income from fee- and commission-based merchandise market section.

Based on the findings on the Malayan banking sector, the consequences of the concentration attack and the market construction suggest that the competitory behavior of Bankss is non needfully related to the figure of Bankss in a market or to their concentration because other factors are besides at work. The determination of this paper, nevertheless necessitate to be interpreted carefully given the full market liberalisation procedure of the banking system is at the initial phase and is an on-going procedure. Recent research has highlighted, nevertheless, that the relationships between competition and banking system public presentation are more complex ( Vives, 2001 ) . Although leting new entry to the market may bring on Bankss to act in a competitory mode when there are few Bankss in the market, a well-developed fiscal system besides appears to be of import, possibly because Bankss face competition from other fiscal houses and markets ( Boot and Thakor, 2000 ) . Furthermore, few limitations on the activities that Bankss can set about is of import to contestability may besides be related to increasing competition over the fiscal system.