Analysis of contemporary trends in data science in the Retail Banking Industry
The retail banking industry is oriented to offer financial products and services to individual customers rather than small industry, farmers or corporations. One product that they deliver is credit, for example, through credit cards, mortgages, or loans. The industry faces significant opportunities and challenges because of new competitors, such as fintech companies, more strict regulation about data protection and cutting-edge technology like big data. Therefore, it is crucial to find innovative ways to develop new products and services to be attractive to new customers and consolidate loyal bonds to existing ones.
There are several areas to improve and transform through data analysis that can boost the competitiveness of the sector. Vadlamani Ravi and Sk Kamaruddin (2017) and Hossein Hassani, Xu Huang and Emmanuel Silva (2018), discuss some of them, namely, security and fraud detection, risk management and investment banking, customer relationship management, operational analytics and human resources analytics.
Additionally, according to Kent Eriksson and Cecilia Hermansson (2018) and Mary Wisniewski (2017), the use of transdisciplinary knowledge can lead to generating insights. Concepts of psychology can enhance the development of ground-breaking products, such as applications to teach savings to customers.
However, Srivastava, Utkarsh and Gopalkrishnan, Santosh (2015) pointed out that the application of data science presents some issues as the information that is used today to feed machine learning algorithms was collected in the past without standard procedures and many times with denormalised data and unknown data dictionaries.
The primary purpose of this paper is to describe the current trends of data science in the retail banking industry. Mainly, the aspects of being considered are fraud detection, risk management, investment, privacy, security and operational analytics. There will be an analysis, discussion and evaluation of the application of data science techniques in this sector. Finally, it will be written a conclusion to summarise critical ideas to be taken into account for developing original financial products or services.
Opportunities, challenges and impact in the Retail Banking Industry
Opportunities in the Retail Banking Industry
Fraud Detection
Consumer bank fraud has become more sophisticated, along with the increased need for account security. A 2014 report by Graziella Steele (2014) revealed that “more than 4 million credit and debit cards were reissued at the cost of about $40 million” due to a data leak from well-known retailers such as Neiman Marcus and Target during the vacations, which suggests the need for banks to protect their services and better control who is accessing the information.
There are sophisticated mechanisms to prevent fraud like behavioural analysis intertwined with customer’s transactions data on a predictive model. It can be used historical transactions and consumption capacity of the customers to predict a possible fraud. In research conducted by Srivastava, Utkarsh and Gopalkrishnan, Santosh (2015) the net count of credit transactions, net debit transaction and transaction time was analysed using a predictive model to visualise spending trends per month and per year in order to establish a pattern of fraud based on these behavioural analyses.
Risk Management
One key driver for innovation in retail banking attempts to reduce loan defaults and increase on-time repayments. Banks can implement different strategies to create payment reminders for people thus capitalising on concepts of psychology and statistics.
Tobias Baer and Vijay D’Silva (2018) on behalf of McKinsey have done research that reveals from some features of bank transactions like items, time, day and customer, that it is possible to elaborate a personalised path to guide customers to pay. There is an example that shows the transaction of a person who buys different things in the same place the same day but not at the same time. On the other hand, there is another type of behaviour, it shows a person that buy a group of things the same day in the same place at once, and this pattern can be seen repeated frequently. From this research, there are a 30% increase in payments, 20% of decreasing write-offs, 33% reduction of delinquencies for late-stage collection and a 20% loss of customers who fail to pay and are reconverted to customers again.
Investment
An economically healthy nation has distinctive features, including the capability of its citizens to save money. However, Mary Wisniewski (2017) pointed out that 2015 research from Simon-Kucher argues that at least 35% of a typical bank’s customers do not know how to put aside their money. Since the banking sector is seeking innovative ways to preserve and incorporate customers through loyalty programs and attractive offers, generating original saving products is an opportunity to engage them.
There are concrete actions to interest customers in saving money using machine learning algorithms. For example, Mary Wisniewski (2016) discusses the Splurge Alert mobile application (owned by Ally Financial) which uses geolocation technology to prevent customers from buying more than they need when they’re in stores that are ‘risks’ for them overspending. In another report, Mary Wisniewski (2015) described the Savings Coach app (owned by USAA bank) which is designed for millennials, and gathers transaction data in order to determine a recommended saving target.
