Transcription:
Sean Dugan (
All right. Good evening everybody. Hello. Welcome to Digital Mortgage. My name's Sean Dugin. I'm a Black Knight. for those of you who don't know Black Knight, we're a leading provider of integrated technology data and analytics, to the mortgage industry. We've got many lenders here and across the country that use Black Knight's innovative solutions to overcome business challenges across origination, servicing, and secondary marketing. most importantly, we're a proud sponsor of Digital Mortgage. We hope you enjoy this next session titled The Role of Big Tech and Data in Mortgage Lending. My pleasure to announce the, panelists. I'd like to announce Mr. Derek Ochs with Lower, Rhett Damon with Open Door, Ann Marie Pippin with the Federal Housing Finance Agency and Bill Elderton with Roostify. With that, I'm pleased to turn it over to our session facilitator, Mr. Jeremy Sicklick, the CEO of HouseCanary. Jeremy.
Jeremy Sicklick (
Thanks, Sean. I appreciate it. All right. Excited to be here with all of you today. Sorry for, the formal being up here, but, wanted to make it work so everyone could, could have a mic. so today we're excited to talk about the role of, big tech and data in mortgage lending andreal estate more broadly. before we really start, why don't I turn it over to each member to just quickly introduce themselves and their organization, and then we'll dive into four different themes, for the next, 25 minutes or so. So, Ann Marie, starting with you.
Anne Marie Pippin (
Sure. Ann Marie Pippin, FHFA Federal Housing Finance Agency. We're the Prudential regulator over Fannie Mae, Freddie Mac, Federal Home Loan Bank system. we've just opened a new Office of Financial Technology, which I'm excited to talk to you guys about today. I'm an associate director in that office. Also lead the Office of Governance and Strategic Initiatives in our conservatorship division as well.
Derek Ochs (
I'm Derek Oaks. I'm VP of engineering at Lower, we're a digital lender, and, I lead up all the, engineering and development efforts there.
Bill Elderton (
Bill Elberton, CTO at Restify, and at Restify, we offer a POS platform for the submission of application as well as following the mortgage application all the way down through fulfillment and closing.
Rhett Damon (
Ret Damon with Open door, head of brokerage operations, and now, data strategy for brokerage. so glad to be here and I'm, pleased to see this many people. I think we're the last session before, you know food, right? So five o'clock and beer Yeah. So, good to see everyone here. I was convinced that there were gonna be like three people in the audience. So I'm, thrilled.
Jeremy Sicklick (
And I'm Jeremy Siclac, ceo, Co-founder, House Canary. We're, we're focused on big data, starting with valuing properties, but using it more broadly to, enable transactions to happen very quickly. and so the first topic and theme that we wanted to start with, you know we're honored to have Emory Pippin here from FHFA, and we wanted to hear Emory is part of a new organization at FHFA, a new FinTech office. And so Emory wanted to turn it over to you to just share about, you know how regulators like yourself see the role of big data and the kind of innovation that you're looking at, and yeah, what you're doing in the FinTech office that's newly created.
Anne Marie Pippin (
Yeah. Thanks, Jeremy. We're certainly honored to have been asked to come speak, and to talk to you guys and to interact. Stakeholder engagement is gonna be a big part of what our office is gonna be doing, especially as we start, the office was officially launched in July a couple months ago. seems like longer, so we've been very busy since, we announced the office, in July, and along with that, we issued an RFI. RFI is still open. It's open until October 16th. So if you guys have not had an opportunity to check that out, I definitely encourage you to take a look. we're also gonna be holding a listening session, in early October as well. And certainly you can ask me any information about that. Glad to share. as far as the goals of the office, So this is certainly a priority, of Director Thompson, of the head of our agency to really understand technology driven developments in the housing finance sector broadly. And so we have a very broad mandate. We're looking to understand and idenRestify opportunities to understand barriers and to address emerging risks. So, this all falls within our, our mandate as also Prudential regulator, over Fannie Mae, Freddie Mac, and the Federal Home Loan Bank system and really what we're gonna be looking to do is advance agency priorities related to what we term responsible innovation. So, that's not innovation for innovation's sake, It's innovation, that considers strong risk management and controls and governance along the way. It's very important to us. Some of the trends that we've been seeing out there, in the mortgage space is that there's been innovation, there's been financial technology breakthroughs out there, but in bits and in pieces and overall mortgage processes remain pretty long for the borrower. You know we've seen today a lot of great presentations on average, about 45 days from origination to close, for a bar to go all the way through that process. 30 or so touch points along the way with different stakeholders from the processor to the underwriter to the closer and multiple asks for documentation along the way. So, what we're gonna do with the office is try to understand that better, deeper, and be able to share that information with our stakeholders internally in the agency, with our policy shop, with our regulatory shop, and of course inform FHFA Leadership Shop. Certainly I'd be remiss without saying that, we do recognize that our regulated entities are engaging in innovation in various ways. Just some examples. Day one, certainty, the validation, services of Fannie Mae, Freddie Mac's AIM product or asset and income modeler. those are just a couple. And so, you know we're gonna look to be working with our regulated entities to pursue opportunities, again in a responsible, safe and sale manner. So, those are just some of the, some of the high points around our objectives for the new office.
