Fintech is an exciting area in the financial services sector with rapidly growing startups and companies that are disrupting the entire industry, and many fintech companies are targeting niche market segments with innovative solutions. 

One of these segments is SMBs. Working with SMBs has unique challenges such as finding the right leads, credit risk assessments, and fraud prevention. 

The following series of articles will highlight how the right external data can enrich internal data to unlock new opportunities and create competitive advantages. This first article in this series will cover credit risk assessment.

How inaccurate credit risk models are costing fintechs millions

SMBs are increasingly turning to fintechs to provide more agile loan approvals with less stringent background checks than regular banks. This is fintech's great advantage over traditional financial institutions, but it comes with the added tendency to attract high-risk borrowers. 

Before the global pandemic, a fintech company's risk assessment model would typically use data created during a time when markets were stable, economies were growing, and SMBs didn’t face as many uncertainties as they do today. 

But in a post-pandemic economy, using historical data for current risk assessment models will not provide the full picture of an SMB's risk profile. For example, cash flow data is an increasingly unreliable metric for determining risk due to the irregularity in which lockdowns or restrictions are imposed.

This discrepancy between models and reality has led to an increased number of loan defaults and even fraud, costing fintech companies millions in capital.

If you're in the fintech space, especially if servicing SMBs, Tearsheet would like to learn more about your relationship with data in the survey below:

^{Survey: How do you use external (alternative) data?}

How external data can greatly impact the accuracy of credit risk assessment models 

Fintech startups need newer, constantly updating, and relevant data points to enrich their risk models. To achieve this, they must look at alternative sources like third-party data providers and data marketplaces and purchase relevant external data from them, not without expense.

What many fintech companies are aware of, yet do not know where to find, are broader data points such as industry trends, legislative changes, an SMB's over-reliance on a single large client, a build-up of risk in certain products, and so on. This is where external data can come in to fill in the gaps in knowledge.

Some examples of external data signals that help uncover deeper insights and context for lending decisions are: 

• Financial activity data, including the potential SMB’s income, borrowing history, transaction history, assets, and liabilities
• Company data, such as HQ and office locations, years in operation, and company type
• Owner validation, including identity overview, phone and email validation, and employment details
• Web presence data, such as domain creation and expiration dates, number of related links, and global website traffic rank
• Online reviews and ratings, such as Yelp reviews and social media reviews for example
• Location-specific income and financial stability indicators 
• Cost of maintenance vs. predicted revenue of an SMB's day-to-day operations
• Alternative risk scores based on a variety of economic and financial health data that indicate fraud risk
• Internet and social media behavior that points to past fraudulent activity

Creating a novel alternative credit scoring model with external data 

Now that we have aired out the various types of external data, here's how fintechs can use it to its fullest potential. To avoid high-risk applicants, fintechs should import external data signals to expand their indicators of risk. This will create a tiered alternative credit scoring model to identify the levels of loan default risk a business will pose to the lender. 

This alternative credit scoring system can provide 1) automatic exclusion for high-risk businesses, 2) deeper manual review for companies with acceptable levels of risk, and 3) immediate automatic loan pre-approval for low-risk firms.

By enriching data like this, SMB lenders can avoid unacceptable high-risk applicants while selecting quality applicants for loan approval. With the right external data, organizations can retrain existing risk models with new data and reduce default rates by 10% and more. 

External data platforms like Explorium can provide data enrichment to append lenders' internal datasets with external data so that their credit risk models show a borrower's default risk more accurately.

For more insights, download our white paper, The Definitive Guide to External Data for Fintech.