Challenge 1: Upgrading Credit Decisioning

Commercial lending is always impacted severely during economic crises. But this is not just restricted to lenders’ risk appetite or their capacity to fund businesses. It also vitiates their credit decisioning models that rely on historical data – information that can be rendered ‘virtually useless due to the market disruptions of COVID 19’ according to a McKinsey report .

The report also outlines some ‘half-measures’ that lenders may adopt, known as ‘model overlays’. Such an overlay might assign a high default risk to a restaurant because the hospitality sector was severely impacted by the pandemic. But what if a restaurant adopted an omni-channel model and started delivering food instead, enabling it to withstand and even thrive during the lockdowns? A model overlay cannot account for something as fundamental as an entrepreneurial innovation, and upgrading credit decisioning using the latest and most sophisticated models is not just a best practice, it is a necessary measure.

One of the key factors in mitigating risk, assessing credit-worthiness, and arriving at mutually favorable loan terms is the debt-to-service coverage ratio, which in simple terms is the business’ ability to service its debt while meeting its business obligations. While banks can use basic formulae to assess this crucial aspect of a business’ financial health, advanced credit decisioning models, which can parse vast amounts of data, can arrive at a far more precise assessment, not only of the business’ ability to meet its debt, but also its ability to honor its loan terms.

In fact, the McKinsey report recommends extensive data mining to look out for credit signals. This is the basis for creating a credit decisioning model or engine that can account for multiple factors such as the debt-to-service coverage ratio and other estimates of financial health, and even go beyond such signals and look at data sources as diverse as public blogs, sentiment analysis, customer feedback, interviews with relationship managers and more – a process beyond the scope of human operators, but well within the reach of a credit engine that can access public APIs automatically to process vast amounts of data.

McKinsey’s research indicates that lenders who did adopt such models at the right time reaped multiple benefits:

• 5-15% revenue increase – by rapidly identifying creditworthy borrowers, such models accelerated acceptance rates. By automating much of the assessment process and cutting out the paperwork, these models improved both the customer experience and the lender’s ability to provide better rates.

• 0-40% decrease in losses to default – credit-worthiness aside, such models can also predict default likelihood precisely. Lenders, who invariably have to provide for losses by holding onto capital, benefit two-fold when credit losses decrease – they can afford to hold less capital, and they can depend on better revenue from interest.

• 20-40% gains in efficiency – lenders can source data faster, provide straight-through processing for low-risk applicants and better scrutinize prospects that are both promising but high-risk.

Evolution is not only necessary, but also confers tangible, non-trivial benefits – and in the next section, we will discuss how manual documentation and juggling with multiple, criss-crossing platforms leads to near-unaffordable funding delays.

Challenge 2: Preventing Funding Delays

Here’s a problem that every lender is going to face soon: defaults when governments cease support programs initiated during COVID.

Governments around the world loosened regulations and supported both businesses and citizens to an unprecedented degree. The US Small Business Administration, which has funded thousands of businesses, has instituted measures to manage post-pandemic commercial debt. Covid-period economic support cannot continue indefinitely. Banks may be swamped with defaults when businesses go under when they are forced to stand alone and in fact, major players such as Wells Fargo and JP Morgan revealed that they had set funds aside to deal with defaults in commercial real estate loans, especially as regulators have their eye on banks with CRE exposure in their loan portfolios.

Banks need to identify companies that have managed to evolve and thrive during the lockdowns and fund them quickly if they want revenue from interest to outbalance losses to default.

Credit decisioning models that can parse a great deal of data can help quickly identify the right business, but assessing loan documents manually will simply nullify these time savings. Besides, crafting a credit decisioning engine from scratch is not easy, but automating documentation with a digital platform is an easier transition to make, especially because it confers immediate benefits even to borrowers today, who expect detailed dashboards and digitally-streamlined experiences that clearly spell out their debt obligations.