Privacy and security
One aspect of security in the banking industry is customer recognition. This industry offers a different variety of channels where customers can interact with the bank, both face-to-face or remote. In both cases, we need to be sure that the person is, in fact, the customer they claim to be. Technology and machine learning algorithms can be harnessed to create novel ways to authenticate the customer.
For example, biometric technology in the form of facial recognition, heartbeat, type patterns, fingerprints, eyes patterns can be used to establish a trust relationship between the customer and the bank. Biometric Technology Today (2015) discuss an example of the UK Bank Halifax using Bluetooth wristbands to monitor a customer’s heartbeat and with an algorithm create a login to authenticate uniquely in the bank’s applications. This is an explicit use of information and analytics to deliver trustful results to banking and their customers.
Another consideration is the use of blockchain for customer verification. Verifying customers can be an expensive process because it’s necessary to collect multiple pieces of information for comparison with records. This occurs over many contexts from opening an account to appplying for a mortgage or personal loan. According to Hossein Hassani, Xu Huang and Emmanuel Silva (2018) a blockchain will provide a tool to verify the customer in every circumstance that he or she wants to do something with the bank, basically recording his/her information on it. Eventually, it could be available to be shared with other financial institutions.
Operational Analytics
Many of the processes that the banking industry undertakes are repetitive. Machine learning merged with RPA (robot processing automation) provide an opportunity to reduce the operational costs of these repetitive processes.
JPMorgan (2018) found ideal candidates to be automated through machine learning, such as “obtaining real-time price quotations and executing FX transactions, creating accounting entries for inter-company netting and payments, preparing bank account reconciliations, as well as identifying and verifying information in payment advice emails for cash application”. Therefore, JPMorgan developed software that can do all the repetitive tasks. The value earned is that the algorithm performs better than a human with fewer errors, more quickly and less cost.
Challenges in the Retail Banking Industry
There are substantial challenges related to the topics described above. Some of them are:
Select the right model to make predictions. How to apply the correct technique either on fraud detection or biometric customer identification requires statistical knowledge and experience analysing vast volumes of data.
Determine the validity of the outputs of the model. Analysis of the outcomes of machine learning techniques applied to fraud detection or risk management depends on the knowledge of the domain expert, so it is crucial to foster the skill of analysing data to deliver better predictions.
Find reliable and accurate data. To feed the model, it is necessary data, but not always, it is straightforward to get it. For example, the existence of silos of data and lack of procedures to store structured and unstructured data makes it harder to find patterns when it comes to analysing customer’s transactions. It is necessary to find algorithms that can handle to look for information in different places.
Additionally, compliance with regulatory standards will create stricter rules to apply for using customer data, and maybe it will be challenging to join the information of the same person to develop banking strategies for encouraging personal investment.
IT Infrastructure. Sk, Kamaruddin and Vadlamani, Ravi (2016) showed that a specific big data architecture must exist to complete all the complex calculation generated by analytical models, especially to apply mechanisms of biometric recognition.
Privacy and ethics. Underlying the exploitation and analyses of data resides the protection of data. Nowadays, introducing personal and financial people data can be done using mobile apps, bank websites and internal banking networks. However, there is no clarity about how this data would be used or who is going to see and store this information. Therefore, uses like blockchain or biometric technology should be regulated by governments and industries to protect personal information.
Regarding RPA, the challenge is to train the employees in more complex tasks to deliver added-value to their works. Otherwise, it could lead to an increase in the unemployment rate of the country, which is unethical.
Impact in the Retail Banking Industry
There is a noteworthy impact that is the reduction of costs, which leads to an increase in the revenue of the financial sector.
However, people are going to change their financial habits, and they will protect more their personal data. So, this will lead to producing changes in the internal processes of the retail banking industry to be more flexible and respond more quickly to the demands of the customers. Also, the sector must be aware of the current needs of the people to satisfy them and establish a trustful relationship with their clients.
