Cracking the Code: Frank McKenna, Chief Fraud Specialist for Point Predictive, Shares Secrets to Predictive Lending and Fraud Prevention

SPEAKERS

Frank McKenna, Rich Alterman

Rich Alterman  00:04

You're syncing up and tuning in to the lending link podcast, powered by GDS Link. With a modern-day lender can dive deeper into the future of data decisioning and Credit Risk Solutions. Welcome to the show everyone. I'm your host, Rich Alterman, and today we're syncing up with Frank McKenna chief fraud strategist and co-founder of Point Predictive San Diego-based fraud solutions provider which powers a new level of lending competence and speed through artificial intelligence, a powerful data consortium and decades of risk management expertise, estate and technology solutions quickly and accurately identify truthful and untruthful disclosures on loan applications. With over 30 years of experience working in various fraud detection, mitigation and management roles. Frank is a leading authority of all things fraud. He works with lenders to launch solutions that incorporate predictive analytics so they can solve their business problems with greater accuracy. Frank has worked with more than 100 banks, lenders and companies throughout the world, designing strategy solutions and operational practices that help them reduce costs and increase efficiencies. In 2016, Frank started his blog Frank on Fraud to give a Fraud Fighter perspective on the global rise in fraud. Frank is also a board member of the noble, whose mission is to protect the vulnerable, including victims of human trafficking, child exploitation, scams and elder abuse. Frank holds his MBA from Cal State East Bay College of Business and Economics and his undergraduate degree from St. Mary's College of California. In this episode, Frank and I will touch on many facets of fraud, including fraud trends, unique types of fraud, some solutions in the market, and so much more. But before we dive into the interview, please head over to our LinkedIn and Twitter pages at GDS Link that's G D S L I N K and hit those like and follow buttons. If you haven't done so already, please subscribe to our podcasts on Apple podcast, Spotify, or wherever you prefer to listen to your podcast. All right, now let's get synced with GDS Link. Welcome, Frank. I hope you had a great weekend. Thanks for joining me today. Where are you joining us from?

 

Frank McKenna  02:03

Hey, Rich, it's great to be here. I'm joining you from beautiful San Diego, California. We're normally 75 degrees here, but we're about 60. And looks like it might rain today but still glad to be here and looking forward to our conversation.
Rich Alterman  02:16

Well, thanks so much for joining you co-founded Point Predictive in November 2013. Can you take a little time to share more on your background and what you were doing before starting Point Predictive?

 

Frank McKenna  02:26

Oh, yeah, of course. So thanks for that very nice introduction. If I think about my career in fraud, it goes back to right after I graduated from college, I started in banking. So I worked for banks here in California Providian National Bank and Wells Fargo where I manage their credit card and debit card strategies. In 1997, I made the leap into technology. So there's a company here in San Diego called agency software at the time, that had machine learning technology called Falcon that was really taken off at the time. So I joined that company, and really started to get into fraud consulting, where I worked with a lot of different banks, card issuers, lenders on helping them understand how to fight fraud. In 2004. I made the leap from what then became FICO because they acquired that company agency software. And I became an entrepreneur establishing a fraud company called BasePoint Analytics where we tackle mortgage fraud, a company called CoreLogic, that you might be familiar with, they bought that company. And that's when we started Point Predictive, we wanted to find new ways to help auto lenders in particular to fight fraud.

 

Rich Alterman  03:32

Well, you and I are actually alumni I worked when I worked for Teletrack we had been acquired by CoreLogic, we'd have that in common. Also, you know, you mentioned agency Falcon, it's dealt with Larry, spell hog back in the early 90s. And it's funny to me sometimes when AI and machine learning became popular again over the last couple of years, I'm sitting there saying wait a minute. It's been around for a long time. So yeah, it's not as new as we all think it is,

 

Frank McKenna  03:58

they say was the first commercially successful application of artificial intelligence machine learning in the world. So it was an amazing company to be part of and like you miss meet so many great people that have their roots there. And luckily, I was one of them.

 

Rich Alterman  04:13

Well, it's interesting, I can remember when I first was dealing with agency at the company I was working for, that's when they were looking at how they could use neural nets, not only in fraud detection at the point of sale, but actually in lending. So that was something they were discussing back in the early 90s. So here we are, many, many years later. So thanks for sharing your background as one of the things I like to do before I dive in is just ask some personal questions and I had the opportunity to visit your other blog, which is called Frankie Photo, where you go out and take pictures and I liked your motto. Take a walk, bring a camera and find something beautiful. Where's today everyone has a camera with them wherever they go with their phone. They've gotten more and more sophisticated in what they're capable of doing. It's really amazing. So you know what? was the driver behind starting that blog? Is it something that you find time you can keep up to date. And again, the real big challenge here is to help the audience visualize what you would say is one of your most memorable pictures that you ever took.

