This essay is based on the working paper “The Digital Banking Revolution: Effects on Competition and Stability” by Naz Koont.

The rise of websites and mobile applications represents a digital revolution that is transforming business models across a broad range of industries, including banking. Historically, many banks built enduring relationships with clients through in-person interactions at bank branches. Now, websites and mobile applications have become the leading ways that customers interact with their banks. These digital platforms allow customers to conduct a wide variety of banking transactions—including reviewing account balances, depositing checks, and applying for loans—all without stepping foot in a branch.

We know little about how this technological innovation is changing the nature of banking. For instance, these digital platforms have the potential to alter the competitive landscape between large and small banks by shifting cost structures and customer demands. It is not obvious what we should expect. On the one hand, larger banks may be able to invest in developing the highest-quality websites and mobile applications, leading them to capture even more of the market. On the other hand, smaller banks may be able to use this new technology to become more efficient or to become more attractive to customers as the importance of large banks’ extensive branch networks is eroded.

Further, banks provide a range of financial services to a variety of deposit and loan customers. If digitalization disproportionately facilitates the provision of particular banking services to certain customers, it will reshape the distribution of banking services as well as the composition of bank balance sheets. Understanding the resulting implications for customer welfare and financial stability is important for policymakers navigating the increasingly digital US banking landscape.

To assess the impact of digitalization on the banking system, I hand-collect data on banks’ mobile application release dates and find three results. First, after adopting digital platforms, banks expand the geographic scope of their banking service provision (they branchlessly enter new counties to provide services such as loan origination) while simultaneously thinning down their branch networks. Thus, any analysis of the aggregate effects of digitalization on the banking sector must take into account these changes in bank presence across local banking markets, as well as the closure of banks’ pre-existing branches.

Second, banks’ balance sheet growth varies across bank activities after adopting digital platforms. Midsize banks with assets between $10B and $100B grow the fastest, consistent with their offering relatively high-quality digital services while not facing too much demand erosion from their existing branch networks. In contrast, while the largest banks offer even higher–quality digital services, their activities already have been focused on offering transactional convenience through their dense branch networks. Thus, although their existing customers may switch to using digital avenues, these large banks do not see large growth after digitalization. This finding highlights the importance of considering how the effects of adopting digital platforms may vary across banks due to differences in service quality and substitution for their existing service offerings.

Third, banks alter the composition of their balance sheets after adopting digital platforms: they increase their share of uninsured deposit funds while reducing their share of lending to low-income loan customers. These compositional changes indicate that the effects of digital platforms may vary across various deposit and loan market segments.

I next build a mathematical model of the US banking system that allows me to run experiments about how the system would function under different conditions. I pin down the extent to which the effects of digital platforms are due to shifts in banks’ demands versus their costs. The approach also considers the fact that nonbank participants, such as Quicken Loans, offer digital services during this period. Using this framework, I am able to compare the existing digital world to an alternative world without digital technologies. This exercise allows me to evaluate the aggregate effects of digital platform technology on competition, welfare, and financial stability.

My analyses indicate that the adoption of digital platforms has led to a decrease in banking sector concentration and a reduction in the share of banking services provided by banks with more than $100B in assets. This transformation occurs because midsize banks—those with assets between $10B and $100B—experience the greatest relative demand increase after adopting these digital technologies, along with reductions in their variable costs of providing banking services. The growth of midsize banks in terms of market share and geographic reach may be of concern given the weakening regulatory oversight of these banks in recent years, including passage of the 2018 Economic Growth, Regulatory Relief, and Consumer Protection Act, which exempted US banks with under $250B in assets from certain Dodd-Frank Act banking regulations. Moreover, this finding suggests that digitalization may have played a role in the growth of regional banks, which ultimately resulted in regulators classifying their failures during early 2023 as a systemic risk to the financial system.

Banking customers are better off, enjoying improved services due to the introduction of digital technologies. However, high-income borrowers and uninsured depositors see the greatest benefits, due to their valuing digital services more than do lower-income segments of the population (who may not have access to digital tools or who may be unable to qualify for certain digital banking services, such as online loan applications). This point suggests that digitalization disproportionately benefits wealthier segments of the US economy. Moreover, although total bank profits remain unchanged after digitalization, large and midsize banks become more profitable while the profits of the smallest banks fall. This reflects the high fixed costs of developing digital platforms, squeezing the profitability of small players. As a result, digitalization threatens the survival of small community banks in the US banking system.

Beyond the financial stability implications of changes in the size distribution of the banking sector, there are also compositional effects on bank balance sheets that may be of interest to regulators. On the asset side, digitalization tilts banks’ loan originations toward higher-income, more transactional lending. Further, it alters banks’ ability to screen and monitor customers, eroding their ability to do so for low-income borrowers who may rely more on relationship banking and local information. As a result, banks’ expected loan losses associated with low-income lending increase. This suggests that the digitalization of banking may lead to a build-up of credit risks within segments of the banking system that are less well served by digital technologies. This is particularly worrying because banks are uniquely positioned to evaluate risks that are not amenable to the codification that digitalization requires.

On the liability side of bank balance sheets, digitalization alters banks’ funding composition. As a result of digital technologies, there is an inflow of uninsured deposits into the banking sector and a re-sorting of uninsured deposits within the banking sector toward larger banks with digital platforms. These uninsured depositors are more sensitive to interest rates and bank risk, as highlighted by the period of banking instability in early 2023. As a result, both the banking sector as a whole and digital banks in particular have a flightier deposit base due to digital technology, as measured by their uninsured deposit ratio.

Looking to the future, understanding the implications of digitalization will remain important as digital banking becomes more ubiquitous, in part spurred by the COVID-19 pandemic. The research evaluating the aggregate effects of these technologies on the banking sector highlights several trends that interest policymakers. Moreover, the findings relate more broadly to understanding how information technologies are leading a “second industrial revolution” and altering the competitive landscapes of many service industries.

Read the full working paper here.

Naz Koont is a PhD candidate at Columbia Business School.

This essay is part of the Financial Regulation Research Brief Series. Research briefs highlight the policy-relevant features of research on financial systems, including the impact of financial regulations on economic growth, stability, and other factors shaping living standards.

overlay image