A study on profitability and marketability of Taiwanese bank firms before and. after the Financial Holding Company Act

Transcription

1 A study on profitability and marketability of Taiwanese bank firms before and after the Financial Holding Company Act Dauw-Song Zhu Department of Business Administration, National Dong Hwa University, Taiwan. Al Y. S. Chen Department of Accounting, North Carolina State University, USA Yi-Kang Chen Department of Business Administration, National Dong Hwa University, Taiwan. Wei-Hsin Cheng Department of Business Administration, National Dong Hwa University, Taiwan. Abstract The purpose of this paper is to determine whether subordinate subsidiaries to financial institutions can improve their operating performance by establishing financial holding companies (FHCs) in Taiwan. Specifically, this paper uses data envelopment analysis (DEA) to measure profitability and marketability changes among 14 banks with subsidiary FHCs. The DEA models include the CCR and BCC models, both of which analyze the overall efficiency, pure technical efficiency, and scale efficiency of bank firms. Moreover, a bilateral model was used to measure and compare differences in operating performance before and after the FHC Act implementation. Results show that the efficiency scores of were lower than those of ; a bilateral model shows that prior to the implementation of the FHC Act, banks demonstrated higher levels of efficiency than banks subsidiaries to FHCs. Finally, this paper discusses some management implications for the banking industry. Keywords: DEA, operating performance, bank, FHC 1. Introduction Over the past several years, the regulatory framework of the financial sector has maintained the strict principle of keeping business development and operations separately; hence, banking, insurance, securities, venture capital, and bill financing have operated independent of each other. While referring to American and Japanese experiences with financial reform, Taiwan s government has progressively brought forward a number of financial amendments and regulations. First, the Banking Act was amended in July 2000, to loosen the investment and business scope of the banking industry; the 1

2 Financial Institutions Merger Act was then passed in December 2000, to provide taxation incentives and simplify merger processes. Third, the Taiwan Ministry of Finance (TMF) introduced the Financial Holding Company Act (FHC Act) in 2001, to reform the country s financial sectors; it provided a regulatory framework by which financial institutions could perform mergers and acquisitions (M&As) with different business sectors in the financial industry. The aim of the FHC Act was to remove obstacles to the financial sectors cross-use of resources, so they could ultimately realize cost reductions and expand market share advantages through the development of operational, financial, and market synergies. It is generally believed that the establishment of FHCs promotes economies of scale and helps the banking sector compete more efficiently (Lo and Lu, 2006). By cross-selling and thus garnering new business from both existing effective customers and new potential customers, financial institutions could pursue a larger business scope and attain large profits. Setting up holding companies is one way of achieving economies of scale, improving performance, and enhancing competitiveness or achieving combined output. An FHC is a non-business unit whose role is to integrate all financial resource allocations. Mainly objective of FHC is invested in appropriate subjects, while also fulfilling complete financial and business management services for its subsidiaries. Following implementation of the FHC Act, small and medium-size sectors merged into larger financial groups, to expand their capital, increase their service sites, and achieve economies of scale, and thus reduce information cost. From the viewpoint of banks, cross-marketing is profitable in terms of cost control, through the combination of insurance and securities across a variety of financial products and distribution channels; it leads to cost reductions in the form of branch closures, personnel reductions, and achievement-based taxation incentives. According to William B. Harrison, CEO of JP Morgan Chase & Co., he said the synergy resulting from the merger between JP Morgan and Chase Manhattan Bank created a $3-billion pre-tax net profit in 2000, $2-billion of that profit resulted from cost savings as a result of increased efficiencies. Deregulation reforms are therefore rife with opportunity for synergy-based increases in profitability (Sherman & Rupert, 1999). Since becoming affiliated with the World Trade Organization (WTO), Taiwanese financial institutions have quickened their expansion of operation scale and faced challenges from large-scale international competitors. Even though FHCs have merged many former competitors and thus reduced the number of market practitioners, the banks in Taiwan are still too many, too small in comparison to those of other countries. For instance, the three largest banks in Japan, Australia, Hong Kong, and Singapore have combined market shares, within their home countries, of more than 62%, 69%, 89%, and 95%, respectively. In comparison, none of these 14 bank subsidiaries to FHCs in Taiwan has a market share of more than 10%, and the five largest Taiwanese banks represent only 35% of the total market share. It shows that Taiwanese banks, on its own, are too small to achieve economies of scale; therefore, they need to merge, to enjoy the benefits of M&As synergy in Taiwan s small, albeit highly competitive economy. The top five banks, if combined, would be large enough to enjoy regional competitiveness; they need to follow the lead of the Hongkong and Shanghai Banking Corporation 2

