Online Event Archive Recorded April 5, 2022
Learn how you can use this powerful tool, available for free to all NCRC members, to produce a report on mortgages, small business lending and bank branch networks for any city, county or metro area in the nation.
NCRC’s Director of Research Jason Richardson will walk us through the tool using data on the Detroit MSA, highlighting key findings along the way.
You’ll also hear from Phyllis Edwards about how Bridging Communities uses this type of data to hold lenders accountable and work toward a just economy locally.
If you are not currently an NCRC member, join now to gain access to the Fair Lending Tool, along with many other benefits.
Transcript:
0:00
So good afternoon, everyone. Thank you so much for joining us this afternoon. We are indeed excited for each and every one of you to be here. My name is Phyllis Edwards. I’m the Executive Director for bridging communities, and one of the CO leaders of the Detroit NCRC coalition. I have the pleasure this afternoon introducing you to our speaker, Mr. Jason Richards. Mr. Richard is the Senior Director of Research for NTRC. He is a geographer, who has done extensive research on mortgage finance, and its impact on people and communities. He has been with NCRC for seven years, and works with members helping them find data and building tools that support their work. The research team at NCRC also produces original research in support of the policy goals of NCRC and has published many reports are redlining, gentrification, and break bank branches loss. It is with indeed a pleasure to present a Psalm and to introduce the others. Our speaker for this afternoon and our facilitator, Mr. Jason Richards, is Jason, it’s on you.
1:12
Thank you, Phyllis. Thank you very much. Hi, everybody. My name is Jason Richardson. I’m director of research for the National Community Reinvestment Coalition. Today, we’re going to be walking through a tool that is available right now on the ncrc.org website for members of NCRC. little housekeeping here, if you want to take a look through that a little bit more about NCRC is listed here. But really, Phyllis did a great job, I think covering much of this. We’re also joined today, in the chat by Maggie with our membership team. If during our guard discussion, you have any questions about membership, feel free to shoot her a message there. And she can help out with that. Oops, sorry, I told Maggie earlier, I could run this, the, the the slideshow as well. And I knew I should just let her do it. Sorry about that, guys. So what I’m going to do is pull up the ncrc.org website. And if you are a member of NCRC, you can do this right now. And you should be able to sign in on our website. And here is what that would look like, there. So when you’re signed in, you can scroll down here a little bit. And you’re going to locate under Tools, the fair lending tool, this is part of a growing number of tools that we’re developing for members that let you log in and access data or information very quickly, that can help you with your work. When you click on the fair lending tool, it’ll take you to this page. If you’re familiar with Tableau, that’s a data visualization software that we use to create this tool. And we did that in partnership with the tableau Foundation. They help nonprofits, you know, kind of get on their feet data visual, you know, with data visualization and, and tools like this. If you have any questions about it, there’s a button at the top here, you can send us a message or you can just shoot me an email, I’ll make sure that I share that in the in the chat a little bit. You can also get more information about the data that we use for this tool by clicking on the definitions button. And finally, if you click on tutorial, it will take you to a recording of a webinar very similar to this one that we did a few months ago. But it’s always good to do a new one and to answer all the questions that we can from from members or prospective members alike. So today, we’re going to take a look at Detroit. And this is the the title page of the report here, the selection you need to make is the exact geography you want to look at. So today, I’m just going to open this this up and we’re going to type in Detroit. And in the background, what it’s doing is sifting through data from 2018 to 2020 mortgage data, and then putting it into 2020 2019 to 2021 Small Business Loan data and the bank branch location data from 2021 as well. That’s all of the data that’s used in this tool along with some census data, and then some custom data that we’ve created to help everything work together. So it takes a couple of seconds to update. And then if you click on the county button here, it’ll show you the exact counties that are included in this analysis. And you can uncheck individual counties if you want to. For example, if you want to focus just on Wayne County, that’s pretty easy to do. I’m going to leave all the counties check just for demonstration purposes. And then we’re going to click Next page. This takes you to the table of contents. Each image here is a separate analysis. And when you get experienced with the tool, it’s really simple to go in. Select your geography that you want to look at, and then go right to the type of analysis you want it The number of bank branches in the area or the amount of mortgage lending by a specific lender, you can use it that way. Today, though, we’re going to walk through each and every page. And I’m going to describe a little bit about what you’re seeing here. And we’re going to do a little bit of analysis. So this is a basic demographic overview, a little over 4.3 million people live in the area that we’ve selected, they’re spread across a little over 1200 neighborhoods, or census tracts. Median income information is here, as well as the income required to be considered low or moderate income. At the bottom, we have a demographic breakdown of the area population. And to the right of that we look at that a little different way. In the middle, we’re looking at the population in majority minority neighborhoods are those census tracts with, with where more than half of the population is a person of color, or its majority white, on the right, low to moderate income communities, we look at those and the racial demographics there.
6:08
Hold on one second. And now we’re going to go to the next page. And here we look at a mortgage lending overview. We’re looking at lending between 2018 and 2020. On the left hand side, it shows the percent of loans that were home purchase loans, versus those that were refinance, cash out refinance, Home Improvement loans, or home equity loans. What’s interesting is to look on the right and look at it year over year to see how these compare. What you’ll note at the top is that the number of home purchase loans are relatively even hold on one moment please.