Jeremy Sicklick (
Thanks so much, Ann Marie. And one other question. I know in your recently, you guys had a paper that, came out that listed seven technologies that are changing housing finance. And two of the big themes that were highlighted were big data and data science. Would love to hear how you're thinking about the innovation and the risks around those.
Anne Marie Pippin (
And certainly we're definitely seeing that big data is real. And we're seeing, you know a lot of stakeholders within the mortgage process utilize various aspects of big data, be it, structured and unstructured data. something that's very important to us is to understand, how data is being utilized, how data is being shared, how data is being used as an input into models, especially complex models like artificial intelligence and machine learning type models that we're also seeing emerging more and more. We're not just going to be focusing in on one particular type of technology versus another with a new office. So, we're looking very broadly, We're gonna be looking at blockchain, we're gonna be looking at distributed ledger technology and smart contracts. We're gonna be looking at the ways that these types of technologies may disrupt existing processes and really try to understand that from a regulatory standpoint.
Jeremy Sicklick (
Thank you. All right. For the second topic, we're gonna move from big data to big tech. By big tech. When we talk about big tech, we're thinking about the Googles, the Facebooks, the Amazons. And so, you know as housing and, and real estate is one of the largest industries in the US and the world, if not the largest, you know it's up there. It's basically real estate and healthcare. Those are the biggest drivers of GDP. And so if you're at Amazon or Google, of course, when you look for opportunity, this is the big market. And so I'm curious, as we think through big tech firms coming into real estate and think about, are they partners? Are they competitors? Are they a bit of both? I know Bill, you have some experience recently partnering when Restify partnered with Google. would love to hear about that and how you're thinking about partnering and what that looks like.
Bill Elderton (
Yeah, I appreciate that, Jeremy. So, at Restify we have a partnership that we helped get off the ground with Google. In fact, their L D A I initiative, which is their loan doc, AI will be like automated intelligence or, some other form of the AI acronym and really we helped first by feeding them, a lot of data so that they could start to train the models. And we focused on going after the document types and the document sets that are necessary for a primary mortgage. It's not the end all be all, but we, we thought that was a great starting point. And so with a company like Google, as powerful as they are, we had to be careful about giving them too many ideas unless they take that away from us and build it themselves. And so I look at them as a little bit of, you know cooperation as I think people like to say, they're certainly a great partner of ours and they've helped us get those models built and start to improve the efficacy of the data that we can extract from documents as well as their OCR technologies. And we don't see them so much as a, as a competitor in that space from a, we're gonna take over the POS space or we're gonna start offering home loans or anything like that. But what I do think that the big tech companies like an AWS are like a Google are gonna get involved, is they're gonna start to see that their data that they have access to because of who they are and because of the outreach that they have, and a lot of data flows through their systems, if they're going to start becoming a data provider for some of the fintechs, some of the startups, some of the lending institutions. And so I think that's an area that hasn't been explored yet, that I think all of us collectively, especially FHFA in your new, in your new role, is to look at how data flow and what systems it flows through, who has the rights to be able to reuse that data. And, I'm not talking necessarily about PII data, but can Google influence, type data that that comes out because of the amount that they have access to just because it's flowing through a system that happens to be running on GCP? AWS is in a very similar situation. I don't know if a lot of people are aware of this, but AWS has, an entire division devoted towards FinTech and financial services that they don't talk about. They don't market, they don't put out in the public domain. But when you're a FinTech partner or when, when you're a lending institution or another financial institution, you have access to a whole level of AWS that doesn't exist for, for the common user, doesn't realize they're there and it's something that they sign, make you sign NDAs to get detailed access of what AWS can do. But it's the same type of thing where all the data is flowing through these cloud based providers and these cloud based solutions, all of the fintechs and all the startups are all cloud based, and most of them are going with AWS from a startup perspective. And so, watching that data flow through those systems is a great thing because you can utilize the security, you can utilize the firewall systems, you can utilize the routers and such that AWS has established over several years of being a leader provider. However, you have to watch your rights and your contract with an AWS or Google to make sure that you're not actually giving them implicit or explicit rights to that data that's flowing through the system.
