<ul> <li>How to manage the changing auto finance dynamic</li> <li>Developing new channels and partnerships with fintechs without taking on undue ris</li> <li>Using AI and automation to automate processes for better risk results</li> </ul> [toggle title="TRANSCRIPT"] <div class="transcript-scroll-box"> 00:00 So as we know, managing this changing auto finance dynamic is difficult. 00:07 And having all of the potential risks out there to think about on top of that can feel overwhelming. Tech is here to help. But the amount of tech out there can be overwhelming too. So how do we decide which type can help mitigate risks? auto lenders are developing new channels and partnerships within text to overcome digital transformation. But how can lenders do this without taking on undue risk? Also, what areas in the risk landscape are coming up more and more? And how can lenders leverage technologies like artificial intelligence and automation to help automate processes for better risk results? Today we will address these questions and I'm excited to be joined on stage with these two distinguished gentlemen from to captive finance arms. First, we have good until she fiscal strategy So, Yamamoto finance. And we have James Vincent, Vice President of consumer risk because I would like to give them the opportunity to introduce themselves as well as as well as their role their companies, if you 01:18 should. My name is Vijay Patil. I'm Chief Strategy Officer for Yamamoto finance. I'm responsible for risk management with the strategy, as well as servicing operations for Yahoo. Finance, been there for four years was brought in to set up the company to us 60 plus support here's the deal with Yamaha. Prior to that, I was with Vichy with the BC motors credit for for six years, and then HSBC on finance for six years. 01:55 Hi, my name is James incense, I lead the consumer risks. department at Hyundai capital America. I'm Steve about 35 people. 02:05 We manage credit and residual risk for the Hyundai Kia and Genesis brands. Beyond that, we have a team of statisticians who do predictive modeling for loss forecasting and such. We have a team that is dedicated to into fraud, management, prevention and mitigation. We have folks who focus on enterprise and risk management activities. It's kind of the full the full suite, I have a banking background. And so it's sort of my vision as a as a risk manager to kind of take what a federally regulated bank would look like. Like considerably for an auto finance capital company. 02:48 I've been running Captain America for about three years. With my family and from the east coast. at home. I'm the father of five daughters. Wow. Very nice. 11 years old. Six weeks old. So they're Focused on emotional risk management, maybe, 03:06 breaker. 03:09 Amazing. Well, to kick off our panel, I'd like to set the tone for our conversation by asking you both you how your view from, you know, the captain perspective lays the groundwork for how you think about risk management within your business. 03:26 Sure, so a captain who is a little bit different. I worked for an independent company also became Mr. Deasy and then captain. 03:36 So capital, finance companies look at all different. We offer financing for years, with one goal in mind, we offer stability and consistency. So stability, many when there's a recession, we will get them out. We are not going to be clever the day to read the global economy is awesome. And then things go down, we were finding hard time to get funding. So stability is one part of consistency. So we want to make sure we offer a credit programs to our dealers that are consistent. That again, just not this promotion for six months now we offer programs are consistent. Risk Management is outcome. So anything we do has two things in mind. And you're interested in trying to technology looks at from that lens. Are the vendors stable enough to deal with problems that happen have been in the business technology. Sure, every time you see something new and shiny, we all get excited. But we take a step back and say, let's make sure that the shiny object is shiny when it's recession. So because of that we look at the risk management work differently. We take a long term view of risk and uncertainties associated The new thing done me in the finance company? 05:05 I think it's really smart to start with that question. Eugen, I share a very similar perspective. They both and capital finance companies, I imagine it the majority of the folks in the room here have a different perspective, but we each have fun. 05:19 So it's it's good to kind of couch this discussion through that lens. I think if you were to ask me 05:26 what the captives value proposition is, I would almost echo exactly what it was you said. I would say that it's our goal to be a trusted partner to our OEMs. And from my seat on the risk side, that really speaks to consistency. I think if you were to ask in our head of sales, they might say, well, maybe support is our most, you know, our most valued offering and operations, they might say efficiency, but I would say it's consistency of anything that we introduce, whether it's in the product and a piece of technology, that that lowers our consistencies, or increases volatility in our performance is not a value added to our, to our manufacturers. If it increases variability in the customer experience with the