Named the Nation’s Best Newsletter by NAREE
June 2017 Vol 11 Issue 6
MY TAKE By Jaren Pope Brigham Young University P11
SPLITTING THE ATTOM: Streaming Your Own Property Data Playlist P17
DATA IN ACTION Highest Share of Co-Borrowers in High-Priced ‘Millennial-Magnet’ Markets P21
P1 THE RISE OF THE REAL ESTATE DATA BOTS
We’re just at the start of the “fintech” – financial technology – revolution, the growing use of data by the real estate and lending industries. Moving electrons back and forth in new ways will radically change both industries, and like all revolutions there will be both winners and losers. Ten years from now some of today’s largest and most important players are likely to be gone, replaced by – well, that is the big question, isn’t it?
Economics professor Jaren Pope of Brigham Young University shares some early insights from an analysis of bargaining, search and psychology for house prices using certain elements of property data collected from count assessor and recorder offices. Specifically Pope addresses two factors from the preliminary analysis: the possible impact of sellers who own a home outright with no mortgage and the impact of buyers who purchase with cash. P11 MY TAKE: THE SCIENCE OF BUYING LOW AND SELLING HIGH Prospective homebuyers in Q1 2017 were most motivated to move to parts of Colorado, the Carolinas and Florida — along with Washington D.C. — according to an analysis of proprietary pre-mover data by ATTOM Data Solutions using data collected from loan applications on residential real estate transactions. The data is highly predictive of which markets are likely to see a high volume of sales activity closing in the second quarter of 2017. P16 BIG DATA SANDBOX: MARKETS WITH THE MOST MOVERS
P17 STREAMING YOUR OWN PROPERTY DATA PLAYLIST
ATTOM CTO Todd Teta explores how the emerging API Economy is revolutionizing many industry verticals, and how recent and imminent innovations specifically in the world of real estate data APIs will disrupt the way businesses consume public record real estate data. That disruption will come in the form of fast and flexible APIs that can be custom- built on the fly to feed an application with exactly the data elements it needs at each point it needs them — and saving costs while doing so.
P21 DATA IN ACTION: WHERE TO FIND CO-BORROWERS
More borrowers purchasing a home are relying on “co-borrowers” to help them qualify for the loan — nearly 22 percent of all single family purchase originations in the first quarter had multiple, non-married co-borrowers on the loan, up from 20 percent a year ago. But co-borrowers account for more than one in every four home purchases in many high- priced markets that are magnets for millennials.
BY PETER MILLER, STAFF WRITER The Rise of the Real Estate Data Bots
billion and smartphones are everywhere. We’re just 10 years into the smartphone revolution and yet already “the average American spends five hours a day on their phone,” says Federal Reserve Governor Lael Brainard. As much as the mobile phone — essentially a pocket-sized computer — has changed countries with advanced economies, what it’s done worldwide is even more profound: less than two decades ago the UN reported that “more than 50 per cent of the world’s people have never made a phone call.”
Places once technologically remote are now on the grid just as surely as cafes overlooking the San Francisco Bay. Fintech Revolution In a similar sense we’re just at the start of the “fintech” – financial technology – revolution, the growing use of data by the real estate and lending industries. “We can position our clients in front of their customers at the most optimal time,” said Alex Kutsishin, Co-Founder and Chief ROI Booster with Sales
Ten years ago in June something new and different entered the American marketplace. The iPhone was a hit from day one, the first “smartphone” and a hint of things to come. It was not just a communication tool, said Apple co-founder Steve Jobs, “but a way of life.”
Jobs turned out to be a visionary. Apple now has cash reserves of more than $250
ATTOM DATA SOLUTIONS • P1
The Data Economy Traditionally the goal of every business has been to sell more and see profits rise as a result. With artificial intelligence (AI) – an expression which broadly includes such things as robots, machine learning, software, deep learning, 3-D printing, and automation — the path to profitability in the new data economy now involves
Boomerang. “We can analyze hundreds of millions of records, compare them to current market conditions and then give our client information that allows them to better service their existing borrowers, or even future borrowers. For instance, we tell our mortgage clients the moment an existing customer has 75 percent LTV in their home or when a prospect needs to get a call because they NOW are a good fit for a loan product. We do all of this with almost no effort from our clients.” Moving electrons back and forth in new ways will radically change both industries, and like all revolutions there will be both winners and losers. Ten years from now some of today’s largest and most important players are likely to be gone, replaced by – well, that is the big question, isn’t it?
revolutionary tools with the capacity to produce more, extract new efficiencies, reduce costs, and improve margins. It’s the difference between digging a canal with tractors and using spoons.
In the same way that all matter is built from atoms, AI is constructed from
We can analyze hundreds of millions of records, compare them to current market conditions and then give our client information that allows them to better service their existing borrowers, or even future borrowers.”