Jeremy Sicklick (
I mean, you raise a great point. Most of us who built tech companies could have never built them at the scales that we have without an AWS or a Google Cloud. I mean, they are what really enabled a lot of the innovation to happen at real scale. Derek, you're a direct lender at, at Lower, Curious how you're working with Big Tech.
Derek Ochs (
Yes. So, we are one of those financial institutions partnering with Amazon So yeah, we know about that. I totally agree with what Bill was saying. I think like, when, when we look from a lender perspective, I don't think that like, Amazon, Google are direct, like are gonna be good direct competitors to ours anytime soon. Because if you think about like where their core competencies are, its scalability, it's intellectual property in a lot of like their technologies and you know right now, like most lenders, scalability is not the issue. Like the ability to scale up to millions and millions of users is not like the challenge they have, but the challenge that they do have is sometimes, things like computer vision, machine learning, those things where it would take a typical company, you know maybe hundreds of millions of dollars to develop that kind of technology. And even a small business can get access to that now through Amazon, through Google, through some of these big, tech companies. So, in some ways I think it democratizes it a little bit and it allows anybody to play in that field but at the same time, I totally agree that they do have access to a mountain of data and, that's kind of untapped from the financial perspective now. I think the reason they won't be super competitive anytime soon in like the direct lending space is, the challenges in our space are, a lot of them are, are bus complicated business rules and things like that. Those are not the things big tech is good at. That's not because it doesn't scale, it's a very personal, customized solution for a lot of different scenarios and you just don't see a lot of these companies jumping into those kind of products. If you look at the types of products that they, they thrive at, it's things that are reproducible and scalable at a very large scale. So if you're, Amazon, you create a marketplace, the shopping experience is gonna be the same for most people. And that's something you can scale and scale and scale, scale financial sector's a little different. Cause it's so specialized. There's so many different players, so many different niche markets. It's highly regulated. There's just a lot of different, a lot of different pieces that make it make it a little bit different, I think. but yeah, I agree. Like it's in the point solutions I think, where it could get a little more competitive because, you know Google might say, Hey, look, I already have some pre-built models here that I've figured out that, I know how to like do, scanning of a W2 and get all that data out and have a really good model for that. So I'm gonna sell package that up and sell it as a solution. But once again, I think that the way that our vendors will, play against that is the integrations that custom integrations into all these solutions like Encompass and things like that are out there. And that's something that I would assume that Google and Amazon might be reluctant to do.
Bill Elderton (
Yeah. Just one, one quick thing to pile on, Derek. I think it's a great point. If you look at the trend that's happened over at least the last 10 years, companies like LinkedIn and Facebook have started down the path that could require specialization, but then have then open source those solutions to the rest of the community. So, I think about things like Kafka and React Native that were developed by LinkedIn and Facebook, and then they open source those to the rest of the technology community to utilize those, just to avoid what you're talking about, which is going into that specialized arena. They think, you know they contribute more to the society in general through the technology innovation and then making that available to the rest of us.
Jeremy Sicklick (
So let's now talk about some of major applications of big data. you know as we think about big data and what it's done, one of the best examples is the ability to value your home. And, what rent you and the Open Door team have done of creating a whole category of I buying instant buying and, talk us through how you guys are using big data to look at understand properties and get to these really fast transaction speeds.