dealer experience, it's not valuable. So it's very important, I think for to set that expectation and kind of look through through that lens. 06:22 Definitely. And, you know, there are many risk factors to keep an eye out for James, you touched on it a little during your intro. There's credit risk, residual risk, reputational risks. So what is the full suite of risks that you are managing and maybe what are some of what are the top maybe three that are just top of mind right now? 06:41 Well, we kind of go through the full spectrum. Like I mentioned, sort of enterprise rescue, I imagine musics similar. We have to be considered a operational risks. I think, top of mind. For us, as always the two biggest restaurants your credit and residual risks. In our business, but I think from an accounting front of capital, we're spending a lot more time to focus on operational risks. So whether it be data criticism, privacy, compliance type, for instance, they're getting a lot more attention. And certainly we're dedicating time and actually dedicated resources to managing operational risks. 07:25 Really quick, just quick follow up, you know, why? What is it about the environment that you think is, you know, kind of having this operational risks come up more and more finger talking about? 07:40 It's a good question. 07:42 If you think about a capital partner, we are working on whether or not our profit center at the end of the day work cost. And 07:51 so that's how our 07:52 manufacturers look to us. We have to be consistent and deliver on a value proposition and so any any any efficiencies in our processes are effectively operational risks to our broader company. So I think, you know, as margins get tighter, certainly our business as incentives don't go quite as hard as they used to go. The focus on on us finance company is to be rigorous with our operational risk management, our efficiency and our expense. 08:26 Right, that makes sense. And BJ Would you like to answer as well, you know, what are the what is the whole suite of risks that you're managing and what's really top of mind 08:36 right now? So it's the same as what Tim said. Exactly. 08:41 Exactly. That's 08:44 the only thing I would say the top of the mind is fraud risk. We offer credit card program, which is not typical for a company finance company. The signal significant portion of the business is offering credit cards so Capital One is to the most They exited the space, he brought it in house. And oh, that was dominated. So realizing the feeling that properties will be in great prospects. And it's because the process processes enable the value proposition to reverses to a very simple application process. You can buy your motorcycle on a credit card. 09:26 And the process is simple. And then that creates issues because you're much 09:32 more likely to the product. 09:37 And I definitely want to jump in a little bit more later in the conversation about fraud, specifically the tools that you're using to mitigate that. I am curious, you know, it is meant so it's top of mind right now is is during seeing an increase in product or so today and then maybe a year ago or 09:57 so it's a good question. We are not at all to get to the Compare or trend of fraud. We launched the program 2425 months ago. But the last program took it away from cap $1 billion or first focus was to control credit risk. And we really beefed it up, we have got 1200 plus rules, to three scorecards we haven't seen on fraud, we would like to maybe start orienting to the fraud. Let's see what we get. And the first one year of experience, we realize that we would like in terms of controlling authority. So we've we've talked about how video tools went through training for emotional scars, and things of 10:45 that nature. 10:46 Where are you so we kind of got a better understanding of what this risk management landscape scape looks like for you right now. So many just surface level touch on what are those technical Any tools that you are using to manage and changing audit findings dynamic, you know, and the risks to help mitigate those risks that we just touched on. 11:10 So tools to manage the changing auto finance dynamic, I think, 11:17 from the captain perspective, 11:20 to be heard as we're not as aggressively pursuing 11:27 newer tools to manage that dynamic. I think that we're, we're certainly fast followers that, you know, you're going to see those changes actually 11:38 the largest marketplace we follow, surely what our manufacturers do. So, we are though looking to in fact, 11:51 we've joined some partnerships on the commercial side of the business of unnecessarily an era of time. 11:59 fluid in But we've made some partnerships there. We put some technology in place to add some avenues for our dealers to sell cars, which has been very productive in its infancy, and certainly plopper later in the fraud space, in the underwriting space, we continue to look for partnerships and additional tools and technologies are there 12:23 any, you know, any technologies that kind of just really stand out to you right now, even if they're not necessarily, you know, you're not necessarily just grasping for something new. You know, what is maybe something that you think all you that you're getting all the time with? 12:44 Certainly on the front end, and underwriting we are spending a lot of time for Friday verifications. We're