3 (HSBC), which has gone from being Hong Kong s leading bank to becoming the world s third-largest financial institution, after 15 continuous years of performing M&As. Financial problems such as higher non-performing loans and overdue loans result from excessive market competition. The Taiwanese banking industry s long-time lackluster operating performance and lower profitability has made the possibilities of strong performance and a high level of efficiency questionable. Returns on assets (ROAs) in 2006 and 2007, for example, were 0.03% and 0.14%, respectively, which is much lower than those of other Asian countries ( %); additionally, the percentage of overdue loans an indicator of loan quality were 8.16% and 7.82% in 2001 and 2002, respectively. Related literature also point out that Taiwanese banks have been confronted with lower operating performance levels; for example, Lin, Hsu, and Hsiao (2007) analyzed the relative efficiency of Taiwan s 37 domestic banks in and found their average performance to be between 50.4% and 59.1%. The FHC Act aims to mitigate problems inherent in an over-crowded and low market share financial institution arena. Essentially, it allows the creation of operational synergy, through consolidation and increases revenues (output) and diminished costs (input), to achieve the goal of post merger consolidation. Because organizations tend to become more expansive and exhibit a high degree of cross-industry complexity after consolidation, they pose major challenges to the management abilities of managers. The stakes are also very high, given the potentially sizeable impact of operational inefficiency on the overall financial and economic market. Therefore, after combining with financial holding companies, whether or not banks achieve economic benefits and enhance operational efficiency, they can enjoy the aforementioned range of merger-based advantages. This paper focuses, foremost, on analyses of the relative efficiency of banking firms, which is important for managers to monitor, if they are to understand and maintain their companies performance. Furthermore, an assessment of banking performance is important in evaluating competitiveness and formulating appropriate strategies. The purpose of this paper is to measure the operating performance of 14 banks subsidiaries to financial institutions and determine whether they could increase their profitability and marketability by establishing FHCs in Taiwan. However, the operating effectiveness of a subsidiary depends upon synergy among the available resources, as integrated by the parent company. Given that banks account for more than 50% of FHC revenue, we focus on banks and exclude other financial sectors. This paper employs a two-stage production process that uses data envelopment analysis (DEA), a non-parametric frontier approach that allows for technical inefficiency, to measure relative performance in the banking industry. An objective measurement method needs to be developed for banks addressing both profitability and marketability performance. This paper refers to Seiford and Zhu s (1999) bank production process, and it modifies the variables therein to render the process suitable to discussions of FHCs. The remainder of this paper is organized as follows. The second section reviews related prior performance evaluations in the banking industry, and it highlights the salient features of this paper. Section three defines the evaluation process and provides some background literature of the DEA 3

4 approach used to select the input and output variables for our model. In section four, we present and discuss the empirical results of a sample of 14 banks with FHC subsidiaries in Taiwan. Finally, we conclude this paper in section five. 2. Literature Review With increases in competition, bank firms have become more attentive to issues related to efficient performance. Berger and Humphrey (1997) provide a comprehensive review of 130 financial institutions efficiency studies comprising more than 116 papers touching upon the topic of banking. This topic obviously has a long history, emerging from numerous published studies. In their research survey, at least five different approaches were employed to evaluate the efficiency of financial institutions, and they are generally divided into one of two types: non-parametric and parametric frontiers. In the prior literature, DEA is one of the most common approaches used to measure relative efficiency in banking industry: Seiford and Zhu (1999) also point out that DEA is perhaps the most common approach used to measure performance. Since Sherman and Gold (1985) first used it to measure banking efficiency, the use of DEA has spread progressively, to hundreds of theoretical and application-based studies. In any case, several important banking issues can be derived from the relevant literature. First, empirical studies have determined that inefficiency occurs primarily on pure technical bases, rather than on scale or allocative (i.e., Chen & Yeh, 2000; Sathye, 2001). Aly, Grabowski, Pasurka and Rangan (1990) investigated the operating performance of 322 US independent banks; their results indicate a low level of overall efficiency, as well as inefficiencies that are technical in nature, rather than allocative. Yue (1992) evaluates the performance of the 60 largest Missouri banks; the author concludes that pure technical inefficiency was the main source of technical inefficiency. Grabowski, Rangan and Rezvanian (1994) assess the impact of deregulation on the efficiency of 670 US banking firms in and ; their results showed that all of the firms efficiency scores had declined over the entire research period, except those pertaining to allocative efficiency. Such inefficiencies were caused by wasting of resources or not fully utilizing wealth, both of which resulted in pure technical inefficiency. Therefore, the authors conclude that there is no positive effect of deregulation on the operating efficiency of banking firms. Miller and Noulas (1996) measure the technical efficiency of 201 major banks, and their results indicate an average technical inefficiency (i.e., just over 5%) that was lower than that of previous literature. This difference may be due to the fact that many banks were acquired or went bankrupt in the 1980s, and that greater level of operating efficiency were needed among then to perform well. Table 1 highlights the literature related to mergers and deregulation in the banking industry. If a bank has not achieved economies of scale or, rather, experiences diseconomies of scale then that bank contribute to overall inefficiency. That is, inefficiency is mainly attributed to the inefficient utilization of resources, rather than an inappropriate input composition or scale of production. The FHC Act creates the opportunity for banks that lack economies of scale to achieve a more equitable market share, and to cross-use resources to sustain optimal competitiveness. 4

