“The way in which the data is used differently now than how it was previously used is due to the level of granularity in the data that is now available. You are able to understand different spending or cash flow trends for the individual,” Terry McKeown, Practice Manager of Credit Risk at Yodlee, shared with the LTP Team in an enlightening conversation about the way technology startups have changed how financial institutions gather and analyze data to make financial decisions.
Elena Mesropyan: Terry, let's start off with what do you do at Yodlee – your position, experience, etc.
Terry McKeown: I am a Practice Manager of Credit Risk at Yodlee and I finish three years at the firm in August. My role is to support those in the sales organization as well as our Solution Consultants as we speak with our customers to discuss our offerings in credit risk regarding best practices in using our data in credit situations for originations or credit decisions.
My background includes over 20 years in the financial services industry. I have worked at FICO, Ally Bank, and Dun & Bradstreet. In each of those organizations, my role supported either credit risk or banking operations. In my current position, I work with our clients in order to help them get the right product to fit their needs.
Elena: Some estimates suggest that around a third of the United States population is consistently scored within the subprime category. What is the approach for those people, and how do you help them?
Terry: It's a very interesting situation. Yodlee has been aggregating data for individuals for the last 17 years. That’s one of our core competencies. For those individuals that may be in the subprime classification or may not have a lot of "traditional credit " or traditional data that is out there from one of the three credit bureaus, Yodlee is able to gather information from data sources such as a customer’s checking, savings accounts; or investment accounts. The granular level data that is available in these accounts may show a history of payments that is not available through traditional credit bureaus. These customers may not have credit in the traditional sense as a mortgage, an auto loan, or a credit card, but they are still able to show experience in regular monthly payments using these other data sources. That experience could be through utility payments or education, things of that nature.
A view of this data may illustrate those that may have a thin credit file or are unscorable; they are able to still show a history of on-time payments.
Elena: You mentioned utilities. What about other sources of data that have not been explored as much? There just may be a dozen examples of companies that use social media, for example, or other type of data that represents online non-financial behavior.
Terry: Yodlee provides aggregation of financial data, such as checking, savings, investment accounts, things of that nature. We’re not in the aggregation business for social media.
But what we do focus on is gathering information directly from that financial institution. We are gathering this data directly from the source so you can verify information around about that person’s financial health, whether that be through assets or trends in their spending. We do not gather social media data.
Elena: I understand. However, what is your opinion on the matter? Is it something that can be seriously considered?
Terry: I know that there are a number of companies that are exploring that the use of social media data. I can't talk about the relevancy of that data since that is not something that we pull.
What I like to ‘hang my hat on’ a little bit more is that we’re dealing with the information that we’re gathering directly from financial Institutions; it can be verified as a payment or verified as an asset. It is not necessarily something that might be potentially manipulated; we get the information directly from the source.
As far as the information that is coming through social media, I know there are a lot of analytics companies that are interested in that, but I can’t really talk about the performance of that data since that is not something that we are directly involved in.
Elena Mesropyan: So the main point here is that you can verify the information that you pull with institutions. And the fact that it can't be manipulated by a person or by a network of individuals makes the data credible. Yes?
Terry McKeown: Yes. It’s like the information from a credit bureau; it’s coming directly from the financial institution. It’s information that’s showing the balance, or what a payment is.
A lot of the information that we’re gathering is actually taking that to the next level. We’re gathering it directly from that financial institution. So we’re able to verify those assets on the account and the transactions that are occurring within that account. We’re showing things that are more along the lines of cash flow, because you’re looking at the transactions at a granular level – each transaction versus reviewing the data at a higher view of just understanding the balance. We’re able to understand what comprises that balance. It’s information that’s coming directly from those financial institutions.
Elena Mesropyan: Do you have adjustments in the assessment model for people with very different financial histories? What is the approach for such variations in the consumer market?
Terry McKeown: Our method of gathering the data would be the same across the board regardless of the individual’s profile because that is not what we are looking at. We are not looking at their profile when we are gathering the data; we are just looking at being able to gather that data and making sure that as we gather that data, it is being able to be provided in a format that the lender is able to consume and make credit decisions, or the financial institution or the financial advisor is able to learn more about that individual.
When a lender receives this data, they are able to get a lot more insight on these individuals. If they are new to the country or new to credit, or they don't have a thick credit file, the lender is able to see that there view transactions that are coming through to illustrate the history of regular payments. Although they may not have a credit card, they may be making their telephone payment or their cable bill on a regular basis. Other types of utility payments for those that may have a thin file are able to show that they are still very creditworthy and give you a little bit more information to assist the lender in that credit decision.
Elena Mesropyan: What have you seen changing in the recent years as financial technology companies came into play? What changed in the handling of data and the way institutions assess data as opposed to before the time when such aggregators were in place?
Terry McKeown: I think there is still a keen eye on making sure that the data is in a format that is consistent. We understand where the data is coming from, how it’s being used. So, I think those types of consistencies are still there.
The way in which the data is used differently now than how it was previously used is due to the level of granularity in the data that is now available. You are able to understand different spending or cash flow trends for the individual. Instead of just understanding what their credit payments may have been as an average over a 12-month period, you are able to better understand what kind of payments they made this month, or what their cash flow is this month or even this week. So the granularity of the level of data that Envestnet | Yodlee is able to gather allows for a deeper level of insight.
In addition to the granularity of the data, Envestnet | Yodlee is able to provide this data in a timelier manner than traditional data sources. A lender can be notified of potentially high-risk indicators that occurred this week versus having to wait until the bureau updates with a delinquency that may have occurred last month or the month before. That being said, the timeliness in the gathering of the data and the ability to apply that has changed.
