The impact of FinTech firms on bank financial stability

FinTech firms have received increasing attention in recent years owing to their rapid development and expansion across economies. The rise of FinTech firms has been welcomed by numerous observers who believe that new technological innovations in the financial industry have great potential to transform financial services by offering less expensive transactions, thus making these services more convenient and secure (e.g., [7, 8, 20]. Similarly, Philippon [39] argued that FinTech firms in the financial sector offer digital innovations and technology-enabled business model inventions, which significantly contribute in improving financial services to the wider communities.

An important model of FinTech innovations in the recent period is P2P lending, which has attracted great attention owing to its rapid expansion in emerging markets, such as China. The P2P lending transaction volume in China is the highest in the world, reaching nearly $550 billion in 2017. Importantly, the market diffusion rate is also ranked highest in China. Further, Huang [23] argued that the main drivers of P2P lending in China are a large supply of funds, greater market dispersion rate, and increasing demand for financial products.

Most importantly, P2P lending replaces the traditional banking system through an electronic marketplace that enables the brokerage of consumer loans between lenders and borrowers [24, 31]. However, notably, P2P lending in China is continuously evolving in a relatively underdeveloped legal and regulatory environment. Huang [23] and Milne and Parboteeah [32]² documented that the P2P lending platform does not merely work as an intermediary to pull funds from retail investors and lend money to individual borrowers and small and medium-sized enterprises (SMEs); instead, it offers other value-added services, including checking the solvency of borrowers and loan ratings, managing payments, and providing investment advice to clients. Similarly, Tang [45] reported that U.S. P2P lending is a substitute for bank lending but complements small loans. Likewise, Fuster et al. [18] reported that FinTech firms in the U.S. offer more efficient mortgage lending services than other lenders—irrespective of clients’ level of financial access.

Despite increasing attention among scholars and practitioners of FinTech firms, the empirical literature on this research topic is scant. A recent study by Chen et al. [8] investigated the value of FinTech innovations using patent-filling data from 2003–2017. The authors used machine learning to identify and classify innovations in their underlying technologies. Their analysis showed that most FinTech innovations offer substantial value to innovators. Specifically, the authors stated that the Internet of Things, robo-advising, and blockchain are the most valuable innovations for the overall financial sector. Further, the authors highlighted that financial industries can avoid the negative impact of innovations by investing heavily in their own innovations.

Lee et al. [28] investigated whether the development of the FinTech industry influenced cost efficiency and technological adoption in the Chinese banking industry during the 2003–2017 period. Their evidence confirmed that Chinese state-owned banks not only operate with a less efficient technology but also have the lowest cost efficiency. However, the authors reported that FinTech development not only enhances the use of technology by the banks but also improves their cost efficiency. This dual benefit is more clearly observed in instances of market-supported service innovations.

In another study, Li et al. [30] examined the risk spillover between FinTech companies and conventional financial institutions during a period of rapid technological advancement. Using the U.S. financial and FinTech firms’ stock returns and the Granger causality framework, the authors investigated pairwise risk spillovers across quantiles. The main findings from the study indicated that FinTech firms’ risk spillover to financial institutions positively correlated with an increase in the systematic risk of financial institutions. Additionally, using a sample of 41 banks and FinTech firms in Indonesia, Phan et al. [36] examined whether the growth of FinTech firms negatively influences banking performance. Their main results demonstrated that FinTech firms’ growth negatively influences banks’ performance.

However, in a recent study, Sheng [44] explored the impact of FinTech firms on bank lending to SMEs in China. Using provincial Chinese banks’ lending data for the period from 2011 to 2018, the author confirmed that FinTech firms have significantly contributed in facilitating banking sector credit to SMEs. Noteworthily, FinTech firms’ influence on banking lending to SMEs is much stronger for large banks than for small banks. Evidently, although the literature on FinTech has been growing recently, it is still sparse.³

The literature clearly shows a strong relationship between the growth of FinTech firms and the banking sector’s performance. For instance, Li et al. [29] reported a positive effect of FinTech firm growth on the U.S. banks’ share prices. From the emerging markets’ perspective, Phan et al. [36] demonstrated that FinTech firms’ growth reduced Indonesian banks’ performance, whereas Lee et al. [28], Sheng [44], Chen et al. [9], and Hu et al. [22] reported that FinTech firms’ development in China increases bank efficiency and reduces risk-taking, internal cash flow, and credit supply to SMEs. The findings of this strand of literature are mixed, which can be attributed to the differences in study contexts and topics. The other issue of whether FinTech firms increase or reduce bank financial stability remains unexplored. For example, some scholars (e.g., [3, 4] have theoretically argued that increasing competition may increase or decrease banks’ financial stability. Given this argument, we posit that increasing competition in the financial sector—due to FinTech firms’ rapid growth—may lower the banks’ profits from lending, thereby adversely affecting their share prices. Consequently, banks are forced to make risky investments, which will eventually lead to a reduction in their financial stability. Alternatively, due to the increasing presence of FinTech firms in the financial system, banks may be forced to adopt FinTech services in their banking business, which may eventually help them operate efficiently and maintain their customer base and revenues, thereby maintaining their financial stability. Given this backdrop, determining FinTech’s impact on bank financial stability is not possible without empirical investigation. Therefore, we empirically determine the relationship between FinTech firms and bank stability in the context of an emerging market—Malaysia, which is an ideal research context for this study for the following reasons: Malaysia is one of the fastest growing FinTech markets⁴ (see Fig. 1). Further, it is a prominent country with a dual-banking system that comprises Islamic as well as conventional banks [47]. Given the above arguments, we propose the following hypothesis:

