Corporate credit continues to be strongly linked to vintage and past credit history. In a world overturned by the pandemic, this does not cut it anymore since past models are no longer predictive of future performance. In fact, what matters most is the behaviour now.
Before Covid (BC) and After Covid (AC)
Commercial lending has been relationship driven from the early days. Regulatory mandates on information sharing as part of due diligence have indeed bring order to decision making. However, as with interpersonal relationships, it takes time and effort to build trust and gain access to non-regulated information that is also equally crucial for lenders to make more accurate decisions on whom to lend and how much. While credit bureaus have helped bridge the information gap in the last two decades, they have only increased the dependence of lending decisions on the particular sliver of insights stemming from credit history.
Unfortunately, to borrow from the mutual fund industry, past credit performance is not predictive of future credit performance. Throw in a pandemic of such epic proportions that the passage of time is itself cleaved into “Before Covid” (BC) and “After Covid” (AC) eras (witness Kotak Mahindra Bank‘s FY20 Annual Report). What we get is the scenario where financial institutions are left with credit decisioning models that have little value in the AC era since all their accumulated knowledge has been from the BC era. Moreover, since the AC era has only recently emerged, there is not much credit history to speak of on which to build afresh.
As companies begin to step away from the more traditional principles and practices of collecting customer behavioural insights. In this post-pandemic AC era, the need of the hour is for new credit decisioning and risk assessment models that incorporate an understanding of a client’s business that goes beyond its use of prior credit facilities.
Stress factors have varied wildly across sectors due to a multitude of reasons. For instance, the travel & logistics sector being hammered due to plummeting patronage. The MSME sector as a whole suffered from depressed expenditure in and poor collections from their BC-era client base.
Business is no longer “as usual” as expansion plans have been curtailed bringing down associated need for new credit and deferment of capex projects. The focus on controlling costs has brought to the fore actions pertaining to rationalizing use of working capital. The magnitude of such change too varies from industry to industry. Such a business climate favours certain types of credit products that result in specific risk accumulation.
Supply chains have become upended due to not just the pandemic but also global trade policies and regional geo-political developments. Business entities that heavily ride on cross-border trade be it for import of raw materials or export of finished goods / services are impacted significantly. For instance, beauty & wellness salons use a number of Chinese equipments and are primarily dependent on footfalls, both of which got impacted significantly in the past year.
It is becoming increasingly pertinent for businesses to utilise advanced tools that are capable of providing more intuitive and tangible information about customers. The need of the hour is holistic information about the business from the present that aids in decision making across the credit lifecycle. In fact, such information should encompass the various aspects of the business which together reflect the functioning health of the business. Much like an individual’s behaviour can be assessed by doctors for symptoms of underlying malaise, business behaviour provides crucial inputs.
The insights that are generated from behaviours is referred to as behaviour intelligence. Companies behave in different forums differently but do exhibit patterns that are useful for risk assessment. For instance, a company that has frequent tiffs with the law as witnessed by the number of cases disposed or in pendency can be considered as a riskier proposition than one which has had a low case count.
Compliance behaviours provide a number of fairly accurate risk indicators. Corporate filings with the MCA and ROC reveal structural information that is necessary in such behavioural models, but unfortunately such information is not diligently digitized in a manner that is easily readable by a machine. Artificial intelligence (AI) techniques can help extract usable information even from such notoriously difficult sources though.
Mandated filings of returns with tax authorities (e.g. GST) and labour bodies (e.g. Employee Provident Fund) provide meaningful insights in to the health of a business. For instance, frequent delays in TDS filings could mean liquidity issues, compared to prompt timely payments.
Most of all, lending institutions such as banks that also provide business operations products are sitting on a treasure trove of behavioural data. Core banking systems have been recording transactions for many years. By mining such transactions, modern AI algorithms can unearth behavioural insights that are able to systematically build up a true picture of a client’s business. By then comparing the business between the BC and AC eras, new stress factors can be rapidly surfaced resulting in both timely risk alerts as well as emergent opportunities for new business.
In summary, lenders need to look beyond past credit history to make credit decisions. Business behaviours have to be understood holistically, especially in this post-pandemic era. The good news is that lenders already have the necessary data in-house, with external data augmenting the richness as necessary. Technologies too have rapidly evolved to accelerate this new dawn of decisioning models. What remains is for the leaders in the lending institutions to change their mindset and adopt such innovation.
The author is Rangarajan Vasudevan, Founder & CEO, TheDataTeam