6:54
Sorry about that I hit mute. The refinance lending is where it gets interesting. In 2018, you see a relatively small number of refinance loans, this dramatically increased in 2019, and 2020. This isn’t that different than what we see nationally right now. So that really isn’t surprising. But the growing share of the market dominated by refinance lending, you know, could lead to some other issues. We cover this a little more extensively in the report released last month on mortgage lending. And you can find that on ncrc.org as well. Now, let’s go to the next page. Looking at lending my neighborhood is interesting, because it gives you an idea which neighborhoods are not getting investment. On the left hand side at the bottom, we’re looking at low to moderate income neighborhoods, a little over 1.5 million people live in these communities, that’s about 336,000 families, about 31% of all families in the area. Yet, if you look below there, you’ll note that only 19% of home purchase loans 10% of refinance and a little under 13% of cash out refinances are being made in these areas. That’s a pretty big gap. We’d like to see a little more investment in low to moderate income neighborhoods. On the right, looking at majority minority neighborhoods, roughly 20% of families live there. Yet only 7.8% of home purchase loans were made there between 2018 and 2020. That’s a very big gap. We’re gonna take a look at maps in a minute that are going to kind of give us some suggestions about where this gap is worse. Let’s click next. All right, home purchase, lending from 2018 to 2020. is what we’re looking at here. This is a dot density map, all of the dots indicate a loan that was made during that study period. But they’re not precisely placed on the property. They are, you know, so they are randomized a little bit just for privacy reasons. But it still gives you a pretty good idea. If we zoom in here, you’ll be able to see areas where there’s sufficient investment or clusters of lending, such as this region here. Royal Oak, Huntington woods. Compare that with areas to the south. This is within the city of Detroit, you see large sections without lending. And it’s a little confusing. So if you zoom in, you can actually see satellite imagery of the area in question. This helps you kind of investigate a little bit sometimes you’ll see a gap in lending. And when you zoom in, you realize that’s a park space or it is an industrial facility like an airport. Now, the green color here indicates neighborhood median income. darker greens are low income and lighter greens are upper income those white areas you see there are areas of missing data, probably because the population in those tracks is so low. We can also look at the data by neighborhood race by clicking the button in the upper left hand corner. Here the darker blue colors indicate more minority Any population, the darkest blue are at least 50% minority. And, you know, other shades are indicated by the legend here, generally, you’re going to see the same patterns appear where higher minority areas are also the site of lower investment. Okay, let’s go to the next page here. Okay. All right. And here we’re looking at the total number of loans made, originated and purchased loans. I’m gonna explain that in just a minute. But let’s so let’s take a look at home purchase lending first. ere we go, and this looks at the total number of loans made by the top 20 lenders in the area. Here we see Quicken Loans has made 16,190 loans during the study period, most of which they originated themselves. But it’s important to understand that purchase loans exist. So purchase loans mean loans bought by a lender, where they did not make the credit decision, but they still have to report that loan under Honda. You’ll note here that some very large lenders don’t originate most of their loans. In this case, JPMorgan Chase, Wells Fargo are two major lenders, but they actually don’t originate a large number of loans compared to some of the others. The challenge with purchase loans is that there’s a loophole in Honda that allows lenders to remove demographic data such as race and borrower income. And then we can’t see much about those loans to analyze. Now, we covered this a little more extensively in the report we did this year on mortgage lending. And we’re asking the CFPB to help us close that loophole so we can get access to that data. For the rest of this report, we really focus on loan originations. But I’d like to start off by talking about purchase loans, because I want you to understand that they are important, and they’re a sizable part of the market. Let’s go on to the next page.