Our sample comprises 26 Islamic and conventional banks and 301 bank-year observations. We use 2003–2018 as our sample period because prior to 2003, FinTech firms in Malaysia were extremely limited (see Fig. 1). We use the Bankscope and Fitch Connect databases for bank-level financial data. We manually collect data on FinTech firms from FinTech Malaysia.⁵ Further, we collect bank-level board governance data from the annual reports of the respective banks. We use the World Development Indicator database for industry- and country-level variables.

Our primary measure of financial stability is the Z-score, with a higher Z-score indicating higher financial stability [27]. The Z-score is defined as follows: \({\text{Z - score}}_{{i,t}} = \left[ {{\text{ROA}}_{{i,t}} + {\text{CAR}}_{{i,t}} /{\text{SDROA}}_{{i,t}} } \right]\), where ROA is the return on assets, CAR is the capital-to-asset ratio, and SDROA is the standard deviation of ROA for bank \(i\) at time\(t\). The Z-score shows the number of standard deviations that a bank’s ROA would have to fall below its expected value to deplete equity and render the bank insolvent. Thus, a higher Z-score indicates greater financial stability. Following Fang et al. [16] and Safiullah [43], we further construct a relative measure of financial stability (hereafter, the RZ-score) using the stochastic frontier analysis technique. We estimate relative financial stability or the RZ-score using the stochastic frontier model, as in Fang et al. [16]. The stochastic stability frontier model and translog specification for the stochastic stability frontier model are specified as follows:where \(\ln Z - score_{it}\) is the logarithm of the Z-score for bank i in year t, \({y}_{g}\) refers to \(g\) th output, and \({w}_{m}\) is \(m\) th input price. To estimate the stability frontier model in Eq. (1) with this translog form, standard symmetric restrictions are applied to the second–order parameters or translog portion of the model as \({\gamma }_{gh}={\gamma }_{hg}\) and \({\delta }_{mn}={\delta }_{nm}\). Furthermore, to ensure price homogeneity, following Fang et al. [16], the Z-score, price of deposits \(({w}_{1})\), and price of physical capital \(({w}_{2})\) are normalized by the price of labor (\({w}_{3}\)). \(\alpha\),\(\beta\),\(\gamma\),\(\delta\), and \(\theta\) are parameters to be estimated; \({v}_{it}\) contains the two–sided error \(v\sim\) N (0,\({\sigma }_{v}^{j2})\), capturing the effects of random error or statistical noise, which are independent of \({u}_{it}\). Finally, \({u}_{it}\) is a non–negative random variable, representing stability inefficiency. Relative stability is estimated by the following conditional expectation: \({SE}_{it}=E({e}^{{-u}_{it}}|{\varepsilon }_{it}\)). We estimate the translog stability function using the maximum likelihood estimator (MLE). The MLE results are obtained with iterative procedures that comprise an estimation of OLS regression and a two-phase grid search for gamma,\(\gamma\) (for details, see Coelli [10]). We define output variables and input prices as follows: total loans—total amount of customer loans; other earning assets—other earning assets comprising loans and advances to banks, other securities, derivatives (if any), and other investments; non-interest income—non-interest income comprising net gains (losses) on trading and derivatives, net gains (losses) on other securities, net insurance income, net fees and commissions, and other operating income; the price of deposits—the ratio of interest expenses to total deposits; the price of physical capital—the ratio of non-interest expenses (operating expenses minus personnel expenses) to total fixed assets; and the price of labor—the ratio of personnel expenses to total assets.

As suggested by Thakor [46], we use the following two proxies for FinTech firms, which are our variables of interest: the total number of all FinTech firms and number of FinTech firms closely related to banking services. Our second measure of FinTech addresses the possible concern that FinTech firms that are closely related to banking services may have a differential impact on bank financial stability than other FinTech firms.

