How Fintech Improves Financial Performance of Banks in China: The Context of Green Finance and ESG

Abstract

Fintech has become an important driver of digital transformation and sustainable development in the banking industry. However, existing studies report inconsistent findings regarding the relationship between fintech and bank financial performance. This study examines the impact of fintech adoption on the financial performance of Chinese listed commercial banks and investigates the mediating roles of green finance and ESG performance, as well as the moderating role of ownership status. Using panel data from 42 Chinese A-share-listed commercial banks between 2015 and 2024, this study employs panel regression analysis to evaluate the direct, mediating, and moderating relationships among the variables. The results indicate that fintech significantly improves market valuation, operational profitability, and asset utilisation efficiency in Chinese listed commercial banks. In addition, green finance and ESG performance partially mediate the relationship between fintech and financial performance, suggesting that fintech contributes to sustainable financial value creation through sustainability-oriented mechanisms. The findings further show that ownership status significantly moderates the fintech–financial performance relationship, with fintech generating stronger positive effects among non-state-owned banks than state-owned banks. Furthermore, the instrumental variable and other robustness tests confirm the robustness of the findings after addressing potential reverse causality concerns. This paper suggests that the effectiveness of fintech depends not only on technological investment but also on sustainability capabilities and institutional conditions. This study provides empirical evidence on fintech-driven sustainable banking transformation in China and offers practical implications for regulators and commercial banks seeking to promote digital finance and sustainable development.

1. Introduction

The rapid advancement of fintech has transformed the global financial landscape, reshaped traditional banking operations and redefining the role of financial intermediaries in recent years. Powered by breakthroughs in big data analytics, artificial intelligence, blockchain, and cloud computing, fintech has driven significant gains in operational efficiency, strengthened risk management capabilities, and broadened access to financial services, particularly for underserved populations [1,2]. In emerging economies such as China, fintech has emerged as a pivotal force in financial innovation, underpinned by proactive regulatory frameworks and robust digital infrastructure [3]. As Chinese listed banks deepen their integration of fintech into core business models, the implications for financial performance have attracted growing attention from both academic researchers and industry practitioners [4].

Although fintech is gaining prominence, existing literature presents mixed evidence on its impact on bank financial performance. On the one hand, fintech adoption can enhance cost efficiency, optimize resource allocation, and improve profitability by reducing information asymmetry and transaction costs [5,6,7]. On the other hand, fintech investments typically demand substantial upfront capital, technological adaptation, and organisational restructuring, potentially generating short-term financial pressures and raising operational risks [8]. Recent empirical evidence further indicates that, under certain market conditions, fintech adoption may exert a negative effect on bank performance, because the elevated compliance costs and intensified competitive pressures [9,10]. These conflicting findings underscore the complexity of the fintech to financial performance relationship in the banks and highlight the need to examine the underlying mechanisms through which fintech exerts its effects.

In this context, green finance has emerged as an important channel linking fintech and the sustainable development of banks’ financial performance. Green finance refers to financial activities that support environmentally sustainable projects, including green credit, green bonds, and investments in low-carbon technologies. In China, the national policy-driven initiatives have accelerated the integration of green finance into the banking sector. Existing research shows that fintech and green finance are increasingly interconnected, jointly promoting sustainable transformation and resource allocation efficiency of companies [11,12]. Fintech can facilitate green finance by improving environmental risk assessment and information transparency, thereby enhancing financial performance.

In addition, environmental, social, and governance (ESG) performance has been receiving increasing attention as a key determinant of firm value and financial outcomes. ESG performance reflects a firm’s commitment to sustainability, social responsibility and governance quality. Empirical evidence suggests that ESG practices can improve banking profitability, reduce financing constraints, and enhance investor confidence [13]. Moreover, fintech plays a critical role in enhancing ESG outcomes by improving data availability, monitoring systems, and transparency [14,15]. However, the mediating role of ESG performance in the fintech and financial performance nexus remains underexplored.

However, the effectiveness of fintech may vary depending on institutional factors, such as ownership structure. In China, ownership status plays a critical role in shaping bank behaviour and strategic decisions. State-owned banks, which are totally owned by the government, often benefit from stronger policy support and better access to resources, particularly in the context of green finance policy and ESG initiatives. Moreover, other banks operate under greater market pressure. Prior studies indicate that institutional characteristics can significantly influence the financial performances of sustainability and innovation strategies, highlighting the importance of considering ownership heterogeneity [16,17,18].

Despite growing literature on fintech and financial performance of banks in China, several important gaps remain unresolved. First, prior studies have mainly focused on the direct effect of fintech on financial performance. They have paid insufficient attention to the sustainability-related mechanisms by which fintech creates financial value. In particular, the mediating roles of green finance and ESG performance are underexplored in the banking context. Second, existing studies often examine fintech, green finance, and ESG separately. Few integrate them into a unified framework explaining how digital transformation contributes to sustainable financial performance. Third, although ownership heterogeneity is highly relevant in China’s banking sector, limited research has examined whether fintech’s effectiveness differs between state-owned and non-state-owned banks under varying institutional conditions. These limitations indicate the need for an integrated framework that links fintech, sustainability capabilities, and institutional heterogeneity to explain banks’ financial performance.

To address these gaps, this study investigates the impact of fintech on the financial performance of Chinese listed commercial banks. It pays particular attention to the mediating roles of green finance and ESG performance. It also assesses the moderating role of ownership status. From a theoretical perspective, this study integrates Dynamic Capability Theory, Dynamic Natural Resource-Based View (Dynamic NRBV), and Institutional Theory. This explains how fintech-enabled digital transformation contributes to sustainable financial performance under different institutional conditions. Using panel data from Chinese listed commercial banks between 2015 and 2024, this study develops an integrated analytical framework. The framework captures both the direct and indirect ways fintech influences bank performance.

This study is expected to contribute to the existing literature in several ways. First, it seeks to extend fintech–bank performance research by moving beyond a direct-effect perspective and examining the sustainability-related mechanisms through which fintech may create financial value. Second, by incorporating green finance and ESG performance into the same analytical framework, this study aims to provide a more integrated understanding of how digital transformation may support sustainable banking performance. Third, by examining the moderating role of ownership status, this study intends to highlight that the effectiveness of fintech adoption may vary across different institutional and governance conditions. In doing so, the study is expected to provide additional empirical evidence from Chinese listed commercial banks and offer useful implications for digital transformation and the development of sustainable finance in emerging banking markets.

2. Literature Reviews and Hypotheses Development

2.1. Fintech and Financial Performance

The continuous advancement of digital technologies, including big data, artificial intelligence, cloud computing, and blockchain, has established fintech as a primary catalyst for transforming the business models, competitive dynamics, and profit structures within China’s banking industry [19,20]. Fintech enables traditional commercial banks to redesign service processes, enhance risk-control systems, and overcome operational limitations associated with physical branches. These changes significantly affect various financial performance indicators [21,22]. Historically, traditional banks have faced challenges including high operating costs, pronounced information asymmetry, limited profit channels, and inflexible asset-liability structures, all of which have constrained sustained improvements in profitability and market value [23]. The integration of fintech addresses these challenges through multiple mechanisms.

From an operational efficiency standpoint, fintech enables banks to achieve intelligent customer acquisition, automated risk control, and online business processing, thereby reducing labour costs and physical branch operating expenses [24,25]. Big data technology facilitates the mining of user credit information and transaction-behaviour profiles, mitigates information asymmetry in the credit market, reduces the default risk associated with lending businesses, and consequently stabilizes and expands banks’ probability [26,27]. In terms of business innovation, fintech encourages commercial banks to expand into inclusive finance, consumer finance, and wealth management, disrupts the homogeneous competition of traditional deposit and loan businesses, creates new profit growth points, and further enhances accounting profitability and market valuation [28,29].

Empirical studies have consistently demonstrated a positive correlation between fintech and bank financial performance. Analyses of global bank samples indicate that fintech development significantly improves banks’ financial performance by optimizing operational processes and enhancing risk management capabilities [30,31]. Research focusing on Chinese listed commercial banks has shown that the application of fintech effectively expands banks’ interest income [32]. However, some scholars contend that the initial investment in fintech entails substantial capital and technology costs, which may reduce short-term profits and temporarily inhibit performance [33,34]. Despite these concerns, as digital transformation matures and scale effects are realized, the long-term value-added impact of fintech on bank performance is expected to become predominant.

