Why Real-Time Data Matters in Lending

We recently announced a partnership with Method.

This partnership is exciting for a couple reasons – namely that post-COVID, lenders experienced FICO inflation, influenced by factors such as government stimulus, deferred payments, and shifts in consumer spending habits.

For example, a 720 FICO used to signal a certain level of creditworthiness pre-COVID, but the same score today doesn’t mean what it used to. Model drift challenges the reliability of traditional risk models, making it harder for lenders to discern who might default from those likely to repay reliably.

The Fed’s research on this shows that average FICO scores are at a record high, even though delinquency rates for auto loans reached their peak since 2000, and credit card delinquencies increased significantly after 2021.

This drives the need for lenders to detect financial strain early—way before these changes surface in a FICO score and lead to defaults in the loan book.

While FICO maintains that their scores have adapted to reflect current economic conditions accurately and continue to be robust indicators of credit risk, their scores don’t adjust in real-time. Relying on solely FICO can lead lenders to make decisions based on old data – often up to 60 days delayed.

The Problem

Lenders lack live monitoring tools and instead rely on infrequent, ad hoc efforts. This results in a significant gap in continuous, real-time monitoring of model performance and borrower financial health. Such tools are essential not only for forecasting loan performance but also for valuing these assets when selling as securities. The common methods used today are:

• Manual Reviews and Audits: Periodic, sample-based reviews of loan files to assess compliance and credit risk, and any significant changes
• Spreadsheet Tracking: Use of manual spreadsheets to monitor loan performance, prone to errors and not scalable.
• Consulting Experts: Occasional consultations with external financial analysts or credit experts for advice on complex cases.

These methods highlight the lack of continuous, real-time data analysis in traditional lending practices, underscoring the need for more advanced and real-time monitoring tools.

The Solution

Recognizing the gaps in traditional scoring methods, our partnership with Method equips lenders with real-time data and cashflow analytics. These tools are designed to detect early signs of financial strain—well before they manifest in FICO scores and lead to defaults. This strategic integration allows lenders to maintain a clearer and more current understanding of borrower creditworthiness, ensuring that risk assessment keeps pace with rapidly changing economic realities.

We’re not replacing existing models; we’re enhancing them with continuous, real-time insights to mitigate the risk of model drift.

The Future

Lenders expect the inflation of FICO scores to continue for a while (and they have to adjust for it).

This doesn’t mean that FICO will be replaced, instead, it means 3 things:

1. Lenders have an appetite for more real-time insight into performance. On the SMB side, it’s been known that credit score is not indicative of creditworthiness, hence the appetite for cashflow analytics
2. FICO is less important in lender’s models but still the standard for securitization. There’s a good post on this by Alex Johnson: https://workweek.com/2024/01/12/fico-score/
3. An ecosystem of tools is emerging to help lenders react faster: https://workweek.com/2024/03/20/credit-decisioning-os/

As lenders adapt to a landscape where FICO scores may no longer fully reflect borrower risk, the shift towards more agile, real-time analytics tools is becoming crucial. This shift is supported by innovative solutions from companies like Pave and Method, which enhance lenders’ capabilities to assess and manage credit risk effectively in real-time.

‍