Finally, we use control variables following the banking literature on the determinants of financial stability. Specifically, we use bank size measured as the logarithm of total assets. Further, bank age represents the number of years from the year of establishment to the end of 2018; the non-performing loans (NPL) ratio is the ratio of NPL to gross loans; equity capital ratio is the ratio of equity capital to total assets; assets growth is the annual growth rate of total assets; income diversity is the ratio of non-interest income to total operating income; ROA is the return on assets; bank concentration ratio is the assets of the three largest banks as a percentage of total banking industry assets; GDP growth rate is the annual growth rate of per capita GDP in percentage and bank-level corporate governance score; the governance score is the average of eight bank-level board of directors’ attributes—namely, board size, board independence, CEO duality, board busyness, board members’ financial expertise, audit committee size, audit committee chairman independence, and risk management committee size. The board governance score is constructed following the governance literature (e.g., [1]).

We begin our preliminary investigation by presenting descriptive statistics for the variables in Table 1. The statistics suggest that the Z-score ranges from − 0.20 to 2.66 in the sample, which implies that financial stability varies significantly among the selected sample banks. Consistently, the relative financial stability measure (RZ-score) differs noticeably in the sample. Table 1 further suggests that FinTech measures exhibit higher standard deviations, as the number of these firms varies considerably across the sample period. The mean bank age was approximately 25 years, with the lowest being 10 years. The average NPL is nearly 5% of the total loans, with the highest at 73%. Another important indicator is asset growth, with an average of 11%. Notably, non-interest income (income diversity) is nearly 20% on average, which is a good sign indicating that banks are able to generate income from sources other than traditional banking services. However, it is equally important to note that the concentration of banks is extremely high in Malaysia, as only three banks represent over two-thirds of the entire banking system in the country. Overall, these descriptive statistics suggest significant variation in the sample observations.

A pairwise correlation matrix is presented in Appendix A1. The correlation matrix shows a positive correlation between financial stability (Z-score) and FinTech, indicating that the development of FinTech firms contributes to higher bank financial stability. The highest correlation between ROA and financial stability is 0.28. Therefore, multicollinearity is not a problem in this estimation. Finally, for the control variables, income diversity and bank-level governance exhibit a statistically significant positive correlation with financial stability, while bank concentration exhibits a significant negative correlation with financial stability. These significant associations between financial stability and the other control variables suggest the need to control for these variables in our regression model. Notably, none of the pairwise correlations among independent variables exceed 0.44 (between bank size and equity capital ratio), and all variance inflation factor values remain below 10 (untabulated). Thus, multicollinearity is unlikely to be a major concern in our study context.

The empirical results are presented in Table 2. The first three columns explain the contemporaneous effect of FinTech firms on bank stability, while the last three columns provide information on the lagged effect (one year) of FinTech firms on the financial stability of banks. The results show that FinTech companies have significant positive impact on financial stability across alternative models. Notably, the lagged impact of FinTech companies on financial stability is greater, and this evidence is consistent across alternative models. Among the control variables, bank size, capitalization, and return on assets are statistically significant and play an important role in improving the financial stability of banks in Malaysia. Overall, these results suggest that the lagged impact of FinTech companies is greater on banks’ financial stability than the contemporaneous effect. These findings indicate that the banks begin responding and taking appropriate actions to improve their financial stability when they begin to experience the competition from the FinTech companies in the market.

In the next step of our investigation, we classified our sample banks into large and small groups. Table 3 reports the empirical results on the lagged effect of FinTech firms on financial stability by bank size. The results show that FinTech firms continue to positively impact the financial stability of large and small banks. However, FinTech firms’ influence is greater on the financial stability of small banks than on that of large banks. This empirical evidence makes more practical sense because small banks are more proactive due to their size and institutional setup to implement necessary actions to counter increasing competition and changing market conditions.

We further conduct a sub-sample analysis based on bank-level corporate governance. Specifically, the sample banks in Malaysia are divided into high and low corporate governance groups, and the results are displayed in Table 4. Interestingly, the results show that FinTech companies have a greater positive impact on the financial stability of banks with low corporate governance than those with high corporate governance. The effects of the control variables were consistent, as shown in Table 3. The main takeaway from this analysis is that banks with low corporate governance are more proactive in improving their financial stability with the increasing presence of FinTech companies. This suggests that low corporate governance banks adopt more FinTech in their banking services, which could be characterized as a behavior to mask poor governance or take advantage of the weak governance framework of FinTech practices in banking sector.

We further perform a sub-sample analysis by bank type. Table 5 presents the results of FinTech companies’ impact on the financial stability of Islamic and conventional banks. The findings show that FinTech companies have a greater positive and significant impact on the financial stability of Islamic banks than on that of conventional banks. This indicates that Islamic banks are more efficient in adopting FinTech technologies in their services and in managing the competition arising from FinTech firms. We further investigate FinTech companies’ impact on banks’ financial stability by controlling for the GFC period, the results of which are displayed in Table 6. The findings reveal that FinTech companies continue to positively drive the financial stability of banks in Malaysia. We further note that bank size, capitalization, and return on assets are the other potential drivers of banks’ financial stability.