5 Table 1. Survey of mergers and deregulation studies in the banking industry using DEA Issues Authors Samples Mergers Peristiani (1997) 4900 individual mergers in US banks Avkiran (1999) 4 major trading banks in Australia Ralston, Wright and Garden 31 credit unions in Australia (2001) Sherman and Rupert (2006) 200 branches merged into four banks in US Sufian (2007) 6 commercial banking groups in Singapore Al-Sharkas, Hassan and Lawrence (2008) 440 merged banks in US Deregulation Berg, Forsund and Jansen 152 Norwegian banking deregulation during (1992) Elyasiani and Mehdian (1995) 150 randomly selected small banks and 150 mid-size and large banks in US commercial banks Fukuyama (1995) Japanese banking deregulation from 155 in 1989, 154 in 1990 to 153 in 1991 Zaim (1995) Turkish commercial banks deregulation from 42 in 1981 to 56 in 1990 Grifell-Tatjé and Lovell (1996) Spanish savings banks deregulation from 77 in 1986 to 56 in 1991 Second, the previous literature has generally focused on banking M&As and deregulation, and this focus is unlikely to take into consideration how profits and shareholder value can be simultaneously increased. It is generally agreed that information-sharing and the cross-use of resources in post-consolidation will save costs; in doing so, manifold benefits can emerge, including higher cash flow and returns. However, it is not only profitability, in terms of producing revenue, that is vital to bank firms; marketability is also crucial to shareholders wealth, because it is the firm s duty to maximize their interests. Therefore, the earnings per share (EPS), market value, and stock price of bank firms are commonly paid more attention by investors in the stock market. Seiford and Zhu (1999) originally proposed a two-stage production process and used the DEA approach to evaluate both the profitability and marketability efficiency of the top 55 US commercial banks; They determined that the efficiency of a bank is wrought in two-stage: profitability and marketability. This concept framework has been widely adopted by numerous studies in financial areas. Following the bank production process, Luo (2003) assesses a broad sample of 245 US large banks. Ho and Zhu (2004) created a modified two-stage DEA approach that separates efficiency and effectiveness, to assess the banking industry in Taiwan. Lo and Lu (2006) use Seiford and Zhu s model to measure 14 FHCs in Taiwan. Hwang and Kao (2006) applied a similar concept to 24 Taiwanese non-life insurance firms. Nonetheless, to the best of our knowledge, the two-stage DEA has not been used to measure the performance of bank subsidiaries to FHCs before and after regulatory framework reforms. The current study is the first to do so. 5

6 3.1 Basics of DEA and performance models 3. Research Design DEA is a performance evaluation technique based on a non-parametric linear programming approach for frontier analysis; it permits the use of multiple inputs and multiple outputs in the model. It was designed to estimate the relative efficiency of non-profit and for-profit organizations that required a set of homogeneous DMUs to distinguish which focal samples were efficient or inefficient. The basic concept of DEA is derived from the deterministic non-parametric efficiency frontier (Farrell, 1957) and envelopment theory (Farrell and Fieldhouse, 1962); the initial DEA model was introduced by Charnes, Cooper and Rhodes (1978) and is conventionally referred to as a CCR model, which observes a constant return to scale (CRS) and broadens the scope of a single input and output to multiple inputs and outputs. After the CCR model was introduced, Banker, Charnes and Cooper (1984) extended the returns to scale to variable returns to scale (VRS), and this model is generally referred to as the BCC model. The major difference between CRS and VRS is in the proportionality of the input-output relationship: If one unit of input generates one unit of output in a transformation process, it is known as CRS; otherwise, it is VRS. Consequently, these two different approaches are the main DEA approaches used in empirical and theoretical studies (Seiford, 1996). The CCR model is used to evaluate the technical efficiency (TE) under the premise of CRS, and the BCC model uses VRS to measure pure technical efficiency (PTE). Thus, TE calculated from CCR model can be compartmentalized into PTE and scale efficiency (SE). In this paper, the authors presume that 14 bank subsidiaries to FHCs keep the output level constant, and that they also minimize the inputs. The following linear programming formula and conditions of the CCR model measure the TE for the Min s. t. θ j (j = 1, 2,, n): m s + θ j ε si + sr (1) i= 1 r= 1 n j= 1 n j= 1 λ j X ij θ j X ij + si = 0 i= 1,2,..., m + λ j Yrj Yrj sr = 0 r = 1,2,..., s i + r λ, s, s 0 j where θ j is TE, is non-archimedean infinitesimal, λ j is the DMU j weight value. If 6 s i is input slacks, θ j = 1 and all input slacks and output surplus, equal to 0, then DMU j is technically efficient and operating on the CRS frontier. If 1, or some input slacks and/or output surplus, technically inefficient. s i and + s r is output surplus, and s i and + s r, are θ j is smaller than + s r, are not equal to 0, then DMU j is The BCC model, based on the premise of VRS, is used to evaluate the performance by imposing n the additional constraint of = λ = 1 on Eq. (1). The following linear programming formula and j 1 j

