Black home buyers are approved at essentially the same rate for mortgages today that they were in 1990. And for Native Americans, the picture has gotten much worse.
This is Gated Communities, where we talk about everything you’re not supposed to talk about in the mortgage industry. It’s time to learn how to play fair. FairPlay, the world’s first fairness as a service solution, released a report on the state of mortgage fairness in the US.
The research identifies some troubling trends, like the fact mortgage fairness for black people hasn’t improved in the past 30 years. For Native Americans, mortgage fairness has declined by more than 10%. However, female and Hispanic mortgage applicants experienced a steady increase in mortgage fairness over the same period.
Today, we have founder and CEO of FairPlay, Kareem Saleh, to discuss the alarming findings within the report and how AI could help eliminate bias in our current system. Well, thank you, Kareem, for joining us today on Gated Communities. To get started a little bit, I mean, I’ve introduced you to our audience before we do the recording, but in your own words, can you kind of explain what FairPlay is?
How FairPlay Eliminating AI Bias in Lending
Yeah, thanks for having me, Katie. Delighted to be here. FairPlay is the world’s first fairness as a service company.
We allow anybody using an algorithm to make a high stakes decision about someone’s life to answer five questions. Is my algorithm fair? If not, why not?
If not, could it be fairer? What’s the economic impact to our business of being fairer? And finally, did we give our declines, the folks we rejected a second look to make sure that we didn’t say no to somebody we ought to have approved?
Our customers are primarily financial institutions who want the economic, reputational and regulatory benefits of being fair.
And explain this report that came out recently, that kind of measures fairness and had some unique observations about the current market and how it’s playing out today.
The History of Lending Bias in America
Yeah, well, so Katie, as you may know, we have an unfortunate history of redlining in America. And as a consequence of that legacy, Congress passed something called the Home Mortgage Disclosure Act, which requires every mortgage originator in the country to submit certain loan level data to the government every year so that the public can understand if mortgage originators are engaging in redlining or otherwise excluding protected groups from the housing market. And so every year, mortgage originators are required to submit certain data to the government about their lending practices.
And in theory, that should allow us to construct a picture of mortgage fairness in America. And that’s what we did at FairPlay. We went back and we downloaded all of the loan level data in the Home Mortgage Disclosure Act database going all the way back to 1990, all the way down to the census tract level, and then applied a standard test for discrimination to understand what the state of mortgage fairness is today and how it’s changed over the last 30 plus years.
And the results, particularly for Black Americans and Native Americans were really astounding. We found that over the course of about 30 years, Black home buyers are approved at essentially the same rate for mortgages today that they were in 1990. And for Native Americans, the picture has gotten much worse.
In 1990, Native American home buyers tended to be approved for mortgages at 95% the rate of White home buyers. And today, Native American home buyers are approved at 80% the rate of White home buyers. So a decline in mortgage fairness of 15 percentage points for Native American home buyers over the last 30 years.
Yeah, this was definitely a very interesting report. It showed how much progress we have made thus far. There were some nice findings in terms of women achieving some fairness.
Fairness Metrics for Disparity Testing
But interestingly enough, are the findings that we found with Native Americans, which we have not gotten to talk about on this podcast before. So before we delve into these specific findings, I want to just clarify for the audience what an adverse impact ratio is, because the report measures fairness by adverse impact ratio. So can you briefly break down for our audience what that is and how that works before we get into this conversation?
Yeah, sure. The adverse impact ratio is a common test for disparity applied by courts and regulators to understand if one group experiences a positive outcome for something like approval for a mortgage at a higher or lower rate than another group. So for example, in the housing market or the mortgage market, when we apply the adverse impact ratio, we apply it to understand, for example, at what rate women are approved for mortgages relative to men or at what rate black home buyers are approved for mortgages relative to white home buyers.
The courts and regulators have never articulated what adverse impact ratio threshold they consider to be fair. But over the years, there’s been a kind of emerging consensus between industry, government and consumer advocacy groups that if you’re approving one group, let’s say women, at a rate that is at least 90% the rate of the control group, in this case men, then there is a presumption of fairness. If you’re approving women at between 80 to 90% the rate of men, that is starts to be the indication of a disparity between groups and generally folks falling between that 80 to 90% approval rate ratio for the protected group are encouraged to see if they can do better.