 

Frank McKenna  05:13

I think, if I go back, it was 2011. I just gotten a really nice camera, we just sold our company CoreLogic. And I wanted a project, I said, I want to learn how to do photography. And so the blog blogging about it gave me a purpose and a reason to go out, take a walk, find something nice and take a picture of it. So I started that in 2011. I didn't want to just create photos, but I wanted to create photos and have until the story behind it. I started a blog, it helped me motivate me to get out there, give me a purpose around photography. And I did that for you know, five or six years. And my photography really excelled. And I had a lot of people reading that blog. I stopped in 2016, for two reasons. One was because I had a son. And he was born that year. And I just didn't have the time. But I also wanted to focus my blogging on my career. So I started with Frank on fraud. So I kind of just shifted from photography into fraud, which are my two loves, I would still say if I was to describe a memorable picture, it would have been in Vietnam, I was in a small village in Vietnam, it was about two in the afternoon, it got really hot, and then just started torrential downpour. I walked outside with my camera, and I was just going to take a picture of the rain. And just when I got outside, there were these three girls that were all sharing a single bikes, they're all kind of sitting on the seat, driving down in this torrential rain. And I took to take a picture of them and they were all laughing and the rain was pouring down. And it was just this visual of like life in another country. And it was just a beautiful moment. And I still have that picture. And I have it framed because it just reminded me of just how awesome life is.

 

Rich Alterman  06:56

And that was a picture of something so simple, and carefree. And today, life seems so complicated. And yeah,

 

Frank McKenna  07:05

here in the US, especially in Vietnam, it's still like that, though, very simple life. And that's one of the reasons why I like traveling internationally. To get that perspective,

 

Rich Alterman  07:13

I'll hope to have the opportunity to visit there someday, let's put away our phone cameras and take a look at the ever accelerating world of fraud. So Frank, with more than 30 years involved within fraud detection prevention, you've witnessed quite an evolution in fraud schemes. I remember when I ran the fraud department of Bank of Boston back in 1985, we would send our fraud analyst out to seminars where they would learn on how to detect the way fraudsters would fill out applications. And they slanted their letters a certain way they dotted their eyes a certain way. So here are things are now far more sophisticated than that. So when you think about fraud trends, what are some of the things that you've really found the most interesting, and actually some of the schemes that are out there?

 

Frank McKenna  07:55

Yeah, interesting that you say you talked about that application fraud monitoring, because I think I went through those same courses. And I learned, I got my start looking for a lot of those same things. My first job was actually looking for dots at the end of signatures that were indicative of Nigerian fraud rings, it was just a different time, it was a lot simpler than, but some of those things are so interesting. So if I think about the top three trends that really stand out, in my mind, I really think of things that were turning points in the industry that led to big changes. So three examples, I would guess I would give us the first was the stolen mail theft at airports in 1991. It's right when I first got into, into Fraud Management. And in Houston, Texas in Atlanta, Georgia, the thieves were stealing trays of credit cards out of the mail. And they were using those cards at the time in 1991. There was no such thing as card activation. It was you get a card in the mail, you go to store, and you use that card. Wow. And there was no stickers on those cards to call in and activate it. But the fact that fraud rapidly rose because of that mail theft, they adopted that technology. They called CRV card activation value, putting the stickers on the card and it led to fundamental change in the way that we protect our cards in the mail, those stickers and that activation process. That'd be the first thing I think that stood out in my mind. The second would be the rise in counterfeit skimming in 1995. And I don't know if you remember this, but in 1995 the fraudsters figured out how to install overlay machines on top of these credit card readers started skimming the data. And they were able to just think about the antiquated technology of a magnetic stripe and all the data is there in the clear, you can take that off that data off of the skimmer and then just plug it into a new card. And that just cause counterfeit to go crazy, not only here in the US but across the world. That fraud that rise in counterfeit skimming led to sweeping changes and the adoption of chip and pin. So that stood out to me as Well, because that was something that was cost America in every other country in the world a lot of money. But it was enormously successful. That'd be the second thing. And I think the third thing that stood on my mind was the mortgage crisis of 2004 2005. When fraud was running rampant with these liar loans with the straw borrowers with just every type of mortgage fraud, you could think of that fraud crisis led to sweeping changes in mortgage right, its mortgage process is much different than it was in 2005 2006. It's much safer. Those three trends to me were really turning points for the industry that have had long standing, I think good consequences overall. But those are the things that stood out to me. With the

 

Rich Alterman  10:43

years of experience I've been in the industry, everything you say is totally relatable. Funny, today, you use word take one, you say the person take one application, like, what is that? Right? Well, something that would sit on top of a cigarette machine and retails at the bar to apply for a Visa or MasterCard,

 

Frank McKenna  10:59

you'd fill out your name, address, social date of birth, and it would be like a postcard, and there wouldn't even be an envelope, you just send that back through the mail with all your PII. And the banks didn't even think about it at the time. Pretty funny,

 

Rich Alterman  11:11

you know, Point Predictive, when you guys first started the company, you really were heavily focused in the auto space. So maybe you could share a little background on how predictive evolved and why was it auto and you mentioned mortgage? And why did you start in the White House?

