FEATURED ARTICLE: DATA-FED DISRUPTION IN THE MORTGAGE MARKETPLACE
“That's both because of the credit performance we've proven, and also the tech just stripped out so much cost of manually underwriting loans that we don't have to pass through to make those expenses back.” LendingHome’s origination plat- form eliminates traditional loan origination costs by leveraging data, including data on the property itself as well as on the prospective bor- rower and that borrower’s past per- formance with previous fix-and-flip projects. The data allows Lending- Home to “automate huge pieces of the underwriting process” while also constantly testing the accuracy of its upfront property value assumptions once the home flip project is com- plete, according to Humphrey. “We can oftentimes project very accurately what that after-repair value is going to be,” he said, adding that the data-driven efficiencies in the underwriting process can also help reduce approval timelines, which is especially critical in an industry oper- ating on extremely tight deadlines. “Borrowers in this space have to act fast; they're finding a property — it might be off-market — they have to find it, bring it to us and we have to be ready to give them an answer almost the same day, and ultimately close in let's say three, four, five, six days,” said Humphrey. “To give that level of service, to fully underwrite the loan with 20, 30, 40 different data points and the core credit attributes that we're looking at — that's hard to do without tech — certainly as you scale to the billions per year that we're now at.” LendingHome originated more than $844 million in loans in the first three quarters of 2018, up 35 percent from the first three quarters of 2017 and on pace for more than $1.125 billion in originations for the year, according to an ATTOM Data Solu- tions analysis of public record mort- gage data. Although LendingHome’s share of the overall mortgage orig- ination market stands at one-tenth
of a percent—in part because it's only focused on the niche fix-and-flip market segment—that market share has grown 48 percent compared to a year ago. DATA-FUELED DISRUPTION AND DISCOVERY Roc Capital is another recent en- trant to the fix-and-flip lending space that has experienced rapid growth since launching in 2014. “We add value by leveraging our deep Wall Street heritage, high-touch service, a custom end-to-end platform that makes lending fast, efficient and transparent,” said COO Maksim Stavin- sky, who noted that the company’s use of data-powered technology can be applied to many facets of the business. “We are using data and combining or- thogonal sources of data to help in the areas of risk, lead mitigation, fraud pre- vention, verification and sanity-check- ing of borrower-provided info.” Roc’s platform is driving disruption even in traditionally sacrosanct parts of the mortgage origination process.
“Our underwriters are now able to get preliminary title runs for submitted loans even before we get the report
from the title company,” he said, noting that these preliminary title checks are not used to close a loan but are useful as an initial check. Roc has also discovered additional opportunities within the space that can be supported by its data-driven approach, according to Stavinsky. “In addition to using data to support all aspects of our existing business, we also recently used data to get into a new business — insur- ance,” he said, referring to insurance for borrowers. “We worked with an A-rated insurance carrier to create an automated process for providing policies. Whereas our borrowers previously had to contend with a long application and days of waiting for a quote, today we are able to quote a policy in minutes, with minimal infor- mation required from the borrower.” The automated insurance approval process is fueled largely by the same property data Roc uses in other facets of its business, Stavinsky explained. “We have sufficient property-level data so as to obviate the need for the customer to provide any information,” he said. “Since most of our clients don’t love talking about insurance, this has led to great customer satisfaction. Insurance has never been this quick, easy and painless for customers.” tech-fueled startups such as Lend- ingHome and Roc begs the question: will the highly automated underwriting process these two companies and others employ open up a back door into the mortgage marketplace for the loose lending practices that ultimately led to the market’s downfall a decade ago? Both Humphrey and Stavinsky ar- gued that well-performing loans with low delinquency and default rates are critical for the continued growth of their businesses. That’s because both companies are selling the loans that LOAN PERFORMANCE LONG GAME The rapid growth of data- and
As we started to become an operating business and we started scaling, we were still always raising capital and so then we could actually point and say that data and very algorithmic approaches to underwriting this credit means that we can close loans in five days, not 20, but at the same time have two or three times as good loan performance.
We add value by leveraging our deep Wall Street heritage, high-touch service, a custom end-to- end platform that makes lending fast, efficient and transparent. We are using data and combining orthogonal sources of data to help in the areas of risk, lead mitigation, fraud prevention, verification and sanity-checking of borrower-provided info.
was an intentional choice the company made to ensure the loans originated by its platform performed well, position- ing the company for long-term growth. “We wanted to play the long game, not the short game,” he said. “I think when you look at a space like fix-and- flip that a lot of investors had thought of as these are localized country club loans, these are kind of a weird dark corner of mortgage. We brought that transparency and that clean look to the space, and we've really brought broad, large-scale capital to bear.” BETTER THAN GUT INTUITION The quality of a loan’s performance depends heavily on the quality of its underwriting during origination, but that performance can also be impacted by how the loan is serviced, particularly in the event of delin- quency or a problem the borrower is facing that may lead to delinquency. The importance of smart, data-driv- en servicing is a lesson LendingHome and Roc have both learned in the laboratory of fix-and-flip lending. “Internally, data is useful in risk management to monitor delinquen- cy,” said Stavinsky of Roc, which services the loans it originates. LendingHome did not start out servicing its loans when it began orig- inating in 2014, but it soon discovered it could boost loan performance by ap-
they originate, and investors buying those loans won’t keep buying if the loans don’t perform as expected. “We transparently open up a really complete set of data to the end loan investor who buys that loan, whether that's a massive asset manager with hundreds of billions of assets or it's an individual peer-to-peer inves- tor with maybe 50 or 100K on our platform,” Humphrey said. “They get to the full historical platform perfor- mance of LendingHome’s now well over $3 billion in loans.” Loan performance also comes into play when startups like Roc and LendingHome raise venture capital, both Stavinsky and Humphrey noted. “As we started to become an oper- ating business and we started scaling, we were still always raising capital and so then we could actually point and say that data and very algorithmic approaches to underwriting this credit means that we can close loans in five days, not 20, but at the same time have two or three times as good loan performance,” Humphrey said. “That's where I think investors start to see the true value of the tech and data and un- derstand it's not just (we’re) doing well as a lender. That's a real sustainable, competitive advantage.” Humphrey noted that, before origi- nating any loans, LendingHome spent nearly a year building its lending plat- form from the ground up. He said that
LENDINGHOME LOAN GROWTH
ESTIMATED MORTGAGE ORIGINATION DOLLAR VOLUME
Q1 to Q3 2017
Q1 to Q3 2018
10 think realty housing news report
february 2019 11
Made with FlippingBook Online newsletter