J SWRILL:SX#RBIZ2ñlX Department of Finance National Yunlin University of Science & Technology +ầx Doctoral Thesis Exploring Performance in the Taiwanese Banking Industry — Applications of SFA and NDEA 7S GRIT SAY AL — SFA 88 NDEA Z/§H xa Bao-Ngoc Tong BE W5 St : AR REM Set Advisor: Cheng-Ping Cheng, PhD. Lien-Wen Liang, PhD. rh# EBl 113 # 6ö B June 2024 il Abstract This dissertation makes significant contributions to the literature on measuring performance in the banking industry, both theoretically and empirically. The thesis explores the efficiency of Taiwanese banks, with a focus on the impact of mobile payments, and Environmental, Social, and Governance (ESG) practices.
Utilizing network data envelopment analysis (NDEA) and stochastic frontier analysis (SFA) models, the research consists of three key studies that provide insights into modern banking operations optimization. The first study investigates the influence of new mobile payment services on Taiwanese banks' performance using a two-stage NDEA model. Results reveal the profitability potential of mobile payments and the superior explanatory power of the NDEA model. The second study explores the impact of ESG practices on Taiwanese banks' performance employing a two-stage NDEA method.
Findings suggest a positive correlation between ESG and profitability, with smaller banks benefiting more from focused ESG activities, especially in governance aspect. The third study assesses the impact of digital banking on Taiwanese banks' cost efficiency using a stochastic cost frontier model. The analysis indicates a positive relationship between digital account openings and cost efficiency, with private banks showing proactive digital transformation. The thesis underscores the significance of assessing bank performance in light of emerging fintech and ESG trends.
It employs DEA and SFA methodologies to comprehensively evaluate Taiwanese banks' efficiency, highlighting the potential of mobile payments and ESG integration. The findings suggest that banks should expedite the adoption of electronic banking and ESG principles to enhance efficiency, reduce costs, and align with regulatory and customer demands. This strategic integration can not only drive operational improvements but also contribute to broader sustainability objectives, such as achieving Taiwan Net-Zero Emissions by 2050. Acknowledgements I would like to express my deepest gratitude to my supervisors, Professor Cheng-Ping Cheng and Professor Lien-Wen Liang, for their unwavering guidance and support throughout my PhD journey.
Iam immensely grateful to Professor Cheng for accepting me as his student, assisting me during my initial days, and introducing me to Professor Liang. Your continuous support made my transition into this new academic environment much smoother. Professor Lien-Wen Liang, thank you for your invaluable help with my paper writing and for your insightful feedback. Your patience and encouragement, especially during times of slow progress, were crucial to my development as a researcher.
I would also like to extend my sincere thanks to the people in the department office. Your help and support throughout my studies have been invaluable. To my parents, thank you for your financial and emotional support during this time. Your unwavering belief in me has been a constant source of strength and motivation.
My time at the National Yunlin University of Science and Technology has been incredibly memorable. The experiences and opportunities here have enriched my knowledge and skills, contributing significantly to my academic and professional growth. il Table of Contents NINN ssc cra sina i ISS ain SR Si Sa aS SNE 1 Peer Wy LEC MEG sóc ssx6ts055564655661563506601368103890864601/8568i-)3X88534GSXHS4S2S2338SM/0,531G350380386.38008383586 1 14BIE Of Combet os sacsssssesacsnssnasassesnses vans nanan 616ã053052683g802S151453854GRGE4ES835/.4G8555846438186555218 11 Listet Tables s::sxsz;s152170150138960123500101590GEB1SSERGEIEEEIIIGNSEREEEIS-ELSEEBSATEHGKIEAIBHIXĐTEERSSE4ãSĐ381388ĐĐ3E V Listot Figures ;issossssrtsistnieiiatE0EG513 S041 0VS015EDASYGENIAESIEESAIBSHIIQSITDDRIEMSHEASIB381iáfAsstussHi Vil CHAPTER 1: INTRODUCTION sisecssscssssecssessvee semen emercmieneaeanennreeen 1 1:1: Pettotmante measuteniont oc0ssccseccvmnmers mcr 1 1. and SEA in bankang industry ::-::.
ESG and Niobile:payTniEfiE:.ssssssssssgsssssetieetiiatiisvogtl05564553419309E5ES4GCZNE24090088 6 1. Taiwanese banking: HIđUSỈTV:s;ssssssszsxecss22061100002216615615616981195513058150091380050898653H868883 8 CHAPTER 2: USING NETWORK DEA TO EXPLORE THE EFFECT OF MOBILE PAYMENT ON TAIWANESE BANK EFFICIENCY.---¿--¿55-++<++<<+<<<++ 12 Poly TBO GUCHONE oo sccsscrscrrmenrmerneseennencomaumnnsencincmmemnen ene EN 12 2. Mobile Payment and Data Envelopment Analysis (2EA). Empirical Results oo.