Alex Kutsishin Co-Founder and Chief ROI Booster, Sales Boomerang
Rise of Equity Rich Homeowners Equity Rich* Homeowners Pct Equity Rich
ATTOM DATA SOLUTIONS • P2
Machine learning drives our algorithms for demand forecasting, product search ranking, product and deals recommendations, merchandising placements, fraud detection, translations, and much more. Though less visible, much of the impact of machine learning will be of this type — quietly but meaningfully improving core operations.”
Jeff Bezos CEO, Amazon
explains The Economist. “Technology giants have always benefited from network effects: the more users Facebook signs up, the more attractive signing up becomes for others. With data there are extra network effects. By collecting more data, a firm has more scope to improve its products, which attracts more users, generating even more data, and so on.” • The coming data-based economy is transformative – and disruptive. A few years ago you might have hailed a cab, rented a hotel room, or paid for cable TV. Today you are likely to use Uber and Lyft to get around town, find rooms through Airbnb, and subscribe to Amazon Prime, Google’s Chromecast, Hulu and Netflix rather than a traditional cable service.
tiny bits of information known generally as data or data points. These bits have little value individually but pulled together, arranged and re-arranged, they can produce dramatic marketplace advantages. “Much of what we do with machine learning happens beneath the surface,” said Amazon CEO Jeff Bezos, in his 2016 shareholder letter. “Machine learning drives our algorithms for demand forecasting, product search ranking, product and deals recommendations, merchandising placements, fraud detection, translations, and much more. Though less visible, much of the impact of machine learning will be of this type — quietly but meaningfully improving core operations.”
• Data equals power, more data equals more power, and more power equals more profits. • Traditional metrics relating financial value to physical size, revenues, and even profits may not apply in the data economy. Tesla lost money last year and produced fewer than 85,000 cars. General Motors churned out 9.8 million vehicles worldwide and had a 2016 profit of $9.43 billion. As of May 31, Tesla — which is widely perceived as a data and technology company — had a market cap of $53.64 billion versus $50.01 billion for GM, an auto producer with roots dating back more than a century. • The use of data creates a network effect. “This abundance of data changes the nature of competition,”
The basic rules for the new data economy look like this:
ATTOM DATA SOLUTIONS • P3
manufacturing output and productivity are increasing.”
The Rise of Non-Bank (Often Digital) Lenders 5-Year Change Q4 2016 Purchase Loan Originations (Top 5)
The initial efforts to transform manufacturing involved such basic tasks as using a robot arm to move parts and weld. In a similar sense, we’re only at the starting point of the AI revolution. There’s more to come and what comes will transform real estate and finance with the same zealous efficiency that re-made the factory floor. • BlackRock said in March that “asset managers who simply use the same techniques and tools from the past will limit their ability to generate alpha and deliver on client expectations.” Now, says the company, it wants to harness “the power of ‘human and machine’ to efficiently and consistently deliver investment performance to our clients.” • “Digital lending,” said National Public Radio earlier this year, “is expected to double in size over the next three years, reaching nearly 10 percent of all loans in the U.S. and Europe. There are now some 2,000 digital startups, many of which are using artificial intelligence to analyze the troves of data created every day.” (see chart on this page) • “If not already the buzzword for 2016, AI will be the buzzword for 2017,” said Trulia in January. The company says it has “created and deployed a Recommender Engine that
Caliber Loans Wells Fargo
Fairway JP Morgan Chase
There is no reason and no evidence to believe that real estate and mortgage lending are somehow a “special case,” immune to the changes — and disruptions — which AI and data are now creating throughout the economy. Data — and how it’s gathered and used — is the transcending technology of our time, one that will create quantum leaps in terms of speed, accuracy, savings, revenues, globalization, and efficiency. Like smartphones in 2007, it will rock our world. Manufacturing Parallels For the past few decades “automation” has generally meant changes on the factory floor, the use of software and robots to endlessly repeat given tasks. The result has been a workplace revolution, one where production increased while 7 million U.S. manufacturing jobs were lost. Between 1970 and 2010 manufacturing blue collar
employment shrank from 25 percent of the workforce to just 10 percent.
“Yet,” said The Washington Post in November, “American factories actually make more stuff than they ever have, and at a lower cost. Manufacturing accounts for more than a third of U.S. economic output — making it the largest sector of the economy. From that perspective, it’s hard to argue that American manufacturing today is anything but a success.” While U.S. residents are very aware of the internal job market, what’s generally not understood is that the employment situation worldwide is equally in flux. Automation and technology are moving back and forth across national boundaries with electronic speed. As a result, says Bloomberg, “factory jobs are shrinking everywhere” and yet
ATTOM DATA SOLUTIONS • P4
AI Applied to Fintech The AI leap from factories to finances is likely to be a lot shorter than may be realized. As Mark Cuban said earlier this year: “artificial Intelligence, deep learning, machine learning — whatever you’re doing if you don’t understand it — learn it. Because otherwise you’re going to be a dinosaur within three years.” Consider the problem of loan applications. From the borrower’s perspective applying for a mortgage is a huge hassle with tons of paperwork and lots of nosy questions. The AI alternative is to use an online form, provide some basic information and bingo; you’ll have a solid sense of your borrowing ability in minutes. (see “Mortgage of the Future Will Be Designed for Digital Natives” in April 2017 Housing News Report) Hyperbole perhaps, but what if he’s right?