 

Frank McKenna  11:25

When we sold BasePoint Analytics, we wanted to find a greenfield opportunity that nobody else at all was involved in. So we actually started the company with no set agenda other than finding that opportunity, and filling that opportunity. So we did research. We talked to banks, we talked to lenders, and we said where do you need help. And we pretty quickly figured out that auto lending was a huge gaping hole for risk. There were no consortiums, there was no use of AI or machine learning. Lenders were not communicating with each other about fraud, and they were suffering the consequences. dealers were committing fraud. And once they get shut down at one lender, they go to another lender. And they could run these fraud schemes for 678 years without suffering any consequences because they could just go to the next easy target. So we said, hey, we've got experience in machine learning, we got experience in consortiums, the auto learning industry is in kind of desperate need of something. And by the way, this is a huge part of our GDP 3% of the US GDP is on auto sales, and growth for something to be unprotected like that told us there was a big opportunity. So we decided with conviction to go after auto lending and kind of take all the learnings we had from other industries to help lenders solve the problem.

 

Rich Alterman  12:37

Yeah, it's interesting. I was reading in the paper yesterday, our local paper, a local dealership actually had 100,000 Plus car is driven off a lot by a fraudster. It hit home because I've written the paper right after I finished writing my notes for today's podcast. So Frank, you mentioned straw man fraud, and I think it's a term that's seems to be somewhat unique to the auto industry. Can you explain what it is? What is straw man? Right? You mentioned it earlier.

 

Frank McKenna  13:00

Yeah. So straw man fraud. It's an interesting name. You know, a straw man is like a scarecrow, right? It's something that you put in the field. And it's, it's not a real person, it looks like a real person. strawman fraud is designed to look like a real person or real borrower to a lender, it's very common to have this type of fraud for mortgage lending and auto lending, where somebody goes into the dealership, that is not intending to drive that car on that car, like a real person, they're either going to be a front where they're going to go into an auto dealership, say the cars for them, but it's actually for somebody across the country that they're, they have good credit. So they're going to stand in, and then turn the car over to that person. So that like a straw buyer of the car. Now there's lots of reasons that people would do straw borrower fraud. One is to help somebody that they know get a car who has bad credit. The second is a little more insidious, they might be doing it for fraud perpetrators who want to take those cars and ship them overseas. So they're using these third party and people they'll pay the a straw man can make anywhere between 500 to $1,000 by buying the car and turning it over to one of these fraud rings. And they can do that over and over again. So they can use their good credit to get these cars. We think it's about a billion dollar a year problem. Just an auto though you're looking at a multibillion dollar industry just around having people purport to have a loan for them when it's actually for somebody else.

 

Rich Alterman  14:26

So it sounds similar to where people will sell their good trade lines. Right. And let somebody piggyback on their credit report become an authorized user. It's a topic we talked touched on a couple months back fact I think it's when we were interviewing Tom Algie who recently joined you guys. I think that is one of the topics so yeah, really interesting. But yeah, it was funny to read this article. It's like, hey, that's exactly what Frank was talking about. The other day.

 

Frank McKenna  14:49

I had somebody last week who got a job offer, read my blog on straw borrowers and he reached out to me said should I take this job? The job was really interesting. It was working for this guy. Jumping, I was going to pay him $500 They were going to have him go online buy cars, and flip the cars to them right after they're gonna be $500. He didn't know what he was getting himself into, I just strongly recommended that he reconsider, because he was probably going to be engaged in some fraud that he didn't know about.

 

Rich Alterman  15:17

Well, I'm glad I'm glad he reached out. The fact that he was reaching out in a loan would have been the answer for himself, right. So when we think about different types of consumer loan products, such as credit cards, unsecured personal loans, and auto loans, mentioned mortgage, can you kind of highlight what you see as some of the differences that lenders face? And with each of those different products, we just mentioned, straw man issues, and then more like, what are the similarities, we kind of almost have drew a Venn diagram of all the different products and where they overlap and where they're, they're different.

 