looking for ways to automate and waive stipulations simulations. It's a massive pain point for with customers and for dealers, the amount of time that we spend on the dealership, I think anybody ever like spent four hours with dealer, but that's sort of the typical wait time. And so if we can smartly waive an automated stipulations, whether it be proof of residency and other types of verifications, maybe issue a customer interview for using alternative data to kind of verify that customer. We're spending a lot of time and she's looking for ways to do that. 13:30 Again, we're going to touch on alternative data as well. But first, I want to hear from you, you know, what tools you're using Yamaha to manage the changing auto fix dynamic, and you know, the risks associated with that as well. 13:43 So most of the more than two thirds of the logic is a lot more expanded and color strategies, tools and technological politics 13:52 that the studies use are looking at our channels 100% of population That is true. So we are looking at Consumer Direct the approval channels and processes on the product side. 14:11 More than the technology that too, we're trying to invent something building our own internal fraud score 14:20 for understanding for business, as well as looking for some tools on 14:28 the shelf that can help us when you get across. 14:33 Maybe what's one of the newest strategies that you're also looking at or maybe thinking about work with in the near term. 14:46 We will start as the crowd is productive. We can have enough time 14:54 for us with Mr. carnivale and we will continue using multiple tools Okay, 15:01 um, you know, I, we talked about a little bit of alternative data. And you know, there's obviously benefits and pitfalls that come with leveraging it. So what are some of the use cases for alternative data that you're seeing within your businesses first and foremost? 15:20 For us, the challenge is most 50% of our business to 80% offer to the door applicants do not have 15:31 the visit. And it's given demographics, right, look at the products, they attract the preposterous and pepper bytes that attract them that Google offers, and they will always tend to have, you know, history or very likely history. So one of the very immediate needs to see is underwriting that. So FIFO FIFO, or no credit, less credit customers. We use the continent data for that. Okay, 16:02 I've touched on a couple of them already. 16:05 underwriting for me is the number one priority for use as alternate data. But the rub is kind of a catch 22 for us is that we're expected to underwrite on a consistent basis. And so our, our customers in our in our owners don't fully appreciate small games that we make the other items, and we'll continue to do it, it's our job, you know, it's our job to get better and better and better at what we do. So that's why I'm so excited about using alternate data for stipulation waivers and kitchen clearing, because it has a very tangible impact, very measurable impact to the customer, to the dealer and to our OEM partners. So we consider and on the capital, those are our three customers and future customer or dealer and our OEMs are equally our customers. And something that didn't impact the customer experience. It's very, very powerful. And it's very marketable for us. And that's why we're excited about it. We're spending time exploring those avenues. 17:10 And what are some of those, you know, major benefits that come with using alternative data? 17:18 So for us, it's about influence. If you don't use data, we can't do any of that because I cannot interact with customers using traditional tools. So the biggest benefit is that on the backhand side, we have a bigger challenge on skips. So, Michelle, how many of you know the motorcycle market, the skidplate motorcycle, almost 50% 17:46 in our whole position, hundred customers, by 1999, right? One of 17:53 our products, it's not 99 it is 49 for the products that we sell from the position Listen. 18:03 And the challenge for us is how we'll find the customers who are still booking but the sample products and missing documents sources of information. 18:17 We haven't yet implemented alternative data on the back end. But I think there's an exciting opportunity there on the collection side. 18:24 I think 18:26 traditionally and in the auto space, on the consumer side, your ability to work out a customer and your tools and have are quite limited. So the more information you know about those customers, the you can expand your tools that you can work on a different sort of payment methods, if you understand the customer's payment history, their payment priorities. It allows you 18:47 to kind of have to 18:49 talk to the customer. I think the last you know, gentlemen said previously used to just call the customer talking to everybody in the car and now you have a very gentle warm conversation. Please record does not work nearly as well as he used to. So we're constantly looking for ways to improve our classrooms by our apprentices. 19:09 You kind of segue into my next question, which was also can't always just talk about the benefits, also pitfalls that come with using alternative data. You're welcome, what maybe what are some of those challenges? 