12:00
area. So now we’re just looking at originations, and we’re still looking just at home purchases in the upper left hand corner that met that menu is still check there. We can also narrow it down just to the top 20 banks, top 20 credit unions or top 20 mortgage companies. For right now I’m going to leave it alone though, because I want to look at kind of them all mixed together for a minute. So at the top, we still have Quicken United shore mortgage when incorporated in Flagstar, you’ll note that out of the top five lenders here, first of all four of them are not banks. And Flagstar operates much like a mortgage company in some ways. Interestingly, Lake Michigan credit union is here. It’s one of the few areas that I see with a credit union in the top 10 lenders. The other thing want to take a look at are the type of loans they’re making conventional versus FHA, VA and RHS. If you’re not familiar, RHS is a Rural Housing Service Loan guaranteed by USDA. It’s only available in rural areas. That’s why you don’t see very many of them. But those government loans are important for a lot of reasons. And typically, banks don’t do a whole lot of government backed lending. In this case, only Flagstar I told you, they operate a lot more like a mortgage company in some ways. They’re doing a lot more FHA lending. But generally, these are your top 20 banks and you don’t see a lot of that kind of activity. But if we look at mortgage companies, they’re much more likely to make a large portion of their loans using FHA lending. Let’s go on to the next page. And VA lending shows those patterns as well. Representative Katey Porter, her office recently released a report on va lending and predatory aspects of it including loan churning, if you’re interested that that report is online right now at Katie Porter’s website. So let’s take a look back at home purchase loans, again, amongst mortgage companies and banks. This is a scatterplot. It lets us look at those individual lenders and see how they’re doing in lending to LMI borrowers in LMI neighborhoods on the left hand side. So on the bottom, we’re looking at the percent of loans to low to moderate income neighborhoods. On the left hand side, we’re looking at the percent of loans to low to moderate income borrowers. The dots represent individual lenders. And if you hover over them, you’ll get some more information. So generally speaking, you want to you want to be as far in the upper right hand corner here as possible. That means the largest percentage of your loans are going to low to moderate income borrowers, or in low to moderate income neighborhoods. In this case, Washington Mutual 50% of their borrowers were low to moderate income and 32% of their loans were in low to moderate income neighborhoods. But they aren’t one of the bigger lenders. Let’s take a look at some of the bigger ones. Flagstar also 47% in LMI neighborhoods 25% or 47% LMI borrowers and 25% in LMI neighborhoods on the right hand side We’re looking at lending to minority borrowers, or lending in minority neighborhoods. So at the bottom, this percentage represents the percent of loans to minority neighborhood. And if we look over here, Huntington 20% of their loans went to a minority borrower, and 19% and a majority minority neighborhood. Both of those figures are pretty good compared to the rest of the lending activity, we see. If you want to, you can use this opportunity also to change this over to refinance and see how lenders are performing there. And you see some very different patterns emerge, it’s not unusual to see some lenders that are performing very differently in refinance loans versus those that are that are in home purchases. Let’s go back here. Now we’re going to take a look at those individual lenders. And we’re going to talk about how they’re doing in terms of loans to borrowers of particular groups or incomes. Again, these are the top 20. Lenders and home purchase lenders in the area. And you can see clearly that there are a number of them not collecting demographic data. When it says no data, that means that there was no indication of the race or ethnicity of the borrower on the application. The gray bar behind this indicates the average figure that makes it easy to look at this really quickly and see who is doing above average in terms of loans lacking demographic data. You can also look over at individual groups and see how they’re doing for the larger minority groups in the area. This is something I want to bring up because if any group that has more than 1% of the market share. So for example, if we had run this report in Honolulu, Hawaii, you would see Hawaiian and Pacific Islander on here as a category. Because there’s such a small portion of this particular population, they don’t show up on this chart.
16:57
What’s interesting here is looking at seeing who stands out. So some lenders typically stand out with some certain communities Goldstar mortgage Financial Group 22% of their loans had a black applicant, United shore 9% of theirs had an Asian applicant, nobody’s doing particularly well at lending to Hispanics, however, except for caliber home loans. This kind of information is useful when you’re going into a meeting with a bank or a lender and you want to talk about their data. This kind of gives you a little bit of way of a way to gather some intelligence before you go into that meeting. What’s especially interesting if you want to narrow it down to a specific type of lender. So if you just want to if you’re going in for a meeting with a bank, you can look at the top 20 Bank lenders and easily compare the bank you’re meeting with probably with with the rest of the market and see how they perform. Now, if we click on this button, it says borrower income, it changes this chart to one where you look at the loans by borrower income, low and moderate income are on the right hand side. And this gives you an idea of who’s doing better in each of these categories as well. So there’s a lot of performers in this market that are doing a little better than average. But there’s a few they’re really kind of at the bottom of the rankings also in some ways. And let’s go to the next screen. So one interesting fact about the new Hunter data that started about three years ago is that it includes a lot of information on loan prices, not just the interest rate the borrower gets, but the individual closing costs, including origination fees, lender credits, discount points and a slew of other information on exotic features of loans, you know, and so forth, and so on. So we wanted to take a look here at the differences in in cost to the borrower. And the first thing that ought to jump out at you is that conventional loans are a lot cheaper. These are they were still looking here at home purchase loans made by banks in the area. The average conventional loan closing cost was just over $2,600. Meanwhile, the FHA borrower paid just south of $5,000. VA loans are also pretty expensive. In part, this is due to additional fees that are part of the closing process. And in the case of FHA loans, borrowers were responsible for an upfront more mortgage insurance premium. And that leads to most of the cost here. But the fact is that I want to point out there’s a difference in the cost between these programs. However, FHA and VA loans, especially are very important to communities of color, especially black and Hispanic borrowers. And let’s take a little closer look at that and see why. Alright, so nationally in 2020, Black borrowers, homebuyers used FHA loans for about 58% of their home purchases that’s nationally. In this study area, we’re still looking at home purchases, and let’s uncheck lender type. So we’re going to look across all lenders, black borrowers, used FHA, or VA loans for a little over 50% of all their purchases in in 2018 through 2020. Hispanic borrowers used it quite a bit yet last, but there aren’t very many of them. As we saw from the chart earlier, white borrowers and Asian borrowers are much less likely to use government insured loans. So what’s the effect of that? Well, on the right, another scatterplot. At the bottom, we’re looking at the closing costs paid by the borrower. And on the left hand side, we’re looking at the interest rate. And the effect is that black homebuyers in this area paid a little bit more in interest, and quite a bit more in closing costs than the average white borrower, Hispanic or Asian borrower. And we can actually take a little more detailed look here. borrowers are asked if they are Hispanic, Asian or Hawaiian or Pacific Islander, they’re given the option now further identifying themselves. And if though, if there is enough of that community in an area, this tool will track it and give you some more detail. In this case, a number of borrowers identified themselves as Mexican. There were 1551 home purchase loans to a person who identify themselves as Mexican over the three year period, and just over 26% of them use them. Sure. That’s not that surprising, you’ll get the area population. But if you look down here, this is interesting. Asian Indian borrowers and Asian Indians in the past two or three years have really shown up in the data as a as a as a, a large portion of Asian borrowers in general, however, they’re much less likely to use government insured loans. And as a consequence, they’re also paying lower interest rates, and they’re getting in lower closing costs than other groups.