Frank McKenna  15:48

I've gotten this before. So I think, if we had been involved in trying to help lenders, and banks stop each of those types of frauds you mentioned, so from my perspective, there are differences. And the differences are kind of four things, the first thing would be the data they collect. So when you apply for a personal loan, you're going to provide a little bit different data than if you're going to get an auto loan, or a mortgage, right, a mortgage, they're going to ask you for a lot more auto loan a little bit more on a personal loan, maybe just your name, address, phone, and your income and your employment. So what that lender asked for, that's different. So the data that lenders collect is different, the rates of fraud are also quite different, right? If you look at a credit card fraud rate, it might run, you know, seven basis points, or an auto fraud rate might run 50 basis points, you're gonna have these different levels of risk that associated with each of the types of fraud, and the process to approve, right? When you get a credit card, it's instantaneous. When you get a mortgage, it's 90 days, it's so you have a lot more time that might lapse between when you're actually submitting an application and actually having a fund. And the I think the last thing would be the different types of fraud, you know, credit card fraud is very transactional, most of the fraud occurs after the loan has been originated, you know, counterfeiting, law, stolen non receipt, all of those types of fraud that happened after origination, a auto lending fraud is going to be you're going to lose your money right up front, because it's not transactional, it's right at the time when they get that car. So the different types of fraud that you're gonna experience are also different. But there's a lot of similarities to if you think about the similarities, I'd say they're more, they're more alike than they are different. To be honest, I think you're always gonna have ID theft, whether it's an auto loan, mortgage, loan, credit card loan, personal loan, you're always gonna have identity thieves trying to get access, you're always going to have an element of synthetic identity with fake profiles, you're always going to have an element of people lying about their income, because income is core to understanding that you can afford that loan. And you're always going to have some element of employment fraud. And you're always going to have some element of exploiters people that are first party fraudsters that are not stealing identities, but they're getting those loans, those credit cards, those auto loans, those mortgages, with the express intent of defrauding you. They're first party fraudsters, all those span each of those types of channels, and they all need to be accounted for.

 

Rich Alterman  18:13

One of the things I've always wondered is when consumers had a limit on their exposure on a credit card, you know, really to that $50 fraud, you know, you are a victim of fraud, did it make consumers maybe a little lazy, in your opinion, and really checking every single transaction on my credit card. And I always wondered whether the lender should actually implement, like a point program for consumers if they actually help identify rapidly fraud that's on their card, I went through seven cards with one provider in a year where I was buying, like my airline ticket. And I was amazed at the things that they didn't catch, where I was flying at the exact same time as someone else who would use my card who had a totally different name, and he was going to some country that I wasn't going to, and that that didn't trigger a signal. So it almost seems like there's a way to incentivize consumers to get a little more engaged on their own loans and whatnot from a fraud perspective. Just,

 

Frank McKenna  19:11

ya know, 100%. I totally agree with that. I think consumer participation in the fraud detection and prevention process is critical. I think, going forward, I think we have to get consumers involved in protecting their accounts, by the way, that $50 having been in work with a lot of credit card banks and an investigator myself, I've never heard of a bank actually charging $50. So I think it's there as a as something they could do, but nobody's ever done it. I think there is an element. I think there is a big element of first party fraud in the credit card disputes area I know because I used to be an investigator and what I found is about 40% of the claims that I got on a daily basis. If I called the customer they either committed the transaction fraud themselves, and they admitted it to me and dropped their claim. or they knew it was somebody in their family. So I think the element of first party fraud is probably bigger than the negligence because it's very simple process to get these charges removed, you touched

 

Rich Alterman  20:10

on income. So maybe I'm just going to jump in there for a second. So when we think about employment income verification, Equifax as a work number was probably one of the first out in the market, but GDS Link, we've seen just explosive growth. from an employment standpoint, yeah, permission solutions, like Argyll and Pinwheel nine permission solutions. Experian, and Transunion, have rolled out solutions recently, then you have a company like Powerlytics, which is a estimator based on zip plus four. So I know that that Point Predictive offers some interesting solutions around employment, and income. So maybe just kind of talk about what you guys have built in. And I know you have a whole database of fake employers and kind of how all that works.

 

Frank McKenna  20:50

Yeah, it is interesting to see the focus on income and especially income automation. And I'm really happy to see it, I think the ecosystem has grown a lot. Tim and I, in our companies having looked at this problem for the last, I'd say 15 years, since we started looking at it in mortgage, we think, here Point Predictive, we found the solution that that really works. And what we're doing is we have a product called income paths. It's basically if I think about an ecosystem, and all the products you mentioned are fantastic, by the way, they all offer unique value, what we do, I think the way I described the income passes, we can do an income validation in less than a second, for a fraction of the price that you might have to pay for an expensive database check. So it's, it's a lower cost solution that runs right up front, there's no borrower friction. So the borrower doesn't have to consent. They don't have to provide pay stubs, those are always like, those are always hurdles for the customer. There's none of that. But we do is we use our proprietary data. And about eight different sources of data that we've found have been really powerful and actually predicting how much a borrower should make or probably does make. These would be things like, you know, IRS data, census data, salary, data, job postings on the internet, you think we kind of scour all these sources of data, to try to pinpoint what we think a borrower should make? And we give a lender an indication in less than a second of whether this borrower is being truthful or not. And then they can kind of decide what they're going to do with that? Are they going to go out and request a pay stub? Are they going to go hit the work number? Are they going to hit the pinwheel, the plat and all those other solutions? It's a top of the waterfall check. That gives them an instant response. So they can decide what do I do with this?

 

Rich Alterman  22:27

Yeah, I think so you can combine that with if you think about the Vantage, your FICO score, right? So how do I look at stated income versus estimated income? And based on maybe that FICO score, Vantage score, some custom score other factors that kind of becomes a decision tree? Of do I need to now do further income verification. So somebody maybe with a really high score, and a slight deviation from the stated income versus someone with a bad score with that same deviation? Do I need to do the same thing?

