**Preface to "Sustainability, Digital Transformation and Fintech: The New Challenges of the Banking Industry"**

The banking industry in the XXI century faces an increasingly competitive, complex, and fast moving business scenario. Financial and banking institutions must contend with multiple challenges tied to new regulations, legacy systems, disruptive models, new technologies, new competitors, and a restive customer base, while pursuing revolutionary strategies for sustainable growth. Banking institutions that can address these emerging challenges and opportunities to effectively balance long-term goals with short-term performance pressures will amply stakeholder rewards, as well as the market. Along this line, this book tries to assist banking industry researchers and practitioners to identify current concerns and relevant topics for an equilibrium between banking business needs and the needs of the society where these institutions are embedded. These needs are closely linked to sustainability, corporate social responsibility (CSR), financial inclusion, and banking literacy. For this purpose, the book comprises a selection of papers addressing some of the most relevant challenges and opportunities for the sustainability of international banking institutions. Papers in this collection cover the most recent lines of banking research and are all novel propositions that deepen the analysis of business strategies in the banking industry. In total, a selection of 28 papers form this book, covering topics such as the digital transformation of the banking industry and its effect on sustainability, the emergence of new competitors such as FinTech companies, the role of mobile banking in the industry, the connections between sustainability and financial performance, and other general sustainability and CSR topics related to the banking industry, such as social disclosure and online communication through social media. Contributors to the book represent a wide spectrum of nationalities from all over the globe. Papers come from the United States, Australia, the United Kingdom, China, Sweden, Russia, Spain, the Czech Republic, Italy, Norway, Lithuania, Nigeria, Romania, Korea, Portugal, Colombia, Saudi Arabia, Poland, Vietnam, Bangladesh, Thailand, Hungary, and Taiwan. Therefore, this book provides an unprecedented opportunity to reflect upon the most current research of the banking industry from an enriched multicultural perspective. The book is a Special Issue of the MDPI journal Sustainability, which has been sponsored by the Santander Financial Institute (SANFI), a Spanish research and training institution created as a collaboration between Santander Bank and the University of Cantabria. SANFI works to identify, develop, support, and promote knowledge, study, talent, and innovation in the financial sector.

> **Andrea P´erez** *Editor*

### *Article* **Digital Bank Runs: A Deep Neural Network Approach**

**Marc Sanchez-Roger 1,\* and Esther Puyol-Antón <sup>2</sup>**


**Abstract:** The introduction of Central Bank Digital Currency (CBDC) could represent a deep structural change to the financial sector, and in particular to the banking sector. This paper proposes a Deep Neural Network (DNN) design to model the introduction of CBDC and its potential impact on commercial banks' deposits. The model proposed forecasts the likelihood of the occurrence of bank runs as a function of the system characteristics and of the intrinsic features of CBDC. The success rate of CBDC and the impact on the banking sector is highly dependent on its design. Whether CBDC should carry any form of interest, if the amount of CBDC should be capped by account or if convertibility from banks' deposits should be guaranteed by commercial banks are important features to consider. Further, the design of CBDC needs to contribute to enhancing the sustainability of the financial system, hence a CBDC design that promotes financial inclusion is paramount. The model is initially calibrated with Euro area system data. Results show that an increase in the financial system risk perception would trigger a significant transfer of wealth from bank deposits to CBDC, while the wealth transfer to CBDC is to a lesser extent also sensitive to its interest rate.

**Keywords:** CBDC; digital currency; bank run; banking; central bank

Puyol-Antón, E. Digital Bank Runs: A Deep Neural Network Approach. *Sustainability* **2021**, *13*, 1513. https:// doi.org/10.3390/su13031513

**Citation:** Sanchez-Roger, M.;

Academic Editors: Andrea Pérez and Hirofumi Fukuyama Received: 31 December 2020 Accepted: 28 January 2021 Published: 1 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Central Bank Digital Currency (CBDC) can be understood as a new form of central bank money that is different from the two types of money currently issued by central banks, which are physical cash and reserves. CBDC would represent a digital form of money, denominated in the national unit of account that is a direct liability of the central bank and intended to serve as legal tender [1,2].

The concept of public access to Central Bank deposit accounts is not by any means new, with the first public bank in Europe being the Taula de Canvi de Barcelona in 1401 which offered the possibility of opening deposit accounts to the population [3]. A walk through the first types of central bank money suggests that broadly accessible central bank deposits were the original form of central bank monetary liabilities. Neither banknotes nor commercial bank deposits were the common form of money over the last centuries, with the current limitation for the general population to access central bank accounts being introduced less than a century ago [4].

The introduction of CBDC could have important implications regarding the sustainability of the financial system, and in particular about financial inclusion. Indeed, Central Banks seem to consider financial inclusion as one of the most relevant reasons which support the introduction of CBDC [5]. This shows the strong link between CBDC and sustainability and leads us to consider financial inclusion as a key parameter when designing a CBDC.

We note that several countries and regions have concluded or are currently engaged in experiments and pilot tests to better understand the implications of introducing CBDC, including Ecuador, Ukraine, Uruguay, Bahamas, Cambodia, China, the Eastern Caribbean Currency Union, Korea, and Sweden [6]. In particular, the experiment to introduce CBDC

in Uruguay was part of a financial inclusion program [5], which shows again the relevance for Central Banks of developing a CBDC which contributes to promoting the sustainability of the financial system. In the case of Bahamas, CBDC was officially launched in October 2020 as announced by the Bahamas Central Bank [7].

In this paper, we assume that the introduction of CBDC does not eradicate the fractional reserve system, with commercial banks still offering deposits [8]. However, other authors claim that the direction of travel is to fully phase-out commercial banks' deposits into CBDC, hence leading to an economy based on "public money", also known as sovereign money [9]. It is out of the scope of this paper to analyze the trade-offs of allowing CBDC to coexist with commercial banks' deposits versus fully phasing-out banks' deposits.

The issuance of CBDC only accessible to a subset of the economy has been also explored by the current literature, including the concepts of "retail CBDC" which could be used by households and non-financial businesses, and/or "wholesale CBDC" designed to be used by financial corporations [10]. In this paper, we embrace the comprehensive version of CBDC, which could be understood as a universal CBDC accessible by the general public, including households and non-financial businesses, as well as financial corporations.

Differences in the structure of legal claims and the Central Bank's records are also important features. As per the current literature, we can identify three potential CDBC architectures, including indirect CBDC, direct CDBC, and hybrid CDBC [11]. In the case of indirect or synthetic CDBC, the claim is not on the Central Bank, and this form of CBDC could be understood as private sector financial corporations or e-money providers, issuing liabilities matched by funds held at the central bank [12]. In the case of Direct CBDC, the claim is on the Central Bank, while the Central Bank also handles the payments. Hybrid CBDC provides a direct claim on Central Bank; however, there are intermediaries to handle the payments.

The CBDC infrastructure is also a relevant issue to address [11]. As the authors of that paper describe, conventional infrastructures store data several times in separate physical locations, with data stored in multiple nodes and controlled by one authoritative entity. On the other hand, DLT-based infrastructures differ from conventional infrastructures given that the ledger is managed by different entities, without an authoritative entity, and in a decentralized model. The analysis of the potential tradeoffs is out of the scope of this paper. Note that this paper focuses on studying a system where the CBDC is introduced under the form of deposit based CBDC, disregarding a digital token currency.

The literature on CBDC covers a wide range of topics, from monetary policy or financial system stability to technology and computer science [13–17]. In particular, the current literature on CBDC shows a significant number of papers produced by institutions such as Central Banks, the International Monetary Fund (IMF), or the Bank for International Settlements (BIS), while a strong contribution from mainstream academia is also increasing. This highlights the relevance of the topic and the potential system-changing implications of CBDC for the entire financial and economic system.

Despite the interest in researching the uses and advantages of CBDC, the new form of money is not exempt from criticism. While one could perceive some central banks more inclined to consider the acceptance of CBDC in the future, other central banks consider that its introduction would not represent a substantial improvement [18]. In particular, the authors of that report [18] highlight that (i) it is difficult to see what CBDC would be able to contribute that it is not yet covered by the payment systems already in place, (ii) issuing CBDC would make a National Central Bank a competitor to commercial banks in some areas, and (iii) the introduction on CBDC would lead to risks to the financial stability including an increase of systemic bank runs.

Other literature suggests that the introduction of CBDC does not need to have a negative impact on banks' lending rates and that a well-designed CBDC will not threaten financial stability [19,20]. However, the literature analyzing the impacts of CBDC on the banking sector is still at its early stages. Despite the relatively limited number of academic articles on the subject, a passionate debate is building up between defendants and opponents to the introduction of CBDC. This paper focuses on the implications of CBDC for the banking sector, and in particular, it focuses on bank runs.

The traditional banking business model, in which maturity transformation is at the core of its activity, is intrinsically fragile due to the threat of a sudden and significant number of deposit withdrawals taking place in a short time frame, which could lead to a shortfall in liquidity deriving into insolvency [21]. This mechanism, known as a bank run, has been widely studied in academic literature [22–24].

This work explores the concept of Digital Bank Run, understood as a bank run that takes place in a commercial bank where depositors (retail and wholesale) withdraw their deposits and place them in digital currencies. In particular, this work focuses on the different designs that could shape CBDC to understand and minimize the risks for the financial sector. The design of CBDC is still under discussion, and the final characteristics of it will determine its failure or success. In line with this, the literature has identified several important features of CBDC that is worth exploring, such as (i) adding limits to the amount of CBDC that could be stored by account, (ii) allowing CBDC to carry positive or negative interest rates, and (iii) warranting commercial banks' services to offer full convertibility from banks' deposits to CBDC, amongst others [25]. In this paper, we explore the likelihood of digital bank runs, testing different CBDC designs, varying its interest rates and under different levels of systemic risk.

In line with the above, understanding the potential effects of each of these features is paramount in assessing the impact of the introduction of CBDC on the financial sector. This work contributes to this complex task by designing a scalable Neural Network framework based on Deep Learning techniques initially calibrated with system-wide Euro area data. To the best of the authors' knowledge, this work represents an innovative approach to assessing how the different designs of CBDC could impact the banking sector.

We note that the ultimate goal of this work is to show the adequacy of Deep Neural Networks (DNN) as a tool to analyze the impact of the introduction of CBDC, rather than obtaining numeric conclusions from the specific scenario under analysis in the results section.

After this brief introduction, this paper is organised as follows. Section 2 comments on the impacts of CBDC on the banking sector, focusing particularly on the phenomena of bank runs. Section 3 presents the Deep Learning Neural Network model and describes the different scenarios and data used to calibrate the model. Section 4 presents the results, followed by the discussion section. Finally, Section 6 concludes.

#### **2. Theoretical Background**

Understandably, the introduction of CBDC could entail significant structural changes in the banking sector, which will vary very much depending on the features and the design of the CBDC.

The introduction of CBDC could lead to a disciplining effect on banks, which means that banks would be under the constant threat of bank runs [8]. In line with this, it could be understood that CBDC will have an impact on commercial bank business models, with expected flows of commercial bank deposits into CBDC. This could lead banks to prevent a loss of deposits by increasing interest rates offered in customer deposits, which could lead banks to seek alternatives to maintain profitability such as raising lending costs and increasing fees, ultimately leading to a reduction on banks' balance sheets [2]. The introduction of CBDC could facilitate the transfer of deposits from commercial banks to Central Banks since CBDC could be perceived as a risk-free option to store and protect wealth in economic stressed scenarios [26].

The current literature has briefly explored the steps that could be taken to limit the risk of bank runs as a result of the introduction of CBDC by adding frictions discouraging large amounts of deposit transfers from commercial banks to Central Banks. These frictions include limiting or capping the maximum amount to be deposited in CBDC accounts,

adopting a flexible approach with different tiers regarding the interest paid on CBDC deposits, or imposing fees on large amounts deposited in CBDC accounts [27]. Other proposals include limitations such as removing the requirement for banks to convert deposits to CBDC [25]. The authors of that paper suggest a set of core principles for CBDC, including (i) CBDC paying an adjustable interest rate, (ii) CBDC and reserves to be distinct and not convertible to one another, (iii) no guaranteed on-demand convertibility of bank deposits into CBDC, and (iv) central bank can only issue CBDC against eligible securities such a government bond. The authors suggest that by following these principles the introduction of CBDC should not necessarily impact neither banks' provision of credit to borrowers nor banks' provision of liquidity to depositors.

Next, we briefly describe some of the key features of CBDC. The model proposed in this paper allows the simulation of the potential impact of the different combination of features and its potential impact on the distribution of wealth between cash, commercial bank deposits, other financial assets and CBDC. In other words, the DNN proposed in this work allows us to understand the transfer of funds from commercial bank deposits to CBDC under different scenarios, and hence to assess the likelihood of a bank run depending on the design of the CBDC.


phasing-out (or switching instantly) commercial bank deposits into CBDC. However, this paper does not go down this route since the matter of full migration of commercial bank deposits into CBDC is out of the scope of this analysis.

Our work concurs with most of the current academic literature in the field of CBDC stressing the importance of the design and features of the CBDC [25,29]. Academic literature shows that prediction algorithms based on artificial intelligence have a wide range of applications and generally the results obtained are superior to those obtained through traditional statistical methods when applied to financial analyses [30,31]. In order to allow testing of the different CBDC features in a safe environment, this work presents an artificial intelligence model, based on deep learning, that allows the development of this task. The next section briefly explains the methodology applied and describes the different scenarios.

#### **3. Materials and Methods**

Financial prediction analyses are of great practical and theoretical interest. However, they are notoriously difficult, primarily driven by the non-linear and complex interactions in the data. Most of the mathematical techniques used in the field of computational finance for financial prediction use parametric and non-parametric statistical techniques [32]. An important limiting factor in the performance of statistics-based techniques in computational finance is the uncertainty inherent in any financial transaction, which leads to less accurate statistics-based financial models.

During the past few years, DNNs have achieved enormous success in many data prediction fields such as speech recognition, computer vision, or natural language processing, to name but a few. In this paper, we apply deep learning methods to forecast the likelihood of the occurrence of bank runs as a function of the intrinsic features of CBDC and also external factors such as systemic risk. The proposed deep learning-based model has several advantages over traditional statistical methods, which include (i) input data can be expanded to include all items of possible relevance to the predictor model; (ii) fewer assumptions than statistical models, which allow the model to be more generalizable; and (iii) non-linearities and complex interactions among input data are modelled by the deep learning models, which can help increase in-sample fit versus traditional models.

#### *3.1. Method: Deep Neural Network Architecture*

In this section, we describe the details of the Deep Neural Network (DNN) model we train to assess the likelihood of bank run when the CBDC is introduced. First, we provide the details of the DNN model and then we provide the details of the training and evaluation.

A DNN consists of multiple fully-connected (FC) layers: an input layer, one or multiple hidden layers, and a single output layer [33]. DNNs have a single input layer and a single output layer, and the number of neurons (also referred as units) in the input layer equals the number of input variables in the data being processed. The number of neurons in the output layer equals the number of outputs associated with each input.

In a FC layer, all output activations are composed of a weighted sum of all input activations (i.e., all outputs are connected to all inputs). More specifically, for each hidden unit k, a non-linear activation function *f*(·) is used to map all inputs from the lower layer, *xk*, to a scalar state, *yk*, which is then fed to the upper layer *yk* = *f*(*xk*), where *xk* = *bk* + ∑*<sup>i</sup> yiwik* and *bk* is the bias unit *k*, *i* is the unit index of the lower layer and *wik* is the weight of the connection between the unit k and i in the layer below. In this work, we select the leaky rectified linear unit (ReLU) as the activation function *f*(·) for the hidden layers and the Softmax function for the output layer:

$$f\_{\text{leaky ReLU}}(\mathbf{x}) = \begin{cases} \quad \text{x if } \mathbf{x} > 0\\ \quad 0.01\mathbf{x} \text{ otherwise} \end{cases} \text{ and } f\_{\text{Softmax}}(\mathbf{x}\_i) = \frac{e^{\mathbf{x}\_i}}{\sum\_j e^{\mathbf{x}\_j}} \tag{1}$$

The neurons of the network jointly implement a complex non-linear mapping from the input to the output. This mapping is learned from the data by adapting the weights of each neuron using the back-propagation algorithm.