7 conditions of the BCC model are used to measure the PTE for the Min s. t. ω j (j = 1, 2,, n): m s + ω j ε si + sr (2) i= 1 r= 1 n j= 1 n j= 1 n j= 1 λ j X ij ω j X ij + si = 0 i= 1,2,..., m + λ j Yrj Yrj sr = 0 r = 1,2,..., s λ = 1 j The PTE of DMU j is ω j. If i + r λ, s, s 0 j ω j = 1 and all input slacks and output surplus, 7 s i and + s r, are equal to 0, then DMU j is pure technically efficient and operating on the VRS frontier. If ω j is smaller than 1, or some input slacks and/or output surplus, technically inefficient. s i and + s r, are not equal to 0, then DMU j is pure Therefore, the SE is defined by the ratio of TE / PTE. If SE is equal to 1, then DMU j is scale efficient; otherwise, if SE is smaller than 1, then DMU j is scale inefficient. The inefficient DMU j could improve its efficiency by reducing its inputs when the outputs are constant, increasing its outputs when the inputs are constant, or simultaneously reducing inputs and increasing outputs (Zhu, 2000). This paper evaluates the operating performance of Taiwanese bank subsidiaries to FHCs, based on the concept of the two-stage transformation process of the DEA approach, to design two performance models: profitability performance and marketability performance. Whether it is critical that un-representative input and output variables not be used in DEA to assess relative efficiency is a subject of debate (Paradi and Schaffnit, 2004). To undertake an objective evaluation and select case variables, prior related literature that adopted a two-stage DEA model to measure performance in the banking industry was surveyed. Table 2 summarizes the variables in the related literature that helped the authors in selecting the variables used in this paper. Authors Seiford and Zhu (1999) Luo (2003) Ho and Zhu (2004) Lo and Lu (2006) Table 2. Summary related literatures based on two-stage DEA to measure performance Samples Top 55 US commercial banks 245 US large banks 41 Taiwan s commercial banks 14 Taiwan s FHCs Input variables Assets Employees Stockholder s equity Assets Employees Stockholder s equity Assets Branches employees Capital stock Assets Employees Intermediate variables Revenues Profits Revenues Profits Deposits Sales Revenues Profits Output variables EPS Market value Total return to investors EPS Market value Stock price Net income Interest income Non-interest income EPS Market value

8 Hwang and Kao (2006) 24 Taiwan s non-life insurance Stockholder s equity Business and administrative expense Commissions and acquisition expense Direct premiums written Reinsurance premiums received Stock price Net underwriting income Investment income Figure 1 describes a bank production process adapted from Seiford and Zhu (1999). Stage-1 measures profitability performance, which represents a bank s ability to generate revenues and profits; they consist of three input variables (assets, employees, and shareholders equity) and two output variables (revenues and profits). In stage-2, marketability is the measure of a bank s performance in the stock market, in terms of two input variables (revenues and profits) and three output variables (EPS, market value, and stock price), which is consistent with the existing literature. Therefore, revenues and profits serve as intermediate factors that are outputs from stage-1 and inputs to stage-2 of the iterative process. The analytical process indicates that a bank s performance is a function of the production process of profitability and marketability. Bank Production Process Assets Employees Stockholders Equity Profitability Revenues Profits Marketability EPS Market Value Stock Price Stage-1 Profitability measurement Stage-2 Marketability measurement Figure 1. Two-stage production process for banking 3.2 Sample and data This study examine a sample of Taiwanese bank subsidiaries to FHCs; these 14 banks are, alphabetically: Cathay United, China Development Industrial, Chinatrust, E.SUN, First, Hua Nan, Jih Sun, Mega, Shin Kong, SinoPac, Taipei Fubon, Taishin, Waterland, and Yuanta. According to Charnes et al. (1978), an identical object in different years can be treated as an individual decision making unit (DMU); given that this study examines a 10-year period, then, it contains a total of 140 DMUs (i.e., 14 banks 10 years). The input and output data were reported as 10-year totals and can gathered from the Taiwan Economic Journal Co. (TEJ) database or from the respective companies web sites. The TEJ database is devoted to the accurate reporting of financial industry data and to making it publicly available; TEJ databases are generally considered valid and reliable. Table 3 presents the descriptive statistics of our data set. To verify isotonicity i.e., as inputs increase, outputs cannot be proportionately reduced the authors used input and output data to perform a correlation analysis. Table 4 shows the correlation matrix obtained. In the profitability performance model, the three inputs are positively associated with 8

