What sort of "Ph.D. APIs" will be produced for the Open Banking era?

There are lending business models being devised that have exciting potential to reach new and underserved niche segments of the borrowing market. New niche and narrow lenders can be very effective at matching supply and demand in a focused market segment without major organisational overhead. By staying in specific niches and/or by constraining their proposition very narrowly, they can lower search costs for borrowers and lower due diligence costs for both lenders and borrowers. They can offer online and mobile loan applications in very short time frames. By staying niche and narrow, they can remove friction from a credit application, gain a deep appreciation of the niche sector’s characteristics and risks, increase the speed of capital delivery and improve the focus on customer service. Borrowers are responding positively to these new business models. Borrowers increasingly expect lending services to be comparable among the offerings of competitors and they demand a customer experience that matches that of using a social network. The tolerance for tedious compliance, onerous documentation requirements and unavailable representatives is wearing ever thinner. The emerging niche and narrow lenders can also be attractive to borrowers who have no other choice but to seek alternative lenders. This allows the new niche and narrow lenders to tap into a market with high demand, generating high margins for credit. In the early stages of market development, the loans granted by these new lenders are often uncollateralised. Rapid decisioning on whether to grant credit or not takes place in the absence of hard assets being used as collateral, which removes significant complexity and friction from the process.

Open Banking will also increase the opportunity for new service providers to reach new and underserved niche segments of the borrowing market. Open Banking has the potential to strengthen the capability of these new niche and narrow lenders by creating a streamlined supply of cash flow data about prospective borrowers. Open Banking will alleviate the constraints caused by the distribution bottleneck for this cash flow data, reducing the scarcity of information and expanding the shelf space for examining the cash flow of lending prospects. Open Banking has the potential to reveal a shape of credible borrowing demand that was not visible before. This emerging shared and data-rich environment can address the key supply-side factor that determines whether a sales distribution has a long tail i.e. the cost of inventory storage and distribution. Where inventory storage and distribution costs are insignificant, it becomes economically viable to sell relatively unpopular products; however, when storage and distribution costs are high, niches cannot be reached and only the most popular products can be sold. The potential effect of Open Banking for more potential borrowers is to lower the “search costs” of ?nding new niche and narrow lending services aimed specifically at them i.e. loans that are informed by high quality cash flow data that is sensitive to their specific business prospects and specific market segment. In economics, search costs refer to anything that gets in the way of ?nding what you want. Some of those costs are non-monetary, such as wasted time, hassle, wrong turns and confusion. Other costs are financially measurable, such as borrowing in the wrong way or at the wrong terms (or indeed not borrowing and missing expansion opportunities).

While the new methods of credit analytics being investigated by these narrow and niche lenders are exciting, there could be pitfalls. While these new niche and narrow lenders have the capability to reach new and underserved market segments of borrowers, these new lenders are technology and data reliant. Transactions take place exclusively online on a largely automated platform, with heavy use of technology and “big data”.   They are developing proprietary credit models for loan approvals and pricing in that specific niche. The emerging niche and narrow lenders are pioneering the use of fringe alternative data to underwrite loans when no full credit report is available or as a complement to existing credit scores. This alternative data could include trading activity on social networks, shopping habits, location data, web tracking etc. 

While these new niche and narrow lenders may be pioneering new methods of reaching prospective borrowers that the mainstream, mass-market lenders cannot reach, there may be inferior aspects to the way that these new lenders assess the prospective borrowers as reliable counterparties to a loan. In the early stages of this development, the credit analytics being used by the niche and narrow lenders are proprietary and undisclosed. There could be an overreliance on “big data” analytics as the only form of risk management. Whether fringe alternative data can accurately predict credit events is unproven. Rightly or wrongly, the large incumbent banks still largely rely on the financial history of borrowers and their credit ratings. It is not clear yet whether a rating approach that involves “big data” is superior. There is a risk of putting too much trust in data and software. Not many investors that are funding FinTech start-ups are experts in the lending processes of banks and associated risk analytics. It remains to be seen how the new models of borrower rating hold up.