Conclusion
Data science accompanied by technology, big data and robust privacy policy is an essential player in the economy of the countries. In this paper was presented a wide spectre of applications in the banking industry. Even though the impact in the industry it can be seen as positive there are huge hindrances like lack of data quality, silos of data inside the organisation, poorly defined privacy and security rules that must be addressed to make a profit of the novel technology.
References
Biometric Technology Today, 2015, “Banking sector embraces multiple biometric modalities” published in Elsevier on April 2015, vol 2015 issue 4, <https://www-sciencedirect-com.ezproxy.lib.uts.edu.au/science/article/pii/S0969476515300485>
Graziella Steele, 2014, “Banks target retailers for credit fraud losses”, published in the journal “The Idaho Business Review” on 27 February 2014, <https://search-proquest-com.ezproxy.lib.uts.edu.au/docview/1503985763?accountid=17095>
Hossein Hassani, Xu Huang and Emmanuel Silva, 2018, “Digitalisation and Big Data Mining in Banking”, published in Big Data and Cognitive Computing, 01 July 2018, Vol.2(3), p.18, < https://doaj.org/article/1e2ef0ccb32b4d41b4f2c7f590f32787>.
JPMorgan, 2018 “Demystifying New Technologies in Treasury”, viewed 07 April 2019, <https://www.jpmorgan.com/global/cib/ts/demystifying-tech>
Kent Eriksson and Cecilia Hermansson, 2018, “How relationship attributes affect bank customers’ saving”, International Journal of Bank Marketing, Vol. 37 Issue: 1, pp.156–170, <https://doi.org/10.1108/IJBM-09-2017-0194>.
Mary Wisniewski, 2015, “USAA Launches New App to Help Millennials Save”. Published in American Banker on July 28 2015, 3:57 pm EDT, viewed 5 April 2019, <https://www.americanbanker.com/news/usaa-launches-new-app-to-help-millennials-save>
Mary Wisniewski, 2016, “Ally financial launches financial health app”. Published in American Banker on April 08 2016, 5:09 pm EDT, viewed 5 April 2019, <https://www.americanbanker.com/news/ally-financial-launches-financial-health-app>
Mary Wisniewski, 2017, “Teach customers to save, and maybe they’ll stick around”. Published in American Banker on February 10 2017, 4:11 pm EST, viewed 5 April 2019, <https://www.americanbanker.com/news/teach-customers-to-save-and-maybe-theyll-stick-around>
Mohammad Miyan, 2017, “Applications of Data Mining in Banking Sector”, published in International Journal of Advanced Research in Computer Science, Jan 2017, Vol.8(1), < https://search-proquest-com.ezproxy.lib.uts.edu.au/docview/1901446000?accountid=17095>
Sk, Kamaruddin and Vadlamani, Ravi, 2016, “Credit Card Fraud Detection using Big Data Analytics: Use of PSOAANN based One-Class Classification”, published in Proceedings of the International Conference on informatics and analytics, 25 August 2016, pp.1–8, <https://www.researchgate.net/publication/309638452_Credit_Card_Fraud_Detection_using_Big_Data_Analytics_Use_of_PSOAANN_based_One-Class_Classification>
Srivastava, Utkarsh and Gopalkrishnan, Santosh, 2015, “Impact of Big Data Analytics on Banking Sector: Learning for Indian Banks”, published in Procedia Computer Science, 2015, Vol.50, pp.643–652, <https://www-sciencedirect-com.ezproxy.lib.uts.edu.au/search/advanced?docId=10.1016/j.procs.2015.04.098>
Vadlamani Ravi and Sk Kamaruddin, 2017, “Big Data Analytics Enabled Smart Financial Services: Opportunities and Challenges”, published in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2017, Vol.10721, pp.15–39, < https://link-springer-com.ezproxy.lib.uts.edu.au/chapter/10.1007%2F978-3-319-72413-3_2>.
Tobias Baer and Vijay D’Silva, 2018, “Special Edition on Advanced Analytics in Banking”, published by McKinsey on Payments, viewed 07 April 2019, < https://www.mckinsey.com/industries/financial-services/our-insights/mckinsey-on-payments-special-edition-on-advanced-analytics-in-banking>