21:52
And let’s go forward here. And by the way, I may have failed to mention up at the top are pretty simple navigation buttons, the arrows go forward and backwards in the little house will take you back to the table of contents. Let’s talk about denials. So, borrowers are required to report a reason for a denial. And that sounds like it’s something that they should have been doing all along. But really, it just started three years ago. So we have three years of data now we’re borrowers have indicated the reason for denying. And here we take a look at the total number of applications and we want to compare individual lenders on based on their denial rates to white borrowers compared to black Asian and Hispanic borrowers. So we’re talking about we’re still looking at home purchases, we’re talking about a little under 71,000 applications were made during the 2018 to 2020 timeframe. 10% of those were denied, white borrowers were denied a little bit less just 8.1% of the time. But let’s take a look at the overall denial rate by race. That’s this section here on the left. Black borrowers were denied little over 2.2 times as often as white borrowers. In other words, 18% of black applicants were denied a home purchase loan compared to just 8% of white applicants. Generally, there’s some other variations Hispanic borrowers a little more likely to get denied. And if we look at the reason for denials, interestingly, the biggest reason was collateral, meaning that for the home purchase, the price agreed on by the borrowers was not supported by the appraised value, debt to income ratio and credit. Were the next most common reasons those are pretty normal. The fact that collateral is at the top is a little unique. Usually it’s your debt to income ratio or credit in most markets that I look at. On the right hand side, we look at individual lenders again, and again, it’s the same top 20 that we’ve been looking at. And let’s take a look at an individual lender here. So the white borrowers or denied 12% of the time for Quicken 21% for black borrowers. 10% for Asian 19% for Hispanic. The red checkmarks indicate cases where that group is being denied at least twice as often as white borrowers. So in this case, you’ll see a lot of borrowers are reporting much higher denial rates for black applicants for home purchase loans than they are white borrowers. Yet interestingly, they are reporting better rates for Asian borrowers in many cases than white borrowers. That’s very interesting. I don’t see that very much also, either. Let’s take a look for a little bit. Let’s take a look at business. So business lending unfortunately, we do not get the detail that we get with mortgage lending. For now, with business lending, we do get information on the number of loans made and if those loans were made in low to moderate income neighborhoods or if they were made to businesses that make less than a million dollars in revenue
25:00
There is an aspect of the Dodd Frank Act of 2010, called Section 1071. This instructs the CFPB to collect more detailed data on small business lending. That rule is being written now. And NCRC is involved in commenting on that rule. If you have more interest in that, please let us know. And we can help you, you know, be involved in that effort. But let’s take a look at what we do now. So first of all, on the left hand side, you’re going to see that American Express is at the top of the list now, this is the these are the largest 20 small business lenders that are reporting under CRA. American Express is usually the top lender in most markets, that’s not unusual. Next up are Chase, Citibank, Bank of America, Capital One, most of the bigger banks that you know, the average loan amount, that is something interesting, there’s no way to really differentiate between normal small business loans and credit cards, business credit cards, obviously, American Express, those are all credit cards. So but there’s no there’s no designation in the data to let us know which are which, however, we use a an average loan amount. So meaning that nationally American Express’s average loan amount is under $25,000. So we identify it as a credit card lender. So the blue colors here indicate a credit card lender or what we think are credit card loans. But it’s difficult to tell how much of those are really larger small business loans. On the right hand side, we’re looking at a scatterplot, looking at the percentage of loans going to small businesses, and the percent going to low to moderate income neighborhoods. And again, just like before, if you hover over these, you’ll get more information on the bank. And you can see who’s performing better. In particular, I’m interested in banks and lenders that make loans in low to moderate income neighborhoods. So if I was an organization working with small business owners in an LMI neighborhood, in this community, I might want to talk to a Chemical Bank and see what they’re doing because they are making a sizable portion of their loans to small businesses and an LMI communities. And we can also take a look at a map of business lending. And again, like the mortgage map, this is a dot density map. So you’re gonna see little dots all over the place. And they’re randomly placed within their census track. But it does give you an idea of where lending might be clustering. So for example, this area of Birmingham, oops, sorry about that. sees a cluster here. And if you zoom out and go, sorry about that guys. Down to Detroit itself, we see a generally less lending in this area. But again, these maps do a good job of kind of letting you recon out an area and see where there are gaps, or where there might be clusters that you want to learn more about. And just like before, you can click on the button in the upper right. And you can look at neighborhoods by racial or racial demographics as well. So here, though, we are seeing some concentrations of lending even in high minority neighborhoods. And again, that usually means the presence of some type of an anchor institution. In this case, hang on, what is this? Yeah, Lawrence Institute of Technology, you know, any of these things might be anchors for more business lending.