19:24 Sure. Certainly, there's difficulty in explainability and clarity with regard to alternative data. I think industries are built on traditional credit data. And so when you're talking to a customer, even explaining a decline to a dealer or customer is perhaps not quite as recognizable. Or it may feel 19:52 maybe less, less powerful than telling you the customer back for history in full time mortgage 19:59 starters To say that you couldn't pay off 20:04 your second checking account had an NSF or something like that. So explainability is definitely one. Our industry is not quite nearly as heavily regulated as the mortgage industry. But still we can have that same forever in regards to adverse action, that's a difficulty in thermo for me would be achieving funding. There's, you know, you don't have a full suite of believers, probably in any company here. Particularly at an executive level. I'll have some detractors, maybe they'd be super for these in the past didn't work well. They don't quite believe it. And so it's more difficult, I think, to to convince 20:50 our executives as well as our 20:54 Korean parents, to allow us to kind of invest in some of these new parents 21:01 What are some of those pitfalls for you? 21:02 So 100% I think that's 21:07 one of the other failures that I see. 21:11 Explain that, 21:13 in terms of alternate data is being able to integrate with a traditional IRA. What I mean by that is, we don't have the division of alternate data with a traditional data, which I refer to as critical area. So we can derive it like in a waterfall fashion. So on the right, this was the old fashioned way. When you have new information, you know, everything else, if you don't have that information, then go to this alternate data. And then under that. I also started some tests on this that one of the prime data prime cost of 787 30 customers. If you use alternate data source, you'll see the good separation power, but we've haven't figured out a way to do that yet? Because what do I tell the customer? Right? Same thing appointment. So if my primary underwriting is using the trade information, let me give you a reason based on the traditional sources of information and not tell you how you can take your cellphone bills. So integration is the challenge that we see alongside 22:24 as well. I think that's word work out of that sample. It's much easier for us to use that type of information to to get additional approvals, and that is to 22:35 provide that percentage. 22:38 Yeah, actually, you know, it really just comes to mind. What are you doing to navigate the challenge of integration? 22:46 taking a lot of time. 22:49 I think we're 100% agree. It takes a lot of time and effort. You know, we spend, spend many months or even years of building and training traditional models. It's not as easy just to take a whole additional data set and kind of integrate that data set. So you can lay it on top, you can have a second there. But it's not really too powerful. And I think you run the risk of making decisions that you don't have a long term statistical support on those decisions. So it becomes a risk in and of itself. Even though you're ostensibly mitigating the risk of forward you're also introducing some additional variability in some risk. 23:32 So I wanted to go ahead and segue from alternative data into fraud and fraud detection. I'm looking at our mobile app right now. And there's already a slew of questions, but I'm sure a lot of it has to do with fraud. It's a big topic right now. And it's definitely come up especially all problems in auto finance, and dealerships have come up more and more, especially in our reporting at auto finance news. dotnet. You know, and you did talk about a little bit of so Those practical technology tools that you're using to help mitigate fraud. But you know, can we just go a little bit deeper into exactly what you're doing? Maybe what is that one top technology or tool? And how is it helping you make a fraud? 24:17 So, again, one of the tools, I think it's a comprehensive approach on how to handle frauds. 24:24 tool is definitely part of it. very fortunate to try certain things at times in 24:29 the room as well, has a really good score turns into a synthetic score. We had Italians in that area. We've tried it for 60 days and then working to develop the tools we were using. We had to have the tools LexisNexis as a product, an ID a tool to use it for fraud. So more than two though, we realized that the training to own the ladders is as effective as the tools we must have Going into two to 10. And the third area that we normally don't talk about is dealer. Make dealers, your partner. In fact, in fact, the fact finding. And I remember when we had a pretty high fraud, three to four times higher than the capital ones experience on gamma portfolio. First thing we did is we realized where is it coming from which parts of the country which parts of the business and also these parts of the network, when we had a possibility, possibly we opened our box and said, Guys, how can you help us media? Because it says your interest as a dealer, as well as an all interested capital to make sure there's no foreign transaction because the more foreigners they see from your dealership is going to have issues in the future. So making the other partner that they'll open up they'll talk about the first level of patrol they have. Some dealers will tell you 25:58 some dealers will tell will verify Do 26:01 we have this machinery or fingerprints to be