Given the current context of China’s banking industry, digital transformation has emerged as an essential strategic choice for listed commercial banks. Consistent with Dynamic Capability Theory, fintech adoption enables banks to strengthen operational efficiency, optimise resource allocation, and enhance digital financial capabilities, thereby improving financial performance. Based on this analysis, the following research hypothesis is proposed:

Hypothesis 1:

Fintech has a significant positive relationship with the financial performance of Chinese listed commercial banks.

However, recent studies increasingly suggest that the financial value created by fintech may depend not only on digital capability enhancement, but also on sustainability-oriented mechanisms such as green finance and ESG-related practices.

2.2. Green Finance as Mediator in Fintech and Financial Performance

Green finance, which includes green credit, green bonds, and green investment, has become a central mechanism for the financial industry to support low-carbon transformation and promote high-quality economic development in recent years [35,36]. For listed commercial banks, green finance constitutes both a key component of social responsibility and a strategic method for optimizing asset structures, reducing environmental policy risks, and creating new opportunities for profit growth [37]. Within the digital economy, fintech serves as a foundational support for banks to advance green finance, thus establishing a critical mediating pathway between fintech and financial performance.

Fintech plays a significant role in advancing high-quality green finance. Traditional green finance faces challenges, including difficulties in identifying green project attributes, high supervisory costs, limited transparency in evaluating environmental benefits, and restricted traceability of capital flows [38]. By leveraging artificial intelligence and big data, fintech enables precise screening and dynamic oversight of green projects, thereby reducing identification costs and default risks associated with green credit operations [39]. Furthermore, blockchain technology supports comprehensive traceability of green funds, enhances the credibility of green finance activities, and encourages banks to allocate more capital to low-carbon and environmental protection industries [40].

Conversely, the advancement of green finance can substantially enhance banks’ financial performance. Allocating credit resources to new energy, energy conservation, and environmental protection sectors enable banks to reduce credit risks associated with high-pollution and high-energy-consuming industries, improve asset quality, and lower non-performing loan ratios [41]. In addition, implementing green finance strategies helps banks establish a positive social image and market reputation, increase investor recognition, and raise market valuation. Differentiated green credit pricing and green intermediary services further contribute to stable interest and non-interest income [42]. Latest literature has validated the logical sequence of fintech, green finance, and financial performance. Some authors demonstrated that fintech significantly increases the scale of green credit in banks, while green finance further enhances profitability and market value [43]. Also have article confirmed that green finance serves as a significant partial mediator in the relationship between financial performance [44].

From the perspective of Dynamic NRBV, fintech-enabled digital capabilities support sustainability-oriented banking activities, including the development of green finance and environmentally responsible resource allocation. Consequently, fintech may promote green finance activities and indirectly improve banks’ financial performance. Accordingly, this study proposes:

Hypothesis 2:

Green finance mediates the relationship between fintech and financial performance in Chinese listed commercial banks.

Building on this, these findings suggest that fintech may indirectly improve banks’ financial performance by facilitating the development of green finance and strengthening sustainable banking practices.

2.3. ESG Performance as Mediator in Fintech and Financial Performance

ESG performance has become a central metric for assessing the sustainable development capacity and overall competitiveness of contemporary financial institutions [45,46]. High ESG performance indicates that banks have met environmental protection obligations, provided social public services, and established robust internal governance structures, all of which are closely associated with long-term operational stability and financial growth [47]. As a significant digital catalyst, fintech can systematically improve banks’ ESG performance and serve as a key mediating channel influencing financial outcomes.

Within the environmental dimension, fintech enables online, paperless banking, thereby reducing resource consumption and carbon emissions. Additionally, fintech supports low-carbon transformation initiatives through digital green financial channels, mitigating negative environmental externalities associated with business operations [48]. Regarding social responsibility, fintech overcomes geographical and service barriers, broadens access to inclusive financial services for small and micro enterprises, rural populations, and vulnerable groups, and enhances both social welfare contributions and brand reputation [49]. In the governance dimension, digital intelligent management systems standardize internal processes, strengthen real-time risk monitoring, limit excessive risk-taking and agency issues, and improve both information disclosure transparency and internal governance efficiency [50].

Enhanced ESG performance can further strengthen banks’ financial outcomes. Institutions with high ESG ratings benefit from lower financing costs, increased investor preference, and higher market valuations. Strong environmental and social responsibility performance enables banks to avoid regulatory penalties and policy risks, maintain asset quality, and support sustained growth in financial performance [51]. Recent research demonstrates that fintech significantly improves corporate ESG performance and that superior ESG outcomes contribute to long-term profitability and market competitiveness for banks [52,53].

Dynamic NRBV further suggests that sustainability-related organisational capabilities, including ESG-oriented governance and environmental responsibility, may strengthen firms’ long-term competitiveness and financial sustainability. Fintech may therefore improve financial performance indirectly through enhanced ESG performance. Therefore, this study proposes:

Hypothesis 3:

ESG performance plays a mediating role in the relationship between Fintech and financial performance in Chinese listed commercial banks.

Consequently, ESG performance may represent another important sustainability-related pathway through which fintech-driven digital transformation contributes to long-term bank performance.

2.4. Ownership Status Moderate the Effect of Fintech on Financial Performance

Ownership status is a key institutional characteristic in China’s banking industry, leading to distinct behaviour patterns and resource endowment disparities between state-owned and non-state-owned listed banks [54,55]. Fundamental differences in policy positioning, market competition pressure, internal decision-making mechanisms, and profit incentive orientation between these two types of banks generate heterogeneous adjustment effects on the relationship between fintech and bank financial performance.

State-owned banks benefit from significant advantages in national policy support, capital scale, customer resource accumulation, and brand credibility, which provide them with greater capital strength and a more robust resource endowment for fintech infrastructure development and digital business expansion [56,57]. However, the greater institutions also have social-strategic responsibilities and within-strategic roles within the national economic system, and their operations are constrained by bureaucratic management structures and rigid administrative procedures. Unlike non-state-owned banks, which priorities market profit maximization, state-owned banks must balance social responsibility with policy objectives during digital transformation. This balance often results in weaker market-oriented incentives to rapidly translate fintech investments into tangible performance gains [58]. Consequently, the marginal effect of fintech on financial performance is relatively limited in state-owned banks.

In contrast, non-state-owned commercial banks operate in a highly market-oriented and competitive environment, facing intense industry-wide pressure to achieve profitability [59]. These banks have flexible institutional mechanisms, streamlined decision-making, and strong market awareness, enabling them to quickly adjust their digital strategies and resource allocation in response to market changes. As a result, non-state-owned banks can fully leverage fintech to reduce operating costs, expand customer bases, and optimize risk control, thereby exerting a stronger positive impact on financial performance rather than state-owned banks [60].

Recent literature further supports the heterogeneous effects associated with ownership attributes. This article found that institutional constraints related to ownership diminish the market-oriented efficiency gains from fintech in state-owned banks [61]. Another paper demonstrated that the performance improvement resulting from digital transformation is more pronounced and responsive in non-state-owned banks with stronger market incentives [62].

Institutional Theory suggests that institutional environments and governance structures may influence the effectiveness of fintech adoption across banks. In China’s banking sector, ownership status may moderate the relationship between fintech and financial performance because state-owned and non-state-owned banks operate under different institutional conditions. Accordingly, this study proposes:

Hypothesis 4:

Non-state-owned banks play a stronger moderating role than state-owned banks in the relationship between FinTech and financial performance.

Furthermore, the effectiveness of fintech adoption may vary across ownership structures because institutional conditions influence banks’ strategic flexibility, governance incentives, and digital transformation capabilities.

2.5. Integrated Research Framework and Research Gaps

Existing studies have extensively examined the individual relationships between fintech, green finance, ESG performance, and financial performance. However, the current literature remains fragmented in several important respects. Most prior studies focus on the direct relationship between fintech and banks’ financial performance, while relatively little attention has been paid to the sustainability-related mechanisms by which fintech generates financial value in the banking sector [52,63,64]. Studies investigating the mediating roles of green finance and ESG performance remain insufficient and are often examined separately rather than within an integrated analytical framework.

