Impacts of Data Science on Finance and Banking

Introduction

In the broadest of terms, data science can be defined as the process of deriving insight from data. It is a continuous cycle of observing real-world phenomena; collecting and processing raw data in order to create a model, perform analysis or test a hypothesis; and arriving at conclusions backed by data that aid understanding and help decision-making.

With the ever-increasing prevalence of technology and data in modern society, many companies and industries have begun to realise the need for data science and to reap its rewards. The banking and finance industries are perhaps the most obvious benefactors of these insights, second only to the technology sector itself - although with the recent surge in financial technology (FinTech) start-up companies and advancements, the line between the two is set to become even more blurred.

Over the past decade, data science in the finance and banking sector has become less of a trend and more of a necessity in order to maintain a competitive advantage over competitors and a healthy relationship with consumers.

In this report, I will outline some of the ways data science has been utilised by banks and financial institutions to transform or disrupt the status quo of the industry, and what potential consequences arise both in the present and the future.

Get Help With Your Essay

If you need assistance with writing your essay, our professional essay writing service is here to help!
Find out more about our Essay Writing Service

App-based Banking and FinTechs

Traditionally, in order to open any type of bank account, a person would have to endure a long and arduous process. First, they would have to find an account that suited their needs, which can be a laborious task. Then, they would have to travel to that provider's nearest, often busy high street branch and queue up to speak to someone in person, taking with them a variety of different personal documents. In most cases at least three documents are needed, providing evidence of one's full name, date of birth, national insurance number, proof of address and proof of identity. The list becomes even longer for more complex accounts like joint accounts and some will require a credit check as well. All of this means it can take a substantial amount of time to go from deciding to open an account to actually being able to use it.

However, over the last few years, a rise in digital banking and the formation of challenger banks have allowed the data generated by consumers in the banking industry to be fully utilised. Thanks to the recent introduction of the Open Banking reforms which requires banks to let you share your financial information with authorised providers [1], the creation of purely app-based banks such as "Monzo" or "Starling Bank" has been facilitated. Such banks do not have any branches and can provide near-instant 24/7 support via live chats in the app - something that traditional banks struggle to offer. These live chats use machine learning to check what has been sent from a customer in the chat and generate possible appropriate responses for the customer service agent. By scanning and analysing the language in the written query, the process of submitting requests is sped up and it helps connect users and help team workers more effectively [2]. Also, the machine learning algorithm will learn what responses are most applicable to a given request, further improving the effectiveness of the service as time goes on. By adopting a data-driven approach to banking, these mobile banks can offer features that established banks have yet to provide, such as the ability to set up an account in minutes; instant card locking and unlocking via the app, and real-time tracking of spending through app notifications and smart spending insights. These features have already proven extremely popular, so much so that both Monzo and Starling have won awards for 'Best Banking App' and 'Best British Bank' respectively [3].

App-based banks are examples of a type of company referred to by the umbrella term 'FinTech', which describes the evolving intersection of financial services and technology [4]. It also often refers to the technologies themselves that are having a disruptive influence on existing financial process. Although in the early stages of its life, FinTech has already had a dramatic impact on the global economy. In 2019, 64% of FinTech service consumers worldwide have previously used at least one platform, up from 33% in 2017. There has also been an effect on start-up companies with 25% of global SMEs (small and medium-sized enterprises) have adopted fintech services for use in banking, financing, and financial management [5, see EY Global FinTech Adoption report].

Fraud Detection

Another key area in which data science has transformed the finance industry is fraud detection. Fraud in the financial world can take many different forms, from improper payments and money laundering to terrorist financing and cybersecurity breaches. In the past, fraud detection used defined business rules and rudimentary analytics in order to search for anomalies. This meant that those investigating potential cases of fraud often had to do so after it had been identified and a crime was committed - it was a reactive approach rather than a proactive one. Nowadays, newer technologies and the development of the data science field has led to a complex and sophisticated set of solutions and approaches that have revolutionised fraud detection and prevention. By moving away from standard analysis and adopting predictive and adaptive techniques; using the vast amounts of transactional data society now produces on a daily basis; and employing realtime monitoring and analytics, the tide has started to turn in favour of prevention over reaction.

Fraud detection in a machine learning context is often framed as a classification problem [6]. This is where we try to classify each data point to a certain class label based on the values of the attributes of that instance. For a given output of the algorithm, each instance is assigned to a class and can be plotted to show a visualisation of the different regions corresponding to each class label (see Fig.1).

Figure 1: An arbitrary example of a classification problem in machine learning for two class labels.

