The opportunities for enrichment in SMB underwriting

The challenges with SMB lending

Fintech writer Alex Johnson recently wrote about the SMB lending space, outlining the core challenges with underwriting small businesses;

1. Diversity - SMBs are a heterogenous bunch
2. Data Coverage - credit bureau data for SMBs is not very comprehensive and SMBs churn at a much higher rate
3. Verification - KYB is much harder than KYC, businesses can have many owners, addresses and subsidiaries

As a result of these challenges and the smaller loan values, SMB lending is difficult to do profitably, creating a significant capital gap with "one third of small businesses in the US told the NSBA that they could not secure the finacning they need for thir businesses." he wrote.

Opportunities of bank data

One of the key stratgies that Alex outlines in succeeding at SMB underwriting is utilizing new sources of data like some of our customers like Wayflyer have found.

Many new SMB lenders incorporate bank transactions in their underwriting workflows for a number of reasons;

1. Bank data is the source of truth
2. Bank data is effectively real-time
3. Bank data can be 100% reliable if accessed via APIs
4. Bank data is granular and un-aggregated

Challenges with bank data

Despite these benefits, there is one main challenge with bank data; it is very messy and hard to make sense of.This means that after receiving it from prospective borrowers, it needs to be cleansed and enriched, which is where Ntropy comes in.

As an example, what do the above business transactions say about the risk of lending to a company? How can underwriters make sense of them in their raw form?

The power of enrichment

Ntropy trains language models to understand financial data at scale and in miliseconds.

Our models parse in raw bank transactions from our customers and provide a clean, standardized and enriched output. We enrich each transaction with a logo, website and most importantly a category that can be incorporated into an underwriting model.
⁠
⁠(Visual from Sofia of an example)
⁠
⁠In addition to cleaning and enriching bank data, using our new Insights product, we can also instantly create both P&L and cashflow statements from bank data.

As Alex points out in his post underwriters spend more of their time transforming and formatting data and our enrichment products and Insights are designed to automated much of this. To quote his post below

⁠"the majority of their underwriters’ time (as much as 80%) is spent aggregating, transforming, and formatting all of this data to get it into a structure that actually allows them to make an underwriting decision."

With Ntropy, lenders can drag and drop bank transactions via PDF, CSV, or via our API and get back clean, enriched data and financial statements. This saves your underwriters time, enables them to easily validate a borrowers self-reported financials and standarizes the process amongst a team of underwriters to ensure consistency.