Moreover, previous literature has generally discussed fintech, green finance, and ESG performance as independent research streams, without systematically explaining how these dimensions interact to influence banks’ sustainable financial performance. Existing studies also present different empirical findings on the effectiveness of fintech adoption, suggesting that the performance consequences of fintech may depend on institutional characteristics and organisational conditions [9,65]. However, limited research has explored whether the impact of fintech differs between state-owned and non-state-owned banks within China’s unique institutional environment.

To address these gaps, this study develops an integrated research framework linking fintech, green finance, ESG performance, ownership status, and financial performance in Chinese listed commercial banks. The proposed framework is theoretically informed by Dynamic Capability Theory [66], Dynamic NRBV [67], and Institutional Theory [68,69], which collectively explain how fintech-enabled digital transformation contributes to sustainable value creation under different institutional conditions. Based on these perspectives, this study proposes that fintech not only directly improves bank financial performance but also indirectly influences financial outcomes through green finance and ESG performance, while ownership status moderates the effectiveness of fintech adoption across banks. Therefore, this study develops a more coherent analytical framework integrating digital transformation, sustainability-oriented activities, and institutional heterogeneity to explain the financial performance of Chinese listed commercial banks.

3. Theoretical Foundation and Conceptual Framework

3.1. Theoretical Foundation

Dynamic Capability Theory:

Dynamic Capability Theory explains how organisations integrate, reconfigure, and deploy internal and external resources to respond to rapidly changing business environments [66]. In the banking industry, fintech adoption reflects banks’ ability to develop dynamic capabilities, as digital technologies enable them to improve operational efficiency, optimize resource allocation, strengthen risk management, and enhance service innovation. Through technologies such as artificial intelligence, big data analytics, blockchain, and cloud computing, banks can continuously adapt to market competition and changing customer demands. Recent banking studies further suggest that dynamic capabilities play an important role in facilitating digital transformation and improving banking performance through fintech-enabled innovation and organisational adaptability [70].

From this perspective, fintech enables commercial banks to develop flexible organisational capabilities that improve transaction efficiency, reduce operational costs, and expand digital financial services, thereby enhancing financial performance. Dynamic Capability Theory, therefore, provides an important theoretical explanation for the direct relationship between fintech and financial performance in Chinese listed commercial banks.

Dynamic Natural Resource-Based View:

The Dynamic Natural Resource-Based View (Dynamic NRBV) extends the traditional resource-based view by emphasizing environmentally sustainable capabilities as strategic resources for achieving long-term competitive advantage [67]. This theory argues that firms can create sustainable value by integrating environmental responsibility, green finance, and sustainability-oriented capabilities into organisational strategies and operational processes. In the banking sector, fintech-enabled digital capabilities facilitate sustainability-oriented banking activities by improving banks’ ability to support green finance and strengthen ESG performance. Technologies such as artificial intelligence, big data analytics, and cloud computing enhance environmental risk assessment, improve the efficiency of green finance allocation, increase transparency of ESG Rating, and support environmentally responsible investment decisions. Recent studies further suggest that sustainability-oriented dynamic capabilities play an important role in promoting long-term organisational resilience and sustainable financial performance [71,72].

From this perspective, fintech not only improves operational efficiency but also strengthens banks’ sustainability capabilities by advancing green finance and ESG-related practices. These sustainability-oriented activities may generate long-term financial benefits through enhanced reputation, lower environmental risk exposure, improved stakeholder trust, and stronger operational sustainability. Therefore, Dynamic NRBV provides the theoretical foundation explaining how green finance and ESG performance mediate the relationship between fintech and financial performance in Chinese listed commercial banks.

Institutional Theory:

Institutional Theory contends that organisational priorities, decisions, and flexibility are shaped by institutional environments, governance structures, regulatory pressures, and social expectations [69]. Organisations in different institutional settings display varied strategic responses and performance, driven by these incentives. In China’s banking sector, ownership defines institutional heterogeneity and directly determines banks’ strategic objectives, governance, and operational priorities. State-owned banks are more guided by policy and regulation, while non-state-owned banks are market- and profit-driven. These contrasts shape how each bank type approaches fintech strategies and converts digital investments into financial returns. Prior research demonstrates that such institutional and governance variations are central in determining how successfully banks digitally transform and improve performance [65].

Institutional Theory explains how ownership status moderates the relationship between fintech and financial performance. It shows that fintech’s effectiveness differs between state-owned and non-state-owned banks because of distinct institutional incentives, organisational flexibility, governance, and strategic responsiveness.

3.2. Conceptual Framework

This study develops a conceptual framework, shown in Figure 1, that links fintech, green finance, ESG performance, ownership status, and financial performance in Chinese listed commercial banks. The framework posits that adopting fintech directly improves financial performance by enhancing operational efficiency, digital capabilities, and resource allocation.

Fintech may also influence financial performance indirectly through sustainability. Specifically, fintech-driven digital transformation supports green finance and boosts ESG performance, thereby promoting sustainable financial outcomes. Ownership status is included as an institutional factor moderating the effect of fintech across banks. The framework merges digital transformation, sustainability activities, and institutional differences to explain financial performance in Chinese listed commercial banks.

4. Methodology

4.1. Sample and Data Collection

This study uses panel data from 42 Chinese A-share-listed commercial banks between 2015 and 2024. Data were collected from multiple authoritative sources, including the CSMAR Database, Sino-Securities ESG Database, annual reports, and the Shanghai and Shenzhen Stock Exchanges. Sample banks were selected based on the availability and reliability of their financial and non-financial disclosures.

During the data merging and cleaning process, observations with missing financial indicators, unavailable ESG disclosures, incomplete fintech-related textual disclosures, and duplicated bank-year records were excluded. No delisted banks were included in the sample, as only continuously listed Chinese commercial banks with complete disclosure records during the study period were retained. To mitigate the influence of extreme values, all continuous variables were winsorized at the 1st and 99th percentiles. After applying these screening procedures, the final sample consisted of 304 bank-year observations. The empirical analysis was conducted using STATA 18.0.

Table 1. Measurement of variables.

Variables		Symbol	Descriptions
Dependent Variable	Financial Performance	FP	Tobin‘Q, Net Interest Margin, Return on Asset
Independent Variable	Fintech	FT	Used text mining method to obtain the fintech value
Mediator Variable	Green Finance	GF	Green Credit Ratio
	ESG Performance	ESG	The ESG ratings of banks
Moderator Variable	Ownership Status	OS	State-owned or non-state-owned
Control Variable	Bank Size	Size	Evaluated by the logarithm of total employees
	Leverage	Lev	Total Liabilities/Total Assets
	Cash Flow	CashFlow	Sum of Cash and its Equivalents/Total Assets
	Capital Adequacy Ratio	CAR	Equity/Total Assets
	Non-Performing Loan Ratio	NPL	(Non-performing Loans/Total Gross Loans) × 100%

introduces the variable descriptions and measures; further details are provided in the following sections.

4.2. Variable Measurement

Independent Variables: Fintech

This study uses fintech as the independent variable, measured by a text-mining-based Fintech Index constructed from the annual reports of Chinese listed commercial banks from 2015 to 2024. Following previous studies [73,74,75], the Fintech Index is developed based on the frequency of fintech-related keywords disclosed in banks’ annual reports. Annual reports are regarded as formal and audited disclosures that reflect banks’ strategic orientation, digital transformation initiatives, and engagement in fintech development. Annual reports were selected because they represent formal, audited disclosures that reflect banks’ strategic priorities, technological initiatives, and digital transformation activities, and have been widely adopted in prior fintech and corporate disclosure studies.

The keyword dictionary comprises 118 fintech-related terms, classified into six categories: artificial intelligence, blockchain, cloud computing, big data, online finance, and mobile banking. All annual reports were converted into machine-readable text format before analysis. Word segmentation and keyword matching procedures were conducted using Python 3-based textual analysis techniques. Repeated occurrences of fintech-related keywords within each annual report were aggregated to construct the annual fintech index for each bank-year observation. Consistent with prior literature, a higher frequency of fintech-related keywords indicates a stronger strategic emphasis on fintech development and digital transformation capability within the bank [76]. The textual analysis procedures, keyword classification rules, and word segmentation processes were applied consistently across all sample banks and years to enhance comparability and reproducibility. The full keyword dictionary is provided in Appendix A (

Table A1. Fintech Index Keywords.