If you’re approving a protected group at below 80% relative to the control group, that starts to be a sign of a practically significant disparity. And regulators and courts will often inquire further to see if that difference in outcome is justified by some other legitimate objective. And if it’s not, it can lead to a finding of discrimination or disparate impact.
Do you suggest that CEOs should do this with their own kind of inventory and what kind of applicants that they’re approving and that they’re disapproving themselves to use this formula?
I think we’re headed for a world in which many different kinds of high stakes decisions are evaluated for fairness. And adverse impact ratio is a good first test to understand if one group experiences a different outcome than another group at a practically significant rate. So I do expect, and in many cases in financial services, but as well as housing, there are a number of decisions that have to be made fairly, including marketing decisions, including identity verification decisions, including underwriting and pricing decisions, and sometimes loan modification, collections, other account management decisions, all of which must be made fairly, and average impact ratio is a good measure.
It’s a good place to start the inquiry, let me say.
Absolutely. And I think it’s a good idea to catch yourself before regulators catch you. So whatever you can do to that extent to monitor yourself and what your company is doing would be proactive in that term.
Consumer Demands for Fairness in Lending
I think that’s right, we’re moving towards a world where the expectation is both not only by regulators but also by the consumers, especially the Gen Z and millennial consumers who make up the future of the housing market. There’s an expectation that lenders, landlords, other players in the ecosystem are going to be proactive, not reactive on fairness. And as you said, it’s really better to have a testing regime in place so that if and when you get the question, you can say, hey, we take this stuff seriously, we inquire into it rigorously, and when we find issues, we commit ourselves to fixing them.
Absolutely, good. So the report tells us, well, the main contention for the report is that mortgage fairness is no better today than it was in 1990. And that’s true for multiple groups.
So why don’t you start by explaining some of the factors that have gone into that and why we have seen that as a consequence? Why haven’t we seen much progress since then?
I think it’s largely because we’re relying on the same tools like conventional credit scores to qualify people for housing. And the world has changed quite a bit since we mandated the use of those scores 35 years ago in the housing market. And you’re starting to see even the agencies admit that there are probably better measures of a person’s ability and willingness to repay a mortgage than just a conventional credit score that was developed using mathematics and variables from 35 to 50 years ago.
Alternative Data to Combat AI Bias in Lending
So I think you’re starting to see increasingly the use of cashflow underwriting data. You’re starting to see the use of alternative data, which gets much more to a consumer’s credit trajectory than it does analyzing them at any one moment in time. And I think we’re starting to see the use of new algorithmic fairness techniques, which have as their animating purpose, the goal of doing a better job of underwriting populations that are not well represented in the data.
For example, populations that have been historically discriminated against, or folks who are newly arrived immigrants that may not have as long of a credit history, or young people who are just approaching the financial system and the credit markets for the first time. So I think ultimately when underwriting a person for a mortgage, you’re really trying to make an assessment about whether or not that person is responsible. And it turns out that you can find indicia of a person’s responsibility in all kinds of places that aren’t just a conventional credit score.
Have there been any policy moves so far that you support that kind of expand on the criteria we can use to evaluate someone’s responsibility, for example, using someone’s rental history to contribute to their credit?
Yeah, I think that, you know, there is this whole world of alternative data, including things like, you know, did you pay your rent on time? Do you pay your utility bills on time? To the extent that you’ve had credit in the past, do you make a point of using it responsibly?
There is this whole new exciting world of alternative data that can allow us to paint a much finer portrait of a borrower’s ability and willingness to repay a loan. Of course, we have to make sure that those data elements are used responsibly and in ways that are privacy preserving because, you know, technology can be a double-edged sword. And so in order to make sure that these alternative data sources and new AI underwriting techniques are doing the greatest good for the greatest number, the users of those systems also have to have governance mechanisms in place to ensure that they’re harnessed for the right purposes.
Absolutely. Absolutely. Do you think in terms of using alternative data that you need to do it under specific types of loans or can you do that with any type of loan, like a conventional loan?
You can use alternative data.