ESG and NDEA oo. 49 Si: CO HE HT ON. concrsa renee cecees eee eae mins 59 4. Stochastic Frontier Analysis (SEA)) 0.
Data and EimimliricgÌ es ssccsssiiacasnscncisas ssecnsisoniecvsussnnnsktesaneanalbeneacniasssasnewntindnessaiiene 69 4. Conclusion and Future research.-- --- 5 5 +2 +22 *+*£+s+eE£eveEserseeerreerrereree 83 CHAPTER 5: DISCUSSION AND CONCLUSIƠN. Comparing results occ eee eeeceeseeeeesseeseeeeeseeseceeceeeseceneeaeeaeeaesaeeaeearereeneseess 87 D2. MEMO GO BY.cenccne-noesnoseravencannennanzeutinnannseteensiuennnensanssneansnunsnatentenaensennen tea 88 5.
Summary and ConeÏuSIOH:. Limitations and Future Research. ốc ca CC. 93 1V List of Tables Table.
Variable de tintti On :ssccccsnessescecnaneonsesesanssneornawemanannmmernenasansenremesonaxca 22 Table 2. Descriptive statistics for 19 Taiwanese banks (from 2018 to 2021). Management efficiency of 19 Taiwanese banks with weights (2020-2021) a re ra ry ot apr 581080588 25 Table 2. Profitability efficiency of 19 Taiwanese banks with weights (2020-2021)28 Table 2.
Overall efficiency of 19 Taiwanese banks with weights and CCR GIIIGIGHGIGSš (2021) sscnsscsasnseumassmeaeunesmeneesmamamene enema nnaeReeR OEE ARREST 30 Table 2. Average efficiency scores and weights of Taiwanese banks (2018-2021) 34 Table 2. Efficiency scores—comparing two models (2020-2021). Descriptive Statistics of Efficiency Score—comparing 2 models (2020— “10.
Descriptive statistics for the 14 Tarwanese banks. Management efficiency of 14 Taiwanese banks with weights (2019-2020) ee ee 49 Table 3. Profitability efficiency of 14 Taiwanese banks with weights (2019-2020) 52 Table 3. Overall efficiency of 14 Taiwanese banks with weights (2020).
Average efficiency scores and weights of Taiwanese banks (2018-2020) 55 Table 3. Efficiency scores — Comparing two models (2019-2020). Descriptive statistics of the efficiency score—Comparing 2 models (2019— 2020) sesosaxccensnsnvananscessaiesee ness suncsesunexetauns ssrawan sss sane euceseesassa nis ian Ran NARI me cEADSSCNSNRINOEREA TO a7 Table 3. Results of Tobit regression for efÍICIenncy.- ----c + sssxs+svssssseersers 58 Table 4.
Descriptive Statistic (SEA) sssiccseessscsunancnssnesscrstmsmcreamcannunobentinsterecinsecniaceeons 70 Table 4. Inefficiency variable definition and Descriptive statistic. Correlation Coefficients for Inefficiency Factors of Public and Private HHHKẾ: wsssscussasssres sen avrasiesras <seasins ange38045803868156. Stochastic Frontier Cost Function Estimations for Public and Private bank eres eee 73 Table 4.
Results of the cost inefficiency model for public and private banks. Cost Efficiency Rankings for Public Banks. Cost Efficiency Rankings for Private Banks. Ranking of average Technical Gap Ratios for Public and Private Banks.
Ranking of average Metafrontier Cost Efficiency for Public and Private VI List of Figures Figure 2. Bank's production process ‹.c‹ccccc-cc c6 0 S100 110021014601 01 HH H2 He ke g1 Cà 4 20 Figure 3. Bank’s production process (ESG) ssers. Cost efficiencies for both public and private banks.
Average Technical Gap Ratios for Public and Private Banks. Average Metafrontier Cost Efficiency for Public and Private Banks. 83 Vii CHAPTER 1: INTRODUCTION 1. Performance measurement Business operations and processes involve transformation, adding value to materials and converting them into goods and services desired by customers.
Organizations seek to assess their performance in terms of resource utilization—such as labor, materials, energy, and machinery—and outcomes, including the quality of finished products, services, and customer satisfaction. Managers aim to evaluate the efficiency of these processes using various performance measures (Zhu, 2014). There is a growing demand for organizational functions and processes to showcase their role in contributing to performance (Micheli & Mari, 2014). Performance measurement is essential for not only operational transparency, but also fostering organizational improvement.
