Introduction:
The Banking Industry has witnessed a huge transformation during the last decade due to technological changes. The advent of information systems (IT) since 1980s, 1990s and early 2000s changed the banks in terms of application of IT in different business process form various perspectives such as cost reduction, revenue generation, fraud detection security etc. Similarly, analytics in banking is also meant for transforming such institution.
Financial metrics and KPIs provide effective measures for banks’ overall performance. Banks need to move beyond standard business reporting and sales forecasting, in order to reach their strategic goals. Datamining, multivariate descriptive analytics and predictive analytics help banks to extract intelligent visions and quantifiable predictions that covers all types of customer behaviors such as account opening and closing, transactions, defaults if any and customer exit. With the continuous increase in competition, regulatory changes, fraud and cyber security threats, banks face tremendous pressure to improve operating efficiencies and grow wallet shares to sustain in the market. So, analytics is found to be the solution for better operating efficiency and proper customer engagement and also in mitigating risk as well as optimizing the deployment and utilization of banks resources.
Big Data Analytics:
Big Data Analytics is an extremely huge and varied data sets which are handled, analyzed, managed and validated through data management tools to make informed decisions. The data sets may be unstructured, semi-structured and structured; meta data from internet, social media data; web browser history and responses to surveys; machine data from Internet of Things (IoT) etc. which are in the form of five V’s such as Volume, Velocity, Variety, Veracity and Value. Analytics is an encompassing and multidimensional field which uses mathematics, statistics, predictive modeling and machine learning techniques to find the meaningful patterns and have knowledge of the data so recorded.
Banking Industry comprises of enormous transnational data that is required to be properly managed, scrutinized, evaluated and utilized for the benefit of the banks and its customers.
Analytics in Banking:
Due to technological advancement there is no much interaction of customers and bankers at least to ensure that the current customer is well satisfied with their services so as to retain them. For banks customer acquisition is costlier than retaining old ones. Customer may be requiring varied services such as discounts on purchases, convenience, simplified home buying, personalized services, information and alerts etc. The traditional tools are not sufficient to process the data for all types of decision making. Hence, banks are using data analytics in an efficient manner so as to enhance their customer value with better and faster decisions and also to maximize their revenue.
Benefits of Analytics in Banking:
There are several benefits of Analytics in Banking, like,
- Better understanding of customer behavior and responding to changes in preferences faster.
- Meeting regulatory requirements and addressing the setbacks on real time basis
- Improved product design and overall product portfolio optimization
- Increase transparency
- Develop a risk adjusted view of performance
- Manage fraud effectively
- Measure customer and product profitability
- Identify high potential prospects and customers
- Improve the ability to target products and services to prospective customers
- Enhance specific elements of the offer like product pricing, channeling etc.
Stages of Banking Analytics:
The basic aim of Banks is to acquire customers, retain and finally develop them. For this they go with the sentiment analysis, 360-degree customer analysis along with customer segmentation, best offers for them product management and design targeted marketed programs to reach them. These activities are supported by data analytics. This involves a series of stages of maintaining data and processing them to reach the informed decisions of the bankers at regular intervals.
Reporting: This involves building data warehouse and report the current situation. Here only raw information is gathered which is both structured and unstructured and which is collected from various sources.
Descriptive Analytics: This is an actionable insight on the current position. Complex and time series data is considered for applying basic set of statistical and mathematical tools to study the data behavior and draw minor conclusions. For example, customer segmentation and profitability, campaign analytics, value at risk calculations etc.
Predictive Analytics:These analytics predicts the likely future outcomes of the events. Here the big data is considered being real time and from various sources known and unknown. Accordingly, advanced and specialized tools are considered for predicting the future possibilities.
Prescriptive Analytics:These analytics prescribes the action on the predicted outcomes for a situation. Still more advanced techniques are used for prescriptive actions on the predicted outcomes and it promotes self-learning. For example, behavioral probability defaults, loss given defaults, exposure at default modeling, stress testing for mandated and custom scenarios etc.