learns a consumer’s preferences and surfaces new home listings they might like, and have trained machines to see real estate photos and understand their contents (i.e. our trained machines can determine if they’re seeing a photo of a kitchen or bathroom, and what type of features it has, such as hardwood floors or granite counters), to personalize and help streamline the house hunt journey.” artificial intelligence technology tailored for legal work have led some lawyers to worry that their profession may be Silicon Valley’s next victim,” reports The New York Times. It says a recent study estimates that as much as 13 percent of all billable hours can be eliminated just with today’s technologies. • Even attorneys have begun to feel the AI heat. “Impressive advances in
What’s really going on is that lenders are increasingly interconnected with data sources. When a borrower applies for a speedy approval, hordes of electrons are instantly dispatched to the far corners of the financial world. Information — data points — from credit reports, bank accounts, public property records and other sources are instantly gathered, analyzed, and then plugged into various mortgage options. But wait! Aren’t lenders required to collect paper tax returns and pay stubs before approving mortgage applicants? The rules don’t actually say such specific paperwork is necessary. What’s required, according to Dodd-Frank, is “a reasonable and good faith determination based on verified and documented information that, at the time the loan is consummated, the consumer has a reasonable ability to repay the loan.”
From the borrower’s perspective applying for a mortgage is a huge hassle with tons of paperwork and lots of nosy questions. The AI alternative is to use an online form, provide some basic information and bingo; you’ll have a solid sense of your borrowing ability in minutes.”
ATTOM DATA SOLUTIONS • P5
The First American Loan Application Defect Index, “which estimates the frequency of defects, fraudulence and misrepresentation in the information submitted in mortgage loan applications,” was down 20.6 percent in April from the October 2013 peak. Foreclosure activity in April 2017 dropped to the lowest level since November 2005, according to ATTOM Data Solutions. Democratizing The Marketplace When it comes to competition a lot of the marketplace is simply off-limits to newcomers because few people have the dollars to open a new supermarket chain or steel mill. However with AI the nature of competition is different. The capital barriers to entry are low, in large part because companies of all sizes have access to the same essential tools. “Technology is the ultimate equalizer,” said Kutsishin of Sales Boomerang. “You can do more with less which means you can compete with larger firms on a smaller budget. Technology also gives smaller companies access to affordable resources that previously would have been too expensive. With freemium models being the norm you can take a
technology for a test drive before you have to fully commit and spend money.”
To make instant approvals work, lenders need access to huge amounts of data that can be obtained quickly and accurately. The big trick is having back- end systems in place and the ability to access and evaluate massive quantities of data in nanoseconds. All of these programs rely on big data and with each use they gain more data points and thus become more accurate, meaning the process is both additive and effectively without limit. Less Marketplace Risk? Does AI assure success from every transaction? Not at all. There are always events which can cause a transaction to fail. For instance, with a mortgage application the property may not appraise. The borrower’s bank might deny a lender access to its records and accounts. The borrower might go on a credit-bruising spending spree: a 2013 study by Equifax found that almost one- fifth of all mortgage applicants opened new credit accounts before closing. With AI new protections are beginning to appear in the system. For example, if lenders are collecting account data directly from banks, mutual funds, and stockbrokers it means there’s no opportunity for borrowers to hand in “modified” paperwork. While no one claims that real estate and lending can be made risk-free, it’s hard not to notice that risk keeps falling. The economy has firmed in recent years and Dodd-Frank has forced the riskiest loans from the marketplace, but are we also beginning to see AI’s invisible hand at work?
Google search is equally open to the biggest companies and the smallest. A lot of programming is cheap to own and free to try. Access to the cloud via such platforms as Microsoft Azure, Amazon World Services (AWS) and the Google Cloud Plaform is available universally and usually one can try a system without cost. According to Salesforce.com, with cloud computing “you’re not managing hardware and software” and like a utility “you only pay for what you need, upgrades are automatic, and scaling up or down is easy.” “Technology is democratizing everything in favor of startups, from computing and storage via AWS to global brand distribution via social media channels,” said Bryan Copley, CEO at CityBldr, a company that provides estimates of what a builder or developer would pay for a property based on its highest and best use. “Small companies have advantages in today’s economy that they didn’t used to have. With a differentiated product
The cost of technology has such a low barrier of entry that even a very small company can utilize the same tech as the largest companies.”
Ken Bartz Founder & CVO, Monster Lead Group
ATTOM DATA SOLUTIONS • P6
(and without legacy products to drag along) startups can fill market gaps less nimble incumbents haven’t filled.” Ken Bartz, Founder and CVO of Monster Lead Group, a company that mines public record property and loan data to identify leads for lenders, explains that
a canned technology solution available at a reasonable cost.”