9 two outputs. The highest correlation coefficient is 0.878, found between assets and revenues. The lowest correlation coefficient is 0.128, which is found between employees and profits. This outcome can be explained by the nature of resources: When a bank hires more employees to maintain its operations, it is likely to dilute its profits. In the marketability performance model, the two inputs are positively associated with three outputs. The highest correlation coefficient is 0.887, found between profits and EPS. It is interesting to note that the lowest correlation coefficient, 0.213, is found between revenues and EPS. This finding indicates that a bank s EPS mainly comprises profits, rather than revenues. Table 3. Descriptive statistics for the 14 banks subsidiaries to FHCs Mean Std. dev. Maximum Minimum Assets* (I 11 ) 613, , ,947, , Employees (I 12 ) 3, , , Stockholders' equity* (I 13 ) 48, , , , Revenues* (O 11, I 21 ) 31, , , , Profits* (O 12, I 22 ) 2, , , , EPS (O 21 ) Market value* (O 22 ) 76, , , , Stock price (O 23 ) Notes: 1. A financial index with * was in unit of million 2. The currency of financial data is NT$ I Table 4. Correlation coefficients among input, intermediate and output variables I Input Intermediate Output I 11 I 12 I 13 O 11, I 21 O 12, I 22 O 21 O 22 O 23 I O 11, I O 12, I O O O The authors chose the research periods, and excluding 2002 from the DEA analysis, for the following reasons: 1. All 14 FHCs declared their subsidiary companies process consolidations in The FHCs operation licenses took effect in Sherman and Rupert (2006) claim that post-consolidation benefits emerge within four years. Therefore, it would be very interesting to investigate and compare two five-year periods (i.e., one 9

10 before and one after the implementation of the FHC Act in 2001), to determine changes to operating performance. 4. Results and discussion 4.1 Profitability and marketability measurement With regards to resource management, from a manager s perspective, it is generally considered that banks have the strong ability to control inputs, rather than outputs. In this paper, a two-stage DEA model was analyzed under an input orientation namely, input minimization. The TE, which can be compartmentalized into PTE and SE, are reproduced to represent the profitability and marketability performance in Tables 5 6. First, we take profitability performance into account. The mean TE scores before and after the FHC Act implemented were and 0.910, respectively, representing a post-implementation decline -0.4%. Of the 14 banks studied, 11 (78.6%) were technically inefficient in both research periods. From the results of the mean efficiency score, mean efficiency score difference between two-period, and numbers bank on the frontier, then, we could not conclude that there were differences in TE before and after the FHC Act; however, the PTE before FHC Act implementation was superior to that after implementation, in terms of mean efficiency score and numbers bank on the frontier. Second, in the marketability performance model, we found a consistent result with respect to TE, PTE, and SE. Relative efficiency after the FHC Act implemented was less than that before implementation, for each of these 14 banks. Third, the SE is a critical criterion used to identify whether a bank s operation achieved economies of scale or it is under diseconomies of scale to undergo inadequate competition. The t-test indicates that the SE is significantly less than 1, which indicates scale inefficiencies in these banks, in terms of both profitability and marketability performance. This evidence shows that the 14 Taiwanese banks really experienced diseconomies of scale, rather than technical or resource allocation problems; it can therefore be suggested that bank subsidiaries to FHCs require further consolidation to broaden their business scope, by M&As. In this way, the Taiwanese financial service industry could achieve economies of scale and reduce cut-throat competition. Table 5. Efficiency scores of banks profitability performance Banks Profitability TE PTE SE Before After Before After Before After Cathay United China Development Industrial Chinatrust E.SUN First Hua Nan Jih Sun

11 Mega Shin Kong SinoPac Taipei Fubon Taishin Waterland Yuanta Mean Std. dev Mean variation -0.4% -1.5% 1.0% Fourth, it is obvious that profitability performance was better than marketability performance for the 14 Taiwanese bank subsidiaries to FHCs, from 1997 to This result can be interpreted thus: banks perform well in generating revenue and profits, on account of its business management and operating abilities, but that their competence in providing superior financial benefits to shareholders is inadequate. Nevertheless, some banks still act in ways appropriate to good performance in the stock market, e.g., China Development Industrial, Shin Kong, and Waterland. Finally, it was also found that two banks, China Development Industrial and Waterland, had TE in both profitability and marketability performance before the FHC Act was implemented; however, only Waterland possessed TE in both profitability and marketability performance, both before and after the FHC Act implementation. This implies that Waterland had good overall performance in terms of an integrated infrastructure, human resource management, asset utilization, and expenditure control in generating maximum return to shareholders. It can be said that Waterland s resource transformation in production process achieved optimal status between 1997 and Table 6. Efficiency scores of banks marketability performance Banks Marketability TE PTE SE Before After Before After Before After Cathay United China Development Industrial Chinatrust E.SUN First Hua Nan Jih Sun Mega Shin Kong SinoPac