In this study, DNN is used as a multivariate regression model to learn the mapping function between the input vector, which in our case is composed of the interest and risk parameters of each one of the financial assets, and the output vector, which are the final weights or amounts allocated to each one of the financial assets given the interest-risk pairs for each asset. Figures 1–6 show an example of the proposed architectures. Color code used in the figures below is black for Cash, blue for Deposits, red for CBDC and green for other Financial Assets.

**Figure 1.** Deep Neural Network modeling the introduction of Central Bank Digital Currency (CBDC).

**Figure 2.** Scenario 1.

**Figure 3.** Scenario 2.

**Figure 4.** Scenario 3.

**Figure 6.** Scenario 5. Note: This model can also be expanded to include a broader number of financial assets, including holdings of debt instruments or stocks amongst others. For the sake simplicity we have not included such additional financial assets in this paper, and we focus on the transfer of wealth after the introduction of CBDC amongst: (i) Cash, (ii) Household commercial bank deposits, (iii) Corporate commercial bank deposits, and (iv) CBDC.

Regarding the initialization of weights, to prevent the layer activation outputs from exploding or vanishing during a forward pass through a deep neural network, all weights and biases have been initialized using the Kaiming initialization [34].

Regarding the loss function, in regression problems, the typical loss function used is the L2 norm of the residual, which during backpropagation produces a gradient whose magnitude is linearly proportional to this difference [35]. This means that estimated values that are close to the ground-truth (i.e., inliers) have little influence during backpropagation, but on the other hand, estimated values that are far from the ground truth (i.e., outliers) can bias the whole training process given the high magnitude of their gradient. To overcome this limitation, we used the Huber's loss function, which is a robust loss function that behaves quadratically for small residuals and linearly for large residuals [36].

Next, we briefly describe the training and validation processes followed. We split the data set into three subsets: (i) training, (ii) validation, and (iii) test. The training set is used to adjust the weights of the network. The validation set is used to minimize overfitting and relates to the architecture design (i.e., the selection of hidden layers and neurons). Finally, the test set is used to assess the actual predictive power of the DNN. For all experiments, we used five-fold cross validation to obtain the optimal model and a grid search strategy to optimize the hyper-parameters, including the learning rate, and the number of hidden layers.

Finally, the mean squared error (MSE), the mean absolute error (MAE), the mean absolute percentage error (MAPE), and the coefficient of determination (R2) were calculated and tabulated to evaluate the performance of the model.

#### *3.2. Data*

The model has been trained using Euro area aggregate data. Data has been obtained directly from Euro area databases, in particular from the European Central Bank Statistical Data Warehouse. Monthly data from January 2003 until October 2020 has been used to train, validate and test the model.

Regarding the inputs, for each element, an interest-risk pair has been defined. Initially, the following levels for each input have been set:

**IC:** The interest rate of cash has been set at 0%.

**RC:** Risk perception of cash has been set at 0. Note that for simplicity reasons, storage risks are not considered.

**ID**\_**House:** Interest paid by commercial bank deposits to households is obtained as the average interest paid for sight and term household deposits across European countries.

**RD**\_**House:** We use the ESRB Composite Indicator of Systemic Stress (CISS) to measure the risk perception linked to commercial bank deposits (households). In particular, the CISS measures the level of stress in the financial system as a whole [37]. However, any other risk perception metric could be used instead.

**ID**\_**Corp:** Interest paid to corporate deposits is obtained as the average interest paid for sight and term corporate deposits across European countries.

**RD**\_**Corp:** Risk perception linked to commercial bank deposits (corporates). We use the same risk indicator as for household deposits (CISS).

**ICBDC:** Interest CBDC, which can be set at zero, negative, or positive rates.

**RCBDC:** Risk perception of CBDC, which is set at zero in this work.

Regarding the outputs, the following magnitudes are used:

**C:** an estimate of the total amount of cash hold as a store of value. Citizens do also hold cash as a store of value. Studies from the European Central Bank show that on average 24% of European citizens hold cash outside a bank account as a precautionary reserve, based on a large sample of European respondents during 2015 and 2016 [38]. Estimating the share of cash used as a store of value is not an easy task, and a significant number of assumptions need to be made. By assuming that the holdings of cash as a store of value is the total amount of cash in circulation with the amount being used for transaction purposes deducted, held by financial institutions or held abroad, the European Central

Bank estimates that more than one-third of the total Euro banknotes in circulation may be in use as a store of value [39]. The total amount of banknotes in circulation amounted to €1,394 bn as of October 2020, while the amount of coins totaled €30 bn as per ECB Statistical Data Warehouse. Assuming a third is being held as a store of value, in line with the ECB findings, this leads to an amount of c.€430bn. A similar study was conducted in Germany, where the main reasons for storing cash were analysed, these being (i) low-interest rate level, (ii) most common means of payment, (iii) cash works even if technology fails, (iv) no fees, and (v) anonymity [40].

**DHouse:** the total amount of household bank deposits in the Euro area. Household total deposits in Euro area commercial banks amounted to €8,210 bn as of October 2020.

**DCorp:** the total amount of corporate bank deposits in the Euro area. Corporate deposits in Euro area commercial banks amounted to €3,120 bn as of October 2020.

**CBDC:** when CBDC is included in the model (from Model 3 onwards) and given that no real data regarding holdings of CBDC exist, we have trained the model using different initial cases leading to a wide range of CBDC values, allowing us to simulate several scenarios. The DNN proposed in this paper will allow users to modify the CBDC inflow hypothesis easily.

As of October 2020, the split between cash, household deposits and corporate deposits was as follows: (i) cash 3.6%, (ii) Household deposits 69.8%, and (iii) Corporate deposits 26.5%. According to the methodology described, this has been used as an output. Regarding the inputs, as of October 2020 the IC and RC were set at zero, the ID\_House was 0.14% while the ID\_Corp was −0.01%. The RD\_House and RD\_Corp were set at 0.11 according to the ESRB Composite Indicator of Systemic Stress. This inputs-outputs combination was only one of the 315 samples used to train and test the model. In particular, the model has been trained using 251 samples which correspond to monthly values of the different series, while it has been tested using 64 samples. Both the CBDC inputs and expected outputs have been set across a wide range of scenarios as shown in the results section. The methodology proposed allows us to analyze a high number of different cases which should allow policymakers and other relevant authorities to assess the strengths and weaknesses of the different potential designs of CBDC.

#### *3.3. Model Description and Scenarios*

The model proposed is composed of four elements: (i) Cash, (ii) Household commercial bank deposits, (iii) Corporate commercial bank deposits, and (iv) CBDC. This model uses two key characteristics of a financial asset as inputs: (i) Interest (or expected yield), and (ii) Risk (or expected volatility), while the outputs represent the allocation of the total wealth of a closed economy between the different elements or financial assets mentioned above. The elements of the model and the number of inputs per element can be easily expanded following the same logic as presented later in this work. The model assumes a closed economy, with citizens and businesses having direct claims against the Central Bank when holding CBDC. Figure 1 shows a visual representation of the model, assuming only one commercial bank in the system, where the outputs are:

**C**: Total wealth amount or percentage allocated in cash as a store of value.

**DHouse**: Total wealth amount or percentage allocated to household commercial bank deposits.

**DCorp**: Total wealth amount or percentage allocated to corporate commercial bank deposits. **CBDC:** Total wealth amount or percentage allocated to CBDC.

Note that when the outputs are expressed as percentages, the following expression is true.

$$
\sum \mathbf{C} + \mathbf{D}\_{\text{House}} + \mathbf{D}\_{\text{Corp}} + \mathbf{CBDC} = 100\% \tag{2}
$$

Regarding the inputs, we assign an interest and a risk level to each financial asset, where

**IC:** Interest paid by cash.

**RC:** Risk perception of cash.

**ID**\_**House:** Interest paid by commercial bank deposits to households. **RD**\_**House:** Risk perception linked to commercial bank deposits (households). **ID**\_**Corp:** Interest paid by commercial bank deposits to corporates. **RD**\_**Corp:** Risk perception linked to commercial bank deposits (corporates). **ICBDC:** Interest paid by CBDC, which can be set at zero, negative, or positive rates. **RCBDC:** Risk perception of CBDC.

As described above, we assign an interest and a risk level to each financial asset, while the outputs represent the allocation of the total wealth of a closed economy between the different financial assets.

Next, we briefly describe the different scenarios that lead to the design of the DNN proposed in this paper, while also showing the block approach followed, which suggests that it is relatively straightforward to add new elements to the model if data to calibrate the new inputs/outputs is available.

Figure 2 presents a simplification of the system, where cash and commercial bank deposits are the only available financial assets to store wealth. In this model, there is no distinction between household and corporate deposits, and the system is set up with a generic commercial bank that centralizes the total customer deposits outstanding. In this scenario, only cash competes with commercial bank deposits.

Next, we make a distinction between corporate and household deposits, which increases the granularity. Further, when increasing the number of inputs/ outputs we added a hidden layer in the design of the DNN. This model enables a simplified representation of a closed economy where the options for customers remain limited to cash or commercial bank deposits. In this system, changes in the risk perception and interest rates could lead to transfers of commercial bank deposits into cash, and vice-versa. This model was calibrated and trained with Euro area aggregate data, which allowed us to understand the sensitivities of the different types of depositors and cash to interest rates and systemic risk perception. These results are not included in this paper, since the focus of this work is to analyze the impact linked to the introduction of CBDC. However, the DNN used in Scenario 2 and trained with the abovementioned data is available upon request by contacting the authors of this paper.

The next step is the introduction of CBDC, which represents a disruptive event in the system. Figure 4 represents a simple framework to analyze the wealth transfer amongst financial assets when introducing CBDC. This system allows the testing of different CBDC designs including interest rate set at positive, negative or zero levels as well as potential limits to the holdings of CBDC per account. In particular, the difference between the previous model and this one is the introduction of the element CBDC, which is represented by two new inputs and a new output. The DNN is designed with eight inputs, a hidden layer composed of six nodes and four outputs. This is the scenario that has been used as an example in the results section, as we describe later.

Despite the empirical work of this paper focusing on Scenario 3, the next step could be to add a new element defined as Financial Asset (FA), which represents households and corporations' holdings of bonds, equities and/or other financial assets. The introduction of FA opens the door to several substitutes of commercial bank deposits, ranging from cash to CBDC or financial assets. The design of the DNN proposed in this work makes it relatively simple to add new elements as mentioned above.

Figure 6 represents a generalization of the model developed in this paper, where competition amongst banks has been introduced. This model allows banks to set different deposit rates, and each bank to have a different risk perception level. This model would therefore simulate both systemic bank runs, and also idiosyncratic bank runs with the presence of CBDC.

#### **4. Results**

In this section, we disclose and discuss the results of this analysis. As mentioned in the previous section, this paper focuses on Scenario 3. This experiment focuses on understanding the sensitivity of depositors to switch from commercial bank deposits into CBDC under different CBDC designs and system configurations. In particular, we analyze the impact of designing CBDC with positive, negative or zero interest rates. We also analyze the wealth transfer from deposits to CBDC under different levels of systemic risk.

Python 3.6+, PyTorch 1.7+ and scikit-learn 0.22+ are used to perform the analyses. To optimize the loss function we used the Adam optimizer, with the momentum set to 0.9 and the learning rate to 0.07. The models were trained for 200 iterations on an NVIDIA GeForce GTX TITAN GPU and the model with the lowest MAE (on the validation set) was selected. The network training required 2 min per epoch on average and it took 0.52 milliseconds on average to process a case during testing. The source code is available at https://github.com/estherpuyol/CBDC\_model.git.

With regard to the validation results, the MSE, MAE, MAPE and R2 are provided in Figure 7. The loss change during training is shown in Figure 8. The system has a low MAE and MAPE, with also a strong regression coefficient R2.


**Figure 7.** Mean squared error (MSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and the coefficient of determination (R2).

**Figure 8.** Loss change during training.

Once the model has been successfully calibrated, we used the DNN to simulate how the wealth distribution amongst the different financial assets would change with the introduction of CBDC. In particular, Figure 9 shows depositors' sensitivity to CBDC interest rate under three scenarios: (i) zero interest rate on CBDC, (ii) positive CBDC interest rate set at 0.5%, and (iii) positive CBDC interest rate set at 1.0%. The first two rows represent the October 2020 wealth distribution in the Euro area, and the average values since January 2003 respectively. October 2020 Euro area data has been used in all cases, while for CBDC since data is not available, the risk was set at zero and the output followed the hypothesis that increases in systemic risk would lead to an increase in CBDC demand in line with the current literature [41].


**Figure 9.** Wealth transfer sensitivity to changes in CBDC interest rate. Blue numbers represent the model inputs.

Figure 10 shows depositor's sensitivity to changes in the financial system risk under three scenarios: (i) CISS set at 0.11 and in line with October 2020 data, (ii) CISS set at 0.5 representing a medium-high systemic risk, and (iii) CISS set at 0.9 representing an extremely high systemic risk. Similar to the previous table, the first two rows represent the October 2020 wealth distribution in the Euro area, and the average values since January 2003, respectively. It is particularly interesting to note the redistribution of wealth in Case 1 vs. the October 2020 scenario.


**Figure 10.** Wealth transfer sensibility to changes in financial system risk Blue numbers represent the model inputs.

Finally, Figure 11 shows the transfer, mainly of commercial bank deposits given that cash remained broadly unchanged, to CBDC. The results shown in the matrix below represent the total volume of CBDC as a percentage of the total wealth. The colour code also shows the risk of a digital bank run, given that in some scenarios the total holdings of CBDC could reach as high as ca. 80% of the total.


**Figure 11.** Wealth distribution sensitivity to CBDC interest rate and Financial System Risk.

#### **5. Discussion**

The results show that the introduction of CBDC would lead to an initial transfer of wealth from commercial bank deposits to CBDC. Figure 9, Case 1 simulates a simple scenario where the transfer of wealth from cash and commercial bank deposits to CBDC takes place when CBDC is introduced for the first time with CBDC interest rate and risk perception set at zero and the rest of the inputs set as at October 2020 levels. It is

interesting to observe that cash would be slightly reduced from 3.6% to 3.1%, while both household deposits and corporate deposits declined by 3.8 percentage points (pp) and 12 pp respectively, leading to a CBDC amounting to c.16.4% of the total wealth of the system. We also note that an increase in CBDC interest rates of 50bps (Figure 9, Case 2) and 100 bps (Figure 9, Case 3) leads to a moderate increase of total CBDC outstanding at c.17.8% and 19.2% respectively. The sensitivity of deposit outflows to CBDC interest rates seems moderate, with c.2.3 pp of deposit outflows per 50bps of CBDC interest rate increase. However, we note that deposit outflows sensitivity to CBDC is not linear as we show in Figure 11.

Figure 10 shows that an increase in systemic risk could lead to digital bank runs. In particular, the results show that a systemic risk indicator increasing from 0.11 to 0.5 would lead to significant deposit outflows, increasing the total wealth stored in CBDC from 16.4% to 51.9%. In an extreme scenario, with systemic risk as high as 0.9, over 80% of the total wealth would be stored in CBDC, leading to a system-wide digital bank run. We note that deposit outflow sensitivity to systemic risk is very high, which is in line with the majority of the literature on the topic as we highlight in the introduction and theoretical background sections of this document.

Figure 11 shows that deposit outflow sensitivity to financial system risk and CBDC interest rate is non-linear. In particular, sensitivity to CBDC interest rate is moderate when the systemic risk is at low or medium-high levels (with CISS ranging from 0.2 to 0.6), while there is no sensitivity to CBDC interest rate with high or very high systemic risk levels (with CISS ranging from 0.7 to 0.9). This suggests that under a stressed financial scenario, a commercial bank deposit outflow would not be stopped by lowering the interest rate of CBDC.

Furthermore, the results suggest that a maximum amount of CBDC per account set at 30% or 40% of total wealth per account would limit more effectively the potential deposit outflows in the case of system-wide financial stress. Adding caps to the design of CBDC could be somewhat seen as the physic limits that cash entails. In theory, the holdings of cash are only limited by the total wealth of an individual, however, in practical terms, holdings of cash are "virtually capped" by the storage and transport risks. Since CBDC would not entail any storage or transport risks, it could be deemed appropriate to add certain limits to the holdings of CBDC, particularly during the early stages of the introduction of CBDC. However, this remains to be tested and further analyses on that point would add significant value to the CBDC implementation debate.