I think alternative data has a role almost in every credit product that we see. It tends to be the most supportive of inclusion and fairness for populations that have credit bureau histories or credit bureau files that are either messy, missing or wrong. So we typically see the greatest lift for consumers who maybe have thin credit files, no credit files, or who’ve had some kind of credit event in their past, like a bankruptcy or a foreclosure, that they’re now seeking to get past as part of their kind of return to the financial system.
The Role of Policy in Lending Fairness
Okay, great. And I wanna hit on one point that I found interesting. In 2021, we had the fairest mortgage market since the 2008 recession.
So I wanna let you get into why that is. And perhaps, did we do anything right, or was this just a consequence of the pandemic and odd circumstances?
Yeah, I think there’s no question that there were a set of government support measures like stimulus payments, but also foreclosure moratoriums that probably did a lot to support the housing market in particular under very unprecedented economic circumstances. And so, I think it’s a combination of both massive government wealth transfers, plus reduced opportunities for spending during the pandemic, plus a set of other policy measures that were taken to support keeping folks in their homes that ultimately led to a very fair mortgage market in 2020 and 2021, even though of course we had dramatic drops in GDP and dramatic spikes in unemployment. The real question is now that some of those government support programs are starting to wind down, what’s the outlook?
And as we start to revert to a much more normal macroeconomic situation or a much more typical housing market, are we going to give back those fairness gains? And I worry that perhaps we might, one of the things that we looked at in our study was what happens to mortgage fairness rates when interest rates rise? And what we found was that when interest rates rise to the current levels that they’re at somewhere north of 6% or between 6% and 7%, that historically at those interest rate levels, mortgage fairness has dropped almost back towards 2008 crisis levels.
So that the last time interest rates were this high, mortgage fairness for black applicants, which was 84% adverse impact ratio in 2021, could drop as low as 68% given some of the rise in interest rates and the other issues with housing affordability that we have, right? That we’re experiencing now and that appear to be on the horizon.
Can you explain that a little bit? How exactly a rising interest rate would disproportionately impact black Americans and other?
Yeah, well, as the cost of capital gets higher, it means that folks who are already kind of at the margin of the mortgage market. So, you know, folks who might be tipping up against that 43% debt to income cap that exists on conventional mortgages, for example. You know, with the cost of capital, with interest rates rising, right?
It makes it more likely for those folks who are just on the kind of right side of mortgage eligibility to slip below, you know, the common approval thresholds. And so I think that that’s, you know, we spent a lot of time getting folks, especially from low and moderate income neighborhoods or majority minority neighborhoods, mortgage ready the last several years. And, but when the cost of capital rises like this, that mortgage readiness degrades and it hits those populations hardest.
Mortgage Fairness and Native Americans
Absolutely. Okay, I see. And there was a speaking of an alarming drop in mortgage fairness.
There was an alarming drop in for Native Americans. And now their adverse impact ratio is below black applicants, which people of this podcast know is a historically redlined group. So it’s interesting to see this reverse and progress.
What do you think is happening there?
I think it’s a very complicated set of factors. As I’m sure you’re aware, the Native American community has been under a terrible amount of socioeconomic pressure the last 30 years. I’m sure that that has contributed to their mortgage readiness as a community.
I have also heard some mortgage lenders say that it’s hard to obtain security interests on properties that are located on Native American reservations. And as a result, some mortgage lenders are more cautious about lending to Native American home buyers because they worry that their security interests in those homes may not be able to be perfected. And so I think we probably need more investigation into the many complicated set of factors that appear to be really working against Native American home buyers in particular, and whether or not there are some adjustments perhaps that ought to be made to make the mortgage market more available to credit worthy home buyers in that community.
Absolutely. I definitely found that really interesting. And I want to get someone on the podcast who can speak to that a bit more.
I might have some information on that. But another thing that I found interesting was how fairness varies by geographic location. Now, I know that some locations in the United States might have some stereotypes along with them.
But nonetheless, it was interesting to see and to pinpoint where exactly fairness is rising and where it’s dipping. So fairness for black applicants appears to get worse in regions where black homebuyers have the highest percentage of the local population. The same thing for Native Americans.
