Innovation isn’t just about faster transactions or slicker apps—it’s about smarter, fairer decisions. A groundbreaking study from Mongolia offers a powerful insight for lenders and neobanks everywhere: psychological profiling can predict creditworthiness with up to 91% accuracy, even when traditional financial data is missing.
Published in Behavioral Sciences (MDPI, 2021), the study shows how mobile-based microlenders used psychometric evaluations to assess risk among borrowers who lack formal credit histories. The results? Eye-opening—and potentially industry-shifting.
The Context: Credit Scoring Without a Credit Score
In markets like Mongolia, a large percentage of the population is underbanked. Many don’t have stable income, formal jobs, or any credit record at all—making them invisible to conventional credit scoring systems. This challenge isn’t unique to Mongolia; it affects over 1.7 billion people globally, particularly in emerging economies and gig-work markets.
Fintech platforms have a unique opportunity here. By exploring alternative data sources—especially behavioral and psychological traits—they can better assess risk, extend credit, and promote inclusion without compromising portfolio quality.
Inside the Study: How Psychology Beat the Scoreboard
The research team surveyed 1,118 borrowers who had taken out loans through mobile microlending services. Participants answered a psychometric questionnaire measuring traits like:
- Self-control and future orientation
- Big Five personality traits (e.g., conscientiousness, neuroticism)
- Altruism and moral identity
- Attitudes toward money, debt, and risk
These psychological traits were then correlated with real-world repayment behavior—revealing that certain traits strongly predicted credit performance.
Key Takeaways Fintech Shouldn’t Ignore
Traits That Predict Repayment
Borrowers with higher self-control, conscientiousness, and a strong future orientation were more likely to repay their loans. These traits point to better planning, financial discipline, and reliability.
Traits That Signal Risk
Higher levels of neuroticism, impulsivity, and risk-seeking attitudes were statistically linked to higher default rates. These insights allow lenders to proactively screen for risky behaviors—before money changes hands.
91% Accuracy
Using machine learning models, researchers achieved up to 91% accuracy in predicting loan repayment behavior using psychometric data. This level of precision makes it a viable complement—or even alternative—to traditional scoring.
Trust, Altruism, and Morality Matter
The study also found that borrowers with strong moral identity and altruism scores tended to honor their financial commitments—an underappreciated signal in most risk models.
Strategic Implications for Fintech
1. Serve the Credit-Invisible Market
Psychometric assessments unlock access to underbanked segments: gig workers, informal entrepreneurs, and young adults without credit histories. Fintech platforms can now build inclusive lending models—without relying on legacy financial data.
2. Smarter AI-Driven Underwriting
Psychological data can be seamlessly integrated into AI-based credit models. When used ethically, these traits help power personalized lending, dynamic risk scoring, and better portfolio management.
3. Better Portfolio Health, Lower Defaults
Early identification of high-risk behaviors allows fintechs to avoid defaults while still expanding access. It’s a win-win: broader reach with reduced risk.
Use Case Ideas for Fintech Leaders
- App-Based Behavioral Surveys during user onboarding or loan applications.
- Gamified Psychometric Tests that feel engaging while collecting valuable data.
- Blended Scoring Models that combine psychometrics with financial and device-based alternative data.
- Micro-tests for assessing traits like impulsivity, time preference, and money attitudes in real time.
Risks and Considerations: Ethics Comes First
While the benefits are clear, fintech firms must address the ethical and legal challenges of psychometric-based lending:
- Informed Consent: Borrowers must understand how their psychological data is used.
- Data Privacy: Compliance with GDPR, CCPA, and local data regulations is non-negotiable.
- Bias and Fairness: Behavioral models must be audited for fairness and avoid reinforcing social or cultural biases.
The Future: Behavior-Driven Lending
This research shows that psychometric data is more than a novelty—it’s a credible, accurate tool for next-gen credit scoring. As fintechs strive to build inclusive, intelligent, and scalable lending systems, the human side of finance—behavior, values, personality—can no longer be ignored.
Whether you’re a credit decision scientist, product strategist, or risk officer, the message is clear: the next wave of fintech credit models will be powered not just by data—but by insight into human behavior.
Read the full study here:
Effect of Psychological Factors on Credit Risk – MDPI
