HNC Software, the company that made the Falcon Fraud Tool, which is used to evaluate credit card transaction risk so card issuers can block cards if there is a problem, sold the tool to FICO. Then the founders pivoted and started a new company called ID Analytics. The goal was to provide actionable insight into credit and identity risk. That was 2002. Lending Times recently spoke with ID Analytics Director of Product Marketing Kevin King.
Rather than focusing on credit card fraud, ID Analytics focuses on new account fraud. They look for indications of identity or intention fraud in applications for loans, credit cards, wireless phones, etc.
Lenders want to know is if an applicant plans to pay back a loan or take the money and run. So ID Analytics built a data consortium of lenders similar to a credit bureau. Companies come to them to evaluate applicants. When the assessment is returned, ID Analytics holds on to borrower information (name, SSN, address, phone number, date of birth, etc.) and asks companies to provide insight on the results of loans issued. What they want to know is, did it turn into fraud or continue to look good?
This business model worked well and the company’s data set became larger as more industries and credit bureaus took an interest. Compliance, authentication, and credit risk were included in the analysis.
Thirteen years later, ID Analytics had formed relationships with several Fortune 500 companies. New industries like FinTech and alternative payments are interested in the power and predictability of knowing the borrower. As they enter new markets, they understand applicants better than the credit bureaus because ID Analytics’ assessment is current.
In addition to knowing whether an applicant pays all their bills, ID Analytics can see how the borrower behaves. For instance, if a borrower applies for five credit cards in the space of one minute, ID Analytics’ score reacts, within seconds, for a high velocity string of behavior. The company employs 150 people, so it can be more responsive and nimble than a larger credit bureau.
The Birth of the Online Lending Network
In 2011-12, ID Analytics started working with P2P lenders and supported them as the industry grew. Those lenders, leading players now, became concerned when loan stacking emerged in 2015. Loan stacking revealed a blind spot in the world of online lending: It was too easy to get loans from multiple lenders at the same time. Since ID Analytics already had relationships with online lenders, this enabled them to see 60%-70% of marketplace behavior. So they built an online lending network.
The Online Lending Network, founded in April 2016, is a group of lenders who have partnered to solve a set of pressing problems in the online lending industry. These problems are disparate because most providers only contribute to one or two bureaus rather than all of them. More relevant and critical, the soft inquiry credit process that developed as a core business model to improve customer experience leaves lenders blind for the short term. They can’t see what a borrower has done in recent activity when an application is made. The network helps to build technology and provide data assets, which makes it unique.
Unlike a lot of fraud behavior, which is nuanced, loan stacking is black and white and involves multiple unsecured loans piled up against the same asset. It needs a black-and-white solution to identify lender risk immediately. ID Analytics offers a two-fold solution.
Adoption by a majority of lenders enables the most complete coverage and visibility of data. Providers can take that visibility and turn it into the intelligence needed to stop stacking. Participation and speed to market wins the day.
When a borrower presents herself to a lender, the critical questions are, “Is this person real?” and “Will they be able to pay me?” The Online Lending Network essentially provides attributes, black and white insights, that count the number of times a particular borrower has been into an online lender in the past three months. It can also zoom into the last hour of activity. These attributes can be drilled down to categories like small business, P2P, subprime, etc.
Members of the network provide their full top-of-funnel velocity. Each time a borrower requires a loan offer, that information is sent to ID Analytics and, in a sub-second, the application is reflected in the information provided. Numbers in the first few test weeks were promising, but there will be much more in the months to come.
It took just five months from concept to live production to get the network up and running. While slow for FinTech, that’s lightning fast for analytics.
What Data Can ID Analytics Tell About Borrowers?
In the top three online marketplaces, 1-1.1% of individuals requesting an offer have been to another marketplace lender in the past three days. In the last hour, 0.3% visit another marketplace lender. ID Analytics can also see that 3% of applicants have applied for another credit product in the last hour even if that application was outside the online lending marketplace.
Sets of attributes showing how frequently credit is sought are “attributes version 1.0,” but in Q1 2017, ID Analytics plans to launch Attributes 2.0. In this set, borrowers are identified as having gone through a truth-in-lending process and commit to moving forward with a loan. That will require members of the network to reveal two distinct points in loan origination: First, when the borrower requests the origination and, second, when the borrower commits to the loan. Looking at these two points can provide a lot of insight. It’s important to discern whether the applicant is a rate shopper who is responsive or someone who really is opening up too many loans.
The problem with loan stacking could be one of fraud, where the fraudster is aware they can open 4-6 loans in two hours and get the funds without planning on repayment. This type of fraud is what TransUnion is trying to solve.
Another scenario is the unintended consequence of the leap forward in customer experience. A borrower goes to the P2P marketplace and gets $10,000. They say, “That was fantastic; I can do home repairs and go on vacation!” Then they look around and see what all they can do with another $10,000 and take out another loan from another lender that they also intend to repay. But in 24 hours they have borrowed $20,000 and will end up defaulting on those loans because they are in over their heads. This is an equally important problem, but it is not of malicious intent.
ID Analytics is focused on a different aspect of the problem than TransUnion is. TransUnion is focused on fraud, building a set of more analytically-driven tools to put a lot of science into the solution. ID Analytics is solving the problem of stacking by utilizing coverage and visibility. They can tell everybody in their Online Lending Network when a borrower comes, when he was seen at other lenders, and when loans were committed to within the last hour. This transparency is a more straightforward approach.
ID Analytics works with marketplace lenders seeking to understand their problems and provide products that solve their unique challenges. There are at least 15 large players in their network with a majority of those brands reading Lending Times regularly.
Written with Nicki Jacoby