As a McKinsey report states, currently, the average ‘time-to-cash’ from lenders is about three months. But banks and lenders that have embraced the concept of digital credit are now capable of funding within 24 hours

A key benefit of automation is the freedom it confers on expert loan officers to focus on the deals that really count, while low-risk loans can be automated. However, lenders remain wary of digitization, are frustrated with legacy digital platforms that do not seem to provide any tangible benefit, and lack a single ‘owner’ of credit who can drive digital transformation.

But when digital transformation of the credit journey is done right, both lender and borrower benefit. The McKinsey report describes an exemplar of this sort of digitization – a Scandinavian bank that set up a seamless, end-to-end online process for their SME customers by simply leveraging the extensive historical customer data they already had. The platform automated document processing, fast-tracked low-risk applications and drastically cut down on paperwork and collateral review, resulting in major efficiency gains. While the bank did away with its legacy platform, it also focused on creating a minimum viable product for their digital lending solution, reserving more experimental features for later, and did not create any sophisticated credit engine based on big data – as a result, this platform was up and running in less than six months and was servicing the numerous SME’s with whom they had long-standing relationships.

In fact, digitization doesn’t even have to make loan officers redundant – it can put their expertise to use where it is most needed. Some banks have literally created interfaces where relationship managers and borrowers can discuss and complete digital loan applications on a shared screen that allows the loan officer to address any and all of the borrower’s concerns, including the risks due to automation – and this is precisely what borrowers say they want – a digital experience mediated by human expertise.

Such a platform is meant to mature into a fully-digital solution based on an advanced credit engine, and when a lender’s apparatus finally matures, borrowers get cash up to 80 percent faster, decision making time is cut by 30-50% and improved risk decisions set up the lender for greater profits down the road.

And why spend all this money setting up such a platform in-house, when solutions such as Finanta exist for this very purpose?

Challenge 3: Understanding Customer Life-Time Value

Imagine a borrower who has had a long-standing relationship with their bank, trivializing the calculation of the customer lifetime value. Now add the pandemic to the mix, and a new piece of information – the borrower runs a restaurant. No easy formula exists to calculate the customer’s lifetime value now.

In fact, this is much the same reason why banks as big as JP Morgan and Wells Fargo are earmarking funds to provide for losses to default on commercial real estate loans, despite not seeing significant losses yet. This is because of glum forecasts that estimate that office property values could decrease by $800 bn over the next seven years, with demand for office space declining by 13% in less than 6 years. Customer lifetime value is self-evidently connected to default risk, and a global economy emerging from a pandemic throws off most CLV calculations.

This is why relationship banking still matters. Lenders are working with customers directly to minimize CRE loan defaults, and this can be possible only with a full understanding of the customer, with access to proprietary information about the borrower, a long-term relationship with the business, repeated interactions and a clear assessment of the relationship’s profitability over the long run. This, again, has been illustrated already – the Scandinavian bank that set up an online platform in less than six months could not have done so without the data they held thanks to their long-standing relationship with SME customers.

A solution like Finanta also allows a 360-degree view of both existing and prospective customers with a digitally-enhanced platform that can translate years’ worth of a fruitful relationship into data that drives future lending and refinancing decisions. In fact, relationship banking is often based on ‘soft’ information that is difficult to quantify, but it still confers a competitive advantage to lenders because they can make decisions based on long-held trust. Consider the ‘band-aid’ solution described above, wherein a bank assigns a flat high-risk value to a restaurant because of COVID. However, a close relationship with this business would have prevented the lender from making such an error, and a solution or platform that can give a 360-degree view of the customer, containing both financial metrics and information from relationship managers, will inoculate the lender from such errors of judgment.

Such a platform can subsume both ‘hard data’ and ‘soft’ information into a single panoptic view, and can speed up loan allocation. As it helps the lender limit information asymmetry, the platform can encourage cross-collateralization – allowing the borrower to secure multiple loans with a single collateral – and even cross-ownership , where the lender is convinced to buy a stake in the very business it is funding, thanks to a relationship informed by data secured in a digital platform and cemented by trust.