Rhett Damon (
Yeah, sure. you know it's hard because what we are doing, and constantly refining and, and working on all hours of the, of the day is to be able to generate, you know an attractive offer on a seller's home within seconds from the top of a button, something to get the conversation started, right? We have to be accurate enough from a phone, right? Someone, you know communicating with us through a mobile device, to say they want more, they want more information. We're, in the ballpark. And of course, as we've seen even in this macro where home prices just took a dive, it's tough. And so that is where I think, you know Open Door started, back in 2014, a bunch of young kids in San Francisco in their garage trying to figure out how to like buy and sell homes from your phone. But then the models are living, breathing and constantly being tweaked and kind of redesigned. And as we expand, what's interesting is, and it's really fun, you know currently open doors in now 51 markets, when I started a little under two years ago, we were in 2022 markets. So, as we go into a market, it's all heads down trying to understand and learn property values. and we go back as far as we can, right, because it's not just, you know the last five years, the last eight years, the last 10 years. It's as far back as we can go because I think a lot of us who build models understand, you know and a lot of us who have, are old enough to have lived through lots of different cycles. It's important to, to get as much as you can and to train and learn on that, and to constantly, evaluate. So, being part of the expansion team and going into new markets and, you know really evaluating, a Chicago or evaluating a DC or evaluating a, you know a Charlotte, those are really fun, fascinating types of, exercises with lots of data scientists and lots of data that's moving within the ecosystem that's all truly designed to be able to very quickly, offer an accurate home valuation to a seller and that's really, that's kind of the DNA of Open Door at its core.
Jeremy Sicklick (
That's great. I mean, you guys have had a massive impact in creating this whole category that's become a way, a great way for sellers to really be able to have an option to quickly.
Rhett Damon (
Yeah. That's really all it is. It's an option. And, you know I buying still represents a very small percentage of overall transactions. I mean around one and a half percent. So it's a small piece, but it is a solution and an option that solves and fills a need for a lot of home sellers. And then as, as we then, you know iterate on that, we become the principal owner of the home that the seller sells to us, and then we resell that home. We don't, you know we're not in the business of holding, buying and holding. We're in the b business of buying and reselling. And so just equally, you know those valuations and that modeling isn't just about, Hey, what should we offer in the second that that seller says, Hey, what will you offer on my home? We're thinking, what is, what's happening in this market when we are gonna be selling it in a short period of time that maybe, you know weeks, months, but it's not a long period of time. So that's, that's the business that's, that's difficult, but it's also fascinating.
Jeremy Sicklick (
And, and Bill and Derek, just to pile on here, I'm curious, as you look at big data, how do you, how far along are we in terms of using this data for instant loan approvals or getting to, you know this view of instant certainty in your businesses.
Derek Ochs (
I mean, I think, yeah, instant loan approvals are pretty, still pretty far off. There's a, there's a lot of things, in the way there that actually, have a decent time span, you know an appraisal, we can do digital appraisals, but we can't do them, you know finalize them a lot of times. So, there are things that are always gonna take a little bit, maybe not always for the time being gonna take a little bit of time, but definitely, I don't know if it's big data speeding it up, but I think it's the ability to, automate processes that were very difficult to automate before, open APIs, really help a lot with that. One of the first things we did, when, when we kind of build up our engineering team was to develop an open api and now we can offer mortgages as a service. So, people that want to add mortgage products, they can do that. They integrate with our api, so that's like super important. on the, on the data side, more data science. I think analytics is becoming a must have for every business now and not just, Hey, I wanna see a financial report at the end of the quarter. It's, I need to see exactly what my customers are doing at any given point in time, both internal and external customers. I need to know every point along alone how long each stage of this of the loan processing is gonna take. I need to know which divisions are slower than others. I need to know when my user comes in through my onboarding process exactly where is the drop off. So, if you don't have a good funnel graph that can kind of see where your users are dropping off, that's all analytics and it takes time and it takes the data itself. One of the things that we found is we just didn't have the data. So first project is like start to make our systems what we call more observable. So our systems are reporting more and more data back that we can actually crunch and turn into usable data. Even at the high level, executives, need to be do more than just the overall reports they need to be able to drill down. And, that sometimes requires a grain of data that you don't often have. Like, you know if I see that there's a loss in a certain area, the first thing I'm wanna do is to quickly go in and see which loans are giving the loss. And you need to see the whole way down to that loans and the loan officer, so you can go talk to them and see what's happening there. So, analytics becoming super important and you know we've dressed it up with fancy data science, word, but it really is still, it's just another level of analytics and people that really know what they're doing on that side. So, most, I think lenders are starting to at least think about having data teams. We have one at lower. I think we had it before we even built the engineering team. it's that important to our business. We have to see what's going on. It's what gives you visibility into your business so.