“the cost of technology has such a low barrier of entry that even a very small company can utilize the same tech as the largest companies. The problem is exposure; some of these small firms aren’t on the tech radar so they need to search out solutions. If a small firm can identify their problem, there is probably
Does Fintech Really Work? A strange thing has begun to happen. It has become clear that fintech is not just a shiny new thing, it’s actually producing results. Copley writes that his firm “met with five neighbors and told them their homes were best sold together. Leading market valuations priced their homes at a total of $1.9M, but our AI said a developer would pay $3.2M to use the aggregated site to build apartments on. We were wrong — the developer paid $3.5M. That solution was a direct result of the investment we’ve made in machine learning.”
(We) met with five neighbors and told them their homes were best sold together. Leading market valuations priced their homes at a total of $1.9M, but our AI said a developer would pay $3.2M to use the aggregated site to build apartments on. We were wrong — the developer paid $3.5M. That solution was a direct result of the investment we’ve made in machine learning.”
Bryan Copley CEO, CityBldr
2016 Residential Prices by Property Type
Annual Pct Change
2016 Median Price Per Square Foot
Single Family Residence
Multi (2-4 Units)
ATTOM DATA SOLUTIONS • P7
as proximity marketing and mobile marketing become more accessible. Mail as a flat paper medium certainly has a shelf life but as long as the USPS is delivering mail it will exist in some format.” Copley explains that “your marketing should be a mutual reflection of your company and your prospective customer. That being said, even for tech companies, digital will not completely replace analog. If your customer responds well to mail, you need a strong direct mail campaign. If your customer drives around looking for property, signs are effective. And many buyers want to see a property before they buy, so the open house isn’t going anywhere. Many of these analog experiences will become seamless extensions of digital experiences, and vice versa.” The Alien Dreadnought There’s no doubt that the data revolution is about to change the world, as much as the smartphone and more. But the impact of new technologies is always complex and in the case of AI there are looming issues such as privacy, regulation, and security. But the biggest question of all, and surely the most visible, will be the jobs impact. Looking ahead, Tesla’s Elon Musk made an interesting remark last October in his company’s third-quarter earnings call. According to the transcript posted by Seeking Alpha, his goal in the future is to create an “alien dreadnought.”
Consumers, of course, are also AI drivers. If new technologies mean less paperwork and fewer hassles, if they mean lower costs — that’s great, bring it on. “As a customer I would love to get a call from my mortgage banker letting me know that without a shadow of a doubt I have enough equity in my home to get rid of my mortgage insurance (MI) or that based on my current FICO score I can get a better loan,” said Kutsishin of Sales Boomerang. “Or even more importantly, I would forever be grateful to the LO that calls me and tells me I am now qualified for the loan that I was denied for a few months earlier.” Will Traditional Technologies Survive? In fiscal year 2016 the Postal Service distributed almost 81 billion pieces of standard mail, what’s generally known as advertising mail. As it turns out, the largest users of direct mail are financial firms such as banks and credit unions. Speaking to eMarketer, Bob Dixon, Director of Product Technology Innovation for the Postal Service explains that
he’s not worried about the future of the mailbox.
“Mail is still the marketing channel with the highest response rate,” he said. “Without consumers receiving the mail, there is a $900 billion industry that ceases to exist.” If you think about it, mail is a “technology” and so are such things as real estate signs and open houses. So what happens to such traditional marketing approaches in the AI era? “These mediums will always be utilized in some fashion but they must become dynamic,” said Bartz of Monster Lead Group. “A real estate sign in the future may launch a virtual tour on your phone just by having a particular app, or in the vein of predictive analytics your phone may actually know you are looking for a home, the criteria for that home and then alert you when you pass a home with a sign.
“Open houses will still be useful,” he continued, “but the way they are marketed may significantly change
ATTOM DATA SOLUTIONS • P8
2016 Residential Sales Dollar Volume by Property Type
Single Family Residence
$0 2000 2002 2004 2006 2008 2010 2012 2014 2016
As Tesla production capacity increases, said Musk, “what really matters is the factory, the machine that designs the machine … becomes actually of greater significance, much greater significance than the machine itself. That’s where we have most of our engineering team working on. So sort of an internal code name for the factory machine that builds (the) machine is the alien dreadnought.” You can bet that the “alien dreadnought” will include lots of robots, software, and AI in general. And that’s great. But what about workers? “The overarching goal is to get past the limits of human speed,” explains Matthew DeBord, writing in Business Insider.
“A fully automated factory could, in Musk’s thinking, be operated by a few human experts, but otherwise, raw materials would go in one end and finished cars would roll out the other. In between, robots would do everything, at very high speed — speeds too dangerous to risk around frail human bodies.” when job opportunities grew, wages rose, home values went up, and profits increased. The idea was that a rising tide really did raise all boats — comparisons between productivity and wages show they tracked closely between 1948 and 1973 according to the Economic Policy Institute (EPI). In fact, during this period It used to be that an expanding economy was equated with good times, a period
EPI reports that productivity rose 96.7 percent while wages grew 91.3 percent.