able to apply for financing to the dealership? So you get this multiple levels of question, answers from them around practices that they're using. And we met them a partner said, Okay, if there's a dealer, she was doing really well. How can we make sure that those practices are followed by this other leaders? And being a capital, it's easy because they only want capital and 20 of the bags they live. So there's capital that will open processes since they're more open to partner with us. And that has really helped us control the 26:37 channel, James, how are you? Where are you? 26:41 Opening a fraud, and we have 26:43 similar type of products that it was mentioning, we utilize experiential for synthetic ID LexisNexis for ID verification. We have a number of tools for skip tracing and for income verification But I think for us, the way I look at fraud is is very, very difficult to identify fraud at the time of application, the datasets that we have to use is credited. fraud is not a credit event. fraud is an operational risk. And so if we're if we're only looking at our historical jobs and the performance of our customers using credit, then we're not going to be able to easily just pick out the fraud from our jobs, and then they essentially become a credit risk. So to me that's, that's the the biggest hurdle in fraud prevention is disaggregating credit loss from problems. And so as BJ was saying, the way that we have to attack it is through our operations. So it's training the underwriters, which one do we have operation centers in? in Irvine are the ratings and Atlanta and Dallas we spent time actually doing in person training with them. Our we set our fire teams out there despite the cause. Because the difference. We do spend time there dealers, particularly with our parts, dealer groups, you know, we'll send our sales force as well as some of our credit buyers out there to talk about what they're seeing, tell us what they're seeing. And that will tell them some of the areas that are challenging, but it's very, very difficult. And I'd say that the landscape for fraud 28:33 is increasing. I don't know it's increasing. It's certainly chicken chain is always changing. 28:39 Yeah, you know, do you feel the industry has become as or as an industry becomes more advanced with technology that the broad landscape landscape maybe increases or that's causing that change? 28:51 It could be I mean, there could be technologies that make it easier to perpetrate some type of fraud 28:59 for Hyundai and Kia, you know, are the vehicles that that we produce are high quality long warranties. And so our dealers have less service business than, than some other brands. And so therefore, they struggle a little bit more for profitability. And so dealers that are struggling for profitability, potentially could look for ways to cut corners. And so we have to be very, very cautious with dealer behavior. And so we have, we're spending a lot of time to integrate consumer fraud mitigation and commercial fraud mitigation and hazard this 360 view 29:40 of our dealer network. So what we're seeing on the consumer side, what potentially we're seeing on the commercial side with regard to dealer financial is trying to merge into 29:50 a, you know, what about yourself, do you feel that as an industry become more advanced technology, that landscape increases? 29:56 Absolutely, because the reason will grow Use technologies and devices to provide a good customer experience for our customers. And the more get into the experience area, the more 30:10 we leave it open for customers to take advantage of it. Because if you don't design house equal to one person, it is unsafe for the broader customer base applies to prime credit, for example, if somebody is in a prime business, you don't design the program. That's good for 2% of customers. And the program is good for the 90% customer, that they have easy access to credit, you know, verify certain things. And that opens up more doors for four parcels. And that's one two is visited, the more digitally go, the more avenues you open up. And I think we have to again, we have to be ahead of the parser to make sure that we are controlling things ahead of time. 30:57 Right I was actually just thinking to ask you what was some best practices lenders can do to stay ahead of the concert. 31:05 Lenders can't do that, of course, because you don't know what forces will do. So 31:14 yeah, so I think it was mentioned earlier, if you take a narrow view of the 31:19 technology, 31:22 that's about your problem to be solved by having a very comprehensive strategy. So you look at areas of exposure to the product. And then make sure you have processes, tools and technologies and policies in place to handle. So that's something that's just ahead of them and react quickly. As soon as you see a trend. Don't just say this is a trend. Let me get some more data side to do what you can as quickly as you can, to, 31:53 to stop that. What form this product most often take. So, 32:01 for us, I think it's identified that that's number one, because a significant portion of our customers are in credit. And the attorney get in the area immediately don't know, but it's imperative to clear truth. Credit, oh, somebody submitted to you for credit. And that's the that's why TransUnion score is really good score is working for fraternal. And as you can imagine, it's very hard to predict synthetic fraud. So creating a synthetic ID merchandising for a couple of years, and then all of a sudden going and making multiple purchases, which is very similar behavior of a customer of a student coming out of college. Maybe there was smoke grenade or low credit and they buy the car. How do you distinguish this small portions one 2% of the market, from the broader group of customers that they have Listen to the truth like synthetic fraud score really is helpful for 33:06 William James, what form is fraud, you know, typically take for you that you're seeing. 