Bill Elderton (
Yeah, I think all those are fantastic points. I mean, I look at this maybe slightly differently, not spending the last 15 years of my career in the mortgage space, in that I think the mortgage industry in general is behind the rest of the technology space in adopting the data. So If I just think about, you know my closing package is three inches thick, how much information is contained within that closing package and how much of that is actually used by the different vendors along the process flow? I would guess that it's less, less than 1% less, that'd just be a guess, right? But, I look at the transformation that's occurred in other industries that are highly regulated. Just take the banking industry, you know five years ago no one was paying bills online and now nobody pays bills through the mail anymore, right? So, is the mortgage industry in general ready for that type of transformation? The iPhone love it or hate it, you know 11 years ago completely changed the way that that people do just about every activity in their life. And is this industry keeping up with that type of digital transformation beyond just taking an application at submission and making that convenient? What else are we doing? I don't know of any lenders that are currently using geolocation information as part of the, part of the approval process. Whether that makes sense or not, there's a rem of data that's available to us today that wasn't available to us five years ago that wasn't available to us 10 years ago. And whether or not we're actively seeking to use that as part of our decisioning process or part of our regulatory process is I think a gap in the industry today. And, and I look this gap to be closed and have the same transformation that the financial industry, mainly banks and credit unions went through. but are we ready for a five year quick turn on that, like they were, or are we gonna fight it the whole way around? and that's what I think's gonna be really interesting over the next three to four years.
Jeremy Sicklick (
Fantastic. Well, we have about 90 seconds left and so maybe ask each of you just really quickly, 20, 30 seconds as you think about each of us are thinking about how we use data to get more customers and customer acquisition, maybe really quickly how you're thinking about using data and how you do for driving customer acquisition and growth. Maybe start with you Ret.
Rhett Damon (
Yeah, I think for us it's a little different. it's through partnerships within, for example, Zillow, that was recently announced. you know that that is a, a very powerful partnership, but there's also a lot of information that we can now share, that I think is important. And also we have similar partnerships with realtor.com and other brokerage and other portals. So we really look at, because their audiences are large and they allow us to kind of better understand consumers, it's all gonna be what consumers are demanding and what they want. and so we're looking at how we can serve the consumer and the information we can learn about the consumer through a lot of our strategic partnerships.
Bill Elderton (
So one of the things I've heard a couple of times today and one of the sessions I heard that the average loan cost is now up to around $10,000, to process a loan. How can data help that while obviously on the qualification of an applicant and how quickly they can move through the process, or who has to be put off into a special category to be done manually? That's, a big space that I don't think the lead qualification has been fully utilized from a data perspective. And then I also look at, things like, the amount of time spent in processing. So all of this movement around document and data extraction and validation of that document for it to go into underwriting and for it to go into the processor space, how much of that can be pre-qualified or preapproved if you will, before getting there so it becomes more of a checkbox and less of a 20 hours or 60 hours or 150 hours spent on actually processing the loan.
Derek Ochs (
For us, I would say aside from like getting really good visibility into how our users are interacting with the system, I think, advertising analytics, you can't really talk about data and bringing customers in without advertising analytics. So, we basically take, when customers come in, we can take, it comes in from a Google ad, usually, we could take and score that user and we can send that back to Google and say, this particular user gets this score, meaning they're highly likely to convert into a loan, give us more users like that. And Google will match that with profiles of other users and they'll start targeting those users with ads, when they come in. Super important with driving growth.
Jeremy Sicklick (
Well, Anne, Marie and Ret and Bill and Derek, Thanks so much for this. Really appreciate the insights. Thanks everyone.