Dimension	Keywords
Artificial Intelligence	Artificial intelligence, robotics, machine learning, deep learning, neural networks, facial recognition, biometrics, voiceprint recognition, pattern recognition, image recognition, facial payment, virtual reality, augmented reality, knowledge graph, intelligence, smart deposits, smart counters, smart finance, smart branches, smart credit, smart, smart banking, smart marketing, smart home automation, intelligent, smart risk control, automation, natural language processing
Blockchain	Blockchain, distributed ledger, supply chain, internet of things, near-field, quantum, quantum message, quantum communication, data encryption, digital currency, electronic currency
Cloud Computing	Distributed, distributed storage, distributed computing, distributed architecture, distributed database, financial cloud, trusted computing, cloud computing, private cloud, virtualization, privacy computing, cloud, cloud services, cloud service platform, cloudification, cloud architecture, cloud platform, cloud system, cloud disaster recovery
Big Data	Big data, big data analysis, big data services, big data technology, big data models, big data mining, data warehouse, data technology, data model, data mining, data governance, data centre, digitalization, digital transformation, digital finance, digital signature, digital ecology, digital credit card, digital banking, digital marketing
Online	E-commerce, electronification, electronic finance, electronic channels, e-commerce, electronic banking, electronic payment, internet, internet finance, financial technology, FinTech, digital technology, networking, online internet finance, online financial management, online financing, online consumer loans, online internet banking, online internet payment, online transactions, online banking, online payment, online banking
Mobile Banking	Scenario-based, scenario-based finance, program interface, open platform, open banking, platform-based, software development kit, mobile banking, mobile payment, barcode payment, mobile e-commerce, mobile internet, mobile finance, mobile banking, mobile payment, digital payment, application programming interface, direct banking
Total Number of Keywords	118

).

To improve construct validity, the keyword dictionary primarily focuses on specific fintech technologies and operational applications rather than relying solely on broad digitalization terminology. This focus improves measurement reliability and reduces potential bias from symbolic disclosure or marketing-oriented narratives. The Fintech Index in this paper serves as a systematic, widely adopted proxy for banks’ fintech development orientation and their disclosed technological engagement. Robustness analyses, including lagged fintech, alternative DV, and use endogeneity test (Instrumental Variable) to address potential endogeneity and reverse causality concerns. This text-mining approach has been widely adopted in prior studies on fintech and digital transformation in China due to the limited availability of standardised bank-level fintech investment data.

Dependent Variables: Financial Performance

This study adopts multi-dimensional indicators to measure financial performance of banks, following previous studies [32,77,78]. Specifically, Tobin’ s Q reflects market-based corporate value, while net interest margin captures core operational profitability from traditional lending businesses. Return on assets is used to evaluate overall asset profitability. These three mainstream indicators can comprehensively reflect both short-term operational outcomes and long-term market valuation of listed commercial banks. All financial performance from the CSMAR database, this approach better reflects market expectations and firm valuation in China.

Mediator Variables: Green Finance

Green finance refers to financial activities that support environmentally sustainable development, including green credit, green bonds, green leasing, green wealth management products, and other low-carbon financial services. In the banking sector, green finance promotes environmentally responsible investment and facilitates sustainable economic transformation.

However, due to the limited availability and consistency of publicly disclosed bank-level green finance data in China, this study uses the Green Credit Ratio (GCR) as a proxy for green finance. The variable is constructed using manually collected data from the CSMAR database covering the period from 2015 to 2024. Specifically, the Green Credit Ratio is measured as the proportion of the outstanding green credit balance relative to each bank’s total loan portfolio. Green credit balance refers to the outstanding loans allocated to environmentally sustainable projects, including energy conservation, pollution control, clean energy, green transportation, and other environmentally related industries.

This measurement approach has been widely adopted in prior studies on Chinese commercial banks because green credit represents the most observable, institutionally significant, and policy-oriented component of green finance activities within China’s banking system. Dividing green credit balance by total loans helps standardize the measure across banks of different sizes and better captures the extent to which banks allocate lending resources toward environmentally sustainable activities. Compared with other green financial products, green credit data are more consistently disclosed across listed banks and provide relatively reliable comparability for empirical analysis. Therefore, the Green Credit Ratio is considered an appropriate proxy for evaluating banks’ engagement in green finance in this study. A higher value indicates stronger alignment with sustainability objectives and environmental responsibility. Previous studies have similarly employed the Green Credit Ratio to evaluate green finance engagement and risk mitigation among Chinese commercial banks [79,80].

Mediator Variables: ESG Performance

ESG performance is measured using ESG ratings of major banks. The data are downloaded from the Sino-Securities Data Vault. The rating comprises three pillars: Environmental (E), which evaluates banks’ impacts on climate change, emissions, environmental policies, resource usage, and clean energy initiatives; Social (S), which assesses labour practices, customer relations, community engagement, social inclusion, and human rights; Governance (G), which covers board structure, risk management, transparency, compliance and internal controls. Some studies used ESG ratings from the Sino-Securities database to examine how ESG performance influences bank financial performance [81,82]. This measurement is more popular in the field of accounting research.

Moderator Variables: Ownership Status

Ownership status is checked directly on the Shanghai Stock Exchange. Ownership status is captured by a dummy variable, SOE, which equals 1 if the bank is a state-owned enterprise and 0 otherwise. This variable interacts with fintech adoption to examine whether the effect of fintech on bank performance differs between state-owned and non-state-owned banks. This classification has been widely applied in recent studies, with ownership type serving as a moderator in bank governance to analyse the effects of ownership on financial performance [17,83].

Control Variables:

As mentioned earlier, this study incorporates several control variables to account for firm-specific characteristics that may influence financial performance. Following prior studies, the selected control variables (see Table 1) include bank size (Size), leverage (Lev), liquidity (Liquid), Cash Flow (CashFlow), Capital Adequacy Ratio (CAR), and Non-Performing Loan Ratio (NPL), and all these measurements are download from CSMAR database.

Finance and ESG-related factors that may impact financial performance. Controlling for these variables helps isolate the net effect of the main explanatory variables and reduces potential omitted variable bias. Moreover, incorporating multiple dimensions of firm characteristics enhances the robustness and comparability of the empirical analysis across firms, thereby improving the reliability and validity of the study’s findings.

5. Results

5.1. Descriptive Statistics

In this section,

Table 2. Descriptive Statistics.

Variable	N	Min	Max	Mean	SD	25%	50%	75%
TQ	304	0.938	1.040	0.995	0.0240	0.979	0.995	1.009
NIM	304	1.514	4.100	2.473	0.580	2.095	2.345	2.725
ROA	304	0.420	1.319	0.849	0.187	0.727	0.838	0.978
fintech	304	3.838	5.282	4.733	0.314	4.551	4.770	4.967
greenfinance	304	0	0.238	0.0940	0.0600	0.0470	0.0860	0.134
ESG	304	0.672	0.929	0.843	0.0630	0.799	0.840	0.909
SOE	304	0	1	0.651	0.477	0	1	1
Size	304	25.47	31.33	28.39	1.639	27.01	28.55	29.62
Lev	304	0.903	0.947	0.925	0.0100	0.918	0.924	0.933
CashFlow	304	−0.0750	0.138	0.0150	0.0380	−0.00400	0.0140	0.0330
CAR	304	10.63	18.42	13.65	1.665	12.51	13.52	14.57

presents the descriptive statistics for all variables used in this study. Following the data screening and cleaning procedures described in Section 4.1, the final balanced panel consisted of 304 bank-year observations. To reduce the influence of extreme values, key continuous variables were minorized at the 1st and 99th percentiles.

Regarding financial performance, Tobin’s Q (TQ) has a mean of 0.995 and a relatively small standard deviation of 0.024, indicating limited variation in market valuation among the sampled banks. Net Interest Margin (NIM) has a mean value of 2.473 and a standard deviation of 0.580, while Return on Assets (ROA) records a mean of 0.849 and a standard deviation of 0.187, suggesting moderate differences in profitability and operational efficiency across banks. The independent variable, fintech, shows a mean value of 4.733 and a standard deviation of 0.314, indicating a relatively homogeneous level of fintech adoption among Chinese listed commercial banks.

Among the mediating variables, Green Finance has a mean of 0.094 and a standard deviation of 0.060, indicating substantial variation in green finance development across banks. The minimum value of zero further suggests that certain banks had not yet established meaningful green finance activities during the sample period. ESG Performance (ESG) exhibits a relatively high mean value of 0.843 and a standard deviation of 0.063, indicating generally strong ESG performance among the sampled banks.