Model Framework for Analytics in Banking
The key areas where analytics in Banking impacted a lot are:
- Consumer and Marketing Analytics
- Risk, fraud and Anti-Money Laundering / Know Your Customer Analytics
- Product and Portfolio optimization modeling.
Accordingly, a frame work model can be designed with basic drivers / components of banking data analytics being –
Operations and Performance Management:
Operations management is one such driver which involves a series of analytics that can be considered such as supply chain analytics, claims analytics, call center analytics, work force analytics, IT operations, spend and usage behavior analytics. All these focus on product and portfolio optimization that determines prepayments, misbehaviors defaults and cash flows to the banks. These analytics shows better impact on profitability of the banks thereby helps in smooth flow of operations.
Customer Management:
Under customer management of banks we come across market sizing, segmentation and targeting, customer acquisition strategy, cross sell and upsell opportunities, marketing mix and optimization leading to channel performance, campaign and sales effectiveness, customer satisfaction from customer lifetime value (CLV) estimation, digital experience of customers product comparison and attributed sentiment and tracking sentiments in future, brand equity and trends information from social media and digital media and finally real time offers and personalization.
Risk Management:
Risk management analytics modeling involves analysis of various portfolios to forecast likely losses and make provisions for those adequately. It comprises of risk assessment, scoring and rules, credit risk, AML, KYC, loss forecasting, default management, collections analytics, regulatory requirements in relation to Basel and CCAR, trade cancels and settlements etc. Early warning signals of both customers and banks are sent in case of any mis-happenings or finding such preventive actions for protecting from AML incidents.
Regulatory Governance and Compliance:
Due to stringent regulatory environment there is rising cost of compliance and also risk of non-compliance in some cases. Under regulatory and governance compliance analytics proper regulations are followed by the banks and there is a check if any deviation is there in the operations or any issues relating to the customer activities thereby protecting the governance of the banks. This ensures trust on the banks from the customers.
Fraud Management:
Proactive fraud detection is necessary for the banks to secure customers and employees. Fraud Analytics comprises of detecting, preventing and mitigating fraud risk in real time, application and transactional fraud monitoring, real time monitoring of rules and AML solutions. This ensures early warning signals to the banks whenever any deviations in the activities are found to be aroused.
Implementation of Analytics in Banking:
Analytics in Banking can be implemented in the five stages that are discussed.
Prioritize the focus areas:
Banks should identify the areas (i.e., customer, risk, finance, governance or fraud) where data and analytics can show greatest impact and obtain leadership engagement from the start.
Streamlining of data:
This requires integration of high quality of data with the data in the silos across products and lines of business. For example, a single view of customer, his transactions, tastes and preferences, aggregated risk exposure by product etc.
Integration with decision management system:
Analytics is itself meant for taking real time smart decisions. Thus, proper integration with decision management systems is necessary.
Talent hunt:
Finding the right talent (i.e., statistical modelling professionals, big data analysts etc.)for right process ensures the success of that activity. Banks should have a talent plan that builds on both existing internal talent and external sources.
Make connections:
Banks which have already had certain facets of analytics should chalk out a smart plan for connecting the teams across the whole organization which in turn strengthens the existing one or comes out with a more effective one.
Challenges faced by Banks:
With the facets of varied data, software tools and programs to be adopted, outcomes to be analyzed and decision to be made on real time banks lack internal capabilities and capacities.
Cost and Time: Banks budget may not be sufficient to meet the high planning of analytics implementation. Similarly, the time available to integrate the present process to analytics is more and if done they may be risking the competitive advantage in delays.
Expertise of Analytics: Banks may not have expertise staff in analytics or even the understandability of such areas is very less which ultimately hinder the implementation process.
Technology Resources: Understanding of analytics tools and their integration to the present process flow is limited in banks due to lack of expertise and also resources.