Granularity and timeliness of the data have allowed us to expand the view of customers that we are able to provide information on so lenders are able to expand into the “credit-invisible” consumer groups that we weren’t able to view information on before. Now, we’re able to use other forms of data such as alternative data through checking, savings, and investments. We’re able to use that type of data in order to shed light on credit invisible and lift them from that status. We have more information to apply to the credit decision in order to help them out and get them the credit products that they need. I think that’s how things have changed in the usage of data and how we have modified how the industry will be using some of this data going forward.
Elena Mesropyan: How does the risk evolve with expanding opportunities for thin-file individuals? It must be different, right?
Terry McKeown: There is a different level of risk, but I think the same core foundation of making a credit decision still applies. Do they have the funds to make that payment? Do they show a history of making payments in the last X number of months or X number of years? Do you feel comfortable to provide them this loan based on those key foundational elements of what makes a good credit decision?
I think the difference is that instead of using terms like ‘past delinquency’ or ‘number of credit accounts that are open,’ we’re now looking at things like: 'What are their assets? What are we able to show from their income that’s coming into their checking account?, or 'What are we able to show from information on regular monthly payments for some of these non-credit types of opportunities (such as monthly payment for education, insurance or utilities, things of that nature) that are still showing a regular monthly obligation that the customer is paying on a regular basis, which is very much in line with those credit foundations?'
Elena Mesropyan: Since you deal with a lot of sensitive data for people with history/without history, how do you handle security? What does Yodlee do to ensure the security of the data?
Terry McKeown: We take data security very seriously, and this is one of our core competencies. Envestnet | Yodlee has built a trusted and secured partnership with financial institutions by working with them to ensure our services meet their stringent security and compliance requirements thereby protecting them from any fraud and security issues. For over 17 years, our data aggregation platform fuels innovation for financial institutions and Fintech innovators, enabling consumers to get better lending rates, better returns, and more. While achieving these goals, we adhere to leading practices for security, risk, compliance management, and privacy. Yodlee is examined by the US Federal Banking Agencies, per the Bank Service Company Act, for the services provided to U.S. financial institutions. That same Financial Data Platform is leveraged for all Envestnet | Yodlee customers, so they benefit from the full breadth and rigor of Yodlee’s risk management programs. In addition, Yodlee has undergone nearly 200 audits by financial institutions in the most recent 24-month period.
Elena Mesropyan: As Yodlee gathers data and passes it to clients, is there any moment in time that data is being manipulated in any way?
Terry McKeown: Envestnet | Yodlee adheres to a strict user-permissioned model, meaning no customer data is collected without end-user permissions. Protecting the personal information of individuals who use our customers’ products and services is a top priority at Envestnet | Yodlee. This, in turn, creates more consumer trust and loyalty in their bank relationships while providing them with valuable services that help them improve their financial situation. When we do gather consumer permissioned data, we are gathering it in a digital format so that the lender or the financial institution that we’re working with is able to consume that data into their models or their lending platform.
So, more than anything, it’s really just normalizing the data into a format they are able to use immediately.
Elena Mesropyan: I see.
Elena Mesropyan: What do you see happening in data aggregation, and in how financial Institutions handle data and expand opportunities for credit for thin-file individuals? What are some of the trends?
Terry McKeown: One of the great things about where the industry is going is that we are at a point that we are able to gather the data for the customer so that they are able to consume this information in a digital format. By doing this, financial institutions are able to create a more streamlined process. They are able to bring the data in, instead of having to wait for the customer to provide that in a more manual format. We are able to gather more information at a granular level that allows the lender to not just understand the balance but understand what comprises that balance.
If you think of a credit decision that’s been made using a credit score, that credit score is a point in time that indicates the health of that individual on a numerical score. But it gives you that point in time. By understanding the data, you are able to see the trajectory leading up to that score. Is that person going up or are they going down in their financial health? If you are looking at a credit score that might be at 700, is that credit score migrating up from a 680, or is it coming down from 750? Lenders are able to get more information in order to make a more informed decision. They are able to do this with Envestnet | Yodlee data that is much more timely and fresher than what they have seen with more traditional data sources.
Elena Mesropyan: What is the role of artificial intelligence and machine learning in working with the data that you gather?
Terry McKeown: I think there is a lot of insight that is gathered through machine intelligence by going through the data and understanding those trends. Withmachine learning, you are able to better understand trends over time as well as the related risk indicators that might be important to manage over time. Instead of looking at something more traditional such as delinquency rates, you are able to look at other types of NSF or over-limit fees that may show that risk is impending. It allows a complete understanding of the risks involved with an individual.
Elena Mesropyan: Let’s say I am applying for a financial product from a certain institution and that institution is your client. What is the playbook here?
Terry McKeown: The information we gather is customer-permission data – the customer is providing us the permission to gather this data on their behalf. As we’re gathering that data, they would indicate which accounts they would like us to gather that data from.
This is something that, with so many things in today's world, is becoming easier for the consumer. We are assisting so that the data gathering process can be more streamlined and it is helping the customer get financial information for mortgage, an auto loan, or almost any type of loan, whether it’s a consumer or small business, providing that information in a more efficient way that really helps the customer get the loan product that they need.
As a result, we get a win for the lender because they are able to streamline their process, and it’s a win for the consumer because they are able to provide this information in a timelier manner. I think it is a trend that is going to continue. I am looking forward to it and I think it’s going to be an interesting couple of years as this evolves; I’m happy to be a part of it.
Elena Mesropyan: Terry, thank you for this insightful conversation!
Terry McKeown: Thank you very much.