12 Taipei Fubon Taishin Waterland Yuanta Mean Std. dev Mean variation -6.2% -7.6% -0.3% To summarize these results, the banks that were inefficient in either profitability or marketability were inefficient in terms of resource allocation such as Taishin, and tended to be inefficient in marketability performance after the FHC Act was implemented; these banks can improve performance by increasing their EPS and stock price by repurchasing treasury stock from stock market. More importantly, the results of SE analyses indicate that Taiwanese banks cannot sustain an appropriate scale to exert consolidation synergy, further supporting the assertion that banks need to undertake further M&As activities in Taiwan s financial service industry. 4.2 Analysis of performance on profitability and marketability This section presents an overview analysis, including profitability and marketability performance. The BCG matrix developed by the Boston Consulting Group was used to representing a reference point by which to identify each firm s relative performance. Using the TE scores of profitability and marketability, the position of each bank can be plotted on a two-by-two matrix; the banks can be broken into four quadrants: stars, cows, sleepers, and dogs. The quadrants were divided subjectively, based on mean TE scores. Figure 2 and 3 shows the distribution of the banks on the profitability-marketability matrix, before and after the FHC Act implementation. The stars were found to possess TE scores of 1 in each dimension, while the dogs zone is characterized by low profitability and marketability performance. These four segments of banks are described as follows: Stars: These banks possess high levels of efficiency, in terms of both profitability and marketability performance. Between 1997 and 2001, two banks can be found in this zone: China Development and Waterland. However, between 2003 and 2007, only one bank (Waterland) can be found in this zone. Banks that appear here can serve as good operational benchmarks to others, because they are technically efficient and operating on the frontier. Cows: These banks experience a higher level of profitability, but a lower level of marketability. Between 1997 and 2001, five banks (i.e., Cathay United, Mega, Hua Nan, First, and Chinatrust) could be found in this zone. Between 2003 and 2007, six banks were found there: SinaPac, Hua Nan, Mega, First, Taishin, and Chinatrust. The cows zone is characterized by banks that can generate profits, but lose in terms of their performance on the stock market for their shareholders. They must improve EPS and stock price by exercising financial leverages to acquire market attractiveness. 12

13 Sleepers: These banks have high marketability performance, but low profitability performance. Between 1997 and 2001, five banks appear here: E.SUN, Shin Kong, Jih Sun, Yuanta, and SinoPac. Between 2003 and 2007, six banks were classified here: China Development, Shin Kong, Cathay United, Jih Sun, E.SUN, and Yuanta. It is suggested that banks belong to this classification should pay greater attention to profit-making. Dogs: These banks perform in an inferior manner, in terms of both profitability and marketability. Between 1997 and 2001, two banks are included here: Taipei Fubon and Taishin; between 2003 and 2007, only Taipei Fubon was found in this zone. Dogs should pay greater attention both to generating cash flow or profit and to increasing market attractiveness. Figure 2 and 3 provide a visual diagram depicting where banks should focus their attention in achieving better overall performance. For example, before the FHC Act was implemented, E.SUN would need to focus on improving its profitability performance; on the other hand, after the FHC Act implementation, SinoPac would need to focus on improving its marketability performance. Marketability Figure 2. Distribution of 14 banks on profitability-marketability matrix in Marketability Figure 3. Distribution of 14 banks on profitability-marketability matrix in

14 4.3 Bilateral analysis of performance on profitability and marketability According to the purpose of this paper, we needed to measure and compare efficiencies before and after the FHC Act implementation to determine whether bank subsidiaries to FHCs could improve their operating performance. The functional types of input and output in the DEA model were constrained in an unlimited fashion; consequently, it is difficult to assess the distribution of efficiency scores (Charnes and Cooper, 1980). Using a non-parametric approach, we determined that the efficiency scores of the two groups of DMUs were significantly different. A bilateral model was proposed by Tone (1993), who used a comparison method based on a non-parametric approach for banks to determine the statistical difference in efficiency between the pre-fhc Act (Pre-FHC) and the post-fhc Act (Post-FHC) periods in the two-stage banking performance model (combining profitability and marketability performance). Our hypotheses are as follows: H 0 : The two groups have the same distribution of efficiency scores. H 1 : The two groups have the different distribution of efficiency scores. Pre-FHC and Post-FHC periods each comprised the same 14 banks. The authors used a rank-sum-test (Wilcoxon-Mann-Whitney) a non-parametric approach to test the null hypothesis that the two groups have the same distribution. Since the sample size of this study was not large than 30, the T test was used to determine the null hypothesis that the two groups have the same distribution at a given level of significance α. If T T α/2 or T T α/2, we then reject the null hypothesis. The test statistic T can be computed via the following: RS n1 ( n1 + n2 + 1) / 2 T = (3) n n ( n + n + 1) / where RS is the rank-sum statistics, and n 1 and n 2 are the sample size of Pre-FHC and Post-FHC, respectively. In the profitability performance model, the RS scores of Pre-FHC and Post-FHC were 110 and 296, respectively; the RS scores of Pre-FHC and Post-FHC were 160 and 246, respectively, in the marketability performance model. Given level of α with 0.05, it then holds that T = The T scores in profitability and marketability models with Eq. (3) are 4.62 and 2.13, respectively. In this case, the null hypothesis of both the profitability and marketability performance are rejected at the 5% significance level. Table 7 show the results of T test and to interpret that Pre-FHC period outperformed Post-FHC period, which means that banks prior to FHC Act implementation exhibited higher levels of efficiency than bank subsidiaries to FHCs post-implementation. Before the FHC Act was implemented, the focal banks generated plentiful profits and performed well in the stock market, thus maximizing their shareholder s wealth