For the moderating variable, ownership status (SOE) has a mean value of 0.651, implying that approximately 65.1% of the sample banks are state-owned or government-controlled institutions. The control variables, including bank size (Size), leverage (Lev), cash flow (CashFlow), and capital adequacy ratio (CAR), display reasonable variation and are generally consistent with the operational and financial characteristics of Chinese listed commercial banks. Overall, the descriptive statistics suggest sufficient cross-sectional variation in the main variables and provide a sound basis for the subsequent regression analyses.

5.2. Correlation Analysis and Multicollinearity Test

Table 3. Correlation Analysis.

	TQ	NIM	ROA	Fintech	Greenf~e	ESG	SOE
TQ	1
NIM	0.434 ***	1
ROA	0.293 ***	0.277 ***	1
fintech	−0.139 **	0.206 ***	−0.103 *	1
greenfinance	0.0410	0.335 ***	0.400 ***	0.358 ***	1
ESG	0.135 **	0.504 ***	0.255 ***	0.377 ***	0.588 ***	1
SOE	−0.604 ***	−0.233 ***	−0.0420	0.096 *	0.0120	−0.0670	1
Size	−0.169 ***	0.169 ***	0.124 **	0.527 ***	0.501 ***	0.401 ***	0.290 ***
Lev	0.206 ***	−0.264 ***	−0.0360	−0.282 ***	−0.155 ***	−0.172 ***	0.0760
CashFlow	0.0280	0.0900	0.136 **	−0.116 **	0.0190	0.0690	−0.0270
CAR	−0.0790	0.111 *	0.125 **	0.212 ***	0.350 ***	0.228 ***	−0.0530
NPL	−0.0720	−0.0150	−0.252 ***	−0.0290	−0.182 ***	−0.221 ***	0.162 ***
	Size	Lev	CashFlow	CAR	NPL
Size	1
Lev	−0.0350	1
CashFlow	−0.0450	−0.0510	1
CAR	0.103 *	−0.524 ***	0.125 **	1
NPL	0.174 ***	−0.0290	−0.0810	−0.288 ***	1

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

presents the Pearson correlation coefficients among the key variables. Consistent with theoretical expectations, the focal independent variable, fintech, exhibits statistically significant but moderately correlated with the three measures of bank financial performance. Specifically, fintech is negatively and significantly associated with TQ (r = −0.139, p < 0.05) and ROA (r = −0.103, p < 0.10), while it shows a positive correlation with NIM (r = 0.206, p < 0.001). These preliminary bivariate results suggest that fintech’s impact on bank performance is sensitive to performance metrics, warranting further identification through multivariate regressions to control for confounders.

Regarding the mediating variables, Green Finance (greenfinance) and ESG performance demonstrate strong positive correlations with both fintech and various performance indicators. Green finance is positively correlated with fintech (r = 0.400, p < 0.001), NIM (r = 0.335, p < 0.001), and ESG (r = 0.588, p < 0.001), highlighting the synergistic relationship between fintech development and green initiatives. Similarly, ESG displays significant positive correlations with NIM (r = 0.504, p < 0.001) and ROA (r = 0.255, p < 0.001), implying that superior ESG performance is conducive to enhancing bank profitability. These high correlations also hint at potential mediation pathways through which fintech may influence performance via green finance and ESG channels.

For the moderating variable, Ownership (SOE), the results reveal a significant negative correlation with TQ (r = −0.604, p < 0.001) and a positive correlation with NIM (r = −0.233, p < 0.001), yet no significant association with ROA. This indicates that state ownership exerts a distinct influence on different dimensions of bank performance, reflecting the unique institutional context of the Chinese financial sector.

Among the control variables, Bank Size (Size) shows significant positive correlations with most performance measures and mediating variables, confirming the scale economy effect in banking. Leverage (Lev) is negatively associated with TQ but positively with NIM, while Capital Adequacy Ratio (CAR) displays a positive relationship with ESG and ROA. Notably, the correlation coefficients between independent variables are generally below 0.60, suggesting a low risk of severe multicollinearity. Nevertheless, variance inflation factor (VIF) tests are still recommended in the subsequent regression analysis to ensure the reliability of empirical results.

The variance inflation factor (VIF) test was conducted to assess potential multicollinearity among the explanatory variables. As shown in

Table 4. Multicollinearity.

Variable	VIF	1/VIF
CAR	1.650	0.608
Lev	1.460	0.686
NPL	1.210	0.828
Size	1.070	0.937
CashFlow	1.020	0.980
Mean	1.280	0.78

, the VIF values for all variables range from 1.020 to 1.650, with a mean VIF of 1.280. All individual VIF values are well below the conventional threshold of 5, indicating that multicollinearity is not a concern in the current model specification. The corresponding tolerance values (1/VIF) are all greater than 0.60, further confirming that no variable exhibits significant linear dependence on other regressors. These results validate the reliability of subsequent regression estimates and rule out multicollinearity as a confounding factor for the core coefficients of interest.

5.3. Regression Analysis

Table 5. Baseline Regression.

	(1)	(2)	(3)
Variables	TQ	NIM	ROA
fintech	0.017 **	0.487 ***	0.088 **
	(2.57)	(2.91)	(2.55)
Size	0.006	−0.723 **	0.455 ***
	(0.46)	(−2.32)	(7.53)
Lev	0.113	−11.770 **	−1.890 **
	(0.63)	(−2.13)	(−2.04)
CashFlow	−0.055 **	0.790	−0.024
	(−2.12)	(0.97)	(−0.16)
CAR	0.003 ***	0.056 **	−0.014 **
	(3.55)	(2.16)	(−2.39)
NPL	−0.010 *	−0.122	−0.269 ***
	(−1.85)	(−0.82)	(−7.64)
Constant	0.612	30.977 ***	−10.191 ***
	(1.53)	(3.10)	(−5.39)
Observations	304	304	304
R-squared	0.760	0.645	0.875
ID FE	YES	YES	YES
Year FE	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

reports the baseline fixed-effects regression results examining the impact of fintech on financial performance of commercial banks, measured by Tobin’s Q (TQ), net interest margin (NIM), and return on assets (ROA). All models include bank and year fixed effects, with robust standard errors. Consistent with theoretical expectations, fintech adoption exerts a positive and statistically significant effect on all three-performance metrics. Fintech is positively associated with TQ (β = 0.017, p < 0.05), indicating enhanced market valuation. The effect on NIM is the strongest (β = 0.487, p < 0.01), suggesting fintech significantly improves the banks’ interest-generating capability. Fintech also boosts ROA (β = 0.088, p < 0.05), reflecting improved operational profitability.

This regression models exhibit strong explanatory power, with R-squared values ranging from 0.645 to 0.875. Control variables behave as expected: bank size (Size) positively impacts ROA but reduces NIM, while leverage (Lev) and non-performing loans (NPL) show adverse effects on NIM and ROA. These results provide robust initial evidence that fintech improves both market-based and accounting measures of bank financial performance, setting the stage for subsequent mediation and moderation analyses.

To verify the robustness of the baseline results, this study replaces the original financial performance measures with ROE and NIMW. The results are presented in

Table 6. Robustness Tests (alternative DV).

	(1)	(2)
Variables	ROE	NIMW
fintech	0.013 **	0.581 ***
	(2.27)	(3.63)
Size	0.054 ***	−0.394
	(5.80)	(−1.35)
Lev	1.031 ***	−2.189
	(7.23)	(−0.38)
CashFlow	0.012	0.949
	(0.50)	(1.20)
CAR	−0.001	0.066 **
	(−0.73)	(2.52)
NPL	−0.026 ***	0.031
	(−5.63)	(0.21)
Constant	−2.401 ***	11.854
	(−8.71)	(1.34)
Observations	304	304
R-squared	0.899	0.585
ID FE	YES	YES
Year FE	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

. The findings show that fintech remains positively associated with financial performance under alternative measurements. Specifically, fintech has a positive and significant effect on ROE (β = 0.013, p < 0.05), indicating improved shareholder profitability. Similarly, fintech significantly enhances NIMW (β = 0.581, p < 0.01), suggesting stronger interest-generating capability and intermediation efficiency.

The consistent coefficient signs and significance levels confirm that the main findings are robust to alternative proxies of financial performance. Among the control variables, Size and Lev positively affect ROE, while NPL negatively influences profitability. In addition, CAR shows a positive relationship with NIMW.