Lending Disparity by Geographic Location
In many counties, this was in the South and the Great Plains. So why exactly? I’m hoping you can kind of maybe either give your opinion or what you think is happening in those locations to make it disproportionately more unfair towards them.
Yeah, it’s a great question, Katie. So, just to summarize the finding, what we found was that we basically clustered states into one of three categories. Let’s start with black homebuyers.
So we found that black homebuyers, basically the mortgage market could be broadly bucketed into the three categories of states for black homebuyers. The first set of states was Approaching Fair. And the second set of states was Unfair.
And the third bucket of states was Really Unfair. And so what we found was in places that were experiencing high levels of GDP growth, and that had moderately sized, relatively wealthier black populations, the mortgage market was Approaching Fair. Now that is to say, black homebuyers were not exactly approved at parity with white homebuyers, but they had adverse impact ratios in the high 80s.
On the flip side, we found that there were about six states in the deep South and it didn’t matter how good the macroeconomic environment was. It didn’t matter how low unemployment was. It didn’t matter effectively what was happening in the broader economy.
Arkansas, South Carolina, Louisiana, Mississippi, and a handful of others were persistently unfair to black home buyers. That is to say that, and those tended to be states that had relatively large black populations and where the median income was about 13 to $14,000 less than that group of states that were approaching fair in the mortgage market for black home buyers. So it feels to me like there is some ongoing residual discrimination from parts of the country that had a legacy of slave owning that has resulted in black home buyers being approved at persistently low rates even today in the 2020s.
And do you think this is kind of residual discrimination or more so redlining that has happened throughout the generation so that it’s more systematic?
Implicit Bias in AI Lending Models
My guess is that it’s very little to do with humans deliberately discriminating and much more to do with new algorithmic decisioning processes that rely on data from the past to make their judgments. And that data from the past has encoded things like redlining and other measures of exclusion for black home buyers in particular in the deep South. And so, I think we have unfortunately a history of discrimination and the machines that are trained on data from the past are learning that discrimination and embedding it into the digital decisions that they’re making today.
And so I think that as an industry and as a society, one of the things that we’re gonna have to grapple with as these algorithmic systems take over higher and higher stakes decisions in people’s lives is how are we gonna correct for that data bias so that we ensure everyone’s got a fair shot at the American dream.
Don’t miss the largest regional mortgage show in the nation. The New England Mortgage Expo returns to Mohegan Sun in Connecticut, January 12th and 13th. See us at www.anymortgageexpo.com.
Start your year with the best connections in the industry, dozens of sessions, scores of exhibitors. It’s where success is written every hour. www.anymortgageexpo.com.
Absolutely. And I think that’s an important point, that it’s not necessarily the humans in these locations, but it just has a deeper history of redlining, a darker history. And unfortunately, algorithms and data and the standards that we set a long time ago carry that forward to newer generations.
So it’s very hard, and it will take a lot of innovation to kind of revamp that and correct those mistakes and those data biases. So for Native Americans, it was also kind of a similar thing where you have, most Native Americans that were impacted by unfairness were in locations with the highest percentage of Native Americans. And also that a lot of these, they were also in areas where, in areas of higher mortgage unfairness, were largely coincident with the Navajo and I think it’s Hopi reservations in those states.
So I thought that was quite interesting. Do you think that that has any effect on the unfairness that we’re seeing? Is it just that they are present in those locations or that people, maybe it’s a discrimination thing where people see them as that in their locations and therefore think they’re not worthy?
Yeah, I mean, if I had to guess, I would say there’s likely a very similar dynamic playing out with Native Americans as is experienced by Black Americans, which is, you have reservations which have been systematically under-invested in, which are in some sense, you know, legally apart from the communities that surround them. And as a result, you have islands of economic disadvantage in the form of these reservations where lenders are for various reasons reluctant to do business. And so, you know, it’s kind of the same observation that we had for black home buyers.
You know, there are states with large concentrations of Native American applicants, like New Mexico, like Mississippi, like North Carolina, like Arizona, where no matter how good the macroeconomic situation has been, fairness for Native American home buyers has been declining, it’s been on a downward trend. And so, you know, I think it’s worth considering as part of our broader domestic policy on Native American communities and tribal lands, evaluating whether or not there might be measures like special purpose credit programs, second look underwriting programs, other forms of credit assessment that might allow more worthy Native American applicants to qualify for homes.