 

Frank McKenna  22:56

You are so 100%? Correct on that, and I think a lot of banks, hundreds of companies miss the point that this is part of the risk. Here's something interesting in our analysis shows that if a borrower misrepresent their income, and you use that alone, and you say, I'm not going to prove the loan, because I think they misrepresented their income, that has zero predictability for default, whether you lie about your income alone, if you just lose that in isolation, you can't really predict the fault. However, when you combine it with like your what you were talking about your approach, if you look at somebody who has less than stellar credit, and they misrepresent their income by $10,000, and they're making $40,000 a year, that is substantially different than a super prime borrower that misrepresented income by the same amount. So it's all relative. And that's what makes income complex is you have to create that ecosystem and that good understanding of risk and apply that friction when it makes sense and take the friction off when it doesn't.

 

Rich Alterman  24:00

And for our young listeners, what we're talking about here is one of the season credit which is a character, right? Am I telling you the truth or not telling the truth? So there's still something that's around. So going back to Otto for a second, your firm's getting ready to release your annual report by the lending fraud trends. And thanks for letting me get a sneak peek. One section that report stated that fraud experts at Point Predictive believe up to 1 million more fraudsters were activated in 2020. After stimulus programs were launched. After the stimulus programs ended those same fraudsters shifted their efforts to steal cars through provisional and financing vehicles. And maybe this is something you've talked on the straw man problem, but maybe you could share some more highlights from the report that's getting ready to be released. Yeah, it's

 

Frank McKenna  24:42

pretty good report. We spent a lot of time on that. I personally spent probably 40 or 60 hours on just putting that together. What I found most insightful from the report. I didn't go into it expecting this but we saw 35% increase in identity theft last year, and it was already high. So We're seeing is identity theft, at least an auto is on this trajectory upwards. And it all ties back to COVID. The people that were committing this PPP fraud, and this eidl fraud and this unemployment fraud, they learned how to commit sophisticated fraud, complex fraud, they learned how to buy data off the dark web and then apply for loans. They just shifted their efforts to auto lending. In fact, we did our own analysis, and it's a part of this report as well, kind of a case study in there. We analyzed about 300 synthetic identities that occurred in Chicago, these are people that hit auto lenders, and the auto lenders last year $67 million. Our research showed that 76% of those synthetic identities, had taken out a PPP loan, the prior year. So they're these fake identities that were used the year prior. So I think synthetic identity, identity theft were primary drivers of that increase. We've seen, just anecdotally, and you mentioned the dealer case that you read about in the paper. There are car dealers now in the United States that are getting three fraudsters a month walking in the dealership trying to steal cars. And these dealers had never seen a fraud before. I don't know what's going on. But these fraudsters are going after dealers right now, and they're going after lenders, and they're using fake identities for driver's licenses. It's crazy.

 

Rich Alterman  26:21

I know one of the key strains of your offering. And you mentioned it is your proprietary data repository. And in looking at your website, it talks about how you hold more than 120 million loan applications and more than 55 million unique applicants. And of course, it's growing every day. And it's interesting going back in time a little bit for you. And I, I remember when I think it was visa rolled out their issuers clearing house for the credit card space where issuers had to report application data, and it was all designed around capturing fraud. So it's somewhat similar. Can you share, you know, what are some of the key data assets mentioned employment earlier that are found in the repository, and perhaps speak to the process, you go through in taking that raw data and turning it into actionable intelligence.

 

Frank McKenna  27:05

So you think about predates proprietary data, there's a lot of awesome data out there, our data is completely different. So you think about like, the credit bureaus have a lot of great information and identities LexisNexis, all these companies provide a lot of value on the identity, what we do is we collect about 85 fields of information on every auto application 125 fields of information on every mortgage application, on a personal loan, we may get about 10 To 15 credit card, we may get seven or eight. But we basically collect this information at the consumer level, we got about 67 million unique consumers that we have in our data that we've seen anywhere from five to six to seven times. So we see them multiple times. What makes it so unique is all these data points, there's like, we know what the borrower is reported as their income, their employment, what they purchased, what type of car was it? What was the sales price, it was a personal loan, what type of personal loan, what were they reporting is their income, their employer, we have this kind of cross industry information. And we tie it back to the default. So every month, the lenders tell us which loans they had fraud, which loans they had default, which loans charged off, and we're able to take each of those individual data points. Right now we're coming up on like 26 billion different data points from the database, and we tie it back to the fraud and the default. So you think about the massive amounts of data that we can then use and feed it into machine learning algorithms, to mine all those patterns of fraud that might be hidden to the human eye. But our intuitive, like incomes a great example, you know, when we see a borrower that's reported the same income, you know, or close to the same income and employment for four years, we know that there's very little risk of income fraud. If we see that borrower again, reporting the same information. However, we see somebody that's changing their income and every application by 1520 30,000, we know that that's a bad risk. Last year, we had one borrower, we saw 120 times who had 120 different incomes, and 18 different employers. So we know that that type of thing is very risky. So what we do is, I think chat GPT is wonderful, by the way, because I think it's educating people on the power of AI, we apply a similar concept we do for fraud is we're just scouring massive amounts of data to try to find hidden fraud and make sense of fraud that's happening across these industries.