Regulators are watching closely to see if there are potential pitfalls in the new methods of credit analytics.  The immaturity of the risk management models of the emerging narrow and niche lenders has attracted the specific attention of the European Central Bank (ECB) in their September 2017 “Guide to assessments of fintech credit institution licence applications” (the Guide). The ECB defines “fintech” is an umbrella term encompassing a wide variety of business models. In line with the ECB’s responsibilities, the Guide exclusively refers to bank business models in which the production and delivery of banking products and services are based on technology-enabled innovation. The ECB aims to allow scope for innovative market participants to contribute positively to the financial sector and seeks to achieve this, in accordance with its mandate to maintain the safety and soundness of the European banking system, by maintaining adequate prudential standards for newly licensed banks. The ECB is quite clear about its role– it is to ensure that fintech banks are properly authorised and have in place risk control frameworks for anticipating, understanding and responding to the risks arising in their field of operations. The ECB is clear that, to ensure a level playing field, fintech banks must be held to the same standards as other banks.

The ECB focuses specifically on the use of “alternative credit-scoring methods and data” by these emerging niche and narrow lenders. The ECB explicitly states that where alternative data sources and credit scoring methodologies are used, regulators will assess whether their use is supported by commensurate risk management and the necessary capital safeguards. The ECB recognises that start-up phase of a fintech bank could pose a greater risk of financial losses which may progressively reduce the amount of own funds available. The regulator expresses concern that the business plan of a fintech bank in its start-up phase may therefore entail an aggressive pricing strategy to gain market share. As a fintech bank learns more about its operating environment, the ECB recognises the risk that it may be more likely to change its business model to respond to market needs in order to maintain profitability in what is often a niche segment. The ECB also identifies that during the start-up phase a fintech bank may face increased liquidity risks, when online depositors might exhibit price sensitive behaviours, being more likely to withdraw their deposits and switch to a competitor paying higher interest rates. They cite a real risk that online deposits accepted by fintech banks are more likely to be volatile and less “sticky” than traditional bank deposits.

Avoiding these potential pitfalls in credit analytics isn’t just a matter of new niche and narrow lenders having sufficient capital resources to solve the problems. Even giants with vast capital resources like Amazon may be pausing for thought about how to achieve effective credit analytics.  Amazon has recently partnered with Bank of America Merrill Lynch to provide loans to merchants. Reporting suggests that Amazon had been reassessing its credit risk exposure and that its total lending total grew only slightly in 2017 after almost doubling in 2016. There is speculation that there had been a deliberate effort by Amazon to slow the expansion in 2017 as the company attempted to better understand the credit risks that come with a large-scale lending practice. Whether this speculation is true or not, the reporting does highlight an important question about non-banks emerging as niche and narrow lenders in an Open Banking ecosystem. Just how much relevant data and specialist knowledge is needed to lend successfully in the long term – and can the emerging niche and narrow lenders efficiently obtain this relevant data and specialist knowledge?

What extra data and processes could the new niche and narrow lenders need to improve their risk management to the level that regulators have come to expect from the big incumbent lenders that operate at a very high scale across multiple borrowing market segments? The new niche and narrow lenders should have a thorough insight into all financial risks relating to their lending, including risk factors they have exposure to - such as interest rates, currencies, inflation and asset prices (because changes in asset prices change the value of collateral for loans). Expected and unexpected losses that may arise from exposure to market risk should also be clear. The forecasts of future market risks being used by the new niche lenders should be derived from reliable and consistent models of market risk. Robust analytics should measure all risks based on real-world stressed market conditions, stressed default probabilities, in addition to expectations of general borrower behaviour after a significant market shock. Estimating unexpected losses from all potential risks is also a must. The emerging niche lenders need to know the systemic links between their borrowers and the peers of the borrowers. This is particularly important for new niche lenders as they are naturally prone to concentration risk (i.e. lending to borrowers with similar business models in the same market sector).  

There will be a temptation for the big incumbent lenders to let these new niche and narrow lenders to struggle to solve these credit analytics problems on their own. An incumbent bank with a long-term perspective that competition is a matter of “product-versus-product” will probably be disinclined to commercialise its risk management models so that can be used by emerging niche and narrow lenders. Some incumbent banks could retain a perspective that the narrow and niche lender has a fundamentally weak product in the long term. It can assume that the quality of the borrowers will be low (e.g. “sub-prime”) and that the niche and narrow market position of the lender will produce an unhealthy and highly correlated population of borrowers. Given the long-term environment of lower-for-longer interest rates, a bank with this perspective of competition will assume that these emerging niche lenders have a business model that is likely to struggle in stressed market conditions. A bank with a perspective of competition that is product-versus-product will also focus on any high correlation of sources that are funding these loans, which results in a less reliable interest rate spread between lending rates and funding rates. A bank with this product-versus-product perspective can also hope that no sophisticated rivals with extensive risk management models will emerge to sell data and processes that will assist the risk assessment and pricing by these new niche and narrow lenders. 