And I think the relation between them there is that these areas are a little bit more segregated. So you have obviously redlined neighborhoods in the south, which affects black applicants. They’re stuck in certain locations, they’re kind of cornered into them a bit.
So they’re less integrated with the rest of the population. And then with these Native American reservations, it’s a similar thing where they have their territory and then there’s everybody else’s. And that kind of perpetuates this thing of there’s them and us.
So it plays a bit into psychology of, are you like me? Should I measure us the same? And how do I know that you are responsible?
And you have your own culture, your own way of being, especially with a Native American reservation, although there may be a lot of similarities, they have a different way of life. So it’s hard to measure us to the same standard. So what do you think of that and basically how segregation plays into this?
My guess is that you’re right, Katie. And in fact, there’s some evidence from the outcomes for Hispanic homebuyers that tends to suggest that you’re right. So for example, one of the interesting features of the Hispanic, the findings from the mortgage fairness monitor for Hispanic homebuyers is that the more Hispanic your neighborhood, whether it’s South Florida or South California, the fairer the mortgage market is to Hispanics.
And I suspect that that is both because of the upward economic mobility that the Hispanic community has exhibited over the last 30 years, but also that there are more likely to be mortgage brokers and mortgage originators owned by Hispanic proprietors serving those communities. And as a result, who have a greater sensitivity to being able to underwrite loans from applicants who have Hispanic backgrounds. And so, I think that what ultimately, what we’re finding is that it really helps to be a mortgage originator kind of steeped in the credit characteristics and the other cultural characteristics of the community you serve, because it may appear as any way from the data to make you more sensitive to the credit worthiness characteristics of the individuals in those neighborhoods.
Absolutely. So I think, yeah, in those neighborhoods, if you happen to be an LO or someone working in the mortgage industry who lives in those areas or those states, maybe that’s something just to be cognizant of. That they’re not getting as much attention as they should.
They are worthy of a home in many cases. Just because they live within a specific area or they have a specific income does not mean that they’re not ready to buy a home necessarily. And moving on to the next group, a little bit more positive, Fairness for Females is now on par with that of the control group and their adverse impact ratio is 99.2%.
Lending Fairness for Women
So what do you think has contributed to this process and how can we emulate this to support other groups?
Yeah, you’re right. One of the few silver linings of our study was the finding that mortgage fairness for women has increased somewhat over the last 30 years. In the early 1990s, women were approved at about 92% the rate of men for mortgages.
And in 2021, women were approved at more than 99% of the rate of men. So almost perfect parity. And I think that, you know, my guess is that that is largely attributable to rising wages for women relative to men, rising credit scores, rising rates of employment for women relative to men.
But I think that, you know, if you look at certain subpopulations of women, so for example, black women, Native American women, et cetera, we still have a lot of progress to be made. And so, you know, I’m very happy to see that women in the aggregate do encounter a fairer mortgage market today than they once did. My guess is that that says something about the upward economic mobility of women over the last 30 years in our country.
But not all women have gotten to rates where they’re approved at parity. And so we do still have some work to do for certain subpopulations of women who still encounter denial rates that are high in the mortgage market.
That’s right. The report does point out that there are some areas in the South and the Great Plains, more likely in rural areas than urban areas, that women are seeing unfairness with their mortgage applications. And I think that’s exactly right what you said about women getting more financial independence and kind of elevating themselves and their careers throughout the decades.
Also, the fact that in marriage, even if you do make less than your husband, you’re more likely to control the finances nowadays. So it’s something that advertisers, I think, realized a while ago when they started targeting their ads towards women, you know, in their commercials and everything like that. They go, well, women have the power of the purse, so he’s not going to buy it unless she says it’s okay.
And they start advertising, you know, all these products towards women. And now it’s kind of the same with home buying, and you have to be really cognizant of what the woman is thinking, what the female in the relationship is thinking. And, you know, she’s going to decide what’s going to happen with the finances.
She also has a say. So being cognizant of that, and, you know, it could be a couple, and she’s filling out the mortgage application, you know. And also, you know, in marriage nowadays, because women have achieved financial independence, she might be the breadwinner.