Benchmarking Data: bench marks and efficiency indicators help a lot in comparing internal performance, but the in analytics it is difficult to set the quantifiable targets due to lack of historical information.
Process Expertise: It is necessary to connect the analytics to operational performance objectives and this can be done through a third-party service provider (as an outsourcing engagement) thereby driving analytics objectives towards process performance.
Factors affecting Successful Implementation of Analytics in Banking:
For successful implementation of analytics in banks there are three most important things to be considered. They are:
Data coverage and Relevance: – It is very important to validate the source and completeness of data as such incomplete and broken data may result in wrong observations.
Suitability of Technology:– Selection of technology should be based on capability, cost and future needs.
Governance Structures:– Right governance structures are to be adopted after clear relevancy to functions.
Banking analytics can be successfully implemented by either pure-play analytic service providers such as Absolute data, CRISIL etc. who support only analytics services; and IT/BPO service providers who engage themselves both in analytics as well as BPO services such as Accenture, Cognizant, IBM, Genpact, hp, Infosys, TCS, WNS etc.
Analytics in Indian Banking Scenario:
Analytics in Indian Banking is still at its nascent stage and so the value creation potential is also very low. The market size of Finance and Banking sector in Analytics is growing at a faster pace than any other sector as it was USD575 million, USD756 million and USD1030 millionand USD 6.32 billion in 2020 during 2016, 2017, 2018 and 2020 respectively and is projected to reach USD11.02 billion by 2026. Delhi and Bangalore are showing increasing growth rate with respect to market size. Let us see trend of analytics in Indian Banking system through a few cases.
Several Indian banks have already initiated data analytics initiatives. Some examples are as under-
A. HDFC Bank was able to scale to large datasets and build models using high-performance parallelized RevoScaleR algorithms. The Bank provided a seamless loan application experience as well as quick turnaround on loan dispersals for customer Demographic, geographic, and other data are utilized to augment loan applications and credit analysis. Instead of data manipulation, preparation, and governance the Bank may focus on models, develop algorithms, streamline updates, and create new innovations in customer services.
B. Axis Bank automated the assignment of scores based on risk and other propensities in their consumer lending business. The Bank expected that the scorecard driven underwriting would result in better accuracy, consistency and lower credit costs. Analytics is utilized in their marketing campaigns. The Bank is also concentrating on data from Internet and mobile banking, social media etc. towards improved customer service by understanding their customers’ challenges. The Bank plans to incorporate fraud analytics as well.
C. ICICI Bank equipped 2000 managers with design thinking and data analytics skills. The Bank has been utilizing data analytics for multiple activities and it was declared winner in the ‘Best Use of Data Analytics’ category at the Retail Banker International Awards 2018- organized by Retail Banker International, an online publication that provides news on banking and finance from across the globe.
D. Kotak Mahindra Bank’s analytics platform extracts data from its core banking system and the relationship management system, combines them and puts in place algorithms which are working on certain thresholds. The bank implements various significant events like change in demographics viz. marital status, city, transactional pattern viz. significant credit/ debit, premature withdrawal of term deposit, stop in any regular activity viz. auto-pay, standing instructions/ECS, drop in transaction value or volume throughputs across a period vis-à-vis historical trend and payments to beneficiaries like securities, builders, dealers These details are shared with Relationship Managers of the Bank for identifying business opportunities. The Bank has identified that a meeting led through cues from data analytics tools is 3 times more effective than the random meetings scheduled by a Relationship Manager.
Conclusion:
Banks should transform massive volumes of organizational data into actionable insights and strategies. Business Analytics or Big Data Analytics provides comprehensive capabilities to help banks to perform customer profitability analytics, manage risk and improve operational efficiency. Sophisticated predictive and prescriptive analytics improves banks’ profitability, compliance, sustainability and competitiveness. However, implementing these analytics is a challenging aspect looking at the basic realties such as functional silos, talent crunch and technological resources or infrastructure. But there is no other way out for the banking industry that can create value than the tool Analytics.