The increasing adoption of digital innovations in the financial system is pushing the academic discussion about its potential benefits or drawbacks to look for a solid foundation of empirical evidence. While prior studies have investigated the effects of information technology (IT) adoption on different banking outcomes (e.g. Beccalli 2007, Koetter and Noth 2013), findings so far have been inconclusive and, apart from a few exceptions (Pierri and Timmer 2020, Kwan et al. 2021), have not yet been tested during periods of crisis.
In our recent paper (Branzoli et al. 2021), we exploit the Covid-19 pandemic – an unpredictable event that is likely to have enhanced the importance of digital prowess as a source of competitive advantage – to analyse variations in credit across Italian banks associated with different ex-ante levels of IT adoption. We find that IT-intensive banks increased their lending to non-financial corporations (NFCs) more than others in the months following the outbreak of the pandemic; the increase was economically sizable even when nationwide mobility restrictions were lifted and public health conditions improved.
Measuring banks’ IT adoption
We measure banks’ level of IT adoption using unique data on IT-related costs reported in the income statement and survey information on the use of digital technologies at the bank level. These are expenses incurred for the purchase of hardware (e.g. personal computers, servers, mainframes) or software, the compensation of IT specialists (e.g. computer support engineers) and the outsourcing of IT services to external providers. IT costs are normalised by the total operating costs of the bank. Figure 1, Panel A, shows the evolution of the IT-to-total costs ratio over time and across percentiles.
To assess whether a greater share of IT costs is related to a higher degree of IT adoption, we explore the relationship between banks’ IT expenditures and the use of digital technologies. We combine data on IT costs with bank-level survey information on the status of digital transformation of the Italian banking sector.1 More specifically, we ask banks to indicate which financial services they offer online (e.g. loans, payments, asset management), if any. Respondents are also asked whether they have innovative projects under way, which technology underlines them (e.g. big data, biometrics, artificial intelligence) and for what purpose (for instance, improving consumer profiling or credit risk evaluation). Controlling for a rich set of bank characteristics (including size, funding structure, and profitability), we find that our measure of IT adoption is actually related to banks’ degree of digitalisation and propensity to innovate: the higher the IT expenditures, the greater the likelihood of offering digital services and engaging in innovative processes.
Panel B of Figure 1 displays credit dynamics in Italy before and after the pandemic outbreak, based on banks’ degree of digitalisation: since the beginning of 2020 credit drawn by high tech banks (i.e. those in the top quartile of the distribution of the IT-to-total costs ratio) increased by 11%, twice the rate recorded by other lenders.
Figure 1 IT costs distribution (Panel A) and credit dynamics across banks’ tech levels (Panel B)
Notes: The top graph shows the evolution of the 25th percentile, the median, and the 75th percentile of the distribution of the IT-to-total costs ratio in each year. The bottom graph displays lending patterns across banks with varying degrees of IT adoption. All banks in our sample are split into three groups according to their IT-to-total costs ratio: low tech if they fall in the bottom quartile, medium tech if they stand between the second and third quartile and, high tech if they are in the top quartile. The total amount of credit per each bank is normalised to 100 based on the amount of outstanding credit in December 2019.
We also investigate the dynamics of credit and its allocation across NFCs. Using a difference-in-differences identification strategy, we find that the effect of IT on credit growth was larger for borrowers hardest hit by the pandemic. NFCs located in the areas of the country most affected by the pandemic2 experienced a greater increase in lending from higher-tech lenders. We find positive variation in credit for businesses operating in sectors deemed non-essential during the lockdown and consequently forced to close their physical locations. Small and medium-sized enterprises (SMEs) – more exposed than larger firms to liquidity shortfalls – have benefited the most from the growth in loans fuelled by technologically advanced banks.
Digital versus physical channels
Whether technology is reducing the effect of distance on lending decisions is under debate (Petersen and Rajan 2002, Basten and Ongena 2020, Keil and Ongena 2020). In our analysis, we study the role of geographic proximity (between lenders and borrowers) in influencing the effect of technology adoption on credit during the pandemic. Figure 2 plots the physical and digital reach of Italian banks at the eve of the pandemic: at comparable technological levels, the dispersion of branch diffusion reflects a high heterogeneity in banks’ business models. In exploring the relative importance of these two dimensions, we find that banks able to serve their customers through both traditional and digital channels showed the highest credit growth from March 2020 onwards; in other words, we provide evidence that brick-and-mortar locations still matter, when combined with a strong digital presence of the bank.
Figure 2 Distribution of physical versus digital channels
Notes: The horizontal axis shows the IT costs ratio. The vertical axis presents the percentage of provinces in which the bank has a branch. Size of dots correspond to total assets in millions of euro. All computed in 2020.
We shed light on the impact of technology adoption in lending during the Covid-19 pandemic. Our results suggest that banks with a higher degree of pre-pandemic IT adoption have granted more credit to NFCs as the crisis started to unfold. Greater digital capabilities might have helped banks handle a larger-than-usual number of loan applications, improve workflow through automation, and streamline approval processes. We also show that, even under severe physical restrictions, customers still valued the possibility of having face-to-face interactions with their bank. Our analysis paves the way for future research on the long-term consequences of digitalisation in banking. As the trend towards digital uptake is here to stay, banks need to adapt to changing customer preferences and anticipate shifts in competition. Implications for business model innovation will definitely lie ahead.
Authors’ note: The views expressed here are those of the authors and do not necessarily reflect those of the Bank of Italy.
Basten, C and S Ongena (2020), “Online mortgage platforms can allow small banks to improve their inter-regional diversification”, VoxEU.org, 15 August.
Beccalli, E (2007), “Does IT investment improve bank performance? Evidence from Europe”, Journal of Banking & Finance 31(7): 2205-2230.
Branzoli, N, E Rainone and I Supino (2021), “The role of banks’ technology adoption in credit markets during the pandemic”, Working Paper.
Keil, J and S Ongena (2020), “It’s the end of bank branching as we know it (and we feel fine)”, Working Paper.
Koetter, M and F Noth (2013), “IT use, productivity, and market power in banking”, Journal of Financial Stability 9(4): 695-704.
Kwan, A, C Lin, V Pursianen and M Tai (2021), “Stress testing banks’ digital capabilities: Evidence from the covid-19 pandemic”, Working Paper.
Petersen, M and R Rajan (2002), “Does distance still matter? The information revolution in small business lending”, Journal of Finance 57(6): 2533-2570.
Pierri, N and Y Timmer (2020), “Tech in Fin before FinTech: The importance of technology in banking during a crisis”, VoxEU.org, 9 August.
1 The Regional Bank Lending Survey (RBLS), conducted by the Bank of Italy on a yearly basis, involves a large sample of Italian banks representing 90% of the deposits of the whole banking system.
2 The severity of the pandemic is tracked using data on hospitalisations, deaths, and changes in residential mobility at the province level.