 

Rich Alterman  29:34

So you mentioned as you're working with the lenders that are contributing, you know, they tag is a fraud loss, or is it a default loss? And, you know, when I talked to a lot of lenders around fraud solutions, a problem that they raise sometimes is they may not be sure they're not always sure whether it is a fraud loss or not. Does your with your solution? Do you have the ability to do like a retro study where a lender can report other losses to you, you can come back and say, Well, really these losses here were actually fraud. And they became a first payment default, but you're classifying it as a payment default. It really wasn't. It was really a fraud issue.

 

Frank McKenna  30:11

Yeah, this is the trickiest thing. We, we spend a lot of time doing this, we actually do two things. Actually, I say three things, we do a lot of aggregate analysis across the industry. So we do analysis through forensic reviews, we have our own fraud team that will go through applications and our reports, and even use third party sources to take all of the early payment default and first payment default for a lender or at least a sample of them, and try to help them understand what's fraud and what's credit. And what we found, on average, anywhere between 30 to 70% of those early payment defaults, have some lie on the initial application. So somebody lied about their income, their employment, their identity, there's something that's untruthful, so it's a high degree of those early payment defaults. So what we do, we offer what we call a retrospective test, where a lender can actually send us all the data, and we're actually able to kind of identify for them how we perform and how the solution will help them solve both fraud. They knew about fraud they didn't know about. And so we go through, it's called the retrospective test, we'll actually do a forensic review on a sample for them as well. We did this for a lender last week, who said, you say 300 of these high scoring loans are fraud, you know, which are fraud. After review, we actually sent it to our fraud analyst. And we gave them a number back. It wasn't 100%. But it was around 50% of those early payment defaults that we said we're fraud had fraud in them. So we try to help a lender navigate that fine and blurry line between fraud and credit. And it's it's not a simple black and white answer either.

 

Rich Alterman  31:51

Right? So you mentioned your fraud team. And your website talks about not only artificial intelligence, but natural intelligence. So can you maybe elaborate more on what is your service around that natural intelligence piece, and how it can play?

 

Frank McKenna  32:05

Yeah, just been a career anti fraud professional, I felt like it was really important. We had our own fraud team as a company. So we actually ended up hiring fraud analysts from the industry. These are experienced investigators that know how to find fraud. So we have a team of for fraud analyst. We have a team of three fraud consultants, fraud analysts are kind of do low level reviews and fraud consultants actually work with customers to help improve their strategies. But this team, we call our solution ai plus ni AI is the artificial intelligence. And AI is a natural intelligence, we think you need both natural intelligence, in all cases feeds the artificial intelligence. That's how even chat GPT works, right. It's all of the all of the the way they review the interactions online and the data that came in what people wrote. And they can mimic that. So we use our fraud analysts to feed classify fraud, to help the data scientists fine tune the model, reduce false positives. We think that natural intelligence and having the human aspect is so important in the as we develop AI solutions,

 

Rich Alterman  33:12

if I was to hire you as a consultant, and I was working on my fraud strategy as a lender, and you kind of think about a checklist, what are some of the key points that you would want to determine, as you were looking at our current processes, what data sources we're using? What are the things that you'd really want to understand as you would help us craft the best fraud strategy for our type of business?

 

Frank McKenna  33:37

This is my passion is and I've done this, this is a, I was a fraud consultant for many years, and I worked on fraud strategy, we call it a fraud situation analysis. I've done about two and a 50 times so I think about my process is is kind of a structured process I would always follow. The first is really just understanding your business. Like I'd always spend a little bit of time understanding of lenders or banks business. But one of the gaps I find with a lot of banks and lenders is they don't take a loss based approach to how they invest their money. They tend to have people want a certain tool, but they haven't looked at it in their overall picture. So the first thing I would do is look at, I'd sit down with a bank and say where are your losses coming from? Where are your pain points? Where are you investing? And how do we fill those gaps, you have to take a loss based approach. Because improving your losses is how you get the ROI approved and how you get it prioritized with the executives and with the technical technology teams. So you have to start there, look at where your losses are and then plug in the right solutions to fill in the gaps. My other advice always to banks is create as many use cases with the single product as you can because you're going to have to go to finance to get the approval. So if you want to fraud tool, look at it. Its impact to stop fraud, but then look at the other use cases in the organization. Can I use that frog tool too? dues stipulations on the front end, to reduce friction to auto approve more, because then I'm going to get somebody in sales and marketing and the product owner saying, Hey, I've got a use case for this as well. And then those benefits can be included in your ROI. So look at that complementary nature. And if you have other parts of the organization that can benefit, get those involved as well. And the last thing I, I'd say lenders, and banks and finance companies really need to look at is how orthogonal, the value is of that new solution, you don't want to buy the same solution that gives you the same result, because now you have two solutions, do the same thing. Look, the complementary nature of any solution that you're evaluating, and you'll oftentimes find that I hate to say this, because sounds so cliche, but one plus one can equal three, we've actually done analysis with our score and another lender score. And we've actually found, the other lender score in our score makes our score better, and their score better. And it gives them a almost a doubling effect, because it can come out from two different angles. That's how I would kind of at a high level recommend to a lender is just kind of looking at that kind of four step process of what's my losses, where are my gaps? What's the orthogonal value? And how do I get other parts of the business in that business case?