In line with the above, some authors consider that a cap on CBDC would reduce the effectiveness of it [18]. Somewhat linked to that point, new questions regarding the scope of CBDC would need to be explored at some stage, including whether the introduction of CBDC would represent just a first stage before moving into a financial system embracing the concept of "sovereign money", whether it should be designed to compete with commercial bank deposits, or if it should be conceived as just a mere substitute of cash.

#### *Limitations and Future Lines of Research*

This methodology is not without limitations, with the lack of CBDC data being the most notable. As described above, the model has been initially trained using real Euro area aggregate data for all components but CBDC. In the case of CBDC, different assumptions have been made, which leads to a certain degree of subjectivity. On the other hand, the model proposed allows other researchers to modify such hypotheses concerning CBDC initial calibration.

Next, we discuss some of the potential implications of CBDC for the banking sector. The following points constitute hypotheses to be tested and do not express the authors' views on the subject. The goal of the remainder of this section is to stimulate the debate regarding the introduction of CBDC and the potential impact on the banking sector and on bank runs in particular.

In the scenario described in this work, commercial banks would compete for deposits with Central Bank, leading banks to increase interest on customer deposits, hence eroding its margins. After the introduction of CBDC, and as per the results obtained in Figures 9–11, banks would need to reduce their balance sheet and increase lending rates, potentially tightening the access to credit. However, under this hypothetical scenario banks could use several strategies, including cross-selling techniques to retain deposits, and other measures such as offering lower loan rates to corporate clients holding deposits in the bank, or lower mortgage rates to retail customers using banks' deposits and payment services. Exploring this area would help to measure the impact on banks' accounts and indirectly the potential spillover effects on the availability of new credit.

Furthermore, banks would try to replace some deposits lost with other forms of funding, such as wholesale funding, leading to a higher cost of funding. In turn, higher wholesale funding reliance, and depositors' choice to transfer funds to CBDC could lead to higher market discipline in the banking sector. Some authors question whether the need for a deposit guarantee scheme would still be justified after the introduction of CBDC [42]. As described on the European Commission website, a fundamental principle underlying the Deposit Guarantee Scheme (DGS) is that they are funded entirely by banks and that no taxpayer funds are used. In line with this, a DGS would in part mitigate digital bank runs under stressed scenarios.

This work uses Euro area data to train, validate, and test the model. The authors would like to suggest the development of similar analyses with data from different geographic areas and regions as an interesting future line of research. Results obtained from Euro area data could differ from results in regions where there is not just a common monetary policy, but also a single fiscal policy.

The study uses a model based on a closed economy where only domestic use of CBDC was possible. However, we would encourage other researchers to elaborate on the subject of whether CBDC should be used exclusively in the national territory or whether the CBDC should be open to international use too.

As described in the introduction of this work, designing a CBDC that contributes to the sustainability of the financial system is relevant. In particular, some authors consider that features enhancing financial inclusion could limit the effectiveness of CBDC as a system of payments [43]. The relatively scarce literature on CBDC and financial inclusion suggests that further empiric analysis on that topic would be relevant for the current literature on CBDC.

So far, the literature is strongly focused on the impact on banks' funding costs, which as highlighted by many authors, could lead to disintermediation of the banking sector. However, it would be interesting to explore further whether the Loss Given Default of debt instruments and banks' cost of equity should remain unchanged or increase/decrease after the introduction of CBDC and the potential spillover effects mentioned above. Moreover, macroprudential and policymakers' options remain in place to contain a potential increase in banks' funding costs. Such options could include banks' debt being eligible for quantitative easing programs launched by central banks. Furthermore, it could be understood that in a hypothetical scenario where the public has the option to keep money in the form of CBDC at the Central Bank, the need for strict regulation over the sector declines. Capital, liquidity and loss absorption requirements could be loosened, leading to lower regulatory costs for banks, particularly on the funding side.

In line with the above, another question to explore under the hypothesis of CBDC introduction is whether banks should need to maintain a capital structure in line with Basel III, and complying with bank-specific capital and loss absorption requirements such as SREP (Supervisory Review and Evaluation Process), MREL (Minimum Requirement for own funds and Eligible Liabilities), or TLAC (Total Loss Absorbing Capacity).

It is also particularly important to design a solid transitional framework should CBDC be introduced. The transfer of significant amounts of commercial bank deposits to CBDC would add additional pressure to the stability of the banking system, hence raising the questions of whether and under which conditions the Central Bank may temporarily replace commercial banks' deposits lost with Central Bank lending. A phase-out calendar matching the maturity of the outstanding loans in the balance sheet could be adopted, with Central Bank lending in place during the phase-out period.

#### **6. Conclusions**

This paper develops a Deep Neural Network (DNN) design to assess the potential impact of the introduction of CBDC on the banking sector, and in particular, focuses on the link between CBDC and the bank run phenomena. This work represents an innovative method to assess the implications of the introduction of CBDC, allowing the simulation of the transfer of wealth between different financial assets depending on the design of a CBDC. The transfer of flows from one asset class to another can be used as a proxy to understand the dynamics that one could expect in the event of CBDC introduction. Below we conclude by highlighting some of the findings or relevant points highlighted in this work:

First, this paper describes the use of a DNN design, representing a multivariate regression model that learns the mapping function between the input vector, which represents the interest and risk for a sample of financial assets, and output vector which represents the wealth allocation into the different financial assets available. Several scenarios are described, and the model is finally calibrated using 315 real data samples corresponding to Euro area aggregate data including cash and loans volumes, interest rates and systemic risk metrics. The calibration of CBDC to be used in the training phase of the algorithm represents an important limitation linked to this type of technique. On the other hand, using different hypotheses vis-à-vis the calibration of CBDC data allowed the authors to draw conclusions on the impact of CBDC in a financial system.

This work considers that different designs of CBDC would lead to a wide range of outputs in terms of success and acceptance of CBDC, and the ability to increase financial inclusion through the deployment of CBDC is, by all means, one of the characteristics that this new form of money should pursue.

Second, the results point out that the introduction of CBDC would lead to an outflow of commercial bank deposits into CBDC, particularly given the "risk-free" nature of CBDC, while commercial bank deposits have an inherent (low) risk perception. Interest rates on CBDC are an important but not leading factor when analyzing the risk of commercial bank deposit outflows, as per the results shown in Section 4. However, as discussed, it is highly likely that the outflows from commercial bank deposits to CBDC would become barely interest-sensitive in the event of severe financial distress. The likelihood of digital bank runs seems strongly driven by the overall financial risk perception, modelled through the CISS indicator in this work. Other levers such as the full guaranteed convertibility of bank deposits into CBDC or limits on the CBDC accounts could be used to limit bank runs at this stage with potentially higher success rates. However, the higher the number of frictions introduced in the CBDC design, the lower the utility of it. Further research would be required to fully understand the trade-offs of suboptimal CBDC designs to partially protect the banking sector.

Third, the DNN architecture used in this paper to model the introduction of CBDC is highly scalable, allowing us to build a series of different blocs or new elements on top of the basic system analysed in this work. The authors would like to invite other scholars to continue assessing the implications of the potential introduction of CBDC through Deep Neural Network designs. The authors consider that after proving that DNN constitutes an adequate and valuable method to analyze the potential different designs of CBDC and its implications not just for the banking sector, but for the broader economy, this technique could be used more frequently for analyzing problems of a similar nature.

Finally, the potential introduction of CBDC could represent one of the most disruptive changes introduced in the financial system in a long time. Hence, a scrupulous analysis of the trade-offs derived from introducing CBDC is paramount. This paper contributes to the research in the segment of the interaction of CBDC with the banking sector by proposing an innovative method to assess the likelihood of a sudden and significant transfer of funds from commercial bank deposits to CBDC. This is a relatively new concept to most of us but threatening to be a source of intense debate and concern in the years to come: Digital Bank Runs.

**Author Contributions:** Conceptualization, M.S.-R. and E.P.-A.; methodology, M.S.-R. and E.P.-A.; software, E.P.-A.; validation, M.S.-R. and E.P.-A.; formal analysis, M.S.-R. and E.P.-A.; data curation, M.S.-R. and E.P.-A.; writing—original draft preparation, M.S.-R.; writing—review and editing, M.S.-R. and E.P.-A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The database and the code used for this study is freely available at https://github.com/estherpuyol/CBDC\_model.git.

**Acknowledgments:** This research has been conducted using a GPU generously donated by NVIDIA Corporation.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Beniamino Callegari and Ranvir S. Rai \***

School of Economics, Innovation, and Technology, Kristiania University College, Oslo 0107, Norway; Ben.Callegari@kristiania.no

**\*** Correspondence: Ranvir.Rai@kristiania.no

**Abstract:** Organizational ambidexterity is widely recognized as necessary for the economic sustainability of firms operating in the financial sector. While the management literature has recognized several forms of ambidexterity, the relationship between them and their relative merits remain unclear. By studying a process of implementation of ambidextrous capabilities within a large Scandinavian financial firm, we explore the role of top-down reforms and bottom-up reactions in determining the development of sector-specific innovative capabilities. We find that blended ambidexterity follows naturally from the attempt to correct the tensions arising from harmonic ambidextrous blueprints. The resulting blended practice appears to be closely related to the reciprocal model of ambidexterity, which appears to be a necessity rather than a choice, for large firms attempting to develop innovative capabilities. Consequently, we suggest to re-interpret current taxonomies of ambidexterity not as alternative blueprints, but rather as stages in a long-term process of transition.

**Keywords:** economic sustainability; organizational ambidexterity; blended ambidexterity; innovation process

#### **1. Introduction**

Organizational ambidexterity has gained considerable attention in management literature since it was first introduced by Robert Duncan in 1976. Originally referring to the ability to use both hands adroitly, the term has been increasingly used in reference to organizations attempting to simultaneously manage both exploitation and exploration to ensure long-term sustainability and competitiveness. Exploitation extends upon current knowledge, seeking greater efficiency and incremental innovation. Exploration, on the other hand, entails the development of new knowledge, experimenting and reaching for more radical innovation [1]. Pursuing and maintaining such opposing business logics and concurrently alleviating related tensions is at the very core of organizational ambidexterity.

The relevance of ambidexterity has been further highlighted by present demands for a transition towards more sustainable business practices. While entirely novel business models can be pioneered by new entrants, spinoffs, and other dedicated experimental market niches, meeting the sustainability challenge also requires incumbents to gradually transform their activities. Such transformation, however, must be conducted while simultaneously maintaining their operations functional and their customers satisfied; while more sustainable options must be explored, exploitation of current opportunities must similarly continue. In other words, from an incumbent perspective, the sustainability transition implies the development and application of ambidexterity.

Extant literature suggests multiple paths to ambidexterity; structural [2], contextual [3] and sequential [4,5]—usually treated as mutually exclusive pathways. However, as recently proposed by Foss and Kirkegaard [6] and as evident from this case study, various modes of ambidexterity may be present at once—ambidexterity can be "blended". Blended ambidexterity may portray a messier yet more realistic picture of "ambidexterity in the

**Citation:** Callegari, B.; Rai, R.S. Blending in: A Case Study of Transitional Ambidexterity in the Financial Sector. *Sustainability* **2021**, *13*, 1690. https://doi.org/ 10.3390/su13041690

Academic Editor: Andrea Pérez Received: 31 December 2020 Accepted: 2 February 2021 Published: 4 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

happening" as companies attempt to lessen some of the negative aspects inherent to any specific ambidexterity mode.

In our study, we focus on the dynamic issues related to the implementation of blended ambidexterity at a large financial organization. More specifically, we examine the issue through a case study of a Norwegian bank—currently undergoing a digital transformation towards becoming a technology-based provider of financial services. In line with previous research [7,8], we investigate how the organization is transitioning between different ambidexterity modes over time in accordance with various challenges and opportunities facing the firm. Our objective is to highlight the unique implementation challenges faced by a top-management team using blended ambidexterity. In particular our theoretical framework leads us to believe that the organization will suffer from significant tensions and paradoxes as they transition from different modes of ambidexterity over time. Our empirical study focuses on exploring the evidence relevant to these transitional changes, and the relative managerial response. We find evidence of a specific transitional trajectory of ambidexterity accompanying the implementation of new practices. The transitional trajectory, and the factors underpinning its course are examined and discussed as part of a transformational strategy for the organization. Additionally, the emergence of specific ambidextrous practices through time have implications for the question of the sustainability of banking business models in the long term.

The article is organized as follows. After this introduction, Section 2 describes our theoretical framework of reference. Section 3 illustrates our methodology. Section 4 contains our findings, which are discussed in Section 5. Finally, Section 6 concludes.

#### **2. Literature Review**

Smith and Lewis [9] have argued that sustainability is indissolubly linked with the concept of ambidexterity, the ability to balance activities dedicated to exploiting current opportunities and to exploring new methods, business models, and resources, an ability that appears to be vital for the long-term survival of firms [8,10–13]. A successful, sustainable firm is one able to exploit current opportunities and efficiently manage today's demands while also being open to the exploration of new possibilities [13]; to pursue new knowledge while using existing knowledge optimally [12]. In short, a successful firm is more and more commonly described as ambidextrous [4,14–16]. However, not only are there significant differences within the academic community regarding the conceptualization of ambidexterity and its antecedents, there is also substantial heterogeneity in its practical implementation within actual firms.

The various positions can be classified along two main axes. The first regards the understanding of explorative and exploitative activities as either complementary or opposite. The differentiation view argues that explorative and exploitative processes compete for the same resources, their simultaneous implementation involving significant trade-offs [8,17]. The integration view argues instead for the presence of significant complementary benefits between the two processes [3,18,19]. Consequently, the differentiation view supports implementation strategies based on separate exploration and exploitation activities across business units, while the integration view argues that the processes can, and should, coexist within the same unit. The second axis relates to the temporal dimension of ambidexterity. Exploitation can follow exploration in a more or less rigid alternating sequence [20] or the two activities may always be available for managers and employees to decide when to pursue them [3,21].

Combining these two axes, Simsek et al. [20] propose a classification of ambidexterity in four categories: harmonic, partitional, cyclical, and reciprocal. Harmonic ambidexterity refers to the simultaneous pursuit of exploitation and exploration within a business unit, resulting in processes or systems enabling individual employees and managers to autonomously decide about how to divide their time between conflicting demands [3]. Cyclical ambidexterity, based on the literature on "punctuated equilibrium", refers to the sequential pursuit of exploitation and exploration within a business unit, alternating between

long periods of exploitation and short bursts of exploration. Partitional ambidexterity refers instead to the simultaneous pursuit of exploitation and exploration across business units, what is commonly referred to in the literature as structural ambidexterity [8]. Finally, reciprocal ambidexterity refers to the sequential pursuit of exploitation and exploration within and across business units.

The latter configuration is certainly the most complex, requiring the establishment of structurally independent units characterized by different, and differently evolving, strategies, structures, and cultures [22]. While each unit operates at least partially independently, they are interdependent with respect to the achievement of ambidexterity. In this reciprocal context, the outputs of exploitation from one unit become the inputs for exploration activities conducted by other units, and vice versa [23]. Such complex configuration implies ongoing information exchange processes, collaborative problem solving activities and resource flows between middle management and the senior management team [24], requiring the presence of a clear shared vision at all operative levels shared vision [25].

While the categorization proposed by Simsek at al. [20] is comprehensive and clear, it is not surprising that empirical realities can be messier. Studying the operation of two business units operating under the William Demant Holding corporate umbrella, Foss and Kirkegaard [6] find evidence of "blending" modes of ambidexterity: the co-presence of structural and contextual ambidexterity, or harmonic and partitional, following Simsek et al. [20] classifications scheme. Specialized organizational units coexist and cooperate with units pursuing both exploitative and explorative activities. Foss and Kirkegaard argue that such configuration is not the result of muddled strategies or faulty implementation processes, but rather the outcome of intentional attempts to mitigate some of the negative aspects inherent to any specific ambidexterity mode, such as the potential loss of specialization advantages under contextual ambidexterity and the problem of diffusing the results of the explorative units activities within exploitative units under structural ambidexterity [14]. The resulting blended approach could also alleviate paradoxical pressures created by the demands of contextual ambidexterity on the individual, through the creation of dedicated supporting units.