In contrast, some big incumbent banks who are focused on developing their presence in emerging ecosystems may see an opportunity to sell some of their credit analytics expertise to these emerging niche lenders. They will seek an effective and efficient methods to commercially distribute their credit risk management analytics and models to as many emerging niche and narrow lenders as possible. By now, most IT professionals and line-of-business managers “get” APIs. They know that APIs enable them to quickly and easily add capabilities to their own products and services and that offering APIs is a terrific opportunity to expand audience and gain mindshare. But now the API Economy is seeing a new generation of APIs--ones that are smarter and better. Dubbed “Ph.D. APIs”, these are APIs that package and deliver the expertise of a team of doctoral students and researchers. Making use of Ph.D. APIs means companies can add highly sophisticated (and often expensive) features that will benefit the business and the business’s customers. APIs enable companies to add capabilities quickly and easily. The great opportunity is that, as the APIs get more sophisticated, the capabilities that can be added to the business become more valuable, increasing the return on investment exponentially. In the emerging Open Banking era, there could be many Ph.D. APIs for credit analytics that can add highly sophisticated features to the lending processes of new lenders that will benefit the new lenders and their customers.

An incumbent bank with a long-term perspective that competition is becoming a matter of ecosystem-versus-ecosystem is far more likely to be inclined to commercialise its risk management models into Ph.D. APIs. A bank with this strategic perspective could move to develop, operate and support a credit risk analytics platform to be used by emerging niche and narrow lenders. It could aim for revenues from incumbent banks, fintech banks, peer-to-peer lending platforms and regulators. A sophisticated lending bank could spearhead a credit analytics platform with its existing experience and talent. The market could be served with world-class interactive analytics for a fraction of the cost of emerging niche and narrow lenders developing and supporting a solution in- house. Users of the platform would stream data into the platform and receive processed analytics via APis. This would make investments in new lending to the “long tail” more robust against risk and potential losses and ensure optimal profitability. It would help identify systemic concentration and maintain standards in financial data and analytics regarding industry risk and profitability. Aside from the API usage fees that a sophisticated lender yields by offering Ph.D. APIs for credit analytics to new niche and narrow lenders, the Ph.D. API provider positions itself into this emerging value chain with an opportunity to envelop new lending activity in the newly served niches into its own balance sheet growth targets, once it has become comfortable with the quality and durability of lending demand from the “long tail” of the borrowing market Offering Ph.D. APIs gains a vantage point for future expansion into underserved segments as well as new revenue streams without the credit risk.

In crude conclusion, building an ecosystem can offer an advantage when the effectiveness of the loosely coupled ecosystem exceeds that of a traditional organisation. Three distinctive entities - the multi-sectoral and deposit funded incumbent bank, the emerging niche and narrow lenders and the borrowers must be better off with a business ecosystem for it to be more attractive than a traditional go-it-alone business model. For the traditional bank aiming to act as an ecosystem orchestrator, the value proposition is to innovate faster and cheaper around their core business capabilities than rivals. Developing internally focused risk management models buried inside traditional banks into Ph.D. APIs can help the ecosystem to reach potential untapped market microsegments of borrowers with acceptable levels of risk. A focus on Ph.D. APIs will help traditional banks to focus more deeply on their core competencies that are difficult for new entrants to imitate or substitute. It may be more valuable for incumbent banks than focusing their API strategies on Young Millennials in Starbucks with Mobile Phones, as these Young Millennials are as yet having a very limited impact on the industry’s balance sheet. For the emerging niche and narrow lenders, the value proposition of these Ph.D. APIs is that they offer a scalable analytics foundation for their own work that allows them to better focus on their own data gathering expertise within niches and to reach a vastly expanded market in the “long tail”. For borrowers across the spectrum, the value proposition is faster innovation, increased customisation and lower search and transaction costs.