 

Rich Alterman  36:15

But you mentioned the use of multiple sources, right? There may not be one solution that fits all the problems, do you recommend a B testing. So where I may take a percent of my applications and say, Hey, I'm going to actually introduce a new solution provider, and let's say a Point Predictive, they're not currently using you, but they want to do some testing, maybe it's easier to do it on a former flow basis. So as part of the consulting, would you recommend to your clients that they do some level of champion challenge or a B testing as part of the evaluation process,

 

Frank McKenna  36:46

there's two routes you can go one is the retrospective to pilot to full production. And that's kind of the way I look at that is like you do a back test of old loans. And then when you're satisfied that you're going to get a use case, you can run a pilot program, which is like, like you said, it's either reckless run 10% or less run 100% for a period of time, and look at the results, you can do it that way. And there's a variety ways, I think it depends on the lenders urgency. So one of the benefits to a approach you mentioned, which is like an AV test is you can be very methodical and you can be very careful. And you can see the value on a small percent. However, if you have a fraud problem, the cost of like an AV test can be substantial, right? Because you're losing fraud on the other 90%. It's a case by case basis, it's wonderful to have that AV testing component. But for every lender may be different. They may want to go with a retro approach, they may want to go with a pilot approach, or maybe an IV AB approach. So having those kind of different options is always,

 

Rich Alterman  37:50

always a plus. Yeah, and then having the flexible technology, like we offer at GDS Link, to easily be able to accommodate all those different use cases, obviously becomes critical. But also, you know, you're not all fraud solution providers, like Point Predictive can actually support retro, right, some of them really only can support a forward flow analysis shares, their structure, maybe how they're buying data, whatnot. So you guys are a little unique there. So I was reading one of your blogs this weekend. And I came across something that I really need you to explain to me because it was kind of getting me to scratch my head a little bit. And you talked about zombie debt reassignment boosts synthetic identities, and you know, for our audience can explain what what is Zombie Debt reassignment and how are fraudsters taking advantage of it?

 

Frank McKenna  38:37

Yeah, I don't blame you. It's a head scratcher. I actually took me a good day or two to figure out what I was seeing online. So it started off, I started, I'm always researching fraud trends. And I started to see this term Zombie Debt reassignment appear on YouTube, on telegram on the dark web. And I was like, What the heck is this? And I learned and I'll give you the summary very quickly. zombie debt is debt that's charged off a consumers credit. It's non collectible, it's usually three to 10 years old. You can go on sites like debt catcher, which are kind of these bad debt sites. You can buy these debts for pennies on the dollar. So some like you can buy a mortgage that charged off 10 years ago, that's to another consumer, then you own that debt, presumably for the rights to collect on that debt, although you can't. But what you can do apparently, is you can send a letter to the original creditor saying you need to report this correctly. I assume this debt My name is Frank McKenna. I paid it in full reported to the credit bureau that that mortgage for $100,000 has been paid in full by me satisfactorily. That means it goes on my credit report. Having a mortgage that's paid off makes my credit score go up and I looked like a good credit risk. That is Zombie Debt reassignment turning charged off bad debt into a better credit score. And it's blows my mind that you can do this, but apparently you can see it all over the internet now. Well, this is

 

Rich Alterman  40:03

a good segue into one of the final things we'll talk about. And you mentioned this in your blog. And I think we all know that the fight against fraud and interpretation of fraud is clearly a cat and mouse game, fraudsters developed new techniques to industry find ways to combat those new techniques. And then fraudsters develop new techniques or approaches. And my question is, with fraudsters demonstrating such a strong skill at adapting, what do you feel it will take to really win this war? Because we are at war.