What happened between 1973 and 2015 was very different: Productivity grew 73.4 percent while wages rose just 11.1 percent, according to EPI.
What if this trend continues?
In San Mateo, Reali has just received $5 million in Series A funding. What does Reali do? It’s a real estate brokerage that offers “data-driven insights across the complete lifecycle of buying, owning, and selling a home. Reali’s technological platform and app-driven approach fits today’s mobile lifestyle, and makes
ATTOM DATA SOLUTIONS • P9
the entire process more accessible, transparent and financially rewarding than ever before.”
of economic, legal and regulatory reasons. Furthermore new automation technologies will both create some totally new jobs in the digital technology area and, through productivity gains, generate additional wealth and spending that will support additional jobs of existing kinds, primarily in services sectors that are less easy to automate.” In effect, artificial intelligence is a two-edged sword. It promises great productivity and lower costs but it can also displace workers. It’s a trade-off that worries some people. “I think we should be very careful about artificial intelligence. If I had to guess at what our biggest existential threat is, it’s probably that.” Who issued this warning? According to the Daily Mail the speaker was Elon Musk, of course. Unquestionably, the AI genie is out of the bottle but maybe our job worries are misplaced. Perhaps we’ll wind up as we did a century ago when we traded horses for motors. A new technology changed the world back then and for most everyone it was an improvement that led to remarkably better lives.
Go back to that online loan application, the one that provided a solid estimate of borrowing ability in just a few minutes. Did the borrower speak with a loan officer? Did an underwriter review the file? Across America, says the Bureau of Labor Statistics, there are almost 306,000 loan officers, individuals who typically earn $63,650 a year. That’s nearly $19.5 billion in wages — or costs, depending on your perspective. With automation a lot of those jobs and salaries are at risk. How many? The consulting firm PWC estimates that as many as 38 percent of all U.S. jobs could be lost by the early 2030s. However, the firm also says “in practice, not all of these jobs may actually be automated for a variety
Reali says it “offers lower fees and cash back with a full refund of the buyer’s agent commission (typically 2.5 percent) that is paid back to the buyer. The seller’s commission is reduced to just 4 percent, saving tens of thousands of dollars in certain markets. Full mortgage brokerage service and warranty are offered to both homebuyers and sellers. Reali has made significant progress since its initial launch in late 2016, driving substantial benefits and cost-savings for its customers. On average, each transaction through Reali saved home buyers $31,335 off the cost of their home purchase.”
I think we should be very careful about artificial intelligence. If I had to guess at what our biggest existential threat is, it’s probably that.”
Elon Musk Founder, CEO, and CTO of SpaceX
ATTOM DATA SOLUTIONS • P10
The Role of Bargaining, Search and Psychology for Housing Prices
Jaren Pope is an Associate Professor in the Department of Economics at Brigham Young University. Jaren’s primary research areas are behavioral, environmental and urban economics. Much of his work has been focused on using property value information and quasi- experimental hedonic techniques to understand how households value environmental and urban amenities.
BY JAREN C. POPE, ASSOCIATE PROFESSOR, DEPARTMENT OF ECONOMICS, BRIGHAM YOUNG UNIVERSITY.
Understanding the role of key housing characteristics in explaining housing prices is of critical importance to real estate professionals. Two residential homes in the same housing market will sell for very different amounts if the two homes differ on key characteristics such as location, the time at which they sold, their size (both house and lot), the quality of their construction, their age, and many other characteristics of those homes under consideration.
According to economists, the functional relationship between housing characteristics and housing prices is the “Hedonic Price Function” (see Rosen (1974) for the seminal paper in economics). Economists use regression techniques to estimate this functional relationship in an effort to understand the relative contributions of specific housing characteristics to the sales prices of homes in a housing market.
ATTOM DATA SOLUTIONS • P11
of a home if a focal round number is within the price range over which the bargaining between a buyer and seller occurs, and thus may act as a natural solution to the bargaining process. Another possible influence on housing prices outside of measurable housing characteristics and psychological phenomenon might be differences in bargaining and search costs for buyers and sellers. For example, if a seller is time- constrained and is not able to invest as much time and effort into the bargaining or search process, they may end up selling their house for less than another seller that can invest more time and effort.
A real estate professional armed with this information could then make more informed real estate investment decisions since it could help to predict the value of homes in the market. While the large academic literature that estimates the hedonic price function has done much to further our understanding about the relationship between housing characteristics and housing prices, there is still substantial variation in housing prices that cannot be explained by these models.
housing prices beyond spatial, temporal, and measurable characteristics of homes. One area of research has been to consider potential psychological phenomenon that influence the final sales price of a home. For example, Sydnor, Pope and Pope (2015) document that there are large spikes in the distribution of final negotiated house prices at round numbers, especially those divisible by $50,000 (see figure 5 from their paper and reproduced for convenience below). This work suggests that round numbers can act as “focal points” in the bargaining process. The magnetism of these numbers can influence the final price
Recently economists have begun to explore other features of the housing market that may lead to differences in
Figure 5 - Sales Price Histogram. $375K-$525K. This figure provides a histogram for houses whose final sales prices were between $375,000 and $525,000. The histogram groups final sales prices into $1,000 bins (rounded down).