28:27
And let’s go to the next screen. Finally, we’re going to take a look at bank branches. We also just released a report on the declining number of bank branches in the US. And they have continued to decline for the last 10 years. It looks like they’ve actually accelerated enclosures since the beginning of the pandemic. This data comes from June of 2021. It’s only current as of that date. And I can already see a couple of interesting things we want to talk about. On the left hand side, we’re looking at the top 20 Bank networks in the in the area by the number of branches. So if you hover over this Comerica Bank had 134 branches, that’s a little over $23 billion in deposits, or 14% of the area’s total deposits. What’s interesting here is as you go down, you see a clear drop off after the top 10 banks or so that had in this case, Flagstar with 60 branches, you drop down to First State Bank with only 14, that’s a little bit different. I’d be interested to see where those branch networks are actually located at especially because on the right hand side, we’re looking at the percent of branches and minority neighborhoods at the bottom, and on the left of the percent of branches in low to moderate income neighborhoods. And what’s really interesting is that a lot of these smaller branch networks have nothing located in minority neighborhoods at all. Now we’re talking about a relatively small number. I’m guessing these are probably more rural areas, given the geography and then some branches. America is very good. About locations of minority neighborhoods. However, when I say very good, they only have 20% of their branches in all of Detroit are in minority neighborhoods. So we’d want to take a look closer at the total demographics and see where the Comerica Bank is located. Maybe before we judge. Let’s go forward to the next page, which is an interactive map, which is going to let us do just that. So these are all of the banks active in the area. Now obviously, this is a bit much. So let’s zoom in and take a closer look at branches there. And again, the background here indicates the median neighborhood income. And let’s take a look at Comerica, click this select Comerica, and zoom out a little bit to see where their branches are. Generally, this is a pretty good distribution with the exception of I do see a pretty big gap here in Detroit itself, outside of a couple of say, well, they’ve got a cluster looks like downtown. Let’s take a look at it by race. And go back to Comerica. And, you know, there’s a pretty big area here that that has no Comerica branches, but generally looks to me, like, you know, they’re pretty well distributed. And this lets you kind of scope out and see if there are any real empty spots that you might want to engage with this bank and ask them about. And you know, in the past, we’ve worked with local organizations, and coalition’s to encourage banks to open branches in hard hit areas. So this is something that if you’re not a member of NCRC, you can talk to Maggie with our membership team about. And, you know, we can discuss ways in which we might be able to help you talk to a bank about opening a branch in some locations. But this tool is going to give you a lot of clues about where that might be and where you want to where you want to take a look at. Alright, so now I’m going to go back to the table of contents. And I’m going to stop sharing my screen. Phyllis, if you want to lead us into q&a. I think there are already some questions in queue.
32:13
Yeah, absolutely. I think somebody might have answered the first one. But how often is this data updated?
32:21
Several times throughout the year it gets updated. The 100 data comes out once a year, usually late in June, it does take us some time to get it cleaned up because it’s usually between 12 and 14 million rows of data and then loaded into the system. So if you look at the definitions button on the fair lending tool, it will give the vintages of different data that are included. Small business data is typically released in the like December, I think late December usually. And bank branch data comes out sometime around September. So whenever this new data is available, we go in and we you know, we download it, and then we update the tool with it.
33:02
Great, thank you. Next question is, is there a data on areas of LMI interest that are being gentrified?
33:10
So it’s yes. And we’ve done several reports on gentrification, excuse me. We’ve done several reports on gentrification that look at, you know, look at exactly what you’re talking about the challenge with 100 data is that it’s happening. First of all, it’s mortgage data. And it’s no, you’re not getting it for about 12 to 18 months after it actually occurs. But yes, this data can show clusters of mortgage lending in LMI are high minority areas. And then if you dive into it, it can tell you those borrowers are not LMI, or they’re not minority. And that’ll give you a clue if the if the activity might be gentrification. But I would encourage you not to use hunda as a way to try to judge gentrification, it’s just a part of the puzzle. You really do need to talk to people living in the area to confirm if gentrification is really occurring.
34:00
Thank you, Jason. The next question is, will we consider conventional loans differently than FHA to qualify? Who is acquiring these loans?
34:10
Well, there’s, I don’t like to give the impression that FHA loans are inferior in some way, FHA and VA programs exist to fulfill specific needs. However, the cost of FHA loans is much higher. So if you if you if you have a situation where black and Hispanic borrowers disproportionately are being steered into those loans, then that’s a problem. If there are not, you know, in my view is that there should be conventional loans that can service that population at a lower cost during times when when credit is flowing really easily like it is right now. Essentially, FHA was intended to be a counter cyclical program that would only expand when the conventional market couldn’t meet demand, the fact that FHA has become a de facto first choice loan For Black and Hispanic and LMI borrowers, that is problematic because then they are Bert bearing most of the burden for the cost of the program also. That said, I think all my mortgages personally have been FHA. So, I mean, it’s great to have when you need it.