So, you know, it’s really been a big mix up. But that is interesting to see how in other demographic groups, there is still progress to be made. It’s not necessarily viewed the same as white women who are achieving this.
So what do you think is happening there? If you have any opinion on that, is it just taking longer for people to realize that women in all sorts of demographics and situations have more financial power than they did back then?
I think women are an emergent economic force. Over the course of the last 30 years, the purchasing power and, as you point out, the ability to direct the finances of their family units may have increased. And so I suspect that a great deal of the gains that we see are owing to some of these cultural and other economic factors that have buoyed the economic status of women over the last few decades.
You do point out that there is still, however, a persistent gap in mortgage fairness between urban and rural areas. I was interested to find that that was the case, not only for women, but we see it for black homebuyers and others, too. Cities appear to be persistently fairer to protected homebuyers than rural areas.
And why do you think that is? Do you think there’s a cultural lag in some of these rural areas and where they might be?
Maybe this issue that we were talking about earlier, which is, you know, loan originators may be more sensitive to or more willing to go the extra mile for applicants who resemble themselves. And so if you are, you know, a member of a protected group that’s not well represented in a particular rural area, it may be that loan originators have a harder time, you know, meeting you with a product, a mortgage product that would be suitable to your needs.
Yeah, absolutely. And, you know, it’s a complicated issue. And for some of these things, we can only speculate why they happen.
How FairPlay Combats AI Bias in Lending
More research needs to be done. But nonetheless, these findings are really something to think about. Let’s get into FairPlay and what FairPlay does to achieve fairness.
So FairPlay uses advanced AI fairness techniques to reduce algorithmic bias for people of color, women and other historically disadvantaged groups. So can you walk me through some of the services that you offer Fairness Analysis, Second Look, Optimizer? Go ahead and give us a brief explanation for those services.
Yeah, so our product basically has two core modules, bias detection and fairness optimization. Bias detection basically asks, you know, is my algorithm fair and if not, why not? That’s the basic kind of fair lending testing and reporting that mortgage originators do and other consumer and small business finance companies do.
But the thing that we’re best known for is our fairness optimization module. And those are tools that allow financial institutions, landlords, insurance companies to leverage fairness in ways that enhance the profitability of their business. And we typically do that through one of two methods.
The first is by tuning your algorithm to be fairer in ways that preserve its accuracy. And the second way we do that is through second look credit underwriting programs, where, for example, before declining somebody that would not have ordinarily been approved, lenders send us all of their declined applications or applications that they’re about to decline for re-underwriting using a model that has been tuned to be fairer to populations that are not well represented in the data. And what we find is that something like 25 to 33% of the highest scoring people of color and women that would normally be declined for loans actually perform as well as the riskiest folks that most lenders approve.
So we like to say that second look underwriting is good for profits, good for people, and good for progress. It’s that rare occasion to that you can find an opportunity you might have overlooked by using a kind of fair second look.
So would you recommend some of these people in these geographic locations where unfairness comes at a higher rate to definitely look into using these services?
Yeah, I think that if you are in an area where you know that low and moderate income or majority minority applicants have struggled to get approved, it can be both economically worthwhile, but also great to do from a customer satisfaction perspective to re-underwrite your declines and make sure that you didn’t say no to somebody who would have paid you back. And as I said, we often find that if you… I’ll just give you one example.
So one of the variables that often gets used in credit underwriting is what is the consistency of the applicant’s employment? And if you think about it, consistency of employment is a perfectly reasonable variable on which to assess the credit worthiness of a man. But consistency of employment, all things being equal, is gonna have a discriminatory effect on women between the ages of, say, 18 and 45, who take time out of the workforce to start a family.
So maybe before you decline somebody for inconsistent employment, you ought to run a check first to see if they resemble good applicants on dimensions that you didn’t heavily consider. And so what we find is that oftentimes, an applicant might exhibit slightly inconsistent employment because they took time out of the workforce to start a family, but they exhibit great stability of residence. And the number of professional licenses that they’re accumulating is increasing and that they haven’t been aggressively seeking credit, so they don’t have a high number of inquiries.