ATTOM DATA SOLUTIONS • P12
household’s financial position and in turn, its search and bargaining costs.
Certain elements in property data collected from county assessor and recorder offices might be used to proxy for differences in search and bargaining costs across buyers and sellers.”
In a preliminary analysis, a standard hedonic regression model was estimated, controlling for location of home, housing characteristics, time of sale, and a host of other characteristics that determine house price. Then an indicator variable was added to denote those houses that were owned outright by the seller. The analysis suggested that homes that were owned outright sell for approximately 3 to 4 percent less than other comparable homes! A similar exercise was performed for buyers that payed completely in cash rather than using some form of financing to purchase the home. Again, the analysis suggested that these homes
The same may be true for buyers that are unable to devote as much time and effort to search and bargaining as other buyers. If one were to have data on the time and effort spent by buyers and sellers in the housing market, this could be used to supplement the standard method economists use to estimate the functional relationship between housing prices and housing characteristics. Linking direct measures of time and effort for buyers and sellers with housing prices and characteristics is likely a daunting empirical effort;
however, there may be another possible solution for tackling this problem.
Certain elements in property data collected from county assessor and recorder offices might be used to proxy for differences in search and bargaining costs across buyers and sellers. For example, using mortgage and deed data provided by ATTOM Data Solutions, one can determine if a seller owns their house outright rather than having taken out a loan to pay for the house. This key characteristic might proxy for the
ATTOM DATA SOLUTIONS • P13
Share of Single Family & Condo All-Cash Sales
50.00% 45.00% 40.00% 35.00% 30.00% 25.00% 20.00% 15.00% 10.00%
Get all the Facts Before You Buy
For example, collecting information about buyer and seller characteristics and then linking that information to existing housing datasets such as those provided by ATTOM Data Solutions, could allow for a deeper understanding of the role that bargaining, search and psychological characteristics have on explaining and predicting housing prices.”
with cash buyers sold for 5 to 10 percent less than comparable homes.
For example, collecting information about buyer and seller characteristics and then linking that information to existing housing datasets such as those provided by ATTOM Data Solutions, could allow for a deeper understanding of the role that bargaining, search and psychological characteristics have on explaining and predicting housing prices. This could prove to be of great importance for savvy investing strategies in the future.
Criminal & Sex Offenders Former Local Drug Labs Nearby Hazardous Sites Quality of Schools Property / Loan Information
While far from conclusive, this preliminary evidence is suggestive of the importance of heterogeneity in buyer and seller search and bargaining costs. It is also suggestive of how real estate professionals and academics might benefit from digging deeper into the role of search, bargaining, and psychological phenomenon to better explain and predict the value of a house.
Get My Free Report
ATTOM DATA SOLUTIONS • P14
P r Public Records T a Tax Assessor D e Deed F c Foreclosure P m Plat Maps C I Cost of Living Index
L s Landslide E q Earthquake F i Fire N h S h Sinkholes P b Parcel Boundaries Natural Hazards
ATTOM Table of Data Elements
M I Mortgage Loan
F I Flood S p Spills H h Health Hazards A v Assessed Values
P f Pre-foreclosure
O w Ownership S c Schools C r Crime
E r Environmental Risks N c Neighborhood Characteristics C o Criminal Offenders
S f Superfund Sites D I Former Drug Labs
B f Brownfields A q Air Quality H c Home Condition
R p Registered Polluters U v UV Index B p Building Permits
U t Underground Storage Tanks R d Radon H v Home Values
F t FCC Towers
D g Demographics
P c Property Characteristics
P v Pre-mover
U S Utility Score
MLS Analytics M s
Public Records Environmental Risks
Property Characteristics Neighborhood Characteristics
Natural Hazards Health Hazards
ATTOM DATA SOLUTIONS • P15
BIG DATA SANDBOX
Prospective homebuyers in Q1 2017 were most motivated to move to parts of Colorado, the Carolinas and Florida — along with Washington D.C. — according to an analysis of proprietary pre-mover data by ATTOM Data Solutions using data collected from loan applications on residential real estate transactions.