Overall, the robustness test results further support the conclusion that fintech development improves the financial performance of Chinese listed commercial banks.

The findings (

Table 7. Robustness Tests (1-year lagged fintech).

	(1)	(2)	(3)
Variables	TQ	NIM	ROA
L.fintech	0.019 **	0.374 *	0.076 **
	(2.36)	(1.70)	(2.01)
Size	0.003	−0.649 **	0.485 ***
	(0.18)	(−2.10)	(6.73)
Lev	0.010	−11.827 **	−1.070
	(0.04)	(−1.98)	(−1.06)
CashFlow	−0.055 *	1.064	0.054
	(−1.71)	(1.17)	(0.31)
CAR	0.003 ***	0.051 *	−0.018 ***
	(2.65)	(1.76)	(−2.99)
NPL	−0.012 *	−0.291	−0.272 ***
	(−1.81)	(−1.58)	(−6.67)
Constant	0.798 *	29.792 ***	−11.735 ***
	(1.69)	(2.91)	(−5.23)
Observations	265	265	265
R-squared	0.768	0.698	0.883
ID FE	YES	YES	YES
Year FE	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

) indicate that lagged fintech remains positively associated with financial performance. Specifically, L.fintech has a significantly positive effect on TQ (β = 0.019, p < 0.05) and ROA (β = 0.076, p < 0.05). In addition, the coefficient on NIM is positive and marginally significant (β = 0.374, p < 0.10). These results suggest that the positive impact of fintech on financial performance of banks may persist over time.

To address potential reverse causality, we use the one-year lagged fintech index as the core independent variable. The results confirm the positive effect of fintech on performance, although the statistical significance for NIM declines from p < 0.01 to p < 0.10. This attenuation is expected, as the contemporaneous relationship between fintech adoption and interest income generation is inherently stronger than the lagged effect, while the reduction in sample size also contributes to higher standard errors.

Overall, the consistency in coefficient signs and significance levels supports the robustness of the main findings and further confirms that fintech development has a positive contribution to the financial performance of Chinese listed commercial banks.

To mitigate the potential influence of extreme outliers, which might distort coefficient estimates and inflate standard errors in linear regression models, all continuous variables are winsorized at the 5th and 95th percentiles before re-estimating the baseline model. The results presented in

Table 8. Robustness Tests (winsorized at the 5th and 95th percentiles).

	(1)	(2)	(3)
Variables	TQ	NIM	ROA
fintech	0.019 ***	0.580 ***	0.115 ***
	(2.73)	(3.36)	(3.31)
Size	0.005	−0.353	0.242 ***
	(0.51)	(−1.49)	(5.38)
Lev	0.037	−13.649 **	0.137
	(0.20)	(−2.37)	(0.15)
CashFlow	−0.048	0.783	0.134
	(−1.57)	(0.87)	(0.79)
CAR	0.003 ***	0.037	−0.017 ***
	(2.85)	(1.39)	(−3.19)
NPL	−0.017 ***	−0.212	−0.332 ***
	(−2.84)	(−1.35)	(−11.05)
Constant	0.720 **	22.129 **	−6.016 ***
	(2.11)	(2.47)	(−3.73)
Observations	304	304	304
R-squared	0.763	0.642	0.875
ID FE	YES	YES	YES
Year FE	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

are fully consistent with the main findings. Across all three specifications of financial performance, the coefficient of fintech remains positive and highly statistically significant at the 1% level (β = 0.019 for TQ, β = 0.580 for NIM, β = 0.115 for ROA, all p < 0.01). The magnitude and significance of the estimated coefficients remain stable and are even strengthened relative to the baseline model, suggesting that the positive relationship between fintech adoption and bank performance is not driven by outlier observations. The models maintain strong explanatory power with consistent R-squared values (ranging from 0.642 to 0.875), and control variables behave as expected. These findings reinforce the reliability of our core conclusions and rule out outlier bias as a confounding factor.

To address potential endogeneity concerns and reverse causality between fintech adoption and bank financial performance, this study further employs an instrumental variable (IV) approach as a robustness test. Reverse causality may arise because banks with stronger financial performance are more likely to possess greater financial resources, technological capabilities, and digital infrastructure to invest in fintech development. Therefore, baseline panel regression estimations may suffer from endogeneity bias.

Drawing on prior fintech research, this study constructs an instrumental variable equal to the product of the number of post offices at the bank’s location in 1984 and the number of internet users in the previous year [84]. Historical post office data and internet user data were obtained from the China Statistical Yearbook [85]. The historical number of post offices reflects the regional communication infrastructure and early information transmission capacity, which may influence the regional foundation for subsequent digital financial development. Meanwhile, lagging internet users capture the degree of regional digitalisation and internet penetration. Therefore, the interaction term provides exogenous variation in fintech development while being less directly associated with current bank financial performance, thereby helping alleviate reverse causality concerns.

The first-stage regression results reported in

Table 9. Robustness Tests (Instrumental Variables).

	(1)	(2)	(3)	(4)
Variables	fintech	TQ	NIM	ROA
tool	0.450 ***
	(6.17)
Size	0.414 ***	−0.022	−2.263 ***	0.433 ***
	(3.59)	(−1.40)	(−3.29)	(6.66)
Lev	1.387	0.052	−15.133	−1.938 **
	(0.74)	(0.22)	(−1.51)	(−2.05)
CashFlow	−0.004	−0.050	1.067	−0.020
	(−0.01)	(−1.40)	(0.69)	(−0.14)
CAR	−0.002	0.003 **	0.033	−0.014 **
	(−0.15)	(2.09)	(0.56)	(−2.57)
NPL	−0.152 ***	0.003	0.593 *	−0.259 ***
	(−2.92)	(0.37)	(1.90)	(−8.77)
fintech		0.097 ***	4.824 ***	0.149 *
		(4.82)	(5.58)	(1.82)
Observations	304	304	304	304
R-squared	0.823	−0.651	−2.674	0.457
ID FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

show that the instrumental variable (“tool”) is positively and significantly associated with fintech adoption (β = 0.450, p < 0.01), indicating that the instrument has strong explanatory power for fintech development among Chinese listed commercial banks.

The second-stage regression results further demonstrate that fintech remains positively associated with financial performance after controlling for endogeneity concerns. Specifically, fintech has a significantly positive effect on Tobin’s Q (β = 0.097, p < 0.01), Net Interest Margin (β = 4.824, p < 0.01), and Return on Assets (β = 0.149, p < 0.1). These findings are generally consistent with the baseline regression results, suggesting that the positive relationship between fintech adoption and bank financial performance remains robust after addressing potential reverse causality issues.

In addition,

Table 10. Robustness Tests (F-statistic of Instrumental Variables).

Statistic	Value
First-stage F-statistic	38.121
Weak IV test	Passed

reports the weak-instrument-variable test results. The first-stage F-statistic is 38.121, which is substantially higher than the conventional threshold value of 10, indicating that the instrumental variable does not suffer from weak instrument problems. Therefore, the IV estimation results provide additional evidence supporting the robustness and reliability of the empirical findings.

Table 11. Mechanism Tests: Mediation Effect (Green Finance).

	(1)	(2)	(3)	(4)
Variables	Greenfinance	TQ	NIM	ROA
greenfinance		0.080 ***	2.962 ***	0.316 **
		(2.69)	(4.03)	(2.29)
fintech	0.022 *	0.016 **	0.423 ***	0.081 **
	(1.71)	(2.47)	(2.60)	(2.35)
Size	0.006	0.005	−0.742 **	0.453 ***
	(0.21)	(0.46)	(−2.47)	(7.71)
Lev	0.049	0.109	−11.916 **	−1.906 **
	(0.11)	(0.62)	(−2.27)	(−2.09)
CashFlow	−0.082	−0.049 *	1.034	0.002
	(−1.26)	(−1.86)	(1.28)	(0.01)
CAR	0.001	0.003 ***	0.052 **	−0.014 **
	(0.51)	(3.51)	(2.05)	(−2.41)
NPL	−0.025 **	−0.008	−0.049	−0.261 ***
	(−2.01)	(−1.53)	(−0.34)	(−7.60)
Constant	−0.218	0.629 *	31.623 ***	−10.122 ***
	(−0.25)	(1.67)	(3.31)	(−5.44)
Observations	304	304	304	304
R-squared	0.784	0.769	0.665	0.878
ID FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

presents the mediation effect analysis examining whether green finance serves as a channel through which fintech influences financial performance of banks. The stepwise regression results confirm a significant mediating pathway.

