Complexity in fin-tech means greater need for ID security
The recent Money20/20 exhibition area was packed with ID and alt-data source exhibitors.
Equifax Assistant Vice-President of Product Management Jeff Knott said the industry has long recognized the limitations of standard identification methods and has worked to complement them by developing additional tools to help financial services companies make wise lending decisions.
“We help service and enable the financial services industry by providing tools to help them get a clear picture of who that person is,” Mr. Knott explained.
Equifax provides these services in 19 countries, including the United States, Canada, Russia, and Mexico, all locations where they are ranked as the top service provider. They are the second-most popular choice in India, Brazil and the United Kingdom.
That reach allows Equifax to analyze data on more than 600 million consumers, 80 million businesses and 200 million employees. They generate 60,000 credit score updates every second.
Given the turbulent period we are emerging from, those tools are welcomed by an industry working with a growing percentage of people whose lives have been influenced by job or home loss and relocation. In those situations, several methods can help provide a more detailed picture of the person submitting an application, Mr. Knott said.
“Utilities, telecommunications records, bank information, employment history, job tenure, sources of income beyond wages through tax transcripts are all things to really build out that perspective of the individual.”
Advances in technology, especially in machine learning and data aggregation, have been especially helpful in understanding the consumer beyond large homogenous groups separated by a few general characteristics, Mr. Knott said.
The advanced use of technology in this area is still relatively new, Mr. Knott added. So far the industry has focused on the more straightforward lending cases as they build out those initial processes. More data and the ability to interpret it are required before more complex cases can be analyzed with the same accuracy.
One key factor in how much that data will be used in the future is to what degree the consumer will allow their personal data to be used, Mr. Knott suggested.
“We want to be good stewards of the data and make sure it is not misused.”
The goal is to bring value to the consumer, he added, not to simply market to them.
A constant challenge is to protect both companies and consumers from fraud, Mr. Knott explained.
Abnormal transaction velocity raises several questions, Mr. Knott said.
“Does the activity seem out or sorts? Manufactured? Is it a typical habit the individual engages in?”
Another challenge facing the industry is the authenticity of the data they collect on an individual, whether it is directly submitted or gleaned from social media behavior, Mr. Knott said.
Pay stubs and other employment-related documents have been faked for decades, so added assurances are welcome when accepting such submissions from an individual, Mr. Knott explained.
“Trusted, verified data is key. Just because technology is involved and the document is being uploaded it doesn’t mean it is a trusted document.”
Given the growth of social media activity, it is not surprising identity verification and underwriting methods making use of it are being developed. Tread carefully, Mr. Knott advised.
Graduate students are an interesting demographic. They know their online activity is being monitored, so they develop multiple online personas. Their professional identity, the one they post on their Linkedin profile and official Facebook account, shows the world their academic accomplishments and volunteer activity. Their secondary social media accounts, which often use their first and middle names, are where they plan the next kegger.
For that reason, social network data cannot be viewed as a silver bullet, Mr. Knott said, but it can be used in combination with other methods to be a healthy predictor.
“At face value self-reported data is risky.”
Mr. Knott said he is paying close attention to the marketplace lending landscape, as its ecosystem is still being formed and consolidation has yet to occur.
“It is clear this is no longer just available to the Wall Street investor.”
He also looks forward to the continued development of tools to identify and interpret alternative data sources. As they are refined, their accuracy will continue to move down from the macro.
“The tools to get to an individual level are out there.”