An additional issue regarding the classification of ambidexterity regards its evolution over time. The question of how transitions between exploration and exploitation occur in contexts of sequential ambidexterity is still open [19]. Furthermore, due to the limitations linked with any specific approach to ambidexterity, Nosella et al. [7] and O'Reilly and Tushman [8] argue that the most advantageous strategy would involve systematically shifting between ambidexterity modes over time, according to current challenges and opportunities facing the firm. The possibility of blended ambidexterity further muddles the issue, as it is possible that empirical reality could be more correctly described as a shifting mix of ambidexterity regimes, morphing through time. However, the factors dominating the evolutionary shift of the ambidexterity mix through time are yet to be explored.

The link between ambidexterity and sustainability is in evidence within the corporate sustainability discussion that started in management literature in the mid-90s [26], gaining progressively more focus and attention since then [27–29]. In general, corporate sustainability refers to the ability of the firm to both make optimal use of environmental, physical, human, and organizational resources in relation to current knowledge and the existing set of opportunities and constraints as determined by the relevant constellation of stakeholders, while simultaneously cultivating their development over time; a description close to that of the ambidextrous firm [30]. Only through ambidexterity can the sustainable firm guarantee its continued survival by generating value and improving the lives of its stakeholders through time.

In their theoretical review, Sulphey and Alkahtani [31] argue that there is substantial overlapping between corporate ambidexterity and sustainability. The two are argued to complement each other, with sustainability providing the aims and general perspective, and ambidexterity the means and practices. The argument is similar to the one developed by Hahn et al. [32] in regard to the relationship between ambidexterity and corporate social

performance. Ciasullo et al. [33] provide empirical evidence for structural ambidexterity as a key factor for the realization of sustainability among Chinese multinational firms. Similarly, Gomes et al. [34] illustrate how ambidexterity is an important determinant of environmentally sustainable production in the area of quality management. Using a qualitative approach, instead, Minatogawa et al. [35] come to the same conclusions in their study of a small e-commerce company.

The relevance of the topics of sustainability and ambidexterity in the banking industry context is well known. Financial services offered by banks can be considered transformative services [36], indispensable components of the social infrastructure, whose changes and development over time can trigger wide-range socioeconomic change. On a macroeconomic scale, the composition of their investment portfolios affects both the pace and composition of economic development, while the development of new financial instruments can either support or undermine financial stability. Technological changes, such as digitalization, can also bring substantial social consequences. Wang and He [37] illustrate how digital access to banking services decrease the vulnerability to risk of poor farmers in China, while Yeo and Jun [38] show how P2P lending is changing credit conditions for borrowers with low-to-mid credit ratings.

As service firms, banks rely on intangible assets and knowledge to provide a superior customer experience [39]. Innovative activity is required for the economic sustainability of banks and the maintenance of their competitive advantage [40]. The banking industry has witnessed fundamental changes and greater instability that heighten the importance of successfully attaining ambidexterity to increase their performance [41]. Following the intensified competition brought by the gradual process of deregulation that has taken place in the last decades, bank units are expected to constantly improve existing products while also reducing costs by optimizing current operations. To meet these challenges, banks have integrated advances in information and communication technologies to introduce process technologies, increasing internal efficiency and improving productivity. The need to merge digital innovation with banking operations, however, have led to the emergence of new challenges for the development of ambidexterity in banking.

Studying the implementation of sustainable business models in banking, Yip and Bocken [42] notice the practical relevance of short-termism: the tendency to evaluate innovative projects predominantly on their ability to generate value in the short-term. In this sense, digitalization is often preferred because of its potentially calculable benefits, and its relative quick scalability. However, Marabelli et al. [43] document how large banks, with activities dispersed over large territories, facing the challenge to conjugate the need for innovation with the constraints of the short term, are likely to adapt ambitious ambidextrous organizational approaches, despite the additional managerial and organizational costs involved. Ambidexterity is found to require a managerial approach based on trust and decentralization in boundary activities, confirming the importance of managerial leadership [44,45] and a trust-based culture [21,46] as antecedents. Although somewhat animated by short-term considerations, the digitalization process is also promoting the development of ambidextrous capabilities, thus improving the general ability of financial firms to engage in the transformation processes required by sustainability (new references mentioned before). Thus, although superficially separated, the development of ambidexterity can be understood as an intermediate step in the gradual transformation of financial services towards more sustainable configurations.

In conclusion, the literature reviewed reveals that sustainability in the banking industry requires the integration of ambidextrous business models. The questions of which type of ambidexterity is preferable and how should it be implemented remain open, however. Furthermore, actual implementation of new operational routines and innovationsupporting tools appears particularly challenging in the heavily regulated and risk-averse environment of large-scale banking. As low-interest rates persist and digital-based innovative competitors enter the market; however, the challenge of ambidexterity must be met if traditional banking is to retain its central position within the global financial infrastructure.

#### **3. Methods**

This paper introduces a study of an on-going digital transformation in a large Norwegian bank. More specifically, we examine the issue through a case study of a large financial organization—DNB—currently transitioning from a traditional bank business model towards a technology-based provider of financial services. DNB is Norway's largest financial services group and one of the largest in the Nordic region in terms of market capitalization. They offer services to corporate, retail, and securities markets as well as the public sector. In addition to being a world leader in shipping, the bank has a strong position in the energy sector and seafood industries, reflecting traditional Norwegian economic strengths. As one of the largest financial institutions in the Nordic region, our case company is an excellent exemplar of an incumbent undergoing digital transformation while intending to retain its dominant market position.

We conducted semi-structured interviews with key stakeholders across operations to gain in-depth insight into organizational members' reasoning and reflections. This allowed us to comprehend the logic through which they viewed the world [47]. Furthermore, indepth interviews provide an effective means of obtaining rich insights into the phenomenon of interest, as they provide access to detailed contextual information and individual insight that cannot be obtained from surveys [48]. The informants were identified first through expert sampling, and then through snowball sampling, ensuring that all protagonists of the process have been interviewed, and their perspective integrated.

The respondents consisted of 12 managers directly involved in the company's innovation initiatives. They ranged from VP-level to middle management responsible for strategic innovation projects or programs across all DNB areas: New Business (5 respondents), Corporate Market (1 respondent), Consumer Market (2 respondents), Key Accounts (1 respondent), and IT (3 respondents). Within these areas, there are several divisions—referred to as "Innovation divisions" in New Business, "IT divisions" in IT, and "Business divisions" in the remaining areas.

Interviews were conducted until information redundancy was achieved [49]. We used semi-structured interview guides (see Appendix A), allowing for deviation from the sequence in order to follow interesting lines of inquiry and go deeper into relevant, emerging, topics. The approach was deemed particularly appropriate to this case study, since the company was undergoing a complex transformation process, leaving many informants to face significant, persistent uncertainty, which showed in their difficulties to engage directly with our questions. The questions covered the company's existing practices at the time of interview, personal experiences with and interpretations of innovation capabilities and tools, processes and resources, culture and values related to innovation initiatives. Interviews lasted for 45 to 60 min; they were digitally recorded and transcribed verbatim.

To structure respondent explanations and interpretations, we inductively examined our data by following the procedure suggested by Gioia and co-authors [50,51], a method considered particularly appropriate for research on strategic change and sensemaking [52]. In the first stage, we developed "open" codes by uncovering initial concepts from respondents' statements. Then, we confronted our results and integrated them, prioritizing adherence to the respondents' opinions in order to solve controversies. In the second stage, we classified these into higher order themes through axial coding based on the relationships among the initial first-order codes, translating the empirical results into relevant theoretical constructs. Finally, we gathered similar themes into aggregate dimensions that provided an overall structure to our narrative. We performed this procedure iteratively, going back and forth between codes and data until consensus among the authors emerged. Figure 1 shows the final data structure. The vertical order of the aggregate dimensions offers a broad timeline of the process of implementation of ambidexterity capabilities, starting from original situation and the initial plan, moving gradually towards a blended approach directed towards a reciprocal outcome. This narrative emerged during reflexive discussion between the authors and was subsequent used to organize the exposition of empirical

results. The following discussion provides a reconstruction and a potential explanation of the phenomenon in theoretical terms, illustrating which elements are consistent with existing literature, and which ones can be interpreted as an original contribution.

**Figure 1.** Data structure.

#### **4. Results**

#### *4.1. Banking Sustainability*

Most respondents contextualized the current organizational reform process affecting DNB as a reaction to a situation of complacency. The secure oligopolistic position held by DNB within the Norwegian financial market ensures comfortable profit margins, superior to most European counterparts, margins that are not currently under significant threat from the local competition. The management however is conscious of the fact that such rosy situation is unlikely to last indefinitely:

"I don't think we're going to be out-competed, but I probably think the margins within some of our most profitable business models is going to come under increasing pressure then in the years to come, and that requires us to innovate our existing business models, we need to become more efficient in streamlining our operations." —Division Director IT Division.

Lack of profitability pressures and steady profits have led to somewhat underdeveloped innovative capabilities:

"We are lacking some parts of what I would call innovation competence, such as pure business model innovation, experiment design and execution. We have some individuals, but not enough for a company of this size. When you look at traditional business development, how to treat an idea, how to actually build up partnerships, how to build services and products, we have that in house, but using it for innovation is a different issue." —Leader, Digital Innovation, B2C.

Furthermore, innovative resources and competences are not only relatively scarce, but also unevenly distributed and isolated:

"Our innovation capabilities are kind of spread between business developers with different types of roles. And then you have other sides of the innovation capabilities, IT and people of operations in different service roles, like service designers or literally IT architects." —Senior innovation coach.

The interviews underlined how DNB lacked company-wide systems for knowledge integration and sharing [53]. Respondents reported how they often struggled to access and share specialized knowledge across departments in an effective manner:

"DNB actually does have a lot of insight, but we have difficulties in terms of access across divisions (because we are legally not allowed to) and the systems that we use." —Senior innovation coach.

In addition to knowledge, all managers interviewed have described IT resources as similarly scarce and difficult to mobilize, mostly tied up in the maintenance of inherited old IT systems. The challenge appears to be particularly pronounced in connection to innovative activities:

"Our IT department is at all times 100% sold out and over capacity, so that IT with today's setup does not have any capacity to assist in innovation work at all. It is due to the management model, and partly because IT has a rather heavy responsibility related to security, compliance, the Financial Supervisory Authority and other things." —IT Division Director.

#### *4.2. Implementation of Contextual Ambidexterity*

In reaction to these and other commonly perceived shortcomings, DNB has developed the "DNB Way of Innovation" strategy, in order to develop harmonic ambidexterity capabilities across the entire organization. This aim has been pursued through a series of top down reforms, supported by attempts directed towards a comprehensive organizational innovation culture transformation. Several top-down initiatives have been implemented on a company-wide level in order to become more innovative through the integration of more explorative methods, tools, and processes:

"I can't really say that we have a clearly defined innovation framework or strategy, but we do have some new initiatives (innovation boards). Moreover, we have something called 'The DNB Way of Innovation' with certain practices and tools that our innovation coaches should master quite well." —Section Leader, New Business Payments.

DNB is on a transformation journey from being a sheltered, oligopolistic financial institution to a global competitor in the digital innovation arena. This means that DNB must become more forward-looking and able to embrace failure as part of becoming more innovative. However, such a mindset shift is entirely at odds with traditional banking, a business model well-known for its conservative approach and strongly institutionalized aversion to failure. One of the informants reflected on this:

"We are supposed to depart from a culture in which we make 'safe' decisions towards a culture where failure is allowed, and initiative is rewarded." —Section Leader, New Business Payments.

Although the aim is clear, the path remains challenging, however, with significant tensions emerging within and between departments [54]:

"There are a lot of people at the IT department at DNB who do not dare to fail, because there has been no room for failure in the past. It has been a fear-based culture. Even though our CEO says we should dare to fail the organization is rigged towards risk aversion." —Leader, Accelerate Innovation—New Business.

#### *4.3. Ambidexterity Paradoxes*

Since DNB has only recently developed an innovation focus, with most of the processes still to be fully implemented, few employees have actual experience with innovation management practices. As many new tools and processes for innovation are introduced "from above", many informants still feel like their innovation projects are outcompeted on resources against regular projects. One of the respondents commented:

"Some of us are supposed to think innovative according to our scorecards yet we often get tied up to existing projects. I tell my employees to spend twothirds of their time on existing projects and one-third of their time on thinking forwards, but this is not how it works out in reality." —Section Leader New Business Payments.

The lack of dedicated resources and structures supporting innovative efforts from the bottom-up hampered the first attempts to apply the new strategy:

"I think the greatest potential for DNB is to innovate more on the things that we already have. We have a lot of new initiatives [... ] Yet, with few resources, time and capacity on continuous improvement there is a need for a better balance." —New Business, Accelerate Innovation.

One of the respondents reflected further on this part:

"We are often saying that 'hey, we would like to explore some of these opportunities' but then when you look how we invest, everything is core [... ] that is not necessarily the problem, but if you say that you don't want to just do that, and then you do it, that might be a problem." —Senior innovation coach.

In response to these and other challenges, the company has established a new division called "New Business". This division works as a support department for the entire DNB Group, contributing its innovation expertise to other divisions focusing on traditional business areas. Furthermore, dedicated innovation coaches have been hired, to provide support to bottom-up innovation processes. The collaboration, however, is more difficult than expected, as the divide between those who innovates and those who do not becomes more pronounced:

"you sit with one foot on the gas and one on the brake. So, then I kind of think it's either go or no go. Either you feel that you are allowed to work with innovation as an employee, or you are one of those who feel that you are hired to make sure things do not go wrong." —Division Director of Innovation Division.

#### *4.4. Blending Towards Reciprocal Ambidexterity*

In order to alleviate some of the tensions described above, DNB is currently undergoing a second reform wave. One of the informants noted:

"The competence in innovation in the bank is still extremely low [... ] That is why we have started this enterprise innovation program and created a separate division for design and innovation." —IT Division Director.

Another solution to the lack of competence has been to prioritize external relationships. Explorative units engage in comprehensive collaborations with universities, other financial organizations, digital startups, and talented young programmers. Exploitative units absorb the knowledge so attained through internal training and seminars. By collaborating with innovative partners, introducing innovation coaches, and launching more standardized exploration processes, DNB is supporting the implementation of its Way of Innovation with decidedly structural changes. DNB has purchased "Luca Lab" to bolster the innovative capabilities of its DNB Accounting business unit. In addition, DNB has partnered with "11: FS" to develop a consumer finance solution ("Foundry").

The data reveals that, despite the initial harmonic plans, during the process of implementation the organization has blended different modes of ambidexterity [6], introducing structural solutions both within and across departments. The co-presence of various ambidexterity modes resulting in distinct management challenges and paradoxes across units and departments. The company struggles to obtain mechanisms for linking and integrating different modes of operation. In this regard, previous studies [22] have emphasized that separate units may embody distinct strategies and operating logics and even cultures, yet need to be coordinated through a shared vision animating the senior management team's actions:

"We notice that in very many of the discussions we have, there is a cultural difference. I believe in creating a common culture, common goals and a common understanding. So, even if someone is working with an agile approach and others are doing more waterfall-type work that it is perfectly fine as long as we are working towards some common organizational goals." —Leader, Digital Innovation, B2C.

Achieving such synergistic fusion, however, necessitates that different units work interdependently in terms of ambidexterity:

"There is a gap in terms of implementing that innovation strategy and framework and process in the organization. My impression is that there is little awareness of the innovation strategy amongst managers and employees in general in the organization at large." —IT Division Director.

According to most respondents, the main challenge currently faced by DNB in the process to further develop its innovative capabilities lies in the fact that significant amounts of resources are spent on the idea generation and early development phases of new products and services, before any attempts towards validation are made, resulting in significant sunk costs.

"Now, people have a thousand ideas, and everyone works with them for a year, and then you come to the IT team and do not get the resources to make it. We need to become much better at prioritizing in early stages of projects and say no to others, so that the things we actually choose to do we do properly with the right resources." —Division Director, Open Banking.

This has led to the implementation of practices aimed at earlier idea selection:

"One of the things that we try to work with in the innovation team is the number of ideas we are able to kill, the so-called 'kill-index', which is quite important. We have to teach the business areas that bad ideas are bad quicker." —New Business, Accelerate Innovation.

The reform involves the creation of dedicated testing units and resource pools, accompanied by a concerted effort to promote a testing culture: "So we are really pushing the testing culture in the whole of DNB, building up a testing environment where you can actually, everyone and I am not talking about Private Market only, but that all DNB, IT, everyone is using the same platform, in terms of sharing their learnings."—Section Leader, Private Market—Digital Sales.