In short, this is the type of solution that will account for shocks like lockdowns, economic crises and more – at its simplest, it merely fuses multiple sources of information to enable relationship banking at a much deeper level.

Challenge 4: Regulatory Compliance

Regulatory compliance is a minefield that commercial lenders are now having to face. We have earlier pointed out that commercial lending has not traditionally been regulated at the state or federal level in the US – but regulations are now cropping up from state to state.

The US Truth in Lending Act (1968) has for decades protected consumers from predatory lending practices, and most importantly, serves as a unifying rubric for state-level compliance. The federal Office of the Comptroller of the Currency clearly spells out the disclosures that lenders are required to make, and also states that no disclosures are required for commercial loans, or even federal student grants that might be taken to learn a trade before setting up shop.

Commercial lending thus doesn’t have a single source of regulatory guidance, and regulations can ironically end up hurting the very businesses they seek to protect. Consider that lenders have had decades to fulfill every single obligation as per TILA, and in fact, one of the first online lending platforms, Quicken Loans, disrupted the home mortgage market simply by operating from a single office staffed by loan officers well-versed in both regional and federal compliance requirements. But smaller commercial lenders cannot pivot so easily, considering that the very regulations that they now have to fulfill are new, requiring them to train staff or hire new compliance officers. They might simply scale back operations, leaving smaller businesses with few borrowing operations. Such businesses may also be unable to fulfill the stricter underwriting norms of larger lenders.

Another problem is to understand how these regulations may affect capital requirements in case of loan defaults (especially as regulators are closely monitoring lenders with high exposure to commercial real estate lending). JP Morgan and Wells Fargo can earmark more funds but smaller players may not be able to do so.

AThe third is to address compliance pressures in a volatile interest rate environment. Banks stand to benefit from today’s record-high interest rates, but cuts can be mandated any time (consider that the Fed is fighting inflation now, keeping rates up, and this state of affairs may not persist) and new macroeconomic shocks can affect the economy as a whole. Interest rate expense management has become crucial as federal liquidity injections during the pandemic are now being withdrawn. Today, a bank’s deposit holdings can even get depleted simply by social media scares as customers withdraw large amounts using easy digital modes – a new type of bank run, as it were.

Regulators are hence demanding updates on loan methodologies, plans for interest rate normalization and more, just to address a future when interest rates will dip from their current peaks. This is why many banks are embracing forward-looking analytics to manage risk, are developing dynamic interest rate hedging strategies, stress testing for multiple risk factors and more – in short, doing what regulators are asking of them as it is, really, the need of the hour.

There is really no silver bullet to the problem here – as this section makes clear. But no matter what solution – or combination of solutions – a lender adopts, there is a self-evident need for a basic, end-to-end digital platform that can operate within these parameters and actually derive and configure loan terms.

We mentioned above that focussing on a minimum viable product is essential to success. Lenders must evolve to meet new and shifting regulatory constraints – one basic example cited in another blog is that a lender cannot just cut off a relationship with a ‘brown’ company just because green lending is on the rise – precisely because green lending is also under strict regulatory scrutiny. No single platform can solve every problem, but a powerful digital platform can definitely enable decisions made once sophisticated models have addressed the plethora of risk factors, starting from credit decisioning and ending at regulatory compliance – this is what Finanta brings to the table.

Conclusion

No lender or borrower is in a privileged position today. Traditional lenders face challenger banks and peer-to-peer lending platforms but can still leverage the years of trust built around relationship banking, and can even adapt to new regulations by refining strategies honed over decades. They can develop in-house solutions based on years of historical data and they are experienced in dealing with interest rate risk and more.

But consider the brick-and-mortar branches that were set up around the world to allow anyone to access their accounts. Think of Finanta’s end-to-end platform as the 21st century equivalent of this last-mile provision of funding to any and every borrower. Think of it as a digital infrastructure that lenders can use on top of their human expertise. Above all, think of it as the first major step to grapple with the many, many unresolved problems in the commercial lending space.