*Historical percent of pre-movers that purchased home within 30 days of estimated settlement date
ATTOM DATA SOLUTIONS • P16
SPLITTING THE ATTOM
Streaming Your Own Property Data Playlist
BY TODD TETA CTO, ATTOM DATA SOLUTIONS
Two decades ago assembling a playlist of your 10 favorite songs was a bit of a nosebleed. First you would go to the (brick-and- mortar) store and purchase 10 different CDs, each containing one of your favorite songs. Or alternatively mail-order those same CDs through the now-defunct BMG Music Club — remember that! — and wait for them to arrive.
that track before ejecting and inserting the next CD (the recording functionality of tape cassettes had an advantage over CDs in this regard, allowing for the creation of those now wonderfully nostalgic mix tapes). Fast forward five years to 2002 and those same 10 songs on your playlist could be downloaded individually from the Internet thanks the advent of iTunes and iPods, followed soon by other data storage devices that allowed users to more quickly and selectively store music files.
Fast forward another five years to 2007 and those same 10 songs could be streamed over the Internet on-demand thanks to the rise of services such as Spotify and Pandora. And in the last 10 years the availability of that streaming has exploded thanks to those streaming services creating a multitude of Application Programming Interfaces (APIs), allowing developers to quickly integrate the streaming technology into third-party applications.
Second, pop each CD into a CD player, advance to the appropriate track and play
ATTOM DATA SOLUTIONS • P17
SPLITTING THE ATTOM
The API Economy A parallel trajectory has occurred in other industry verticals and has led to the rise of third-party API companies specializing in creating what TechCrunch calls “critical connective tissue” between content providers and developers to deliver applications providing integrated and data-rich experiences to end-users. Some of the emerging leaders in this new API economy include companies like Twilio (communications), Stripe (payments), and SalesForce (CRM). This is creating an environment of connected applications that can be built and modified quickly and leverage best-of-breed components and data. Meanwhile a similar — if admittedly slower and less sexy — evolution is taking place in the world of property data.
The majority of public record property data is still delivered to enterprise clients in bulk files via file transfer, somewhat akin to the CD delivery method for music 20 years ago. There has been some progress in recent years, much of it around leveraging APIs. Real Estate-Centric APIs There are a growing number of real estate-centric APIs that have been built over the past 10 years, providing developers access to very specific nuggets of real estate data. Those include the Zestimate, WalkScore, UtilityScore and ClearCapital’s property appraisals. ATTOM Data Solutions has its own set of public record property APIs that include a basic property profile, property history, AVM, and sales comps among others.
But frankly these static, predefined APIs are limited and do not make the breadth of ATTOM Data Warehouse available to our customers. Returning to the music delivery metaphor, it’s as if just a handful of well-worn music genres were available for streaming on Spotify; maybe just classical, rock, rap, jazz and alternative. But the entire spectrum of musical genres with all their nuance and color — brostep, drone folk, gauze pop, medieval rock, and vegan straight edge are just a few of the more creatively named genres on Spotify — are not available to stream. Clients would have to license Spotify’s entire music library to get at the eclectic set of genres and individual songs important to them — if those songs didn’t fit into any of the five pre-determined genres available for streaming.
There are a growing number of real estate-centric APIs that have been built over the past 10 years, providing developers access to very specific nuggets of real estate data. ATTOM Data Solutions has its own set of public record property APIs that include a basic property profile, property history, AVM, and sales comps among others.”
ATTOM DATA SOLUTIONS • P18
SPLITTING THE ATTOM
even true among companies operating in the same space. For example, one insurance company might be more focused on flood risk while another is more focused on earthquake risk, depending on their geographic footprint. Just a few years ago we might have been able to get away with the standard response to such clients: “To get all the data elements you want you’ll need to license the data in bulk.” And 10 years ago there would have been good reasons for that response. The technology to deliver highly customizable property data via API in large volumes was simply not available. increasingly avoid. The technology is now available to deliver highly customizable property data via API in large volumes (more on that in a minute). Furthermore it’s not in the customer’s best interest, nor is it in our best interest as a data provider to always default to delivering data in bulk files. There will always be some clients who want the data in bulk. Those clients want the full dataset, and they want to retain the ability to fully control, analyze and manage that dataset — something the just-in-time transactional nature of APIs does not accommodate well. But many clients do want the data on a just-in-time transactional basis — even if that means millions of transactions in an hour. Additionally, many of those clients Setting the API Bar Higher But that’s a response we want to
There are more than 1,000 potential data elements available for each of the 150 million residential and commercial U.S. properties housed in the ATTOM Data Warehouse.”
While somewhat absurd on many levels, this musical metaphor illustrates the archaic delivery methods still used for public record property data despite quickly evolving technology and client expectations. The Brostep of Property Data There are more than 1,000 potential data elements available for each of the 150 million residential and commercial U.S. properties housed in the ATTOM Data Warehouse. The common ones include elements such as property value, sale price, square footage, beds, baths and owner name. There are existing APIs already built to deliver these commonly used data elements as building blocks for applications.