35:18
Thank you. Next question. How do you code the race and ethnicity of the borrows borrowers here? How does our debates and ethnicity coded?
35:26
So that’s a more complicated question than you might think there is a we did publish a report on exactly that question. And it’s a white paper that shows how we do it. However, if you’re not familiar with 100 data, there are Every borrower has up to five different ethnicities in five different races, they can identify as, in addition, the CO applicant has five different ethnicities in five different races, they can identify as, so as it sounds like that can make things pretty complicated, though, I’m actually gonna copy and paste this into the chat right now. This is the white paper that we’ve published on how we do it. But I want to lead off by saying this is just the way we choose though this, this methodology emphasizes people who are part of racial subgroups. For example, if you identify yourself as Hispanic, and then in the next ethnicity field, you identify yourself as Mexican, we will identify you as Mexican. In either case, we all know you’re Hispanic. If you are, you know, if you’re doing research on a particular subgroup, let’s say you want to understand more about Puerto Rican lending in New York City, you would want to change your methodology slightly to make sure you’re identifying all people who indicate Puerto Rican, wherever they do. So there are a surprising number of people who who answered with three and four different racial choices. And then you add in CO applicant data to that it makes it even more complicated. But take a look, if you have any questions about our methods. The easiest way I would explain it here is that if the attendee is is anything other than or the I’m sorry, the applicant is anything other than a non white Hispanic person, then they are going to get flagged as being a minority borrower, for our purposes.
37:18
Thank you, Jason, one of the followers on one of the participants is noticing that the information on the website, the numbers are different than the numbers that have been shown on the tool. Is there a reason for that? Or?
37:31
No, probably, it’s due to the filter I put in place on where I limited it to just home purchase lending. So as long as you select the Detroit Metro area with all the counties, and then you change the filter to home purchase, that it should match what I showed there, but if you find any discrepancies, just screenshot them and email them to me, and I’ll try and figure it out.
37:51
Okay, thank you. Can you compare closing costs interest rates for white and non white borrowers?
37:57
Yes. And since I don’t do it in the tool here, but that’s data we can easily produce for members. Just give me a call, and we can put it together in a spreadsheet for you. But generally speaking, white borrowers are paying less because they’re generally using conventional loans, which are cheaper. And oh, I’m sorry, Phyllis, I mean, interrupt, I apologize. But But even in, you know, there are studies where they’ve looked at at similar datasets. So for example, just conventional loans. And they still find racial disparities in lending, even when accounting for things like the down payment amount, so forth, and so on. Now, of course, public home to data does not include credit score. And this is something we think should be part of the public data set. We think that that data can be anonymized safely. And then that would let us do a lot more precise work on understanding who is being denied and why. However, regulators do have access to credit score now. So we encourage the CFPB to take a look at closing cost across the and normalize for credit score and other factors to look for examples of racial discrimination in mortgage pricing
39:04
to thanks, Jason, where’s the US Bank in this matrix?
39:12
It depends on the metro you look at I don’t have it closed out of the browser right now. So you can go in there, though, and take a look and see us bank performs well in some markets and less well, and others. So it’s important to look at what what market you’re talking about.
39:30
And can you clarify, again, how you distinguishing between credit cards and small business loans? Yeah,
39:37
it’s not a perfect system. I’ll be the first to admit. So I wouldn’t use that as gospel. But there is no hard and fast way there is no way in the data for us to tell the difference. What we get is a database that has the lender and the total number of loans that they made, the average amount of the loans and some other data, but it doesn’t indicate to us what type of loan it’s not Like with mortgage lending, where I can tell you this is a home purchase loan, this is a refinance, etc. For now, the business lending data does not do that, in lieu of that we’ve just kind of established a rule, or not a really a rule, just kind of, you know, our method, if the national average loan amount is under 25,000, we consider it to be a credit card lender. And the only reason we do that is that if you want to speak to a lender about their small business lending, they’re going to want to be compared to a peer group that’s roughly similar. This lets you kind of narrow the field a little bit, but I wouldn’t go in there and just, you know, immediately assume that all of the loans or credit card loans, or all of them are small business loans, there’s probably a mix with most of those lenders of the two.
40:44
And Jason, does this report track redlining?
40:48
Well, yeah, I mean, the there there’s evidence of redlining throughout this, we do not have a link directly to the redlining maps from the homeowners Loan Corporation. We’ve done a lot of works with work with those maps. And if you go to our redlining, hub@ncrc.org forward slash redlining, you’ll see a lot of that work. So the redlining maps of the HLC covered about 200 cities total. And you can look at a variety of you know, we do see a pretty sizable impact from redlining into mortgage lending. Today. Most neighborhoods that were redlined in the 1930s today are high minority and low to moderate income. So they track very closely with the lack of mortgage lending or small business lending in most cases, too.
41:36
So is there a way to search by lending institutions, even if one is not in the top 20 In a market?
41:43
Not yet, that’s a that’s a we can do that, obviously, for a member if you call us and talk to us. But we’re working on a new tool that I’ll join the fair lending tool that lets you dive a little more deeply into individual lenders. And then we’re revamping a branch locator tool to also let you pull up information on specific branches to help you with branch closures and things like that.