The informants have also mentioned the lack of a rewards system for innovation efforts: current incentives in the bank follow traditional practices. In this regard, O'Reilly and Tushman [55] emphasize the importance of rewarding cross-unit accomplishments tied to growth targets instead of specific unit goals. As incentive structures and operative realities drift further apart, the opportunities for new paradoxes to arise increase.

These findings suggest that DNB is transitioning towards reciprocal ambidexterity, with different units operating partially independent—attempting to obtain the synergistic fusion of complementary streams of exploitation and exploration as depicted by Simsek et al. [20]. According to Brix [19], the main difference between reciprocal ambidexterity and partitional ambidexterity is that reciprocal ambidexterity is characterized by processes of transferring knowledge back and forth between two separate units. Simsek et al. [20] assert that this type of ambidexterity is more likely to emerge in complex environments—which certainly is descriptive for DNB as an organization.

By transitioning between multiple modes of ambidexterity several paradoxes have appeared to the surface. Structural reforms and company-wide change processes take time to complete, and unforeseen challenges are almost inevitable in a large an organization as DNB. Unclear mandates for innovation, uneven distribution of resources, especially IT, and the lack of a distinct reward system for innovation exemplify some of the paradoxes and inconsistencies that have emerged in a transitional ambidexterity process. We illustrate the trajectory path of transitional ambidexterity at DNB in Figure 2.

**Figure 2.** Suggested trajectory of transitional ambidexterity at DNB; from harmonic to reciprocal. The figure is an adapted version of organizational ambidexterity typology, as introduced by Simsek et al. [20] (p. 868).

#### **5. Discussion**

The implementation of the DNB Way of Innovation strategy is ongoing and, given the experimental, reflexive, and flexible approach undertaken by the top management in its realization, its long-term results cannot be fully appreciated today. What the DNB case can teach us, however, is how, while ambidextrous capabilities can be developed and an ambidexterity strategy planned, its implementation will confront management with certain realities imposing a specific direction to organizational reform, no matter the original intentions. Not only are these findings relevant for our understanding of the development of ambidexterity through time, it also matters for our analytical application of the various types of ambidexterity discussed in the literature. The emergence of specific ambidextrous practices through time have also important implications for the question of the sustainability of banking business models in the long term.

In this regard, it is notable how neither present profitability, nor a strong market position, have blinded DNB management to the need for developing an ambidextrous stance in order to ensure long-term economic sustainability. All our informants, no matter how critical of the bank or its strategies, agreed on this point. A similar consensus was obtained in regard to the satisfactory performance of the exploitative activities of the firm, and its much less adequate explorative capabilities. Consequently, to achieve an ambidextrous stance, a comprehensive reform of the innovation processes of DNB was needed; this is what the DNB Way of Innovation is attempting to deliver.

The key aim of the new model is to achieve higher organizational explorative capabilities by fostering the development of individual innovative ability among the general workforce and, over time, within every business unit [56]. Employees are asked to devote an increasing amount of their time and attention to explorative activities in addition to routine exploitative tasks. Management is asked to support this shift by integrating a variety of new tools and processes developed to support bottom-up innovation processes, by allocating resources to such initiatives and by supporting the gradual diffusion of innovative expertise and an innovation-friendly culture. While digital technologies are involved in most innovative projects, the strategy itself is agnostic regarding content: any proposal able to satisfy the requirements defined by the strategy and to gain the relevant managerial support is welcome. The aim is to improve the general explorative capabilities of the firm, rather than achieving a specific technological or market objective; a choice not without consequences.

Although not labeled as such, the initial strategy has clearly been one of fostering contextual, or harmonic, ambidexterity among several business units. This aim has been pursued through the implementation of new practices and instruments, meant to both support the development of existing innovative projects and ideas, and to foster a more general shift toward an organizational culture more supportive of innovative individual initiatives. The chosen approach reveals an understanding on part of the reformers of the importance of a supporting organizational culture for contextual ambidexterity to be successfully developed, as argued by Wang and Rafiq [21]. However, the implementation of this consistent and theoretically sound plan reveals some inherent limitations.

Contextual ambidexterity encourages bottom-up innovation initiatives. However, especially in the initial phases of a transition towards a more established ambidextrous modus operandi, the burden of such efforts falls almost entirely on the shoulders of the innovative individuals involved. The paradox of ambidexterity highlighted here is that, while individual initiative is encouraged, it is only rewarded in case of success, if at all. The rate of success of innovative projects, however, is always very low; in a context in which explorative capabilities are lacking, and new instruments still being experimented with, the probabilities of success can be described as abysmal. Thus, those individuals open to innovation, receptive to encouragement, and hard-working and risk-tolerant enough to try and kickstart an innovative idea while still keeping up with their daily routine, are likely to encounter uncertainty, frustrations, and, ultimately, failure [57].

This undesirable outcome can be managed, but not entirely avoided. Establishing instruments, practices, and cultural norms able to support reiterated individual failure is a necessary goal on the road of establishing contextual ambidexterity capabilities. In the initial stages, however, the goal of fostering bottom-up initiatives in order to test the approach overrides any other concern, leading, in the best-case scenario, to a proliferation of ideas and new projects to be managed. The newly established and still experimental practices and tools are unlikely to be successful in providing sufficient support to all of them. Selection is necessary and yet, paradoxically, undesirable [58]. Employees are already asked to promote innovative ideas in addition to their normal tasks: to push severe requirements on such ideas is likely to result in very few projects being generated. While the low number of projects may very well be in line with initial managerial capabilities, an optimal outcome even, it may also delegitimize the ambidexterity reform, undermining the proponents in the eyes of top management. Better to let a hundred flowers bloom; the cruel harvest may come later.

Other paradoxes emerge in regard to the issue of organizational culture. If individual employees are to be free to pursue explorative activities, their managers must be willing and able to support such decisions. This requires the adoption of a perspective favorable to contextual/harmonic ambidexterity, and the acceptance of the organizational value of the new tools and practices which such approach entails. However, different business units will have different perspectives on the relative importance of bottom-up innovation processes for the sustainability of their operations. Thus, the single innovation vision promoted from the top will shatter into several different interpretations during the process of implementation. The understanding of what innovation is, how it should be pursued, and what it could achieve was extremely heterogeneous across organizational lines. This is not simply a matter of opinions: the role of management in supporting contextual/harmonic ambidexterity is key [59]. While new tools, practices, and lingo can be adopted as a result of a push from the top, their effectiveness will vary greatly according to the actual support that local management is willing to offer to the new innovative projects, a key factor in the tender early stages of the process.

However, these cultural differences are not created by reform attempts, but rather brought to light. The resulting discussions and attrition play two constructive roles. First, they promote a debate that, in the long term, can lead to a more homogeneous organizational culture. Second, managerial criticism may reveal flaws and limitations of the original strategy, either specific to a given business area or general, an input that, if constructively received, can lead to general improvements in ambidexterity capabilities. The paradox lies in the fact that, in the short term, attempts to foster a homogenous innovation culture will highlight the fractured condition of the organization [60].

Ultimately, these tensions will lead to a struggle for resources, the practical embodiment of the conflict between exploitative and explorative activities [61]. The initial push supporting bottom-up initiative provides enterprising employees with the possibility to allocate some of their work time to ideating an innovative project. Any serious development, however, would require additional resources; however, under contextual/harmonic ambidexterity, resources are not specifically allocated to exploration or exploitation, they are flexibility redistributed according to needs. This means that bottom-up initiatives must constantly leverage their resources into access to more resources, climbing the managerial pole one meeting at a time. While the initial competition will be with other innovative initiatives, eventually explorative and exploitative needs will come to a clash. From an organizational perspective, this conflict is both inevitable and desirable: the ability to constructively and efficiently solve such conflicts is contextual ambidexterity. It is clearly an unfair fight, as the necessity and benefits of exploitation are well known, while the gains of exploration are by nature uncertain. Furthermore, resources for innovative projects often must be gathered across organizational lines, with the resulting organizational culture differences further stacking the deck against the explorative alternative.

The situation is worsened by the early tolerance necessary for the testing phase to begin in earnest. A hundred flowers bloom, but very few give fruit: the early impression of vigor later gives rise to disappointment among the would-be innovators and a general perception of wasted time, efforts, and resources among all involved. Two solutions to this problem can be adopted; in our case, both roads have been travelled. The first is providing more support to the innovators, to allow them to fight more effectively for the resources they need. The second is to create resource pools dedicated exclusively to explorative purposes, thus alleviating the need to compete with exploitative activities. These solutions to the quintessential contextual ambidexterity issue, however, are clearly structural in nature [62]. The former implies the creation of groups and organizational divisions dedicated entirely to explorative support; the latter makes the critical allocation of resources decision an ex ante managerial prerogative, coming from the top rather than being endogenously defined by bottom-up initiatives. While such an organizational response is not necessary, the issues

it addresses are implied by any contextual/harmonic strategy. The choice appears to be either to accept the paradoxes of contextual ambidexterity, and the organizational costs they imply, or to try and fix the situation by blending in structural solutions.

The contextual/structural axis is not the only one that begins to blur, however. Integrating a time perspective, it is possible to see that the DNB Way of Innovation intended to pursue a harmonic strategy, integrating exploration as a continuous focus over time [63]. However, the bottom-up, decentralized nature of contextual ambidexterity, combined with the broad integration of such practices over all business units, implies that the actual ratio of resources engaged in exploration vs. exploitation cannot be planned for in advance. Since most initiatives can be expected to fail in their early stages, most resources can be expected to be spent on exploitative routines. However, exceptional projects will arise from time to time, commanding a significant amount of internal resources. To be realized, such projects will eventually need to access resources across business units' lines, forcing other units to play a supporting role. At the same time, other units will have to pick up the exploitative slack, covering for the missing input for the temporarily explorative unit. Furthermore, as the innovation process develops, the initial unit may de-escalate its involvement, as other units with different capabilities and responsibilities take over the lead.

From a theoretical perspective, this description can be summarized in the following proposition: the implementation of a harmonic ambidexterity strategy within a large organization will lead to reciprocal ambidexterity in practice. The process will be supported by a gradual blending of contextual and structural elements, the latter introduced to diffuse the tensions emerging from the former. The argument can be seen as an extension of Andriopolous and Lewis' [64] argument in favor of a multilevel approach: structural and contextual ambidexterity are complementary strategies, covering each other's limitations. This does not mean that the resulting blend is devoid of paradoxes and tensions: our results already reveal several emergent issues. It is likely that such issues will drive further changes in the strategy. This reinforces our argument further: implementation challenges, and their solutions, drive the development of ambidexterity capabilities, in an evolutionary, path-dependent process illustrated in Figure 3, in which teleology can play only the role of trigger [65]. However, does this process ensure corporate sustainability?

**Figure 3.** Transitioning from harmonic to reciprocal ambidexterity.

Of the various components of the corporate sustainability construct, economic, environmental, social, operational, and strategic, all are involved within the process under study, with one key exception. Strategic sustainability can be seen as the main aim of developing ambidexterity capabilities. Economic sustainability is improved by ensuring that a concern for maintaining the competitive advantage over time is diffused among all organizational units and levels. Operational sustainability is supported by the reflexive practices supporting the development of innovative capabilities. Social sustainability is also improved, through expansion of collaborative explorative initiatives involving external firms and organizations and by integrating customer feedback in the early stages of the innovation process. What remains outside this process, however, is environmental sustainability. The ongoing digitalization effort will have some modest positive consequences on the environmental blueprint of the immediate bank activities, but this is accidental: as discussed above, the ambidexterity process is technologically agnostic. We have found

no evidence of significant environmental sustainability concerns at any organizational level, nor was the issue mentioned as particularly significant for the future of the firm at large. This blind spot will need to be addressed in the future, as the relevance of the environmental aspects of sustainability will come to play an increasingly larger role in the future [66].

#### **6. Conclusions**

The literature on ambidexterity has identified a number of alternative solutions to the need of the firm to simultaneously engage in exploitative and explorative tasks in order to secure its sustainability. Recent research has suggested that these solutions may not be alternatives, but they can be in fact simultaneously pursued, in different mixes. We add to this result by supporting the hypothesis that the limitations inherent in a specific ambidextrous solution, namely, contextual ambidexterity, create conditions that both invite and facilitate the introduction of structural solutions. Thus, a reflexive process of implementation is likely to result in blending harmonic and partitional approaches. Furthermore, we have provided evidence suggesting that, over time, the same process of addressing emerging practical issues may lead to the emergence of reciprocal ambidexterity, the most complex configuration identified by the literature.

Our research contributes to the existing literature by adding a temporal dimension to the analysis of blended ambidexterity, and by integrating the dynamic concept of blending with the static conceptual taxonomies of the ambidexterity construct. From a practical perspective, we argue that, in dealing with an inherently paradoxical phenomenon such as ambidexterity, internal consistency in planning should be sacrificed to operational flexibility and the emerging need revealed by the process of implementation: the resulting hybrid may well be more successful than its theoretical pure counterpart. From a theoretical perspective, we analyze the phenomenon of blending over time, and its consequences for our conceptualization of the phenomenon of ambidexterity. Furthermore, we question the analytical validity of imposing dualistic structure to a phenomenon that appears to be empirically dominated by crossed lines.

The research is characterized by a number of limitations. Firstly, we did not interview lower level employees; A more comprehensive approach in the gathering of data would strengthen the research and open new vistas on the bottom up struggles that characterize any attempt at fostering contextual ambidexterity. Secondly, although clear reciprocal signs are in evidence, the stability of this configuration cannot be taken for granted. On the basis of current evidence, we believe that such configuration may be reasonably assumed to be the final form that ambidexterity will take in DNB, at least in the medium-term. The validity of such assumption should however be tested by future research. Third, while the role of ambidexterity as an antecedent of sustainability is in our view sufficiently established, the actual mechanisms linking ambidextrous capabilities with the development of sustainable business practices should be subject to more direct investigation, something we were prevented from doing by the presently modest attention dedicated by the subject of our study to these concerns.

**Author Contributions:** Conceptualization, B.C. and R.S.R.; methodology, R.S.R; formal analysis, B.C. and R.S.R.; data curation, R.S.R.; writing—original draft preparation, B.C. and R.S.R; writing—review and editing, B.C. and R.S.R.; visualization, B.C. and R.S.R; supervision, B.C.; project administration, R.S.R.; All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** We would like to thank Amir Saed Ali, Arman Muzaffar, and Rehan Yousaf for their help with data collection.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A. Interview Guide**

#### **Resources: People and competence**

How is the composition of people and competence in your division/section when it comes to innovation?

#### **Time and resource allocation**

What are your views on how time and other resources are allocated to activities related to innovation, exploration, reflection and problem solving, in your division?

#### **Tools and frameworks**

Are there any tools or frameworks available to manage and evaluate innovation? **Understanding of innovation**

Do you perceive that DNB and your division understand what innovation is, its significance and how to work on it?

#### **Competition and market position**

Does DNB, and your division, know how the market is structured and how to compete

#### in it? **Processes**

Do you perceive that your division is making sufficient efforts to explore new business or technological possibilities?

#### **Incubation**

Are there effective ways to incorporate new ideas or solutions into the organization and subsequently implement them in products or customer offerings?

#### **Learning and unlearning**

Has the divisions worked to promote, rearrange or change the competence and capabilites in the organization to become competitive in a changing market?

#### **Bureaucracy**

Do you perceive that the inherent organizational bureaucracy is inhibiting how you work with innovation?

#### **Innovation strategy**

Do you perceive that there is a clear strategy for innovation, and is it known among the employees?

#### **Strategy vs. operational**

Do you feel that the innovation activity in your divisions is derived from or linked to one common innovation strategy in the organization?

#### **Culture for creativity**

Are creative solutions or unconventional ideas encouraged—and are they explored or implemented by your division?

#### **Organizational dependencies**

How does your division handle issues of being relevant both technologically and business wise—and how do you consider future opportunities?

#### **Decision basis and rules for innovation**

Do you perceive rules for decision making and evaluation of new ideas to be clear and suitable for enabling innovation?

#### **External relations**

How does how your division collaborate with or evaluate external actors to enable innovation?

#### **Clear decisions and division of responsibilities**

Is it clear and appropriate how the mandate and responsibilities for innovation are assigned in our division?

#### **In general**

What do you think are the most important factors that enable/inhibit innovation in your division?

