Layla Zulfa
The mortgage industry stands at a pivotal juncture, with lenders increasingly embracing a new era of credit scoring driven by advanced models like FICO 10T and VantageScore 4.0, alongside a broader spectrum of borrower data. This push towards modernization and digital transformation, aimed at enhancing efficiency, is met with a critical question: how extensively should lenders leverage these new capabilities? This central tension formed the core of a "Credit Super Session" held Tuesday at the Mortgage Industry Standards Maintenance Organization (MISMO) Fall Summit. Executives from FICO, VantageScore, Experian, Equifax, and TransUnion engaged in an hour-long discussion moderated by HousingWire CEO Clayton Collins, exploring the practical implications of integrating enhanced credit data into the mortgage origination process.
Collins initiated the discourse by posing three fundamental questions that encapsulate the industry’s evolving considerations: "Can we use more data? May we? And should we?" While the first question regarding the ability to incorporate more data is rapidly being answered in the affirmative, the latter two delve into the more nuanced aspects of feasibility, permissibility, and strategic imperative.
The Dawn of Trended Data: A Paradigm Shift in Credit Assessment
A significant consensus emerged around the first question: the capacity to utilize more data is no longer in doubt. Both FICO 10T and VantageScore 4.0 have received approval for use in the mortgage market. Crucially, these new models incorporate "trended" credit information, moving beyond the traditional point-in-time snapshots. This shift offers lenders a more dynamic and longitudinal view of consumer financial behavior, revealing patterns and habits over time rather than just a single snapshot of their creditworthiness on a given day. This development represents a fundamental departure from the decades-old reliance on static credit scoring.
The approval of these advanced scoring models, particularly FICO 10T and VantageScore 4.0, is a significant step for the industry. FICO 10T, for instance, is designed to provide a more predictive score by analyzing up to 24 months of historical credit behavior, including how consumers manage their revolving credit balances. VantageScore 4.0, similarly, has enhanced its predictive capabilities by incorporating trended data and expanding its inclusion of alternative data sources. This evolution aims to provide a more granular understanding of borrower risk, potentially opening up credit opportunities for a wider range of consumers and offering greater precision in risk assessment for lenders.
Navigating the "May We" and "Should We": The Delicate Balance of Data Integration
The latter two questions – "May we?" and "Should we?" – proved to be far more complex, eliciting varied perspectives and highlighting areas where industry-wide consensus is still developing. Panelists largely concurred that an expanded data set can indeed lead to more informed and potentially more accurate credit decisions. However, they also underscored the critical need for lenders to engage in a rigorous evaluation process. This evaluation must encompass not only the potential benefits of enhanced data but also its implications for model performance across diverse borrower segments, the associated costs of implementation and ongoing data acquisition, the readiness of existing technological infrastructure, the acceptance by secondary market investors and agencies, and the ever-present landscape of regulatory considerations.
Anthony Hutchinson, Executive Vice President and Head of Public Affairs at VantageScore, emphasized the operational imperative for lenders. "Ensure that your systems can absorb both of them," he advised, referring to FICO 10T and VantageScore 4.0. This recommendation is critical, as it enables loan officers and other internal stakeholders within lending organizations to "use both and understand both." This transition presents a novel and evolving operational challenge for mortgage lenders, many of whom have historically operated with a singular, dominant credit scoring model as their primary decisioning tool. The ability to seamlessly integrate and interpret outputs from multiple advanced scoring models requires significant investment in system upgrades and staff training.
Justin Demola, Senior Vice President of Mortgage and Housing at Equifax, articulated the rationale for supporting both models: "There is definitely a need for both" scoring systems. He explained that different borrower profiles and distinct stages within the lending lifecycle may be better served by the unique analytical strengths of each model. This suggests a future where lenders might employ a multi-score strategy, selecting the most appropriate scoring model based on the specific characteristics of the loan application or the borrower’s profile.
The Cost-Performance Equation: Optimizing for Efficiency and Access
The strategic considerations surrounding data integration are intricately linked to the persistent pressure on lenders to reduce mortgage origination costs. As Demola highlighted, the fundamental objectives of lenders when engaging with credit data providers consistently revolve around "increasing revenue, reducing costs, and increasing efficiency." However, the pursuit of lower costs does not always translate to superior outcomes across the entire mortgage ecosystem.