42:06
So Jason, certainly this too, provides a lot of data and certainly can be useful to many of us, but get since you sort of like the expert in this in this room as to what data would you like to see that’s not currently being reported?
42:21
Credit score and small business data? That’s a great question. By the way, I, we see no reason that for 100 data that they cannot release a so you obviously don’t need the exact credit score, but they can do what’s called bucketing. And they do this with most of the data and 100, most of the numeric data is bucketed in some way, meaning I don’t get the exact dollar amount of the loan, I get the dollar amount rounded to the nearest $10,000. With credit score, I don’t need your exact credit score, but you can at least tell me if it’s over 720. Or if it’s between, you know, 606 40 or something like that, that would be extremely useful to us, it would give us a lot more tools with which to identify racial discrimination or income discrimination or gender discrimination. And we’ve even found interesting patterns when we look at same sex couples where the applicant and CO applicant are the same gender, and they exhibit different lending characteristics. So I think things like that are the next step. I hope to see that. Aside from that the section 1071 of the Dodd Frank Act has been law for 10 years, and it has yet to be implemented. So we’re really excited about that. And you were pretty heavily involved in those discussions about what that data is going to look like. So if you want to get involved in that discussion, you know, please check in with with our organizing team.
43:45
And I know that the recording is going to be available after that after this presentation. But the presentation itself would also be available, correct?
43:54
Correct. I mean, the recording of zoom will certainly be available. So yes, that will be available after this call. Maggie’s gonna be taking care of that today. She’s recording it, and she’s going to be, you know, handling that recording. So I think if you’re registered, and Maggie, maybe you can answer in the chat, we’re going to send that around to the registered attendees. Correct.
44:16
She said that the recording would be okay.
44:19
Well, the presentation actually, I didn’t go through any slides here. That was actually the fair lending tool. So if you’re a member of NCRC, you can sign in and use that right now. If you are not a member, by all means, chalk talk to Maggie. And we actually do have a sample version. I don’t have the link in front of me right now. But we can we can give you the sample version and you can at least explore it a little bit and see what and see if it works for you.
44:46
Thanks, Jason. And one more question. Can one use the tool to do a similar a similarly situated fair linen analysis, where one compares to similarly situated approved with a similarly Situated denied, with the only difference between with the only difference between the two being race or gender.
45:07
No, not with this tool, but we can help you, you know, just get that data yourself and take a look at it, or we can help sort it. If you remember, just give me a call, let me know we can we can pull them together for you. But I caution you that on the data without the credit score, you know, the constant pushback that you’re gonna get from lenders or banks is going to be that we don’t have the credit score, and that and that is the explanatory factor. So you certainly can, and I can talk to you about that individually. But and unfortunately, this tool doesn’t, doesn’t have that ability.
45:41
So is there going to be my data collected for business loans? And when will this tool reflect new embedded data?
45:48
Section 1071. rulemaking is in effect now I would estimate three to four years before we start seeing data from it though.
45:57
Okay, and I think
46:00
there was a question here that just popped up organizations whose web filter blocks access to zoom. That’s a no, it wouldn’t be posted to WebEx, because it’s I think it’s a zoom cloud recording link, I tell you what, Ricky, if you want to email me about that, I’ll see if I can just download a video file for you,
46:19
Jason. The link is usually a private YouTube link. So it should not be a problem.
46:26
Awesome. All right. Thank you, Maggie.
46:29
And then I think the last question is, can you use this data to create that redlining map?
46:35
You can use this data and overlay it with GIS software over a redlining map. And if you’re interested in doing that, let me know I can connect you with our, the the GIS specialist on our team. And we can kind of explain where those resources are. But it would require you using some sort of mapping software like ArcGIS, or similar product.
47:01
And I think that’s all the questions that I have the with any other questions? I don’t see any hands raised Mikey, if you can see those. If anybody raise your hand, I can’t see him from here.
47:15
Looking through the last year,
47:17
someone raise their hands. If you have any additional questions. Those people that raise their hands, you should type it into the q&a section that you see. And that way. So listen, Jason will see the questions as well.
47:33
I don’t see any at the moment. Hopefully, there’ll be one or two as we wait. So do we want to thank Jason and the team for this great presentation, or two that we can all use in our communities to help better understand the banking, performances and to really have conversation with these banks to do better in communities and to shape the policies and the programs that are being implemented. As you, as Jason says, articulated that there’s something that you need specifically, don’t hesitate to reach out to NCRC him and get the information that you need. That’s what your membership does for you. provide all of these, this additional research that you might not have for yourself, I still don’t see any additional questions.
48:25
Notice, there are a few in the q&a that just popped
48:27
up, okay, Canada to be utilized to assist when applying for a mortgage.