They resemble good applicants on all of these other dimensions. And so maybe for some characteristics of borrowers, consistency of employment isn’t necessarily the best measure by which to assess their credit worthiness.
And you said this could increase profitability, and I think that’s explanatory, you know, when you explain these services, how that can increase profitability. But what about time and time management? Using your services, does that increase time in any way, make it longer to either approve or submit a denial for an application?
How does that affect the loan process? Does it affect the loan process?
Actually, our Second Look underwriting tool runs in something like 800 milliseconds or something. So it’s like literally adds less than a second to your underwriting process. And basically, you know, just before you decline someone, just route them through this other decisioning system real quick and see if maybe their riskiness wasn’t overstated by their traditional credit score or other conventional underwriting techniques.
And so we like to think that this is a very fast ability to check your work and find more good loans that also yield a regulatory and reputational benefit.
Absolutely. Because I think that’s kind of a drawback when people try to consider these things is, oh, is it going to add more time? Do I have to put in more effort, more work?
And no, you don’t. It literally takes less than a second to use this service and to double check that you are being fair. So your products, is this mainly for lenders?
Can MLOs use it? Can brokerages use it? Do appraisers use any of these services?
What do you offer for the wider industry? And give us a few specific examples of how some of these people can use it.
Yeah, so anybody who’s doing marketing, for example, can use it to make sure they’re reaching good applicants from majority minority neighborhoods, low moderate income neighborhoods, other historically disadvantaged groups. You can use it in the underwriting process to get more people approved, whether it’s through desktop underwriter or some of the other nonconventional underwriting programs. You can use it to increase the take rates by offering folks better pricing.
So make it more likely that any offers you generate actually get accepted. You can use it for account management decisions. So like line sizing and other customer relationship management objectives.
And then you can use it in the event that a borrower gets into trouble, right? By offering fair loan modifications and basically being able to meet that consumer where they’re at with terms that they are likely to succeed with and in ways that keep you from losing a lot of time and effort on an impaired asset. So we like to say that we offer fairness solutions for the entire customer journey, from marketing to identity verification to underwriting, to pricing, to line account management and even loss management.
And do you think that technology is, and technology, automation, AI, is the most expedient way to achieve fairness? Is this the path that we should continue on?
Well, I like to say, I think that technology has a big role to play, certainly. You know, I think that, you know, for many years, we have tried to achieve fairness either through blindness, by pretending like we can ignore differences between certain groups, and by keeping humans in the loop, you know, to make subjective judgments. And I think what we’re finding is that there are some things that machines are better at than people.
And if you want to save time and money and achieve more profitable, fairer outcomes, there is a role for technology to play. And you’re really at a disadvantage if you’re not leveraging a lot of these state-of-the-art methodologies to enhance your business, reach new customers, and tell the world or show the world that you’re acting in ways that are good for consumers and good for the communities you serve.
Absolutely. And like you said, this is a way to increase profitability, get more capital flowing in the country. The more homeowners we have, the better the economy is.
This all is free flowing. So thank you so much for coming on the podcast, for sharing your report with us and for telling us a bit about FairPlay. I think it’s really interesting the role that technology can play in making a fairer market.
So thank you so much for coming on and explaining that.
Thank you so much for having me, Katie.
Awesome. Wonderful.
This is Gated Communities, hosted by me, Katie Jensen, for the Mortgage News Network. All episodes are produced by TG. Kudem Peror and Matthew Mullins.
Our head of multimedia is Mike Savino, and our editor-in-chief is Christine Stewart. Make sure you subscribe to Gated Communities so you get future episodes, and be sure to rate and review it so others can find it. The song you heard at the beginning was Wild Side by Saint Society, and the song you hear now is Will You Dance With Me by La La Nia.
This podcast is copyrighted by American Business Media.
OCN is the largest producer of events for mortgage professionals. We bring the action to you. See when we’ll be in your area.
Just visit us at www.originatorconnectnetwork.com. That’s www.originatorconnectnetwork.com.
Learn More About Fairness-As-A-Service with FairPlay
If you’re ready to take the bias out of your AI lending decisions without complicating your current workflows, schedule a free demo of FairPlay today.
