Despite the vast accomplishments of the American credit system, approximately 35 million to 54 million Americans remain outside the credit system. For a variety of reasons, mainstream lenders have too little information on them to evaluate risk and thereby extend credit. As a result, those in most need of credit often turn to check cashing services and payday loan providers, with effective interest rates as high as 500 percent. The lack of reliable credit places them at a great disadvantage in building assets (such as homes, small businesses, or loans for education) and thereby improving their lives. This study offers a feasible market solution to bring those outside the mainstream credit fold within it. Mainstream lenders can use “alternative” or “nontraditional” data, including payment obligations such as rent, gas, electric, insurance, and other recurring obligations, to evaluate the risk profile of a potential borrower. 1 Our findings indicate that alternative data, if widely incorporated into credit reporting, can bridge the information gap on financial risk for millions of Americans.

More concretely, considering that many of these millions outside the credit mainstream are poorer, less advantaged Americans, the information can direct markets toward a faster alleviation of poverty in this country. We examined a sample of approximately 8 million TransUnion credit files with a strong focus on consumers outside of the credit mainstream. The consumers include populations with thin credit files (fewer than three sources of payment information, or trade lines) on payment timeliness, as well as “unscoreable” segments whose risk cannot be determined owing to insufficient information. The credit report files, which contained alternative or nontraditional utility and telecommunications payment information, were applied to models used by lenders to make a variety of credit decisions. The scores, or predictions, of these models were then compared with payment/bankruptcy outcomes observed during the following year. Key findings include: •

Those outside the credit mainstream have similar risk profiles as those in the mainstream when including nontraditional data in credit assessments. The evidence suggests that most individuals in this segment are not at high risk in terms of lending. Using nontraditional data lowered the rate of serious default by more than 20 percent among previously unscoreable populations. The risk profile of the thinfile/unscoreable population—after energy utility and telecommunications data sets are included in their credit files—is similar to that of the general population (as measured by credit score distribution). • Nontraditional data make extending credit easier. Including energy utility data in all consumer credit reports increases the acceptance rate by 10 percent, and including telecommunications data increases the acceptance rate by 9 percent, given a 3 percent target default rate.

• Minorities and the poor benefit more than
expected from nontraditional data.
Including alternative data was especially beneficial for
members of ethnic communities and other borrower
subgroups. For instance, Hispanics saw a 22 percent
increase in acceptance rates. The rate of increase was
21 percent for Blacks; 14 percent for Asians; 14 percent
for those aged 25 or younger; 14 percent for
those aged 66 older; 21 percent for those who earn
$20,000 or less annually; and 15 percent for those
earning between $20,000 and $29,999. In addition,
renters (as opposed to homeowners) saw a 13 percent
increase in their acceptance rate, and those who prefer
Spanish as their primary language saw a 27 percent
increase in their acceptance rate.

• Nontraditional data decrease credit risk and
increase access.
The addition of the alternative data moves 10 percent
of the analysis sample from being unscoreable to
scoreable. Sizable segments would see their credit
scores improve—22.4 percent in the utility sample and
11 percent in the telecommunications sample. Most
remarkable is that two-thirds of both the thin-file
utility sample (60.3 percent) and the thin-file telecommunications
sample (67.7 percent) become scoreable
when alternative data are included in their credit files.
Preliminary evidence strongly suggests that the inclusion
of alternative trade lines in conventional credit
reports improves access to mainstream sources of consumer
credit. In a one-year observation period, 16 percent
of thin-file borrowers whose credit report
included nontraditional data opened a new credit
account compared with only 4.6 percent of thin-file
borrowers with only traditional data in their credit
reports.

• Nontraditional data have little effect on the credit
mainstream.
One worry is that including nontraditional data will be
counterproductive, harming more in the mainstream
that helping those now excluded. The results of simulations
reported here suggest that little will change for
the mainstream population.2

• More comprehensive data can improve scoring
models.
This migration greatly affects the performance of
examined scoring models. For example, in our study, in
one set of calculations we assume that creditors interpret
little or no credit information as the highest risk.
As a result, when fully reported utility or telecommunications
trade lines are added to credit reports, we see a
significant rise in the KS statistic—an industry gauge
to measure the model performance. Specifically, we
see a 300 percent rise for a sample of thin-file consumers,
and a nearly 10 percent rise for the general
sample. In the most conservative case, in which the
general sample is used but unscoreable credit files are
excluded from the calculations, we still find a modest
2 percent improvement in model performance with the
addition of alternative data.

• More data can reduce bad loans.
Including fully reported energy utility and telecommunications
trade lines (i.e., different accounts) in traditional
consumer credit reports measurably improves
the performance of loans for a target acceptance rate.
For example, by integrating fully reported energy utility
data, a lender’s default rate (percentage of outstanding
loans 90 days or more past due) declines 29 percent,
given a 60 percent target acceptance rate. Similarly,
adding telecommunications data reduces the default
rate by 27 percent. These reductions allow lenders to
make more capital available and improves their margins,
capital adequacy, and provisioning requirements.
Such improvements could have further positive economywide
effects.

Leave a Reply

Your email address will not be published. Required fields are marked *