15 Table 7. T test results Performance models RS scores T scores Outcomes of test Profitability Pre-FHC Post-FHC Pre-FHC was surpassed to Post-FHC in profitability performance. Marketability Pre-FHC Post-FHC Pre-FHC was surpassed to Post-FHC in marketability performance. What the causation led bank s profitability and marketability performance getting worse after the FHC Act implemented. There are several interpretations as to why banks could not succeed in generating efficient enhancements after the financial reforms: From the perspective of profitability, according to Sherman and Rupert (2006), any substantial benefits wrought by consolidation will appear about four years later. Because of organizational politics, it is difficult to integrate human resources and information systems, even in the name of sound operations, and so results in efficiency are deferred. Delays in consolidating operations are costly and prevent merged banks and their shareholders from benefiting from these opportunities (Sherman & Rupert, 1999). From the perspective of marketability, although M&As generate real benefits soon after the banks are combined, banks probably achieve lower accounting earnings than could otherwise have been achieved. It results in banks reporting imprecise information to the public; for this reason, investors may undervalue some banks common stock (Sherman & Rupert, 1999). Many banks fail to realize the advantages of size, even after a cross-financial institution consolidation. Manager can fail to integrate bipartite employees into the larger entity and thus achieve superior overall performance, thus slowing or even preventing the bank from developing a unified, more efficient and effective organization that generates profit and value. With regards to SE, much of the previous literature indicates that optimal efficiency and productivity are essential to a bank s success. In the past decade, while Taiwanese banks have been taking part in many M&As, market shares involved indicated that some banks are still vulnerable to diseconomies of scale. This paper suggests that banks should become further involved in M&As or adjust their portfolios, to create advantageous competition. 5. Conclusions The main contribution of this paper is its construction of an analytical framework by which to determine and measure the performance of bank subsidiaries to FHCs; this framework uses a two-stage banking performance model that combines profitability and marketability measurement. When operating in small economies such as Taiwan, bank firms experience intense competition and 15

16 diseconomies of scale that result in lower profitability, i.e. lower ROA. Furthermore, under conditions of bank consolidation or the reformation of a regulatory framework, issues concerning the impact of the FHC Act on banks efficiency are worthy of consideration. This makes it crucial for bank subsidiaries to FHCs to analyze and observe their profitability and marketability performance before and after FHC Act implementation, to achieve optimal competitiveness and keep sustained growth. Prior related literature has widely used the DEA approach to discuss the issues of bank efficiency. Nonetheless, some important topics have remained virtually ignored. First, many empirical studies have indicated that inefficiency is the primary source of pure technical, rather than scale or allocative (i.e., Chen & Yeh, 2000; Sathye, 2001); this paper, on the other hand, argues that it is not a certainty that a bank will always fail as a result of an inappropriate input-output composition. This is especially the case in Taiwan, a small island that consists of too many, too small banks for any to enjoy economies of scale. Second, financial consolidations are rarely considered a topic of vital research, and relatively fewer studies simultaneously take into consideration potential increased profits and shareholder value. This paper therefore focused on measuring the profitability and marketability performance of Taiwanese bank subsidiaries to FHCs. The empirical results of this paper point to a number of managerial implications. First, while banks exhibit relative PTE, the evidence indicates that scale inefficiencies certainly exist in these banks, in terms of both profitability and marketability performance. The result differs from the findings of previous literature, and it suggests that economies of scale are contingent upon further M&As activities among banks or FHCs. Second, profitability performance was better than marketability performance among 14 Taiwanese bank subsidiaries to FHCs within the research period; this finding could be interpreted thus: to become and remain attractive on the stock market, it is crucial to enhance financial performance. Increasing EPS and stock price to maximize shareholder interest is extremely important in achieving marketability performance. Singh and Montgomery (1987) assert that when a company enhances marketability and market share, it can use impact power, together with product price or quantity, to achieve market synergy. Third, it is meaningful to assess the two-stage DEA performance model using a synthetic model that combines profitability and marketability; to this end, the profitability-marketability matrix presented in the current study helps banks more precisely and thoroughly determine their levels of performance. Finally, there is a more important finding: The bilateral result indicates that banks prior to implementation of the FHC Act exhibited higher levels of efficiency than they did post-consolidation. Evidence of improve efficiency following these financial consolidations are somewhat mixed. Peristiani (1997) finds no evidence to support the theory that in-market mergers lead to significant improvements in efficiency; in his research, financial reform activities adversely affect the efficiencies of banks. Zaim (1995) suggests that financial reform seems to have succeeded in stimulating the commercial banks not only to enhance technical and allocative efficiency, but also to execute necessary scale adjustments, to achieve optimal scale. The current study has three limitations. First, this study is subject to the limitations of non-financial data much of which were unavailable at the time of the study and DEA, which 16

