The Case for Using Borrowers’ Employment Data in Credit Underwriting
Scienaptic Reasearch
As part of the credit underwriting process, lenders usually verify borrowers’ income and employment status to ensure that they can repay the loan. However, the COVID-19 pandemic has forced thousands of people out of jobs naturally leading to rapid shifts in employment data.
What is noteworthy is that the pandemic has affected a few industries more than the others. Take a look at the results of a worldwide survey conducted by IMD on the impact of COVID-19 on industries for six months from April to Oct 2020.
Here’s another similar perspective from S&P Global:
The greatest short-term impact is in sectors that are hit by social distancing policies, like travel, tourism, and entertainment. Sectors like oil and gas, auto parts, and engineering also face reduced demand and are bound to face short-term impact. Sectors like pharma, software, insurance, and government are less affected.
In this scenario, banks are being forced to make an unprecedented amount of credit decisions in a short timeframe. Traditional through-the-cycle models aren’t working as expected and risk models are underpredicting outcomes if not breaking down completely. Add to this the CARES Act requirements for payment modifications or forbearance, and the exclusion of adverse payment reporting at the bureaus. This will further weaken risk model performance in the coming months.
Given that, we couldn’t stress more on the need to capture data on the industry or sector which the borrower is employed in or does business in.
Let’s explore why lenders should pursue the collection of industry data along with other employment data. How can this data be collected and used in credit underwriting systems today?
Why Should Banks Capture Employment Data as well as Industry Data of Borrowers?
At the bare minimum, lenders need to know the borrower’s salary, designation or position, and employment history. Apart from this, industry or segment data which the borrower operates in will throw light on which industries will be resilient in the current environment. They will be able to identify the borrower segments who are at much higher risk during the COVID-induced downturn than other segments. Without this data, banks will find it hard to predict business and borrower risk.
Employment and industry data of borrowers will allow banks to underwrite loan amounts more efficiently. They can use it to make real-time analyses on business risk and resiliency, including insights such as:
- Which job titles correspond to the highest risk of poor performing loans.
- Which sectors correspond to the highest risk of bad loans.
- Cash flows of a business based on the industry.
- Resilience of individual businesses to economic downturns.
- Regional and sectoral distribution of loan defaults to predict future charge offs.
- And, much more.
This will provide a decision logic to recommend collections treatment strategies based on the borrower’s situation and the sector situation in general.
Borrowers’ employment data and industry data can be used for credit decisioning across a variety of products, including mortgages, refinance, credit cards, and line of credit.
Sources of Employment and Industry Data
Employment details are usually collected through the credit application form. Lenders may get a verbal, email, or fax confirmation from employers on the information provided in the application. In the case of entrepreneurs or self-employed people, banks verify employment income by checking tax return transcripts (Form 4506-T) from the Internal Revenue Service (IRS). The lender can receive a copy of it directly from the IRS. Some lenders also need this attested by a CPA, the applicable licensing bureau, or a regulatory agency. Lenders also verify the entrepreneur’s business phone listing and address.
Industry classification with a high level of granularity and accuracy should also be sourced from reliable third-party sources like LexisNexis. These data sources should use specific and intuitive taxonomies for industry classification instead of vague ones. It should also have a wide coverage of small businesses, apart from the bigger ones.
These data providers fetch data from employers, the IRS, and other sources. A few examples are
- LexisNexis
- Experian
- D&B
- Thompson Reuters
- MarketStance
- DeMyst Data
- FreddieMac’s Automated Income Assessment (Temporarily suspended), DataVerify, UConfirm, and InVerify for employment data verification
All of these have a huge database of employer-contributed payroll data and/or industry data that is up-to-date. And, credit underwriting systems need real-time data as opposed to data that lags by several months — to speed up the credit decisioning process and make it more efficient. This will help banks mitigate the risk of outdated information in credit underwriting.
To Conclude:
The employment data that banks embed onto their credit underwriting models should include the industry or sector that borrowers work or operate in. This will facilitate more efficient credit underwriting and collections efforts during this pandemic-affected period. This will also improve their prediction power of borrower stability as compared to existing through-the-cycle risk models.