 

Frank McKenna  40:33

I think you nailed it Rich. Earlier in our conversation, you talked about consumers participating in the fraud detection and prevention, that to me, consumers and banks working together, is the way forward, the ability for a consumer to freeze their credit, by the way, is one way, the ability for a consumer to turn their debit card on and off the ability for a consumer to actively notify a bank when they're going out of country. All of that all those things we can think of are going to help us win the war. Because right now, banks are trying to do it independently, without understanding what the consumer is going through. Technology is getting to the point where you can easily interact with the consumer via text or platforms or applications. That's the way to go. I think that is going to be key. I think banks are going to have to work to speed up the process to integrating new technologies, because they are far slower than the fraudsters. And if you go on now, with telegram with dark web, there are 100,000 telegram channels dedicated to fraud, and fraudsters talking to each other in real time, minute to minute, banks don't have the luxury of that. But we're gonna have to find a way for banks to get more comfortable and integrating new and upcoming technologies and talk to each other. Because if they're talking that quickly, and the banks aren't collaborating in the same way, we won't win the war. I think the combination of banks working together with consumers, and then banks getting quicker working with law enforcement and working with social media, all of that is going to be key to winning the war.

 

Rich Alterman  42:06

And the aging population, right is educating. I'm very blessed. I have a 93 year old and a 88 year old mother and 93 year old father, and I'm proud to say that I get calls for them now saying, Hey, I just got this thing pop up on my iPhone. Is it legitimate? Right? And I've trained them at that age two, detect these things. And I'll say, Mom and Dad, if you're calling me and you're asking me, it's not legitimate. But just to lead, right? Right, correct. Yeah. And if you remember, gentleman that was at Greco first advantage started a company, really around helping elders avoid being victims of fraud. And open banking is a way to, I think get at that as children, right? If we get on our parents accounts, and we can see, hey, hey, Dad, why are you writing these checks for $500? It's great. Yeah. Right. And that's the way to do it. So we're getting near the hour. So I'm going to throw out one more personal question, and then we'll we'll sign off and learning a little more about your background, I understand that you come from a very large family, I believe you are one of 12 children. Is

 

Frank McKenna  43:09

that correct? That's right. Yeah. Okay.

 

Rich Alterman  43:11

Can you share some key lessons you learn from growing up with so many siblings? Yeah, that's

 

Frank McKenna  43:16

a really good question. I think growing up with so many siblings, the first thing I learned, is really sharing, and about participating with each other. I was shocked when I went to college, by the way, and we would stock the frigerator with stuff. I had gone to the store, and I bought whatever in there. And if I always expected people to just take it, but on the vice versa, people would never, like if I went and took some milk, they'd be like, Hey, you took my milk. To me, that was the biggest shock was that things don't belong to everybody. Because when you're grew up in a big family, everybody just shares everything. I think that whole concept of just sharing and going you realizing that you can help other people with something I really learned. And I also learned that having a big family, like you said, with your parents, you have such a support network. Now I'm just so grateful. And I think your family is very important in protecting you, by the way from like fraud and scams. I mean, my father, unfortunately was a victim of the elder abuse and $169,000 fraud and I didn't find out about until it was too late. I wished I you know, he reached out to me early in the process like your parents do. So I think that network is family network is just critical.

 

Rich Alterman  44:31

And I just remember the name of the company was EverSafe. A company that was started by Howard Tischler, so for people listening, it's definitely something that you should look out for your elder parents. And you know, it's funny, I say to my daughter, you talk about that your friends and family, and maybe this is negative, but I say, you know, in your life if you have five really, really good friends that would be there for you at the drop of a hat. You're very wealthy person.

 

Frank McKenna  44:56

You're very wealthy person. Yeah, that's right as you get a little older You start to realize that you have that core grit, you start to know who those five are, I think when you're in college and in your 20s and 30s, you have 7080 90 friends, and then it comes down to you got three, and you got your family. And you're happy. Yep, exactly, exactly.

 

Rich Alterman  45:17

Well, Frank, thank you so much for joining us today. This is Rich Alterman. And we've been syncing up with Frank McKenna, chief strategist and co-founder of Point Predictive. If you would like to learn more about Frank's firm, please visit Point Predictive.com or email the experts at info at Point Predictive.com with any questions you have, and I really do recommend for those people listening today or tomorrow, that you do go and sign up for Frank on fraud.com, which is Frank's blog and they abreast of what's going on. Thanks for tuning in today and learning more about fraud trends and fraud prevention in the lending industry. We look forward to having you join us and future podcasts. Make it a great week, and this is Rich Alterman signing off. For those seeking to improve their decision and lending processes don't miss a chance to connect with GDS Link and Point Predictive at FinTech Nexus taking place in New York City on May 10, and 11th. And please stop by GDS Link booth number 742. To meet with our team of solution specialists and learn how GDS Link cutting edge decision solutions can help accelerate your organization's growth. And please be sure to connect with Point Predictive Tom Algie and Jeff Hendren to learn how they are leading the fight against emerging fraud threats. Additional event details can be found in the show notes. Please take advantage of this opportunity to connect with industry leaders and take your lending to new heights. Thanks for listening. If you've enjoyed today's episode, please be sure to subscribe on Apple, Spotify, Google or wherever you listen to your podcast. And be sure to leave us a review. Follow us on LinkedIn and connect with us on Twitter at GDS Link that's at G D S L I N K have a question for the show or have a specific topic you want us to uncover. Hit the link in the description to drop us a note. Thank you for letting us part of your day. Make it a great one.