33:11 I agree that the more egregious types of fraud are with regard to identity, so whether it's synthetic ID or mighty tap, and I do think that our partners here in the room have some very good tools to help me to get that would ask for what occurs more frequently. For us. It's more lately types of frauds of income inflation, in straw purchases, different project addresses, particularly with regard to straw purchase. That's extremely difficult, what's impossible for us, certainly for an automated application to identify, we're not talking to that customer, without relying 100% on our dealer to do that due diligence for us. And as we all know, they often don't and then the identity verse or the the increase of a very expensive and Coverage isn't particularly good for strong identification. Yes, there are tools for I sorry, for income verification, there are tools for income estimation. But they're say they're all sort of mediocre kind of hit or miss. They're, they're more directional. So I would say it's a lobby to credit your friends in the room here to work on some survey, large scale income verification that lenders like. 34:32 That'll be my advantage. 34:34 All right. All right. So you know, implementing these strategies or these technologies, you know, can be risky as well. And sometimes captains need a little help. Right? So what are some of the practices around selecting the right vendor to help deploy these technologies, these strategies that we've been talking about, and the risks associated with that as well. So 34:59 risks around the country selection? Yes, I think that we just have to be very cautious of what I mentioned earlier on the data privacy side. I really like the way that the industry is moving towards partnerships and involvement. And so that's something that, that we're dipping our toes into town. So starting small, not necessarily being an industry leader or being a fast follower. So the model we're trying to follow in a capitalist, numerous partnerships for small pieces of the business, verify that they're working, they're adding value. 35:39 So, for the vendor management, Yamaha has a pretty solid framework. It's probably 35:47 200 questions. But the key to that entire analysis is how stable the vendor is. It somebody is a new vendor, who start a new business and say hi shiny object works really well. It was, what, three months or one year, two months, we'll probably be not open to that. The reason is we want to try something that has been proven out there, back to my stability comment. So that's sad for us how stable vendors are approved on the technologies, and the big names, 36:24 the names, I think 36:28 they come on top of the list, they come back to us and say, This is a tool that will have more faith in that tool than 36:36 having said that, you will not that we want to jump onto it. We will do a pilot for Spring Boot. We know you they're not historical data, make sure it's working. If the vendor is not able to do historical analysis for us, probably are starting to make a case that this tool is not the only way to add value for us is to show that on historical data. It works. Once you do the analysis, we do a pilot, probably a 60, day 90 day 37:06 and then convert that into full time living. 37:12 I do want to take the last few minutes to answer some of the audience questions to throw away. One that is interesting here reads that, you know, type does help us to faster and make more money. But what about operating costs as tech? Does tech tend to increase or decrease operating costs? 37:35 So depends on the value of the quality. Right? So it's very hard to answer this question. Great question, by the way, but if you are going to talk about big data, right, so four years ago, big data was a big day. Everybody wants us to store everything about the customer invested a lot of money, that if you look at the really investment in big data, what's the value that you get out of the big data, telling the customer, the captain, the customer was pretty small. So if you invest a lot of capital model resources in that area, where there's uncertainty around what benefit, you might get the careful, I wouldn't lose that. But there's another area that takes over very predictable. And they thought, Well, if I invest certain dollars on the fraud prevention, I know what I can do, I can sell a certain number of transactions that leads to certain my loss and the catalyst is simple. And I know what I'm getting into by having the goal or outcome of the technology. I was gonna start with that in mind and not not begin with that because in that we figure out what to do with it. So with that, I think the answer changes. So because you have the outcome in mind, you enlisted intelligence as much as you can get benefit from 39:00 James, let's hear your take on that as well. Yes, this tech tech as you increase and decrease operating costs, 39:06 I think the short answer is that increase them. What it does is it shifts the balance sheet 39:14 ecosystem money around. 