First, fintech adoption has a positive and significant effect on green finance (β = 0.022, p < 0.10), indicating that fintech facilitates banks’ engagement in green finance. Second, in the performance regressions, both fintech and green finance retain positive and significant coefficients across all three specifications. Specifically, green finance significantly improves TQ (β = 0.080, p < 0.01), NIM (β = 2.962, p < 0.01), and ROA (β = 0.316, p < 0.05), while the coefficients of fintech remain significant at conventional levels (β = 0.016 for TQ, β = 0.423 for NIM, β = 0.081 for ROA, all p < 0.05).

Therefore, these results satisfy the conditions for partial mediation: fintech promotes green finance, which in turn enhances bank performance, while the direct effect of fintech remains significant. This suggests that green finance acts as a complementary mechanism through which fintech exerts its positive influence on both market-based and accounting measures of bank performance. The findings remain robust to the inclusion of bank and year fixed effects and standard controls.

Table 12. Mechanism Tests (ESG Performance).

	(1)	(2)	(3)	(4)
Variables	ESG	TQ	NIM	ROA
ESG		0.047 ***	3.387 ***	0.192 **
		(3.00)	(7.48)	(2.00)
fintech	0.040 **	0.015 **	0.352 **	0.080 **
	(2.02)	(2.47)	(2.42)	(2.36)
Size	−0.054	0.008	−0.539 *	0.466 ***
	(−1.13)	(0.70)	(−1.88)	(8.00)
Lev	0.461	0.092	−13.332 ***	−1.979 **
	(0.57)	(0.53)	(−3.09)	(−2.18)
CashFlow	0.026	−0.057 **	0.704	−0.029
	(0.23)	(−2.18)	(0.93)	(−0.19)
CAR	0.002	0.003 ***	0.050 **	−0.014 **
	(0.51)	(3.51)	(2.12)	(−2.41)
NPL	−0.055 **	−0.008	0.065	−0.259 ***
	(−2.51)	(−1.41)	(0.49)	(−7.40)
Constant	1.820	0.526	24.813 ***	−10.540 ***
	(1.30)	(1.38)	(2.88)	(−5.75)
Observations	304	304	304	304
R-squared	0.498	0.767	0.712	0.877
ID FE	YES	YES	YES	YES
Year FE	YES	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

presents the mediation analysis examining ESG performance as a potential channel through which fintech influences financial performance.

As the results shows, fintech is found to significantly improve banks’ ESG performance (β = 0.040, p < 0.05), suggesting that digital transformation enables banks to strengthen their environmental, social, and governance practices. In addition, when both fintech and ESG performance are included in the regressions for bank performance, ESG carries positive and statistically significant coefficients across all three specifications: TQ (β = 0.047, p < 0.01), NIM (β = 3.387, p < 0.01), and ROA (β = 0.192, p < 0.05). Meanwhile, the direct effect of fintech remains positive and significant (p < 0.05 in all models). These findings confirm the presence of partial mediation: fintech improves financial performance by enhancing ESG performance. The findings remain robust to the inclusion of bank and year fixed effects and standard controls.

Table 13. Further Moderation Effect (Ownership Status).

	(1)	(2)	(3)
Variables	TQ	NIM	ROA
fintech	0.037 ***	0.770 ***	0.134 ***
	(4.36)	(3.75)	(3.31)
SOE	0.151 ***	1.881	0.215
	(3.85)	(1.56)	(1.00)
c.fintech#c.SOE	−0.039 ***	−0.571 **	−0.092 **
	(−5.27)	(−2.37)	(−2.12)
Size	0.008	−0.691 **	0.460 ***
	(0.70)	(−2.08)	(7.30)
Lev	0.208	−10.389 *	−1.666 *
	(1.17)	(−1.82)	(−1.81)
CashFlow	−0.037	1.061	0.019
	(−1.49)	(1.29)	(0.13)
CAR	0.003 ***	0.053 **	−0.014 **
	(3.23)	(1.99)	(−2.54)
NPL	−0.003	−0.017	−0.252 ***
	(−0.63)	(−0.11)	(−7.08)
Constant	0.411	28.735 ***	−10.760 ***
	(1.16)	(2.66)	(−5.49)
Observations	304	304	304
R-squared	0.790	0.656	0.878
ID FE	YES	YES	YES
Year FE	YES	YES	YES

Robust t-statistics in parentheses *** p < 0.01, ** p < 0.05, * p < 0.1.

shows how ownership status (SOE versus non-SOE) moderates the link between fintech adoption and financial performance among Chinese listed commercial banks. The results reveal clear ownership-based differences in the effectiveness of fintech adoption.

The coefficient on fintech is positive and statistically significant across all three financial performance measures: Tobin’s Q (β = 0.037, p < 0.01), Net Interest Margin (β = 0.770, p < 0.01), and Return on Assets (β = 0.134, p < 0.01). This shows fintech adoption generally enhances market valuation, operational profitability, and asset utilization in Chinese listed commercial banks. The interaction term between fintech and SOE status (fintech × SOE) is negative and significant for all specifications: β = −0.039 and p < 0.01 for TQ, β = −0.571 and p < 0.05 for NIM, and β = −0.092 and p < 0.05 for ROA. This demonstrates that fintech’s positive impact on financial performance is weaker in state-owned banks than in non-state-owned banks.

This finding aligns with earlier studies on ownership heterogeneity and banking efficiency in China. Such as, these studies show that non-state-owned banks are more market-oriented, more flexible, and more innovation-driven than state-owned banks [62,65]. State-owned banks typically focus more on policy-related responsibilities, such as supporting strategic industries, regional development, and financial stability. These obligations may limit their flexibility in turning fintech investment into immediate profit. Previous research also finds that state-owned banks operate under multilayered governance and stricter oversight, which can slow decision-making and digital transformation compared to non-state-owned banks [86].

The results are also broadly consistent with China’s institutional and banking structure. According to the China Banking and Insurance Regulatory Commission (CBIRC), large state-owned commercial banks continue to dominate policy-related lending and financial inclusion responsibilities. In contrast, joint-stock and city commercial banks exhibit stronger competitiveness in digital banking services, online lending innovation, and customer-oriented fintech applications. This institutional heterogeneity may explain why non-state-owned banks are better able to convert fintech adoption into greater improvements in market valuation and operational efficiency.

Overall, these findings show that ownership status is a key institutional factor shaping the impact of fintech adoption in the Chinese banking sector. The results also indicate that fintech value creation depends not only on technological investment but also on organisational flexibility, governance structures, and institutional incentives across ownership systems.

6. Discussions

This study explores how adopting fintech influences banks’ financial performance, measured by Tobin’s Q (TQ), net interest margin (NIM), and return on assets (ROA). It also examines how green finance and ESG performance serve as bridges in this relationship, and whether a bank’s ownership status makes a difference. Using fixed-effects models and examining results in several ways, we find that fintech consistently improves banks’ financial performance, primarily by encouraging greater green finance activity and better ESG ratings. Interestingly, these positive effects are even stronger for non-state-owned banks compared to state-owned ones. Our findings offer new perspectives on the fintech and banking conversation, providing a more detailed understanding than previous studies.

Firstly, our initial findings show that when banks use fintech, they experience clear improvements in both their market value (measured by Tobin’s Q) and accounting measures (such as NIM and ROA). The numbers support what many experts have already advised: digital technologies help banks run more efficiently, reach more customers, and manage risks more effectively [87,88,89]. However, unlike many prior studies that focus primarily on profitability or market value in isolation, this study examines three distinct performance metrics simultaneously, revealing that fintech exerts the strongest effect on NIM. This pattern suggests that fintech’s immediate benefits are most pronounced in core interest-generating activities, such as digital lending and deposit-taking, which directly expand NIM. In contrast, the effects on Tobin’s Q and ROA, while still positive and significant, are slightly weaker, likely because these metrics reflect broader firm value and overall profitability, which are subject to more complex market and operational influences. By comparing these results, which show that fintech doesn’t improve every aspect of performance equally, we find that it tends to have the strongest impact in areas where banks generate their core earnings.

