Consumer lending is a new opportunity.

Innovative entrants will have rare access to customer segments that are not served by the credit cycle if it is reopened. The global COVID-19 epidemic profoundly impacted the economy and ended most of the credit cycles in most markets. These markets will slowly return to normal activity, and a new credit cycle is set. This allows innovative lenders to enter credit markets and gain market share. New entrants, such as insurance companies and utilities, will also be able to enter the market with the resumption.

While banks offer financing solutions for many of the world’s population, many consumers are not being served or underserved. New-to-market lenders can identify gaps in lending coverage and help bridge them. Many potential customers want customized solutions that are more cost-effective than traditional banks. The new entrants can quickly design new offerings and are not hampered by legacy processes. They can go from concept to fully developed offering in months, as opposed to the one to two years it takes for incumbents.

These new-to-market lenders might not have consumer lending operations, and they may not be servicing consumers with credit histories. They may lack the necessary lending infrastructure, credit risk models, and reference data. They will need to develop these capabilities while managing the business’s risks.

Traditional banks could expand their market share, and non-bank financial institutions to be new-to-market lenders. These lenders will be required to manage credit risk and the enabling technologies. Lenders can quickly establish a credit-decision platform and move quickly while still taking on the appropriate level of credit risk.

Use data from many sources.

New-to-market lenders must combine data from many sources to model credit risk. They can compensate for any lack of credit expertise by collecting diverse data, even data they do not own. There are many types of traditional credit behaviour and demographic data for established financial institutions that can be accessed. These data include information about loan amounts, bank deposits, current account information, point-of-sale transactions data, and information about lenders. Nonfinancial companies have other internal customer data sources, such as product usage, interactions with customer-relationship management, call records, email records, customer feedback, and website navigational data.

Lenders can request data from other sources provided they adhere to all privacy guidelines and regulations. External data can be from retailers, telecommunications providers, utility providers, government agencies, or other banks. Partnering with companies with the necessary data may be an option for certain types of lenders. This strategy, which involves a joint venture with companies with complementary data about consumer segments, may be especially suitable for lenders with a regional presence.

One telecommunications company’s approach is instructive. The company created an unsecured cash loan product to serve customers without formal credit. It wasn’t easy because the company didn’t have enough credit information to create the product. The company used its customer-usage data, specifically data on mobile bill payments, to solve the problem. This data allowed the company to create a proxy variable in its credit model training. The target variable performed the same as credit-related information for banks when back-tested. The company then extended credit via a pilot model to prepaid customers, refined using real-world data.

Create the decision engine

The second step is to create the decision engine. This area will give new entrants a significant advantage over existing lenders with legacy software they don’t want to modify. Advanced analytics, machine learning and other tools can build the new decision engine.

Machine learning will allow new lenders to automate up to 95 per cent of their underwriting processes and make more accurate credit decisions. Machine-learning real-time solutions can also improve pricing and limit setting and assist firms in monitoring existing customers and credit lines via smarter early warning systems. Lenders can use straight-through processing to speed up transactions and provide a better customer experience.

Modular design is possible for the decision engine to allow maximum flexibility. This will enable lenders to control strategic processes and possibly outsource other parts. Modular formats can be used to facilitate risk assessment.

This risk assessment method is very different from the one used at large companies. Traditional setups are often one large system that covers all aspects of lending, including creditworthiness assessment and printing documents. This approach is becoming increasingly obsolete as it restricts incumbent lenders’ ability to adapt quickly.

Our experience shows that agile development and implementation can cut down the time to launch a credit engine. This is compared to nearly a year with traditional approaches. One European bank wanted to launch a digital lending division. The bank was limited by its legacy systems and ingrained processes that made it difficult to develop new products. The bank developed a modular credit-decision system that combined parts of existing systems and allowed the team to create new modules. This resulted in a quicker time to market for the digitally launched business.

Create scalable infrastructure

New-to-market lenders have many options when it comes to developing the technology infrastructure. You can begin by identifying your ambitions and market advantage and how current technology and data availability will help or hinder the initiative. Organizations can then plan the best way forward.

Strong customer relationships are a key competitive advantage for companies that want to succeed. However, they may not require extensive risk assessment processes. These companies may be able to purchase turnkey solutions from a trusted solutions provider. Many standard market solutions can be purchased or outsourced. Many offer a complete offering, including credit origination, line administration, automated decision making for credit assessments, customer acquisition, renewals and exposure monitoring. Lenders can see the performance of a portfolio or one customer easily. They can also access credit bureau solutions to enrich their data. While the turnkey approach is faster, it limits customization. Configuring a turnkey solution within a company’s IT infrastructure can prove cumbersome.

Lenders at the opposite end of the spectrum will be competitive because they can rely on a tailored, integrated solution. This could mean building and designing infrastructure from scratch. These complex, tailored solutions require significant time and financial investment. This strategy may require the hiring of talent with specialized skills.

Lenders can choose between fully customized and off-the-shelf solutions. They can also buy individual applications and solutions that can be assembled modularly. This will help lenders gain a competitive advantage in the market. Lenders will be able to customize infrastructure to serve target customer segments better using their credit risk models and solutions. Lenders may also choose a variant of hybrid solution, entailing a custom-built front-end infrastructure–such as the workflow manager–and a standard market solution for back-end elements, such as collateral management or exposure systems.

A second telecommunications company with an 80 per cent subscriber base collaborated with fintech partners to launch a new lending service. This project involved the development of technology to support existing data platforms. The company needed to train existing employees and recruit new talent to run its lending business. Long-term goals include expanding the product range, increasing the infrastructure to support a wider portfolio and collaborating with other financial institutions (by selling credit scoring services).

There are always trade-offs between cost and flexibility when buying or building. How much you spend on maintenance, development, and development is how flexible your solution is.

Over time, monitor and maintain the models.

New-to-market lenders must track key metrics to monitor the model’s performance over time. While each model development is a one-time task, maintaining and monitoring the models is an ongoing responsibility. Lenders can identify problems early by using established metrics to track changes and performance of models over time.

For example, the population stability index measures the current customer base of a lender against the population for whom a risk modelling was first established. The credit-default rate is another indicator of financial health. Metrics based on Gini coefficients will tell if the risk model makes accurate predictions.