#### **References**

1. Atuahene-Gima, K. Resolving the Capability–Rigidity Paradox in New Product Innovation. *J. Mark.* **2005**, *69*, 61–83. [CrossRef]


66. Nizam, E.; Ng, A.; Dewandaru, G.; Nagayev, R.; Nkoba, M.A. The impact of social and environmental sustainability on financial performance: A global analysis of the banking sector. *J. Multinatl. Financial Manag.* **2019**, *49*, 35–53. [CrossRef]

**Florian Diener \* and Miroslav Špaˇcek**

Department of Entrepreneurship, Faculty of Business Administration, Prague University of Economics and Business, 137 00 Prague, Czech Republic; miroslav.spacek@vse.cz

**\*** Correspondence: florian.diener@vse.cz; Tel.: +49-(0)151-4053-2978

**Abstract:** The digitalisation of banks is seen as the omnipresent challenge which the banking industry is currently facing. In this digital change process, banks are facing disruptive innovation that requires adaptation of almost all cooperative processes. Digital transformation in the financial industry is associated with obstacles that seem to hinder smooth implementation of digital approaches. This issue has not been adequately addressed in the current academic literature. The main purpose of this qualitative exploratory study is to identify the main perceived obstacles to digital transformation in both the private and commercial banking sectors from a managerial point of view and to analyse them accordingly. The methodology is based on a methodological approach using a combination of contextual interviews with German board members of banks, inductive content analysis, and the exploration of best-practice approaches. The findings revealed that elements of strategy and management, technology and regulation, customers, and employees receive a high level of attention within the digital transformation. The other main barriers can be found in the areas of market knowledge and products, employee and customer participation, and public benefit. Each main barrier is characterised by several sub-barriers of varying importance for the digital transformation of banks and is described in detail.

**Keywords:** bank; barriers; digitalisation; management; perception; transformation

**Citation:** Diener, F.; Špaˇcek, M. Digital Transformation in Banking: A Managerial Perspective on Barriers to Change. *Sustainability* **2021**, *13*, 2032. https://doi.org/10.3390/su13042032

Academic Editor: Andrea Pérez Received: 28 December 2020 Accepted: 4 February 2021 Published: 13 February 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Over the past several years, digital transformation has received considerable attention in the areas of management, business, information systems, information technology, and marketing. The developments in information and communication technologies in the digital age have significant and varying effects on organisations. Changes in traditional business ecosystems have created new business environments called "digital business ecosystems". Changes in the business ecosystems affect the strategic decisions of the organisations related to the internal and external environment. The size and frequency of these changes are the parameters that make the concept of change more meaningful [1]. The rapid development of technology, as well as a great variety of changes in today's global marketplace, have led to the intensification of a new cooperative adaptation process. This digital transformation and the adoption of new technologies raise a growing number of questions about the changes that traditional companies, strategies, and management practices need to implement in order to respond to them [2]. This response involves the creation of new, innovative business models and/or changes and improvements to the existing business models with the help of digital technologies [3]. The application of these new technologies and their appropriate implementations to improve business performance is an important issue for companies [4,5]. Today, complex transformations affect many dimensions, including strategic direction, competitiveness, business model, decision-making, innovation itself, entrepreneurship, productivity, and customers [6–8]. Based on further continuous development and ever-increasing digitalisation of lifestyles and changes in

customer behaviour [9–12], the world has become more and more informed, transparent, and efficient. Thus, new sales and service markets are coming into being, with continuous changes in technology and customer behaviour [9–12]. As a result, traditional business models in multiple industries are not just competing with each other, but also with new models that are external to their business fields [13]. To satisfy these market-based changes, enterprises have to adjust by reconsidering and reforming their traditional base [14,15].

Industrial companies, various administrations, educational institutions, the financial sector, etc., are all undergoing digital changes, which have a noticeable impact on them. One of the main drivers of digital economy development is the financial sector, which takes the second position, just behind telecommunication [16]. The key underlying process is the digital transformation of financial service systems through financial technology (FinTech)—disruptive innovations by new market entrants that challenge the position of mainstream financial institutions [17]. In particular, retail banks have been at the forefront of technological revolution, characterised by rapid deployment and innovation of digital services, exponential pace of change, and innovative breakthroughs that alter conventional banking practice [18].

The main problem in banking seems to be that traditional financial service providers have not yet implemented comprehensive digitalisation [19–22]. As a result, they often offer an incomplete range of services and are confronted with both strategic and operational barriers within the digital transformation process. In contrast to established service providers, there are new, innovative competitors with new concepts, products, and services [23]—and, above all, with a modern multi-channel approach in terms of distribution, communication and marketing, which approaches customers in a variety of ways. As a result, multi-channel business models have gained substantial market shares, as, for example, in the case of the German company N26 [24]. Recent research confirms the growing and prospective influence of these business models on the finance industry [25]. As traditional companies and their industries adapt slowly and ineffectively to the modern changing markets, there is a high risk of disruption caused by new technologies and business models [26,27]. "If such changes are missed by system-relevant financial institutions, such as large banks or groups of smaller ones, then financial services and the whole economic system will be endangered" [28]. Many existing financial service providers have already recognised the need for basic changes in their business model, and have started to rethink, or rather reform, their approaches [29,30].

This aspect of market-driven technological change, in particular, raises the overall question of how new entrepreneurial approaches, management behaviour, and technology changes in the world of banking, a long-established financial system, as well as how they influence and change banking adjustment, thus leading to the following two research questions (RQ):

RQ1: What are the main barriers to smooth implementation of digitalisation in banking?

RQ2: What are the "best practices" that are applicable in the implementation of the digitalisation process?

Due to a complex adjustment process within the financial system and its all-encompassing entrepreneurial influence, the identification and analysis of obstacles that hinder digital adaptation in the context of an all-encompassing digitalisation is of great institutional importance. For this reason, this paper contributes to the issue of digital bank transformation and identifies obstacles to digital transformation in the sector from the perspective of the management, as the management is ultimately responsible for appropriate bank development and long-term business success. In line with this research objective, the present study identifies and analyses implementation barriers to digitalisation using a methodological approach based on a combination of contextual interviews with bank executives, inductive content analysis, and exploration of multiple best-practice approaches.

#### **2. Literature Review**

Digital transformation is a holistic concept that includes technologies, as well as organisational and strategic changes [31]. Moreover, it is the process that an organisation goes through when it changes from an outdated approach to new ways of working and thinking by using digital, social, mobile, and new technologies [32]. It is driven by the advancement in technology, the appearance of new business models, and changes in expectations of the customer [33]. Several additional definitions of the term "digitalisation" are now commonly accepted. According to Gartner [34], digitalisation represents improvement of existing business models, creation of new revenues, and value-adding opportunities with the help of digital technologies. It can be understood as a complex issue that encompasses several areas like (i) shifts in thinking, (ii) changes in leadership, (iii) technology adoption, (iv) digitalisation of resources, and (v) acceptance of innovation [35]. As became apparent from the preceding classification, the term "digitalisation" should be distinguished from the similar term, "digitisation"; the former rather addresses the impact of digital technologies on the organization, while the latter represents the shift from an analogous solution to a digital one. Digitalisation is organizational renewal through new information and communication technologies [36]. According to Matt et al., digital transformation is a complex issue that proceeds within a framework that includes (i) changes in value creation, (ii) structural changes, and (iii) use of technologies and financial aspects [31]. It is no surprise that digital transformation seems to be blocked by a set of barriers that may hamper or even collapse the whole process of transformation. Digital transformation is considered a driving factor that offers a solution to the challenges currently faced by the banks. The core digital transformation practices, such as leadership, digital trends, digital transformation skills, digital strategies, implementation of digital technologies, and a customer-centric approach, are seen as influences brought to bear on digital maturity levels [37].

The term digital transformation (sometimes nicknamed digital entrepreneurship) is often misunderstood as a straightforward deployment of the latest information and communication technologies. In practice, technological investments entail not only risk, but also require an understanding of the relationship between technological and organisational culture and institutional change within certain boundaries of regulatory frameworks. Digital transformation is far from simple, certain, or predictable. Moreover, it is likely to be disruptive or transformative, with immutable impacts upon associated organisational outcomes related to technical capabilities and behaviours [18].

In the face of the established regulatory standards known as Basel III, banks aim to embark upon new technology standards, like Regulatory Technology (RegTech), which may facilitate digital transition. RegTech is an emerging technological trend that leverages information technology and digital innovations that can greatly assist with a bank's regulatory management process. It is advisable to incorporate RegTech into the digital transformation strategy of a management function, such as a treasury. Integrated adoption would mean that the digital platform can be deployed to support both strategic management activities and enhanced regulatory processes within the treasury. With this arrangement, commercial and prudential objectives are put in alignment [38].

Digitalisation plays a major role in contributing towards the United Nations Sustainable Development Goals. Without transformation of existing businesses, both economic and environmental challenges of the future cannot be solved sustainably [39]. Digital transformations will produce new social groups—partly human, semi-human, or nonhuman—some of which already exist, and some which can be foreseen by extrapolating from recent developments in the field of brain wearables, robotics, and software engineering. Growing dependency on digital services and tools may pose problems for both individuals and organisations [40]. Forcadell et al. [41] argue that digitalisation entails challenges that can hinder the potential benefits and compromise their survival. That is why corporate sustainability plays a significant role in enforcing digitalisation. It may compensate for drawbacks of digitalisation. In particular, the combination of corporate sustainability and digitalisation helps transform the organisational nature of banks by

simultaneously narrowing their boundaries and expanding their scope. El Hilali et al. [42] drew attention to possible ways of reaching sustainability during digital transformation processes. They found that the companies achieved sustainability when effectively mastering customers, data processing, and innovation. On the other hand, they did not prove that the competition played a significant role in enhancing the companies' commitment to sustainability. This opinion was partly endorsed by Ordieres-Meré et al. [43], who confirmed the positive effects of knowledge creation facilitated by direct or indirect application of digitalisation. Technology is reported to disrupt the financial industry, solve friction points for consumers and businesses, and make the overall business more resilient and sustainable. Sustainable financial technology may contribute to the overall stability of the financial system as well [44]. Established technology-based business models can act as a sustainability catalyst to trigger collaborative innovations between traditional financial and banking institutions [45].

Effective risk management, including its diversification, is also ranked among the contributors to sustainable bank development in the global economy [46]. Examination of risk-mitigation strategies in Southeast Asia proved that the banks had become more sustainable when implementing viable risk-mitigation strategies [47].

Nevertheless, when it comes to the barriers to change in the implementation of digitalisation in the banking sector, few resources can be found in the literature. It is evident that the banking sector is changing and institutions have to adapt to new technological developments and customer behaviour. This trend is particularly evident in the increasingly digital user behaviour, as mentioned in Table 1, to which bank executives have to respond.


**Table 1.** Share of payment instruments in Germany.

Note: <sup>1</sup> CAGR (Compound Annual Growth Rate) calculated for 2011–2017. <sup>2</sup> CAGR calculated for 2014–2017. <sup>3</sup> Possibility of bias due to different sample sizes (*n*) and individual transaction levels. Source: Authors' own representation based on Deutsche Bundesbank [9–12].

> Financial technology (companies called FinTech(s)) plays an essential role here. It is an industry composed of diversified firms that combine financial services with innovation technologies offered to financial service providers [44]. Shin and Choi [48] define FinTechs as platforms for the development of sustainable economic growth as well as a prompter of the fourth industrial revolution. These types of companies have several advantages over traditional banks. Typically, FinTechs may also provide a solution for sustainable finance through microfinance or crowdfunding, among others. Moreover, some FinTechs distribute insurance and other financial instruments or provide third-party services. Fin-Techs promise to disrupt and reshape the financial industry by cutting costs, improving the quality of financial services, and creating a more diverse and stabler financial landscape. Their existence is driven by sharing and the circular economy, as well as favourable regulation, and information technology [44]. FinTechs have the potential to unbundle core activities of the banking sector: clearing and settling payments, performing maturity transformations, sharing risks, validating trust, and allocating capital. They have also

created a new paradigm in which information technology represents a meaningful driving force that gives rise to innovation [44].

Hereby, banks are under massive pressure to transform their approaches and business models to a more customer-centric approach in order to remain competitive. The traditional institution has felt the disruption and is working towards changing its business model from product-centric to customer-centric [37]. Similarly Mărăcine et al. [49] suggest that five main areas exist where FinTechs can provide improvements in business models for the banks: introducing specialized platforms, covering neglected customer segments, improving customer selection, reduction of the operating costs of the banks, and optimisation of the business processes of the banks. As digital banking offerings have matured and cost pressures have increased, it has become inevitable to make changes to the operating models of banks. Driven by the sub-optimum performance of the existing business model, the "digital" concept has evolved into more than a channel for accessing services. One of the outcomes was a full-fledged branchless digital bank [50] or challenger bank. A challenger bank stands for a financial institution that can be presented in the plain form of an information–communication system [16].

Sadigov et al. [51] have proved that FinTech development contributes to economic growth by increasing the GDP generated in the financial sector, and indirectly does so by increasing e-commerce turnover and real sector financing, particularly by creating more favourable lending conditions for small and medium-sized businesses.

As has become evident, business models adopted by FinTechs differ from those applied by traditional banks. Nevertheless, these differences do not mean that both types of banks may eventually converge towards a common market by exploiting co-operation strategies. Their business model is intangibly driven, combining e-finance, internet technologies, social networking, artificial intelligence, blockchains, and big data analytics. Moreover, their revenue model is much more scalable than that of a typical bank [44].

Given the lack of literature on banking and existing research that followed a similar approach to identifying implementation barriers, those by Chan [52], Chan [53], Vikneswaran and Anantharajah [54], Kamalulariffin et al. [55], and Yusof and Jamaludin [56] have to be considered; barriers arise in connection with the implementation of new strategies and management approaches. Given that these authors have already properly elaborated and investigated the barriers to the implementation of new strategies, it is important to take their research approaches into account.

In addition, the questionnaires they used have already been partially validated and can, therefore, be a sound basis for this study. For some barriers, however, their questionnaires need to be reformulated or reworded, as they only allow a theoretical approach and do not fully correspond to the specific terminology needed for this work. For example, the study by Kamalulariffin et al. [55] focussed on environmental management in the hotel sector; a closer look at the research findings revealed that this industry is facing a situation similar to that which financial institutes are facing today, with new strategies and business models being pursued internally and by competitors. In particular, new business models are being developed that have never been established in their market before, thus satisfying customer needs in the latest way and, at the same time, endangering traditional business models.

Chan [52,53] already considered internal and external barriers, which he repeatedly validated through his work. These findings can be summarised as (a) implementation and maintenance costs, (b) lack of knowledge and skills, (c) lack of a sense of urgency, (d) the ambiguity of modern banking, (e) lack of qualified consultants, (f) lack of motivation and professional advice, (g) conflicting guidance, (h) outcome uncertainty, and (i) inconsistent support.

Kamalulariffin et al. [55] mentioned the barriers of (a) regulation and government, (b) customer demand, (c) level of competition, (d) cost of greenness at the organisational level, and (e) attitude toward change. Vikneswaran and Anantharajah [54] referred to (a) high maintenance and implementation costs, (b) lack of sufficient knowledge, (c) lack

of resources (time, manpower, equipment, and money), (d) lack of momentum from the company owners, (e) lack of a sense of urgency and ambiguity of guidelines, (f) lack of qualified verifiers or consultants, (g) conflicting guidance, and (h) lack of government regulations and enforcement, as well as (i) difficulty in operating an entity (difficult to balance the quality of service performance). From the content analysis of all the related literature, in summary, 12 pertinent barriers were identified by Yusof and Jamaludin [56]. Chan et al. [57,58] confirmed and extended these results again, which can also be interpreted in relation to banking, the associated digitalisation, and FinTech. Due to the more appropriate and transparent approach of Yusof and Jamaludin [56], their analysis is not considered holistically in this elaboration.

Due to the holistic nature of these works, these results served as a textual foundation for the preparation of the interview questions; consequently, they were derived mainly from the proven works of Chan [52], Chan [53], Vikneswaran and Anantharajah [54], Kamalulariffin et al. [55], and Yusof and Jamaludin [56], which are considered reliable and valid.

#### **3. Methodology**

#### *3.1. Data*

Due to their strong market positioning, the German savings banks and cooperative banks were the focus of this study and, thus, the focus of the data collection process. Both types of banks are equally ranked among the good service providers. They provide the majority of regional and supra-regional branches in retail banking and are the most strongly represented group in banking from a personnel point of view [59–63]. In addition, they offer an almost identical product range to their customers. Although they differ only marginally in their products and services, they differ on an organisational and structural level with regard to their business model [64].