39:16 So, you know, something, 39:20 partnerships that that we would be involved in tend to be heavier on the back side. So the bottom line cost, it could be the same could be less than before and what it does, it's shifting topics. And we're like, like I mentioned earlier, very, very Congress as part of our op x ratio, especially if you are on FX assets. We're very cognizant of it or rigorous around. We're judged by that. So again, as part of the difficult sell when it comes to new technologies and new partnerships and the costs associated. 39:55 One of our audience members is asking Maybe dive into a few examples of alternative data other than bank account information. 40:11 So, let's move 40:15 on to Okay. 40:17 Are you are you leveraging this? 40:19 No. So just be clear, we are not using any alternate data as well using the uniform provided by our data provider. Okay. 40:29 Similar answer. We're currently working to develop a partnership in underwriting using alternative data. For us, we found that that would be a quicker means to define and less expensive than purchasing, modeling story. cetera. So but as far as the types of data transfers, right, 40:58 you mentioned a partnership. underwriting would that be a partnership with a with a vendor, third party use third parties. Okay. And you wouldn't mind. 41:08 I'd rather 41:11 well maybe share a timeline or word or a timeline for when you know this, this would kind of start start functioning in your 41:19 organization. Sure. We're looking to launch in late q2 early q3. Our goal is to make it seamless to our dealers to offer increased approval rates in subprime in this apartment area, so we're looking for ways to increase that value proposition. Okay. Thank you. 41:43 We do not see any any upcoming technologies that we should know about. 41:50 The commodity in general. 41:55 So 41:57 now, that's the end. implementing those technologies, but the people who are a responsible business, pretty sure 42:08 agree with me. These are not new things, right? So automation has been around for for 10 plus years. The nothing new here, what's happening now is the use or application of this technology on different areas of business. And that's where it's becoming like big data, the buzzword nowadays AI ml deep learning. So we will deploy the, to the assessment. Another way to go. Personally not comfortable because machine learning is a self taught model learning. We're not there yet. Applications are limited, 42:51 deep learning. 42:53 I think we use it in fact, we use it 42:57 so we will protect your disease using those technologies. is 43:02 in my job. The new area that I'm personally watching is the distributed ledger technology, which is the Bitcoin currency application was a distributed ledger. But that's something that's new. That's a new that is coming a new font and shapes. But the application of distributed ledger could be enormous. The cost is hard right now, because the computational power you spend a lot of money to compute that. But synchronization, auto synchronization, could could could use that technology. I collections to vote on $120,000. Imagine, you could probably afford to have a smart contract. So those are the areas that are coming up. Those of you watching, but there's no immediate need or 43:52 plan to use it. Okay. All right. Thank you. All right. Final question for you both and I think it's perfectly You know, as with captive buy, to buy companies, how much control do you have over setting and managing your risk appetite? And how does that dynamic impact the tools you have at your disposal as a risk manager? 44:17 I guess I can answer it. First, I'd say that we have 44:21 a very high level of control over setting our risk appetite. 44:26 But the caveat is 44:29 we don't have control over the reactions to the risk appetite. So it's certainly up to us to where to set our credit policies, how people to buy on the risk side, as long as we're in alignment with our, our Treasury, our finance partners, but the expectations kind of everything as we started this conversation, yes, rotation is consistency. And so if we're, if we're very nervous appetite, month to month, day to day, year to year, that's not a good thing. And so we have to establish a risk appetite. improve our processes within risk appetite and then we want to be very vigilant around managing that risk appetite and when we do change it, it behooves us to to not do it in a tabular fashion to get full buy in from all of our customers our partners. 45:22 So, we are a global company and the risk appetite is not not decided that particular region it is a global strategy. But yeah find out so, we this is a new demo USA was quoted. So, Francis. So, because of the AMA finance companies around the world are established companies and the US is leading over. So we do have a lot of say and think clearly, there is appetite for finance. 46:00 Thank you. Unfortunately, we're about a time I could sit and talk about fraud and risk management all day, but hopefully it'll be sticking around and ask you any further questions. But please join me in thanking our panelists. Once more for </div> [/toggle]