17 measures the relative TE values of Taiwanese banking firms. Results will vary according to data, and DMU is included in the analysis model. Second, the cross-sectional data used in this paper to evaluate performance do not reflect multi-period changes in productivity. Future research should investigate performance by using the Malmquist productivity index approach, to explore more specific efficiency variations. Lastly, concerning the constraint of statistic test, the T test and Wilcoxon test did not take into consideration what factors may affect the operating performance of banks. Further research can utilize regression models and control for other variables that may affect the operating performance to further determine what the FHC Act impact on efficiency changes in bank firms. To conclude this paper, the authors find that financial reform such as establishing FHC may not ensure that banks will enhance their profitability and marketability performance. Thus, banks need to make stronger efforts to improve their post-consolidation efficiencies, such as facilitating interdepartmental communication openly, establishing cross-functional teams, and cultivated good relationship with employees and customers to exert consolidation synergies. It is hoped that the non-parametric DEA approach and analysis methodologies used in this study can be adopted for use in performance evaluations in other industries. References Aly, H.Y., Grabowski, R., Pasurka, C., & Rangan, N. (1990). Technical, scale, and allocative efficiencies in U.S. banking: An empirical investigation. The Review of Economics and Statistics, 72(2), Al-Sharkas, A.A., Hassan, M.K., & Lawrence, S. (2008). The impact of mergers and acquisitions on the efficiency of the US banking industry: Further evidence. Journal of Business Finance & Accounting, 35(1/2), Avkiran, N.K. (1999). The evidence on efficiency gains: The role of mergers and the benefits to the public. Journal of Banking & Finance, 23(7), Banker, R.D., Charnes, A., & Cooper, W.W. (1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis. Management Science, 30(9), Berg, S.A., Forsund, F.R., & Jansen, E.S. (1992). Malmquist indices of productivity growth during the deregulation of Norwegian banking, Scandinavian Journal of Economics, 94(Supplement), Berger, A.N., & Humphrey, D.B. (1997). Efficiency of financial institutions: International survey and directions for future research. European Journal of Operational Research, 98(2), Charnes, A., & Cooper, W.W. (1980). Management science relations for evaluation and management accountability. Journal of Enterprise Management, 2(2), Charnes, A., Cooper, W.W., & Rhodes, E. (1978). Measuring the efficiency of decision making units. European Journal of Operational Research, 2(6), Chen, T.Y., & Yeh, T.L. (2000). A measurement of bank efficiency, ownership and productivity 17

18 changes in Taiwan. The Service Industries Journal, 20(1), Elyasiani, E., & Mehdian, S. (1995). The comparative efficiency performance of small and large US commercial banks in the pre- and post-deregulation eras. Applied Economics, 27(11), Farrell, M.J. (1957). The measurement of productive efficiency. Journal of the Royal Statistical Society, Series A, 120(3), Farrell, M.J., & Fieldhouse, M. (1962). Estimating efficient production functions under increasing returns to scale. Journal of Royal Statistical Society, Series A, 125(2), Fukuyama, H. (1995). Measuring efficiency and productivity growth in Japanese banking: A nonparametric frontier approach. Applied Financial Economics, 5(2), Grabowski, R., Rangan, N., & Rezvanian, R. (1994). The effect of deregulation on the efficiency of U.S. banking firms. Journal of Economics and Business, 46(1), Grifell-Tatjé, E., & Lovell, C.A.K. (1996). Deregulation and productivity decline: The case of Spanish savings banks. European Economic Review, 40(6), Ho, C.T., & Zhu, D.S. (2004). Performance measurement of Taiwan s commercial banks. International Journal of Productivity and performance Management, 53(5), Hwang, S.-N., & Kao, T.-L. (2006). Measuring managerial efficiency in non-life insurance companies: An application of two-stage data envelopment analysis. International Journal of Management, 23(3), Lo, S.-F., & Lu, W.-M. (2006). Does size matter? Finding the profitability and marketability benchmark of financial holding companies. Asia-Pacific Journal of Operational Research, 23(2), Luo, X. (2003). Evaluating the profitability and marketability efficiency of large banks: An application of data envelopment analysis. Journal of Business Research, 56(8), Miller, S.M., & Noulas, A.G. (1996). The technical efficiency of large bank production. Journal of Banking & Finance, 20(3), Paradi, J.C., & Schaffnit, C. (2004). Commercial branch performance evaluation and results communication in a Canadian bank a DEA application. European Journal of Operational Research, 156(3), Peristiani, S. (1997). Do mergers improve the X-efficiency and scale efficiency of U.S. banks? Evidence from the 1980s. Journal of Money, Credit, and Banking, 29(3), Ralston, D., Wright, A., & Garden, K. (2001). Can mergers ensure the survival of credit unions in the third millennium? Journal of Banking & Finance, 25(12), Sathye, M. (2001). X-efficiency in Australian banking: An empirical investigation. Journal of Banking and Finance, 25(3), Seiford, L.M. (1996). Data envelopment analysis: The evolution of the state of the art ( ). Journal of Productivity Analysis, 7(2/3), Seiford, L.M., & Zhu. J. (1999). Profitability and marketability of the top 55 U.S. commercial banks. Management Science, 45(9), Sherman, H.D., & Gold, F. (1985). Bank branch operating efficiency: Evaluation with data 18