#### 3.1.1. Interview Process

In total, for this study, 34 interviews were conducted with German bank managers more precisely, bank executives. Two of them had to be disregarded, as they did not fit into the relevant target group; thus, 32 interviews, with an average interview duration of 34 min, were considered for further evaluation. The valid interviews lasted between 22 min in the shortest case and 1 h and 7 min in the longest case; the total length of the interviews was 17 h and 53 min. Two of the 32 valid interviews were interrupted, either due to technical problems or due to interruptions in the person's environment, so in these cases, several recording files were created for each interview; however, this does not affect the validity and substance of the discussions.

These were determined as Table 2.


**Table 2.** Numerical interview data description.

Source: Authors' own representation.

#### 3.1.2. Data Preparation

The data were prepared by transcription according to "simple rules", whereby the audio recordings were transcribed word by word, but not repetitions, word deletions, or non-verbal utterances. Signals of understanding or confirmation, such as "mhm, aha, yes, exactly", etc., were not transcribed. The form of the transcription was based on Kuckartz [65] and Dresing and Pehl [66]. The interviews were transcribed verbatim, but dialectal variants were not transcribed, and slight dialectal utterances were translated in the standard language. Colloquial language was retained. The sentence form, definite and indefinite articles, etc. were retained, even if they contained errors.

#### *3.2. Analysis Procedure*

By means of an explorative interview framework [67], the main goal of this study was to generate impulses for an individual narrative of implementation barriers in digital transformation. In guided interviews, pre-defined questions were asked, but these could be answered very openly by interviewees; the procedure was less strict than in other interview methods. In a semi-structured interview, also called a guided interview, the interviewees are not given specific answers; they can report, comment, and explain freely. The advantage of this method is that, although the interviewer asks concrete questions by means of his or her questionnaire, the interviewee can answer openly and possibly focus the interview on new aspects and expand the entire interview.

Following Mayring [68], a theory-based analysis model was set up for the analysis, which was carried out by summarising and through inductive category formation. The selection of interviewees, as well as the number and scope of the answers given, is of crucial importance in the interview method [69]. As is the case in similar work on expert knowledge, the quality depends crucially on the selection of so-called experts—in this case, the interview participants [69]. Experts are understood to be individuals to whom knowledge of the surveyed topic area is attributed due to their activity and resulting practical expertise, as well as their specific educational qualifications [70]. This usually exceeds the knowledge of people who are unfamiliar with the topic being surveyed [71]. According to Chan [52], these are the managers and executives of an industry. For this study, interviewees are bank experts who are the actual decision-makers (the executive management) of a bank, with budget and personnel responsibility, as well as bank experts with specific knowledge and professional experience in the fields of banking, digitalisation, entrepreneurship, finance, financial technology and innovation, and many more.

A larger sample often leads to more confident and more reliable statements on what to look for [69]. The size of the samples for qualitative analyses is usually smaller than for quantitative analyses. Frequently, more accurate and more representative inferences about the population can be made in the case of large sample proportions; however, interviews will only be carried out as long as new information is perceived. In principle, the sample sizes should be large enough to obtain sufficient data to adequately describe a phenomenon of interest and to enable the research questions to be answered. The aim of this and all other qualitative studies is to obtain saturation of the sample; saturation occurs when the inclusion of additional participants does not lead to new perspectives or information. Glaser and Strauss [72] suggested the concept of saturation in order to achieve an appropriate sample size in qualitative studies. A number of guidelines have been developed for this purpose. Morse [73] suggested about 30–50 interviewees in ethnography for grounded theory. Creswell [74], in contrast, suggested only 20–30, and possibly as few as 5–25 for phenomenological studies, and Morse [73] suggested at least six.

The population is understood to mean the total regional savings and cooperative, and private banks. It can be assumed that at least one decision-maker/expert can be assigned to a bank. However, it may also be assumed that the actual population is much greater, as banks are not authoritarian institutions and their decisions are not made by one person alone; the prevailing ownership and organisation structures have an additional impact on a bank's business orientation.

In the context of explorative inductive content analysis with category formation, one can rely on work that has already been done. According to Mayring [75], the basic principle of inductive content analysis is that categories are derived directly from the respective research material in a generalisation process, without referring to previously formed theoretical concepts. When the terms "categories" and "barriers" are used in the following, they are synonymous and refer to the hurdles of digitalisation. Within the qualitative approaches, the inductive approach has great importance [76]. Its objective is to capture a naturalistic, object-like representation of the investigation material without distortion through presuppositions. This approach is a central process within "Grounded Theory" and is called "open coding" [75]. Within the analysis, this category-building process can be described as systematic, using a step-by-step and line-by-line approach. In this logic, the topic of category formation must first be determined on the basis of theory; i.e., a selection criterion is introduced that determines which material is intended to be the basis for further category definition. Insignificant contents are thus excluded from the analysis. The thematic question of the study is of great importance and in accordance with the main question of this study; it defines the focus of the content. Likewise, within this approach, the category dimensions and the level of detail have to be defined in advance, as well as the analysis units [68].

These were determined as Table 3.

**Table 3.** Category dimensions.


Source: Authors' own representation based on Mayring [68].

The analysis approach to inductive content analysis with category formation follows a predefined process model, which is outlined in Figure 1.

**Figure 1.** Sequences of inductive content analysis with category formation to illustrate the method of qualitative analysis according to Mayring [68,75].

In the analysis, taking into account the level of abstraction and the category definition, a suitable text passage is identified in the test material during the analysis, and a category is constructed. A term or phrase that comes as close as possible to the material is then used as the category name. Whenever a new passage is found in the further course of the text analysis that matches the selection criterion and category, it is also assigned to the same category. This is also referred to as "subsumption". However, if it turns out that no assignment to already existing categories is possible, a new category is inductively formulated from the specific material [75].

After a certain percentage of the material (often at 10 percent to 50 percent), when almost no new categories can be created, this is the moment for revision of the category system [68]. It is then necessary to check whether or not the logic is clear and the degree of abstraction fits the object and the question posed.

In the course of this qualitative analysis, the coded text passages are first paraphrased, then generalized, and finally form the actual category. The result of this process is a series of categories assigned to a specific topic and corresponding text passages in the research material. In the further course of the process, the interpretation is then made with regard to the overarching research question, taking into account the present approach and its findings. Due to the complexity of the interview topic, the respondents preferred to conduct the interviews in their first language. Given the availability of information, the generalised statements of the interviews were first formulated in German and later translated into English for this work. Thus, the internationality of the results is taken into account. In order to ensure the overall consistency of the generalisation, as well as the equivalence of the translations, a re-translation procedure was applied [77], carried out by a professional bilingual translator.

#### *3.3. Structure of the Interview Guideline*

The semi-structured interview guide was divided into a German and an English approach, depending on the interviewee. Both were based on previously extracted theoretical findings, which were used accordingly to interview decision-makers at banks. The barriers mentioned in Section 2 were fully taken into account. Furthermore, due to the different perspectives of the individual respondents, the guidelines were also adapted to each individual situation. They therefore differed slightly and were divided into two categories: banks and financial service providers. However, the basic structure of the interview guide was not changed. First, the interview topic was introduced and introductory questions about the person were asked. In the first section of the interview, the interviewees placed themselves in their respective positions in the company and described their level of knowledge on the topic of digitalisation in banking. This was to determine the suitability of each interview partner in advance. In the further course of the interview, questions regarding the banking sector and digitalisation were asked in detail, taking into account the findings of Chan [52], Chan [53], Vikneswaran and Anantharajah [54], Kamalulariffin et al. [55], and Yusof and Jamaludin [56].

#### *3.4. Conducting the Survey*

The guideline survey focused on interviews with decision-makers at banks. These were identified through personal contacts in the financial and banking industry and active approaches toward banking associations and local banks, as this is where access problems appear to be the lowest [70]. Furthermore, recommendations played an important role in the acquisition of interview partners. Consequently, further contacts with experts were established. Due to the geographical distance from the respective interview partners, the interviews were not exclusively conducted in person, but also via telephone or video conferencing [70].

In order to ensure the clarity of the individual questions, a pre-test with three test subjects was carried out in advance. As a result, the interview framework was confirmed, and no further adjustments had to be made. In order to comply with the applicable

provisions of the General Data Protection Regulation (GDPR), respondents were required to sign a consent form. The interviews conducted were recorded using the memo function of a smartphone. Here, it is important to mention that all respondents were interviewed regardless of ethnic and social origin, age, gender, sexual identity, religion, ideology, or other ethically questionable aspects [78]. Furthermore, all of these interviews were voluntary and the participants consented to the GDPR.

#### **4. Qualitative Evaluation**

#### *4.1. Consistency of Coding*

In quantitative content analysis, the term "inter-rater reliability" is generally understood to refer explicitly to the quantitative quality criterion of reliability. The term is connected to measurement theory and claims to be replicable. Here, a distinction has to be made between a possible agreement in the formation or direct application of an existing category system. Since the formation of inductive categories according to Mayring [68,75,76] is the result of a construction process, the formation of a category system cannot be claimed to be consistent [65]. A coefficient that measures the agreement between two category systems created by two or more different coders using the same data material says little about the quality of the category system. Rather, it could measure something that was not intended to be measured [65]—for example, the logic by which people create category systems. Accordingly, Kuckartz [65] concludes that the demand for agreement among coders thus refers primarily to the application of categories, i.e., the coding of data. However, the classical quality criteria for determining reliability in quantitative content analysis cannot be transferred to qualitative content analysis. Kuckartz [65] justifies this with the fact that, in quantitative content analysis, the coding units are defined before coding. In this quantitative case, relative matches and coefficients, such as Cohen's Kappa, Krippendorff's Alpha, or Scott's Pi would then be calculated [65]. In qualitative content analysis, the material is usually not segmented in advance, which is why Kuckartz sees two ways of identifying the concordance between coders, according to which the present study is oriented:


#### *4.2. Consensus Coding*

Subjective assessment is understood by Kuckartz [65], referring to Guest, MacQueen, and Namey [79], as a situation in which two coders encode a text independently and then compare it subsequently. This process is also called consensual coding [65,80]. Here, the second coder notes questions and problems that arise in the process and discusses them at the end of the individual coding session. Using the formulated category definitions of disputed codes, a coding is agreed in the best case and, if necessary, the category definition is revised. If no consensus can be reached, another person is consulted, who then decides on the controversial case [65].

Since the qualitative data collection resulted in a total of 32 valid interviews, the second coder was provided with a selection of interviews. Due to the large amount of interview material, a complete second coding seemed unreasonable for an external coder. It was agreed that at least 10 to 30 percent of the interviews should be independently coded a second time, as this seemed to be feasible in terms of the time and motivation required. In order to ensure an independent selection of interviews, the principle of drawing random numbers (1 to 34) was applied with the help of a random number generator from Random.org. Interviews 7 and 29 were omitted; the respondents did not fit the target group because their business model was different from that of banks. Thus, a total of six interviews (4, 13, 20, 24, 27/28, 31) were coded by the second coder, which means that about 19 percent of the entire data material was used to verify the

results. Furthermore, based on the sum of valid interviews, one is at least within the methodological revision range of 10 to 50 percent of the categories, which is consistent with Mayring's approach [68,75,76].

Within the coding process, the second coder was first introduced to the developed coding system and the category set, including all sub-categories. For better interview analysis, MAXQDA Analytics Pro 2020 (Release 20.0.8), a qualitative analysis software, was used for the actual coding process.

In a personal meeting, classifications were discussed and definitional assignments were reconsidered. During this process, all six interviews were discussed step by step, or coding by coding. It is worth mentioning that the already provided coding set did not require any improvements and could be used by the second coder without additional modifications, or interpretation difficulties. In addition, there were no problems with the coding. Based on the fact that the second coder experienced the coding system as quite complex during familiarisation with the topic and approach, the first two coded interviews were revised a second time at the end of the coding process to increase reliability. At this stage of the qualitative evaluation, no numerical analysis of coders' agreement was carried out, since the interviews as a whole, rather than individual sections, were the subject of discussion and appropriate review.

#### *4.3. Calculation of Inter-Rater Reliability*

Inter-rater reliability (IRR) is a measure of the level of agreement between the independent coding choices of two (or more) coders [81–83]. Of course, it is expected that the allocation is not arbitrary, but that it is done in such a way that a certain reliability is achieved. In qualitative research, it is important to improve the agreement and to discuss together where there are differences in coding and why these differences exist. The MAXQDA *Intercoder Matching* function enables comparison of the codings of two persons coding independently of each other. It supports determination of the consistency of coding and can be used to establish the deviation of a coder's choices from the ideal or "true codes" ("true codes" are those that garner general consensus among multiple coders). There are a variety of statistics that can communicate a measure of inter-rater reliability, but one of the most common (and most appropriate for the study in question) is Cohen's kappa coefficient [84]. Cohen's kappa is considered one of the most robust measures of IRR and is used widely in science [85]. It is calculated based on the percentage of consistency between two or more coding collections and accounts for the possibility of chance consistency.

For further investigation, it is recommendable to define in advance the segments or citations to be coded. In the present analysis, due to the large amount of data, the selection refers to entire interviews and not to individual segments in order to ensure a holistic approach. Interviews were selected randomly, i.e., six interviews (4, 13, 20, 24, 27/28, 31). Only if this is the case does it make sense to calculate a coefficient to determine the concordance [65]. Based on the selected interviews, the evaluation checked whether the two coders matched in the coding of the individual segments and whether conclusions could be made regarding the reliability of the overall coding. The IRR approach is the comprehensive and typical variant of qualitative coding [84]. Since texts in qualitative evaluation procedures are often not divided into fixed text units, the verification of conformity is carried out by default for each segment encoded by the two coders (evaluation: segments of both documents). A percentage value was defined when two coded segments were considered to be a match. The default value was 90 percent. For this study, the value was set at 60 percent due to the high level of detail of the code set and the number of possible codings, as this allowed for more precise evaluations during the later discussion of the results with Coder 2. With this approach, at the end of the evaluation, for every 60 percent overlapping coded segment, there is a match that can be used for further analysis.

To interpret the Kappa values, ranges from 0.61 to 0.8 are considered acceptable agreement, and from 0.8 upwards, almost perfect agreement [65,86–88]. Further analysis of the relative number of matching codes was also carried out (see Table 4). The "Percent-

age" column shows the percentage of matches per interview. This resulted in an overall percentage agreement of 84.66 percent. It was calculated as follows: Matches/(matches + non matches). In the "Kappa (RK)" column, the result table gives a randomly corrected value for the percentage match [83]. This takes into account the probability of two people randomly selecting the same codes in a document (if they would simply select codes randomly without considering the data material). The calculation only makes sense if the option "Unassigned codes as matches" is selected, which is the case here [83].


**Table 4.** Code consistency between documents.

Source: Authors' own representation based on MAXQDA.

In determining the kappa coefficient, "P observed" represents the simple percentage of agreement. The calculation of "P chance", the random match, is based on the calculation by Brennan and Prediger [89], who have intensively studied the optimal application of Cohen's kappa and its problems with unequal marginal sum distributions. Using this calculation method, the random match is determined by the number of different categories used by both coders. This corresponds to the number of codes in the "code specific result table". The calculation of Cohen's kappa of the randomly selected interviews resulted in a value of 0.68 after a renewed review of the coded sequences with Coder 2, which can be regarded as a substantial agreement and supports the code set (Figure 2).


**Figure 2.** Calculation of the overall kappa coefficient values based on MAXQDA, representing results by Coder 1 and 2.

When classifying the resulting characteristics of the IRR, it should be taken into account that the second coding was carried out under honorary conditions of a scientific research assistant to a professor and not under the conditions of a paid scientific research group. Against this background and the fact that six interviews were double-coded holistically and not just isolated sections, the reported results appear significant for further interpretation.

#### **5. Results**

The analytical approach enabled us to answer the two research questions formulated above, which are answered in more detail in this chapter. The first RQ referred to the contribution to theory, while the second referred to the contribution to practice.