48:35
You can I would use it to kind of maybe give you an idea of who to talk to about a mortgage, it’s a good way, if you want to look at a county you’re in, you can see who’s making loans to people who have similar income or in similar areas that might be helpful. But you don’t want to get into a trap where that kind of reinforces an existing pattern either. So I don’t know if I would, I would really use it as your only input. But if you want to use it just kind of maybe raise, like, you’ll note that some of the mortgage companies in there probably have names that are not familiar to you. And that’s because they typically do business through brokers that and so they don’t really advertise. But if you know who they are, you can go to the website and look for a broker. And that might be helpful. And actually, somebody asked also about the link to the Katey Porter Report. Yeah, let me let me look that up. It’s on her her house website. And I’m sure I can find it in just a second. Yep, here it is. They were focused on something called loan churning. I’ll put that in the chat right now. But really the bigger problem I think with VA lending is that not enough lenders are in the program. major banks avoid it only a small number a relatively small number of Mortgage Lenders are involved in it.
49:51
Okay, um, and what data points do you recommend looking for for an analysis of reverse redlining
50:03
Well mortgage lending first and then comparing it with existing census data. However, I would pay a lot more attention to pricing to be honest with you, you’re gonna see a lot of like loan deserts and I would be looking at something like loans per per number of housing units. All this data is included in Honda. And I’m, you know, I’m happy to remember I can help pull together some basic data for you pretty easily.
50:28
So can you clarify your comment regarding Lupo for some for some banks not reporting rates, and income in the HMDA HMDA data,
50:39
right. So humba allows banks to remove certain data points from loans they purchase. So if if bank of Jason buys 10 loans from the bank of Phyllis, and I don’t have to report that data on borrower income, or borrower race or ethnicity, so I’m free to remove that data. Now, at one point many years ago, that was a really kind of an uncommon step. Because it’s a, it means that they are taking making the effort of going into this data and removing certain items. However, over the past several years, we’ve seen many of the banks and lenders that buy loans, removing data, I think 15 out of the top 20 lenders removed. So 15 out of the top 20 loan purchasers remove 100% of their data. And that creates a pretty sizable gap in our understanding of the lending market. So we’re hoping that at some point, we can get that loophole closed.
51:36
So I think my last question is, you have any info on loans for people with disabilities?
51:43
No. And this is a glaring omission, I think in Honda right now is that we’ve looked, you know, we’ve we’ve been asked this question before, but it just simply is not part of the data tracked by on the on the residential mortgage loan application form. So it would be excellent to get it. But it certainly isn’t isn’t reportable under hunda. And actually, I take that back, I’m not sure maybe on the residential loan application form now, but it still isn’t included. And Honda
52:13
does this to show data on predatory lending in LMI, communities versus bank lending in that same community.
52:22
At a certain level, a tool will let you look at individual counties, and you might see individual lenders that are particularly focused on say, a black lending or black borrowers. And they might be charging a lot more it’ll give you clues about who might be doing that. It would take a lot more detailed work, though, to really tease out whether or not it’s predatory.
52:45
Thanks, Jason. And one of the questions. Can we do a presentation a forum on just hump day researching?
52:52
I’m sorry, what was the question?
52:54
Can you present a forum on just a researcher?
53:00
Sure. Hi, fellas, I am happy to just give me a call.
53:04
All right. We’ll we’ll try to make that happen. And I think that’s all the questions that I saw. I hope I reached everyone’s I’m going back to the make sure that I didn’t miss anything. I don’t think I missed anything. If there are no other questions, I don’t see any in the chat.
53:29
Cat Rackleff in the chat, why do lenders remove data on purchase loans as well why they do it? I mean, there’s probably they’re probably concerned about being charged with some sort of violation. But they’re allowed to do it. Yes. Because they are not making the credit decision. They are buying the loan from another lender. So yeah, unfortunately, it creates a blank spot for us in the in the in the data that we’re hoping to eliminate.
53:57
And this information will be available for the gesture economy club, right? Correct. Okay. Yep. Purchase loans reported when they were first originated,
54:09
if the original lender was 100 reporting entity, which means they need to be meet certain threshold requirements for the volume of loans they make per year, or in their size or their office location. And there there there’s a little more detail to it. That gets way too complicated. But yes, so if I made a loan in during the same year I saw and I reported it under Honda and then the same year I sold that loan, then it would be double reported. Another piece of information we do not have is something called a universal loan identifier. The regulators are able to see the market and see which loans were originated and then were then purchased by another entity. Unfortunately, since we don’t have access to the universal loan identifier, I cannot identify the number of duplicates that might exist, but the data will indicate that the loan was sold during the calendar year and it will indicate what Have entity it was sold to. And from that we know that the vast majority of loan originations reported in Honda that are sold are actually just sold to the GSEs or HUD or, you know, it depends on the program. But so it’s not very many of the originated loans are also reported as purchased. It’s, it should be a relatively small figure.
55:23
So thank you, Jason, and your team for this wonderful informative knowledge increasing presentation. We want to thank the National NCRC for allowing their team to be here and for continuing to work in our communities to provide access to better banking and lending information. We want to certainly thank the Detroit Coalition for already organizing this and bringing this to you. We do have your other suggestions about other future training opportunities. We want to thank all 196 plus people who join us today. We hope this this webinar has been beneficial to you in ways that you have yet to even comprehend. So with that, have a great day in the neighborhood. And thanks, everybody.
Transcribed by https://otter.ai
