Automating loan recoveries: A top priority for banks

Wednesday, 7 November 2018 00:00 -     - {{hitsCtrl.values.hits}}

 

Many argue that performance of banks is more market driven. Better performance is no more a choice and it has become mandatory to sustain competiveness by satisfying diverse interests of stakeholders. Lending is a key function of banks and non-bank financial institutions which directly affects performance. Lending is in the DNA of banking where financial intermediation is the primary role of a bank. Intermediation emphasises mobilising of deposits from surplus units and lending same to deficit units in the economy. This article emphasises on issues and challenges faced by lenders and how technology can be used prudently to reduce or eliminate them. The degree of automation of loan recoveries is discussed in the context of need and feasibility. 



Is loan recovery a problem?

According to media reports, Non-Performing Assets (NPAs) of the banking sector in Sri Lanka reached 2.5% as at end of 2017 which was the best of past five years started showing a continuous increase over the last several months. This is a major concern for banks and non-bank financial institutions given the unfavourable developments in the local and global economic landscape. The sharp drop of lending (as an example Rs. 46 billion in August against Rs. 82.6 billion in June, as reported in Daily Mirror on October 17, 2018) can lead to a further increase of NPAs if substantial recoveries are not done. According to the same report, there has been a 70% increase in rescheduled loans during this year and the NPAs may exceed 5% shortly. This is obviously not a healthy situation for the banking sector and to the economy. 

NPAs affect the liquidity of a bank. It also heavily undermines the image of the bank locally and internationally. It affects the risk bearing capacity of banks. NPAs also increase cost of funds and affect the bank’s ability to lend. NPAs affect net interest income, the profitability of the bank and also the ROI. 



How far you can go with the spreadsheet?

The idea here is not to criticise the use of spreadsheets. I am particularly highlighting the ill effects of over reliance on machine generated reports which are manually monitored. Path dependence is so high among practitioners. So much so the Excel spreadsheet which guides the agenda of most recovery meetings runs in to several sheets and gets highlighted in multiple colours. Is this doable? Is this productive? Can we consider this as rational where your bank has invested millions of dollars in technology and various systems? Obviously, there has to be a better solution for this manual monitoring in the digital age and there is. 



Being proactive than reactive

Identifying the early warning signals is quite crucial to maintain a quality loan portfolio. The new regulations such as IFRS 9 also highlight the importance of using forward looking measures. IFRS 9 which has become effective from this year requires banks to keep provisions for possible bad loans on the basis of expected credit loss model instead the existing incurred loss model. Compliance to this new regulatory requirement will essentially increase the loan loss provisions of banks. In this context, an increase of impairment and NPAs are inevitable. Short term measures such as further lending or rescheduling of loans in respect of defaulted loans will not work. Banks should essentially be pro-active in managing the receivables. 

The stress created with growing bad loans will push banks to further strengthen and strictly adhere to sound lending practices and evaluation standards. However, current and future portfolios demand prudent techniques to sustain loan portfolios. Banks should essentially take measures to be pro-active. With manual systems banks mostly use lag indicators. What really requires at this point in time is introduction of appropriate lead indicators. Lead indicators are input based measures compared to lag indicators derived from outcomes. Readers would easily understand the difference, impact and necessity with the following example. Measuring your body weight on a regular basis with the objective of being slim and fit is relying on lag indicators whereas conscious monitoring of calorie input and calorie burnout is relying on lead indicators. A combination of both types would ease the job. 



Automating loan recoveries

Automation is possible. Easily pluggable loan recovery or delinquency management software to your core banking system can do this. Automation has its own limitations as well. However, irrespective of the degree of automation banks would see a significant improvement of loan recoveries when they introduce loan recovery software. The question is how accurately, when and at what cost? Two major concerns would be implementation time and the level of customisation needed which is directly related to cost and usefulness of the system. Banks should necessarily look at the willingness of the vendor and the cost of customisation. After services in all levels especially the physical presence of the service provider will be crucial for any new technology implementation. 

Some of the functionalities of a fully automated loan recovery system include customer reminders, follow up of payments, age bucket analysis, assigning recovery related jobs to staff, monitoring service level agreements (SLAs), measuring KPIs, escalation of SLA violations to the hierarchy and monitoring of promises to pay. Automation of loan recovery reduces the stress of recovery staff and will enable the bank to use human resources in more productive areas where human interaction is mandatory. 



Predictive analytics 

Banks often use descriptive analytics to explain what has happened to loan portfolios using traditional statistical techniques. Predictive analytics involve statistical and mathematical techniques to predict future unknown events or behaviours based on historical data to support business operations and managerial decisions. Modern delinquency management solutions offer predictive analytics. In addition, prescriptive analytics are also used to suggest best alternatives to improve performance where optimisation techniques and simulations are used. This is being pro-active and use of a blend of lead and lag indicators.

Given the fact that heavy investments are made in technology, it is quite clear that banks should start using some form of automation in their effort of improving loan recoveries where predictive and prescriptive analytics would be the subsequent milestones to achieve in the journey of automation. Exciting times are ahead as a major open innovation initiative is in place to make the most innovative local delinquency management solution even better by adding forecasting capabilities in collaboration with several universities and professional bodies. 



Is integration along the value chain possible? 

These systems can be extended forward to automate legal proceedings that include maintenance of legal diaries, sending letters of demand, follow up actions, scheduling resources for court cases, follow up of writs, auctions and costings of legal proceedings. This can be fully automated where non completion of any activity within the agreed time periods or any service level agreement deviations will be reported to the hierarchy for necessary actions. On the other hand, this can be extended backward up to KYC and customer onboarding process where CRM or customer relationship management solutions will take control of the automation process. 



Is your bank prudent in technology investments?

We often see banks investing, promoting and even struggling with novel digital channels. Offering convenience to customers, increased financial inclusion, cost and time savings and reduced congestion in bank branches are the desired benefits. However, attempting to reach the unbanked community using digital channels is quite debatable and the existing results are not impressive. According to Central Bank information, less than 5% of total volume of transactions is routed through these digital channels. Given the fact that internet banking in Sri Lanka is decades old and the effort towards mobile and branchless banking is substantial we need to revisit the strategies of these ‘nice to have’ technology options. Effectiveness of these technology investments in terms of utilisation and ROI needs to be re-examined. 

Here the big problem ahead of banks is not about investments in alternative channels. The rational question is about how prudently technologies are used in improving the primary role they play in the economy which is financial intermediation. One major aspect of intermediation is lending which covers disbursements, monitoring and more importantly recoveries. Therefore, the key question is whether you use technology prudently in managing the primary functions. The benefits include low non performing advances, better quality of portfolios, improved profitability and less stress on staff involved in recoveries. Benefits of automating loan recoveries with a time tested and fully customised delinquency management system are enormous. Given the unfavourable global developments in the external business environment that can severely affect the credit ratings of banks, it is high time that you seriously consider automating loan recoveries as a top priority to safeguard stakeholder interests. 

(The writer has 12 years of experience as a technology marketer in the banking and financial services industry sector. He holds a BSc degree in Business Management from Monash University, Australia and MBA from the Postgraduate Institute of Management of University of Sri Jayewardenepura. He is a Director and the COO of Avonet Technologies, the CRM and delinquency solutions company from Sri Lanka now gaining popularity for innovative software in the region. He can be contacted through [email protected].) 

Recent columns

COMMENTS