Effective performance measurement enables companies to identify their strengths and weaknesses, recognize top performers, pinpoint areas needing improvement, and establish benchmarks using historical data. However, the term "performance" encompasses a wide range of aspects, including productivity, profitability, efficiency, maximizing shareholders’ equity, among others (Corvellec, 2003). Consequently, there exist diverse methodologies to gauge performance. This study concentrates on two prominent methodologies: data envelopment analysis (DEA) and stochastic frontier analysis (SFA).
These approaches evaluate technical efficiency (TE) by considering firms as production processes. It's important to highlight that SFA and DEA are developed based on distinct perspectives regarding TE. Hence, their results can offer varied perspectives on inefficiency and propose different solutions to enhance a firm's performance. The productivity ratio typically signifies the total productivity of the entire system, evaluating the conversion of total inputs into total outputs.
The Malmquist productivity index, utilized to monitor productivity alterations over time, constitutes the sum of technical efficiency change, technical change, scale efficiency change, and output mix effect (Battese & Coelli, 1995). Conversely, efficiencies derived from SFA and DEA models focus on the optimal level of outputs derived from inputs. These terms, while occasionally interchangeable, do not precisely refer to the same concept. Additionally, TE can be perceived as a component of total productivity.
In essence, while productivity pertains to quantity, efficiency concerns quality. Banks are pivotal in structuring economic development and growth by efficiently channelling funds from borrowers to savers, thereby strengthening a nation's financial and economic system. As the primary sources of financial intermediation and facilitators of payments, banks play a vital role in a country's economic advancement (Sharma et al. By providing financial support to entrepreneurs and businesses, banks facilitate cost reductions through advancements in financial innovations, thereby enhancing production efficiency and freeing up capital for other sectors.
The improved efficiency of the banking system enables higher output to be achieved with the same input level in the market. Therefore, measuring bank efficiency is crucial as bank failures or insufficient liquidity resulting from loan collection issues can endanger the economic well-being of millions of individuals (Burgstaller, 2020). Besides, the heightened productivity within the banking industry yields widespread benefits for people (Košišová, 2020). Consequently, there has been a notable increase in the focus on efficiency analysis of financial institutions, particularly commercial banks, in recent years.
Traditionally, ratio analysis has been the preferred technique for assessing banking performance. It involves examining the relationship between two variables to gain insights into various aspects of a bank's operations, such as liquidity, profitability, and risk management (Paradi, Yang, et al. However, this approach is subject to criticism due to its reliance on benchmarking ratios and inherent subjectivity (Yeh, 1996). While accounting and financial ratios offer valuable information for evaluating a bank's financial efficiency, they often fail to capture the full range of factors influencing bank performance, including assets, revenue, profit, market value, and customer satisfaction (Seiford & Zhu, 1999).
Moreover, the multitude of ratios generated can be contradictory and confusing, lacking an objective means of identifying inefficient units and distinguishing between poor and superior performers (Berger & Humphrey, 1997). On the other hand, frontier analysis, with SFA and DEA as two prominent representatives, offers a more comprehensive and objective assessment of bank efficiency by considering multiple inputs and outputs and evaluating long-term performance. The frontier analysis provides an objective quantification of efficiency, offering numerical scores that incorporate economic optimization mechanisms, thereby facilitating the benchmarking of institutions and the identification of best practices 2 within the banking sector (Berger & Humphrey, 1997). This method can be divided into parametric and non-parametric approaches.
While econometric model like SFA necessitates specific assumptions regarding its functional form and account for random errors, non-parametric like DEA offers a flexible means of specifying the best-practice frontier but do not incorporate considerations of random errors (Bauer, 1990). In summary, traditional ratio analysis, despite its long use in evaluating banking performance, is increasingly recognized for its limitations, particularly its reliance on benchmarking ratios and inherent subjectivity. Conversely, frontier analysis, including SFA and DEA, offers a more comprehensive and objective assessment. By incorporating multiple inputs and outputs, frontier analysis provides an objective efficiency quantification, enhancing the ability to benchmark institutions and identify best practices.
This makes frontier analysis a superior methodology for assessing bank performance, offering greater comprehensiveness, objectivity, and versatility. The next section will delve deeper into two frontier analysis techniques, and their application in investigating banking performance. DEA and SFA in banking industry DEA and SFA are distinguished as one is non-parametric and the other is parametric method. While DEA is an application of linear mathematic programming, SFA utilizes econometric functions.
Hence, DEA is capable of calculating relative efficiency without requiring any predetermined assumptions about the form of the frontier or distributing inefficiency assumptions. Nonetheless, SFA has the ability to differentiate random noise from inefficiency, offering deeper insights into the impact of external factors on performance. Data envelopment analysis (DEA) DEA employs mathematical programming techniques capable of managing numerous variables and constraints.