Eric Lapin of FICO cautioned against a myopic focus on upfront origination costs. He pointed out that credit scoring decisions extend far beyond the initial loan origination, significantly impacting downstream activities such as loan servicing, mortgage insurance, and capital markets transactions. A seemingly cost-effective scoring solution at the origination stage could inadvertently lead to downstream pricing complications if investors or other market participants perceive that the inherent risks have not been adequately factored into the loan’s pricing. Lapin strongly advised lenders to meticulously examine performance data and ensure alignment with investors, rating agencies, and mortgage insurers regarding the models they adopt. This highlights a critical tension between the immediate desire for cost reduction and the long-term imperative of maintaining investor confidence and market stability.
This interplay between cost considerations and performance expectations, coupled with the inherent dilemma of balancing flexibility with standardization, is poised to become an increasingly prominent implementation issue as competition within the mortgage credit scoring market intensifies. The market is moving towards greater choice, but with that choice comes the responsibility of careful selection and strategic deployment.
Unlocking Potential with Alternative Data: A Double-Edged Sword
The integration of alternative data sources presents a similar landscape of both opportunity and challenge. Executives frequently cited rental payment history, utility and telecommunications records, cash-flow information, and consumer-permissioned banking data as valuable tools for more accurately assessing consumers whose traditional credit files may not fully capture their financial picture. This is particularly relevant for borrowers with thin credit files, self-employed individuals, gig economy workers, or those whose financial characteristics deviate from traditional W-2 underwriting patterns.
Susan Allen, Experian’s Chief Product Officer for Housing, cautioned against oversimplifying the integration of alternative data as merely "opening the credit box." She emphasized a critical distinction: "There’s a big difference between accepting more risk as a way to approve more borrowers versus seeing risk differently, calculating it more effectively." Allen further elaborated that a consumer with a thin traditional credit file is "not necessarily a consumer with a thin financial life." She pointed to younger consumers utilizing gig income, peer-to-peer payment platforms like Venmo, buy-now-pay-later products, and extensive rental payment histories as examples of individuals whose financial stability might not be fully reflected in traditional credit reports.
These evolving trends in income and payment documentation are building the strongest business case for modernization. The ability of lenders to identify and approve additional qualified borrowers without materially increasing risk can significantly enhance both credit access for consumers and lender economics. Matias Peterson of TransUnion noted that even a modest increase of 1% or 2% in loan approvals can translate into substantial improvements in a lender’s bottom line, given the significant investment already made in acquiring and processing mortgage prospects. This suggests that leveraging alternative data effectively is not just about expanding access but about more precisely identifying creditworthy individuals who have been historically underserved by traditional credit scoring methods.
Artificial Intelligence: The Devil is in the Details of Implementation
The panel unanimously agreed that the primary hurdle lies in the execution of these advancements. Simply having access to more borrower information does not automatically translate into superior underwriting outcomes. Lenders require robust systems capable of ingesting this diverse data, sophisticated analytical tools to interpret it, and clearly defined policies to govern its influence on decision-making.
Artificial intelligence (AI) introduces another layer of analytical capability, provided that its governance and reliability are rigorously validated. Panelists discussed various AI applications, including document processing, data extraction, advanced analytics, and consumer education. However, Demola raised a pertinent question about the point at which an AI system might effectively begin "acting as a loan officer," potentially triggering licensing requirements and necessitating a careful regulatory review.
Hutchinson underscored the evolving regulatory landscape surrounding AI at both the state and federal levels, urging lenders to proactively involve their compliance and government relations teams when deploying new AI-driven tools. This proactive approach is crucial to navigating the complex web of regulations designed to ensure fairness, transparency, and consumer protection in the application of AI in financial services.
The Path Forward: Data, Decisions, and Digital Readiness
The consensus among the industry leaders at the MISMO Fall Summit pointed towards a clear trajectory: the next phase of mortgage credit modernization is less about the ability to use more data and more about the strategic deployment of that data. The industry is eager to harness the power of enhanced credit scoring and expanded data sets.
The critical competitive question that remains is where within their workflow ecosystems lenders should strategically deploy this enriched data. Furthermore, the preparedness of their internal systems, their relationships with capital markets partners, and their compliance structures will be paramount. The ultimate goal is to transform this influx of information not into additional complexity, but into demonstrably better, more efficient, and more equitable lending decisions. The journey towards a truly modern mortgage credit system is underway, and its success hinges on the industry’s ability to navigate these intricate challenges with foresight and strategic precision.