Second, the mechanism tests highlight green finance and ESG performance as two previously understudied pathways through which fintech affects financial performance in listed commercial banks. The results show that fintech adoption significantly promotes both green finance engagement and ESG performance, which in turn improve bank outcomes. Although prior studies have separately examined the relationships between FinTech and green finance and between ESG and bank performance, limited research has integrated these mechanisms into a single comprehensive framework. The mediation analysis in this study reveals that fintech’s positive effect is partially mediated by these sustainability-oriented practices, suggesting that digital transformation enables banks to align financial goals with environmental and social objectives. This finding differs from earlier work that often-framed fintech and sustainability as separate or even conflicting priorities. On the contrary, the evidence here supports a complementary view: fintech can serve as an enabler of sustainable finance, allowing banks to improve both environmental and financial performance simultaneously [90,91]. This has important implications for policymakers and practitioners, as it suggests that fintech adoption can be part of a broader strategy to advance green and responsible banking.

Third, the analysis of ownership heterogeneity offers new insights into the boundary conditions of fintech’s impact. The interaction term between fintech and state-owned banks’ status is negative and significant across all models, indicating that the positive effect of fintech is significantly weaker for state-owned banks than for non-state-owned banks. This finding is consistent with the view that non-state-owned banks face stronger competitive pressures and have greater operational flexibility, allowing them to leverage fintech more effectively to improve efficiency and profitability [92]. In the context of China’s banking system, state-owned banks are not only profit-oriented financial institutions but also important policy implementation agents responsible for supporting national strategic objectives, regional development, and financial stability. These policy-oriented lending mandates and broader social responsibilities may reduce managerial flexibility and slow the commercialisation of fintech-related innovations [93]. In addition, state-owned banks are often evaluated under more complex administrative and regulatory assessment systems, which may place greater emphasis on stability, compliance, and policy fulfilment rather than short-term operational efficiency. By contrast, non-state-owned banks generally operate under stronger market competition and profit-driven incentives, enabling them to respond more flexibly to digital transformation opportunities and convert fintech investment into financial performance improvements more efficiently. While some prior studies have documented performance differences between state-owned and non-state-owned banks, few have examined how these differences shape returns from digital transformation. The results presented here thus contribute to the literature by demonstrating that ownership status is a critical moderator that determines the magnitude of fintech’s impact. This has practical implications, as it suggests that policy interventions to promote fintech adoption may need to be tailored to different ownership types to ensure equitable benefits across the banking sector.

Fourth, the rigorous identification strategy employed in the study, along with multiple robustness checks, strengthens the credibility of the findings and addresses common concerns in fintech research. The use of lagged fintech variables helps mitigate reverse causality, while winsorizing continuous variables at the 5th and 95th percentiles prevent outliers from driving the results. In addition, the instrumental variable estimation further alleviates potential endogeneity and reverse causality concerns, strengthening the causal interpretation and robustness of the findings. The consistency of the results across these alternative specifications reinforces the conclusion that fintech has a positive causal effect on bank performance. Many existing studies rely on cross-sectional data or limited robustness checks, leaving results vulnerable to endogeneity or sample selection bias. By contrast, this study’s use of panel data with fixed effects, combined with multiple robustness tests, provides a more reliable foundation for causal inference. This methodological rigour distinguishes the study from earlier work and enhances its contribution to the literature.

Overall, this study advances the literature on fintech and sustainable banking by demonstrating that fintech-driven digital transformation creates financial value not only through operational efficiency but also through sustainability-oriented mechanisms associated with green finance and ESG performance. The findings further suggest that the financial benefits of fintech are institutionally contingent rather than universally homogeneous, thereby highlighting the importance of ownership heterogeneity in understanding sustainable banking transformation within emerging economies.

7. Conclusions

7.1. Conclusions

This study examines how fintech adoption affects the financial performance of Chinese listed commercial banks from 2015 to 2024. Using Dynamic Capability Theory, Dynamic NRBV, and Institutional Theory, it constructs an integrated framework that explicitly connects fintech, green finance, ESG performance, ownership structure, and financial outcomes.

The empirical findings show that fintech adoption significantly enhances bank financial performance. Beyond this direct effect, green finance and ESG performance serve as key mediators between fintech and banks’ financial performance, indicating that fintech enables sustainable financial value creation through sustainability mechanisms. Additionally, ownership status strongly moderates the link between fintech and financial outcome, highlighting that the impact of fintech adoption varies between state-owned and non-state-owned banks across institutional contexts. Moreover, by employing an instrumental variable approach, this study mitigates potential reverse causality and endogeneity concerns between fintech adoption and financial performance, thereby strengthening the causal interpretation and the empirical robustness of the findings. Overall, the findings suggest that fintech-driven digital transformation contributes to sustainable banking performance, both directly and indirectly through green finance and ESG-related mechanisms, across different institutional conditions.

7.2. Theoretical and Practical Implications

This study makes three main theoretical contributions to fintech, sustainable banking, and digital transformation literature. First, it demonstrates that fintech creates financial value not only by improving operational efficiency and digital capability, but also by enabling sustainability-oriented mechanisms. The findings show that fintech positively influences financial performance through green finance and ESG performance, highlighting its indirect impact via sustainable financial activities and ESG-related governance practices. This finding advances our understanding of fintech’s role in fostering sustainable banking development in emerging economies. Second, the study advances theoretical integration of fintech and sustainability by explicitly combining Dynamic Capability Theory, Dynamic NRBV, and Institutional Theory into a cohesive analytical framework. By highlighting the critical roles of sustainability capabilities and institutional heterogeneity, the research moves beyond the direct fintech–performance links explored by prior studies to comprehensively explain fintech-driven value creation. Third, the moderating effect of ownership status demonstrates that the effectiveness of fintech adoption depends on institutional and governance conditions. The findings indicate that non-state-owned banks are more capable of translating fintech investment into improvements in financial performance than state-owned banks, thereby enriching the literature on institutional heterogeneity in China’s banking sector.

From a practical perspective, the findings suggest that Chinese commercial banks should not regard fintech merely as a technological investment strategy, but as an important driver of sustainable banking transformation. Banks should strengthen the integration of fintech development, green finance activities, and ESG to achieve long-term financial sustainability.

From a policy perspective, the findings suggest that financial regulators should adopt differentiated fintech development strategies that account for institutional heterogeneity across banks. As research findings show, non-state-owned banks appear to translate fintech investment into financial performance more effectively, regulators may encourage greater flexibility in innovation and digital experimentation among state-owned banks while maintaining financial stability objectives. In addition, policymakers may strengthen regulatory support for the development of green finance, the quality of ESG disclosure, and digital financial infrastructure to enhance the sustainability-oriented transformation of China’s banking sector.

7.3. Limitations

Despite its contributions, this study presents several limitations. First, the measurement of fintech index relies on text-mining-based proxy indicators derived from annual reports rather than direct bank-level digital investment data, such as IT expenditure, fintech investment intensity, or digital infrastructure spending. Although this approach has been widely adopted in prior studies, it may not fully capture banks’ actual technological development levels. In addition, although some Chinese listed banks disclose technology-related expenditure, disclosure standards, reporting consistency, and data availability remain highly heterogeneous across banks and years, limiting comparability.

Second, this study uses the Green Credit Ratio as a proxy for green finance due to the limited availability and consistency of publicly disclosed bank-level green finance data in China. However, green finance encompasses a broader set of financial activities, including green bonds, green leasing, sustainable investment products, and various categories of green credit activities, which were not incorporated into the present analysis.

Third, the sample is limited to Chinese listed commercial banks between 2015 and 2024, which may restrict the generalisability of the findings across different institutional environments and economic cycles. In addition, the study does not fully explore heterogeneous effects across regions, bank sizes, and levels of regional marketisation.

7.4. Future Research Directions

Future research may employ more granular bank-level fintech indicators and broader green finance measurements to improve the assessment of sustainable banking activities. Subsequent studies may also investigate regional heterogeneity, bank-size heterogeneity, and segmented green credit categories to better understand the boundary conditions of fintech-driven sustainable performance. Furthermore, future research could incorporate additional mediating mechanisms, such as operational efficiency, risk-taking behaviour, and digital governance quality, or conduct cross-country comparative studies to examine how different institutional settings shape the fintech–sustainability–performance nexus. Importantly, the improvements in banking disclosure standards may facilitate the availability of more direct fintech-related indicators and enhance the precision and comparability of fintech measurement in future studies.