Not so mainstream — even in the world of public record property data — include traditional public record data elements such as deed type, distressed description and partial interest type, along with non-traditional elements that have been correlated to traditional property data via what we call “ATTOMization.” These correlated elements include flood zone/ risk, wildfire risk and environmental hazard risk. Our increasingly diverse customers and prospects are telling us that they want to integrate many of these lesser-known data elements into their software and applications. Of course, not surprisingly, each client or prospect needs a slightly different combination of data elements. That’s
ATTOM DATA SOLUTIONS • P19
SPOTLIGHT: NORTH CAROLINA
don’t have the internal infrastructure or the human capital needed to ingest and manage a massive nationwide property database on their own. In the past, clients fitting the above description would have to settle for data delivery via the standard, static APIs available — or not get the data at all.
to allow us to create customer-specific “endpoints” that combine a custom set of data elements pulling from multiple data sets in the ATTOM Data Warehouse. Using this functionality we can “shape” data sets to deliver just the specific data elements from each data set that a customer wants to create their custom API.
elements it needs at each point it needs them — and saving costs while doing so. This type of comprehensive and customizable API delivery for property data is long overdue, and ATTOM is excited to be leading the way to making it a reality. This new API delivery will be via the cloud, but it will create a solid foundation upon which our clients can confidently and creatively build innovative new applications that will continue to disrupt their respective industries. We expect that providing this new API solution to our clients will elicit a reaction similar to Chris Pratt’s character Star- Lord/Peter Quill in Guardians of the Galaxy Vol. 2. After laboriously creating mix tapes for years -- 10 songs at a time -- he is given an iPod-like device containing more than 300 songs. He is elated when introduced to this amazingly expanded access to music. At ATTOM, we’re looking forward to elating our API clients by elevating our API delivery to the cloud.
Furthermore, a customer can set up multiple APIs to power different decision
We’re happy to say that will soon change.
This summer ATTOM will be unveiling a cloud- based API solution that will in effect allow our clients to create their own customized “playlist” of data elements and stream that playlist on demand. The entire spectrum of data elements will be available for creating these customized property data playlists, not just the pre-set data elements included in our existing static APIs.”
Elevating API to the Cloud This summer ATTOM will be unveiling a cloud-based API solution that will in effect allow our clients to create their own customized “playlist” of data elements and stream that playlist on demand. The entire spectrum of data elements will be available for creating these customized property data playlists, not just the pre- set data elements included in our existing static APIs.
points in its software or application workflow, helping to improve efficiency and save costs. Take for example a customer that has created an application to identify the best investment property opportunities. The application employs a stair-step workflow to increasingly narrow the pool of potential investment properties, with different data elements needed at each subsequent step in the workflow. With our coming API, ATTOM can quickly create a custom API for each step in the workflow, feeding the application with exactly the data
Todd Teta serves as Chief Technology Officer at ATTOM Data Solutions, where he leverages two decades of experience in technology and product innovation — at companies including Meyers Research, Andersen Consulting, VisionCore and CoreLogic — to lead ATTOM’s technology and product teams
The real power in the cloud API environment comes from its capacity
ATTOM DATA SOLUTIONS • P20
DATA IN ACTION
Highest Share of Co-Borrowers in High-Priced ‘Millennial-Magnet’ Markets
More borrowers purchasing a home are relying on “co-borrowers” to help them qualify for the loan, according to the ATTOM Data Solutions Q1 2017 U.S. Residential Property Loan Origination Report. Nearly 22 percent of all single family purchase originations in the first quarter had multiple, non-married co- borrowers on the loan, up from 20 percent a year ago.
“Throughout Southern California housing affordability continues to be a contributing cause supporting what has been viewed as an extremely tight available listing market year to date,” said Michael Mahon, president at First Team Real Estate, covering the Southern California market. “Increased competition amongst buyers for low available listing inventory and increasing multiple-offer scenarios are driving down use of leveraging mortgages in support of resale transactions while driving increased use of all-cash offers to gain acceptance over competing buyers.” Among the same 35 cities, those with the lowest share of non-married co- borrowers on single family homes purchased in the first quarter of 2017 were St. Louis, Missouri (4.9 percent); Memphis, Tennessee (10.3 percent); Atlanta, Georgia (10.4 percent); Mesa, Arizona (13.3 percent); and Tulsa, Oklahoma (13.7 percent).
Among 35 U.S. cities with at least 1,000 single family purchase originations in Q1 2017, those with the highest share of non-married co-borrowers were Miami, Florida (40.2 percent); Seattle, Washington (37.4 percent); San Diego, California (28.9 percent); Los Angeles, California (28.2 percent); and Portland, Oregon (27.7 percent).
Buying with Co-Borrowers in Q1 2017
Percent of Single Family Home Sales with Non-Married Co-Borrowers on Loan
Click on graphic to view interactive visual
ATTOM DATA SOLUTIONS • P21Page 1 Page 2 Page 3 Page 4 Page 5 Page 6 Page 7 Page 8 Page 9 Page 10 Page 11 Page 12 Page 13 Page 14 Page 15 Page 16 Page 17 Page 18 Page 19 Page 20 Page 21 Page 22 Page 23 Page 24 Page 25
Made with FlippingBook Online newsletter