This type of machine learning is referred to as supervised learning, where algorithms use historical data to learn from and identify patterns that may be of interest to someone investigating fraud. In reality, fraud detection uses a combination of data science methods, from supervised to unsupervised learning - where the goal is to help find and identify previously unknown patterns in the data. Modern fraud detection practices will almost always boast about the use of some form of machine learning. For example, the top three fraud detection software programs - "Signifyd", "Riskified" and "Kount" - all mention the use of various machine learning methods according to G2, the world's largest tech marketplace [7].

Forecasting and Predictive Analysis

Another key application of data science in banking and finance is forecasting when and how markets will move - and predictive analysis is used to do this. Predictive analysis is the use of data, statistical algorithms and machine learning techniques to identify the likelihood of future outcomes based on historical data [8]. This is not something new to the industry - for example, life insurers have been using mortality statistics for decades in order to make predictions for when policyholders may die. However, with the continual increase in the volume and types of data and computers becoming faster and cheaper, this technology is about to enter its golden age.

Many investment banks and financial institutions now employ predictive analysis to model financial markets. One method is to use genetic algorithms - a problem-solving method that mimics the theory of evolution [9] - to find the best combination of parameters and biases in a given trading rule. These algorithms repeatedly test a variety of parameter value sets, each with slight differences. At each step, the algorithm will randomly select a number of parameter value variants and allow these to produce 'children', which over successive generations will 'evolve' the parameter set to an optimal solution. This means that traders, whether that be large institutions or individuals, are able to create models based on initial parameter sets and optimise them using genetic algorithms and historical market data. By making advanced trading and analytical knowledge accessible to people other than the experienced industry experts, there is the potential to disrupt not only the landscape in terms of who trades on the markets but also to the actual financial market itself. Every entity from the large multinational quantitative trading company to the sole trader in their living room has the opportunity to create successful algorithms to help them trade and make a profit, which has the potential to massively change the accepted trends in how the markets rise and fall over time.

Conclusion

We have seen some examples of areas in finance and banking that have changed dramatically thanks to the introduction of data science. The quantitative nature of finance and the vast amounts of data that are involved when considering financial markets, global transactions, personal banking and investment management make this industry a prime target for innovation through data. Many companies now, both new and existing, are moving toward a data-driven business model due to the amount of insight and direction that can be extracted from consumer data. This means that the prevalence of data science in finance and banking will only increase in the future, creating many more jobs for those working with data and allowing more scientific research to be undertaken to better understand how we interact with each other and the wider economy. This insight will continue to optimise society's experience with finance and invigorate innovation in areas such as FinTech - ultimately tending towards a world in which individuals and large institutions can integrate easily, operate securely and deal with a wide variety of financial services seamlessly.

References

[1] R. Bannister, "MoneySavingExpert.com: Open Banking explained," https:// www.moneysavingexpert.com/banking/open-banking/, 14^th Jan. 2019, (Accessed 04/11/2019).

[2] J. McCarthy, "Machine learning and data are powering Monzo's fintech disruption — The Drum," https://www.thedrum.com/news/2019/04/11/machine-learning-anddata-are-powering-monzos-fintech-disruption, 11^th Apr. 2019, (Accessed 04/11/2019).

[3] Best British Banks, "The winners of the Best British Bank Awards 2019," http: //www.britishbankawards.co.uk/previous-winners/, (Accessed 04/11/2019).

[4] D. Nicolacakis, "PwC Q&A: What is FinTech?" https://www.pwc.com/us/en/ industries/financial-services/library/qa-what-is-fintech.html, Apr. 2016, (Accessed 07/11/2019).

[5] M. Hatch, G. Hwa, and J. Lloyd, "EY: Eight ways FinTech adoption remains on the rise," https://www.ey.com/en gl/financial-services/eight-ways-fintech-adoptionremains-on-the-rise, Jun. 2019, (Accessed 07/11/2019).

[6] R. Pierre, "Detecting Financial Fraud Using Machine Learning: Winning the War Against Imbalanced Data," https://towardsdatascience.com/detecting-financialfraud-using-machine-learning-three-ways-of-winning-the-war-against-imbalanceda03f8815cce9, 27^th Jun. 2019, (Accessed 05/11/2019).

[7] G2, "Best Fraud Detection Software 2019," https://www.g2.com/categories/frauddetection, (Accessed 05/11/2019).

[8] SAS, "What is Predictive Analytics & why is it important?" https://www.sas. com/en gb/insights/analytics/predictive-analytics.html, 25^th Aug. 2016, (Accessed 06/11/2019).

[9] MathWorks, "What is the Genetic Algorithm?" https://uk.mathworks.com/help/ gads/what-is-the-genetic-algorithm.html, 3^rd Jun. 2019, (Accessed 07/11/2019).