#### *5.1. Contributions to Theory*

RQ1: What are the main barriers to smooth implementation of digitalisation in banking? From December 2019 to March 2020, a total of 32 semi-structured interviews were conducted with board members in banking. This study contributes to the literary expansion and the first definition of barriers to implementation of digitalisation in the banking market, which can be used for further research. A total of 63 codes were worked out according to Mayring's method, which form the preliminary main category set (see Appendix A). Due to the complexity of the analytical approach, only the steps following the generalisation of Mayring's method can be presented in the Appendix A. Detailed representations are not feasible due to graphical limitations. Corresponding definitions for the respective sub-barriers enable interpretation for future analyses. Main categories represent the superordinate classification level of the respective sub-category set, but are not characterised by an independent definition. Table 5 presents a summary of the set of categories that represent the main barriers to smooth implementation of digitalisation in banks, with respect to savings and the cooperative bank sector.

**Table 5.** Category summary.


Source: Authors' own representation.

#### *5.2. Contributions to Practice*

RQ2: What are the "best practices" that are applicable in the implementation of digitalisation process?

For this study, 34 interviews were conducted with bank executives on the topic of digitalisation in the banking sector, with emphasis on the digitalisation of their own banks. A total of 32 interviews were identified as valid for the case-analytical approach in this chapter and were used accordingly for the practical interpretation of the study results. A similar approach to interpreting interviews had already been followed in an earlier publication by Diener and Špaˇcek [90]. The bank managers were not asked about the problems they encountered in practice, but rather about their best-practice approaches to digital transformation. All interview partners were asked the same question for reasons of consistency.

#### *What concrete measures have you/has your bank taken in the past to keep pace with digital competition and the changing pace of digitalisation?* (Interview Question 10)

This question was deliberately kept general in order to give respondents the greatest possible scope for answering it. As this is also highly sensitive information, data protection had to be guaranteed and data protection rules had to be respected. In particular, the respondents insisted on anonymity. Due to their professional status, the interviewees quoted in this chapter explicitly stressed the wish not to be named or quoted in person. The same applies to their banks. Official interview material would require explicit permission by the bank before publication, which is not feasible in the context of this scientific analysis and would lead to biased answers. For this reason, the original transcripts of the interviews were used for this study and the respondents were only mentioned anonymously. The interviewees are therefore only referred to as "Interviewee" in the following.

The analysis of the interviews showed that all decision-makers were generally aware of digital change in banking and particularly aware of digital changes and the issue of digitalisation in their own banks. Banks are even hiring a board member specifically focused on digitalisation. "*[* ... *] many banks are now hiring a Chief Digital Officer. There is a new position that was actually created*" (Interviewee 20—Section 49).

This move towards more digital orientation and new approaches to customer service may lead to a complete shift in personnel structures in some departments. As a result, employees no longer work in the bank branches that have been known for decades, but are now able to offer a full range of services independently of their geographical location. This development takes into account the efficiency concept of the branch and the increasing competition, which, according to the assessment of the interviewees, will lead to farreaching structural changes within the banking sector in the future. However, with constant digital development, the question arises how a branch without customer traffic can be physically maintained in the future and what justifies the maintenance of cost-intensive branches. These descriptions from the interviewees indicate that they actively try to react to corresponding market developments within the scope of what is economically feasible and to develop further. One interviewee confirmed that his bank is actively addressing new trends and issues in order to take them into account in its corporate focus. *"In particular, we are very active in the field of trend scouting and in identifying and evaluating the strategic relevance of various trends and technologies. [* ... *] We have an innovation lab where we do trend scouting and observe about 170 trends and technology duration"* (Interviewee 26—Section 29). Another manager emphasised that the integration of employees into the thinking process of change is fundamental and that this should be secured and promoted by the necessary freedom of thought and continuous exchange between employees and management. "*[* ... *] we have set up a room of ideas where every employee, from trainee to board member, can say: 'I have a cool idea and would like to present it.' There is a group of supporters who—for example, when trainees come, are perhaps not yet very good at presenting—support the employees in creating a small presentation. And then they can present directly to the board of directors"* (Interviewee 14—Section 139).

The interviews also revealed that cooperation with external partners plays an important role in the further development of banks and that cooperation within the banking association is fundamental, but also leads to a slow-down due to increased structural complexity. Banks today cooperate with partners such as university institutions, which support them in the further and new development of applications, as well as future strategies, and provide them with the necessary know-how for digital corporate transformation. In particular, the participation of the target group at the university level, such as the actual developers of the applications, helps to align and develop the bank's range of products and services to the needs of the customer, which contributes to a tailor-made fit of banking solutions—at least for the younger generation of customers. One interviewee mentioned: *"In two years we have made a lot of progress in this area, and in the meantime we have widened the channels. In other words, we have provided a chat solution to improve customer contact. At the same time, a video service consulting branch is also starting this year. This means a branch in which there are no employees any more, but the branch is looked after centrally from the customer dialogue centre with extended service times from 08:00 to 18:00. [* ... *] What we have achieved for ourselves is to work together with universities, often here in the region, to design things in a way that is appropriate for the target group. [...] Together with a university, we have developed an app in the youth market"* (Interviewee 17–Section 79).

The digital transformation in banking is progressing at an ever-increasing pace. Far away from interface requirements forced by regulatory laws, just a few institutions are creating application interfaces on their own initiative, which enable them to integrate innovative business models and/or products. *"We also support via the Banking-API (Application Programming Interface) such innovations as Google Assistant or Alexa and many other things"* (Interviewee 21—Section 14). However, the main focus is still on realising stronger networking, which ultimately benefits both the banks themselves and the bank customers. It enables a more comprehensive range of products and services. Digitalisation, in this context, implies not only the development of internal approaches and ideas, but also the intended cooperation with other credit institutions, external partners, and FinTechs, which is to be enabled and realised via application programming interfaces "*[* ... *] so that we can also integrate our technical processes, which we have for product closures and transactions, into third-party platforms in order to network more closely with each other*" (Interviewee 21— Section 14). It seems obvious that the management has recognised digital transformation and is focusing on it accordingly in its banking activities.

In order to adopt and pursue new digital approaches, far from merely perceiving trends, managers are responding by elaborating in detail practical approaches that will facilitate and fundamentally enable future digital implementation. This process, however, requires the availability of appropriate resources, as implementation is ultimately only made possible by making them available. On this point, a large discrepancy is evident between small, medium, and large banking institutions—major challenges of a possible holistic digitalisation can be advanced together in a more targeted way. For a small bank, topics such as "quantum computing", which could be highly important within the next 10 to 15 years, simply cannot be tackled today from a financial point of view, since the costs exceed the available resources. It is crucial, though, that banks, regardless of their size, systematically address tomorrow's digital issues today, so that they can *"have the necessary know-how to deal with them proactively at the appropriate time. There is no doubt that banks definitely want to know what is happening and have a clear opinion on it"* (Interviewee 26—Section 29). Knowledge and the ability to react are important here; "*[* ... *] one has to be familiar with the complex issues in good times, otherwise one cannot react accordingly*" (Interviewee 26—Section 33).

Today's bank management attributes an important role model to the employees, as they implement digitalisation in the company in a targeted manner and bring it closer to the customer. It is, therefore, essential that employees are informed regarding technological applications and know how to apply them properly and safely. Employees and customers have to be taken along and introduced to the technology. Since this has to be done holistically and not only for specific target groups of customers and employees, it is important to differentiate between individual groups of both customers and employees and their individual age structures to introduce them to the technology in a targeted manner. Specific educational programmes and events for customers and employees are being introduced in banks to facilitate the implementation of digital approaches and, ultimately, bank digitalisation, as well as to enhance the acceptance and integration of employees and customers. *"Not long ago, we held a digitalisation fair just for our employees. [...] It cost a lot of money. So we did it professionally, with professional providers. Just to show our employees what's available. There were various stands with different focal points. So that you have topics that you can understand. Online brokers were there. There were 3D glasses where you can look at a house that we sell. And things like that. A multitude of things. And for us it is just important to take the employees with us. They know that digitalisation is destroying jobs [...] the trade fair has certainly contributed to the fact that the mood in our company is quite good at the moment. So the employees are going with us. One of the main problems is that some of our employees don't deal with things privately"* (Interviewee 27/28—Section 47).

According to the executive management, internal experts prove to be of great value, as they can deal with a specific digitalisation topic, communicate it to the respective individuals, and support them accordingly. "*For example, we have set up a programme with other partners, called Digital-Tiger, where we have specially trained one employee in each market area as a Digital-Tiger (an expert). He then serves as a multiplier, which also serves to encourage*

*the employees more [* ... *] We actually have seven experts in the entire company, e.g., in the corporate customer area and real estate centre. There is an expert in every branch office who is regularly trained. Then, in turn, he transfers the already existing knowledge and new knowledge or new products to the employees*" (Interviewee 13—Section 45). This development indicates that the relevance of a topic is becoming increasingly important and that the necessary financial and humanitarian resources are being mobilised in order to pursue and promote topics internally. However, this is, without question, highly dependent on the individual perception, understanding, and acceptance of the management on the topic as well as on the general financial situation of each individual bank. It also depends on the availability of trained employees, the actual specialists, as people have to be mobilised for (change) projects and show the will to change and participate. In addition, the respective corporate "change culture" within the bank is crucial, as employees also have to be prepared to be led by specialists and managers and should not be completely opposed to new methods and change. Only through such a culture can new topics or digitalisation be realised. New topics and, ultimately, digitalisation can only be realised holistically and effectively if these framework conditions are met. *"In a traditional company like ours, with people who are very security-oriented and very conservative in their attitudes, this is a challenge in itself. That means accompanying this human resource development or mentality development culture, development process"* (Interviewee 17—Section 29).

Furthermore, the technological framework conditions have to be in place for digital change and have to meet certain prerequisites, but this still fails because of problems such as IT infrastructure, both on the bank side and on the side of the infrastructure provided by the state. *"You first have to create the basis [...] so there's a big hurdle of servers and WLAN or LAN speed. You first have to create this for every branch in principle, for every stationary location [...]. And that is a big challenge, all the more in rural areas. [...] The wires have not been laid in the past. Twenty years ago, no one needed a strong internet connection. So now the question is: Is it possible from a constructional and technical point of view? What does it cost? In principle, that's where it starts"* (Interviewee 10—Section 49). In order to be able to address a complex topic such as digitalisation in an individually customer-oriented way, banks are currently still very much dependent on the cooperative associations and linked to them on a technological level. There is less focus on independent in-house innovation development, which could enable a more tailored transformation on its own. Nevertheless, there are banks that have dedicated themselves to the topic. *"We started a few years ago, for example, to set up an Innovation Lab in our company. [...] Then, of course, we also started to adopt the new technologies and implement agile concepts in our company. This led to the fact that at some point people said that we wanted to create a whole programme and ultimately digitalise the whole company"* (Interviewee 31—Section 57). This approach presupposes the availability of the resources necessary for implementation, but supports the independent development of the corporate identity and can, ideally, address the needs of customers and employees in a precise manner at the same time.

The approaches discussed above represent a variety of elements that, from a managerial perspective, are or can be seen as fundamental for a holistic, fast, and tailor-made digital transformation and at least promote it significantly.

Multiple measures are being taken to promote digital transformation within banks. These include, in particular, measures to introduce employees and customers to new digital processes and technology in general in order to integrate these two groups into the process of transformation. However, funding that is sustainable at the technological level and that leads to a faster technological transformation cycle or to the reduction of regulatory research or its process-optimised (over)fulfilment is only addressed in a limited, global manner and does not address specific issues.

The measures currently taken and the managers' descriptions indicate that technological integration itself seems to be the most effective tool for successful change. This way of thinking is shared by large and small banking institutions alike. It is apparent, however, that small banks expect membership in an association to be the factor that determines

success in further development, and that technological solutions should primarily be made available centrally. However, independent development and implementation of digital approaches are hardly taken into account in small-to-medium-sized institutions (SMEs), and are sometimes not even considered for reasons such as affiliation with an association or the requirement for excessive expenditure. This goes against the self-development approach of applications based on open-source solutions, which can now be realised at low cost and could be used, at least regionally, if an appropriate interface were provided.

These considerations, therefore, contrast with the understanding of management and their arguments against self-development. "Simple issues have to be implemented quickly and directly. Of course, this currently overwhelms many employees. But this will change slowly and continuously in three or four years. Open source, for example, now offers so many possibilities that you can virtually develop software in a very short time. The cost driver has to be assessed quite differently today than 15 years ago. The reason for this is that IT development is now possible in days, weeks, and months and no longer in years. Until then, banks often had three- to five-year development plans" (Interviewee 33—Section 20).

If one considers the prevailing view of the current management, however, it is primarily the situation of a resource-related discrepancy between small and large banks that weakens rural banking structures and increasingly drives these small banks into an association structure and, thus, into increased dependency. Nevertheless, this view cannot be supported on the basis of the above-mentioned low-cost technological approaches, such as open-source technologies and an ever more widespread API and interface structure.

Measures such as staff and customer fairs on technology topics, specially trained staff who supervise digital topics, creative rooms, and innovation labs are approaches that can be implemented efficiently and in a resource-saving manner within banks and can contribute to direct development. The innovation lab, admittedly, is associated with volatile costs, depending on the particular use and programmatic design, which should be carefully examined. However, the increasingly standardised interface programming, APIs, and open-source approaches can also improve and optimise the entire internal bank process structure, as an accelerated and targeted integration of applications of new offers is ultimately made possible. The first approaches mentioned are, from a creative point of view and a resource-conscious leadership approach by the management, especially easy to implement and equally easy to realize.

With regard to the findings on "*Complex technology and increased regulation*", there is a need for banks to catch up in order to implement digitalisation in a competitive and sustainable way so that they can become even more digital in the future. A few individual approaches by banks demonstrate that there are technologies that allow developments to take place at low cost. *"Open source offers many new potentials"* (Interviewee 33—Section 60). The only thing that needs to be done is to find and attract the appropriate IT staff who are specifically trained to deal with these issues. "*In order to develop digital business models that represent a unique selling proposition, not only central but also decentralised IT know-how is needed. Central IT development alone will not be sufficient in the future"* (Interviewee 33—Section 17).

#### **6. Conclusions and Further Research**

In this study, qualitative findings were combined with best-practice approaches from the banking sector with regard to digitalisation and the barriers that arise in this context. Here, for the first time, qualitative derivations were made that require further investigation. The interest of the interview participants, as well as their voluntary and open-minded participation in the study, once again underlines the importance of this explorative approach. The results also show a wide-ranging, still-young field of research, which needs further attention. This is underlined by the problematic nature of the literature analysis and the need to draw on derivations from sectors with similar challenges. Based on the present results, it is suggested that the identified transformation/implementation barriers and the reasons for a prevailing discrepancy between perceived and actual responses to digitalisation should be analysed and scrutinised in detail in further research. It could be

the case that the large number of barriers identified limits the possibilities for interpretation, so a further reduction of the subcategories could be considered. There is also the possibility that other influencing variables can be identified.

In addition, management perceptions of the scale of barriers could be studied in more detail and practical transformation approaches could be analysed in more depth. Studies could also focus on examining banks and their management more deeply in terms of their approaches to digital transformation and identify further best-practice approaches. Furthermore, it is recommended that, in future studies, the results should be examined with a larger number of participants in all methodological approaches to maximise the possibility that further recommendations can be derived on the basis of existing bank structures or their business models. The differences between individual banks could provide insights to obtain more detailed results on barriers to implementing digitalisation.

The development of a comprehensive normative model for scientific and sectoral enrichment would be desirable. It would also be advisable to quantitatively record individual correlations between main and sub-barriers and to implement sustainable aspects in the research approach. An additional topic for further research is the study of the impact of digital transformation on the sustainable growth of banking organisations. This topic does not seem to be addressed in sufficient depth; a thorough investigation of the preconditions for digital transformation, which are prerequisites for sustainable development, is crucial. The development of a questionnaire is indispensable for the implementation of further quantitative studies so that analyses at the main barrier level or even studies of sub-barriers can be enabled. These further approaches could lead to the enrichment of not only the investigation of digital transformation, but also the banking sector in general, and could enable further industry studies, including in other industry sectors.

**Author Contributions:** Conceptualization, F.D.; methodology, F.D.; formal analysis, F.D.; investigation, F.D.; resources, F.D.; data curation, F.D.; writing—original draft preparation, F.D.; writing review and editing, F.D.; supervision, M.Š. All authors have read and agreed to the published version of the manuscript.

**Funding:** No external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** We would like to thank all interview partners for their open participation.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A**






#### **References**


*Article*
