FEATURED ARTICLE: DATA-FED DISRUPTION IN THE MORTGAGE MARKETPLACE
plying its technology to that aspect of the business, according to Humphrey. “I can tell you that within one year of taking servicing back in-house, run over our platform, just about every single metric, whether it was reso- lutions, timelines for disposition of assets, recovery rates and even costs, we excelled against anyone that we had assessed externally,” he said, not- ing that the company’s wealth of data allows them to work more intelligently with a borrower facing a problem. While the investor that owns the loan ultimately makes the decision on how to proceed in the event of a default, Humphrey said that investors have come to rely heavily on the rec- ommendations given by LendingHome “because our data platform performs a lot better than just gut intuition.” Companies like Phoenix are taking a more data-driven approach to servicing in the larger mortgage marketplace, according to Lilly, who noted that the servicing response to delinquency may be different de- pending on the investor. “For many banks, the objective has been mitigation of risk and speed in which assets can be liquidated from the portfolio,” he said, noting that much of the non-performing loan vol- ume lingering from the last housing crash has shifted from banks to pri- vate equity investors who purchased those loans at a discount. “Investors and private equity firms work to maximize the achieved value through a reduction in processing leakage and identifying the optimal balance between condition, timing and value.”
define loss reserves,” he said. “Under- standing what corporate advance rates are at various delinquency levels is key to developing a strategy for retained portfolios. It is also important to understand what advances will be re- coverable versus non-recoverable and what time looks like for recoverability. “For advances that are not recov- erable it is important to understand what events may have been control- lable or uncontrollable and who had the ability to impact those events — sub-servicers, vendors, etc.,” Lilly continued. “Many times, errors or incomplete tasks can lead to indem- nification for events that resulted in curtailment and eventual loss.” SERVICING SCAR TISSUE Marshall with Clear Capital noted that servicers learned the hard way how not to handle a surge in defaults during the last housing downturn. “If you look at the speed at which servicers had to ramp up in 2009, 2010, as a result we saw a lot of blow- back on those processes,” he said. “Whether it was robo-signing scan- dals or unfairly evicting borrowers — all the junk that we dealt with.” The wounds from that experi- ence have forced servicers to adopt platforms and processes that should perform better in a high-default en- vironment, according to Marshall. “The scar tissue from those years will really pay dividends the next time around,” he said. “(It) forced us all to act with more rigor as we manage the
Our goal is to ready a new approach and technology-enabled platform to play offense to the next cycle.
RYAN LILLY
noting that Phoenix has been investing in data-driven tools for dealing with non-performing loans even though de- linquencies have been down in recent years. “Unlike most participants in this space, Phoenix has allocated capital earned during the crisis to reinvest it in REO through the application of auto- mation, machine learning and business intelligence. In the same way, we have been applying this logic to servicing oversight. Our goal is to ready a new approach and technology-enabled plat- form to play offense in the next cycle.” The servicing of loans backed by Ginnie Mae is a focus for Phoenix, said Lilly. That’s because owners of MSR portfolios with Ginnie Mae- backed loans are required to shoul- der the costs of processing defaults in advance of foreclosure, with those advances not recoverable until post-foreclosure — if at all. “Understanding the true cost of Gin- nie servicing will help organizations more clearly strategize efforts to attain the targeted ROI and more accurately
next default boom. If we go through this next cycle, my goal is that the industry can do it with more empathy so that borrowers are not as soured on home- buying as they were in 2008, 2009.” JUST GETTING STARTED Marshall and other experts inter- viewed are optimistic that, if imple- mented properly, data-fed innova- tion can ultimately improve both the user experience and reduce risk in the mortgage marketplace. “We are trying to very thoughtfully apply machine learning, AI. We are trying to artfully use some of these analytic tools to understand collateral change,” Marshall said. “Rather than our innovation used just to generate investment, (we are) keeping humans in the loop and iterating forward in a measured approach so we are intro- ducing true progress and not just in- troducing risk into our client’s shops. That’s something we’ve earned the right to do because we have shown that we care about client risk.” Stavinsky of Roc and Humphrey of LendingHome both believe the success their companies are experi-
We're just getting started; we feel there are endless opportunities to improve on current industry gold standard systems and processes. We can envision disruption in nearly every facet of the business including appraisal, closing, draw disbursement, insurance, just to name a few.
MAKSIM STAVINSKY
encing in the fix-and-flip lending lab- oratory can eventually be replicated in the larger mortgage marketplace. “I would say that we've built a gen- eral purpose residential mortgage platform,” said Humphrey. “While we have our technology and risk model today pointed towards the space of fix-and-flip, we're constantly look- ing at data and (considering) new products and new opportunities. So those new products, new opportuni- ties could be really close adjacencies to bridge — more on the construction side, rental side, things like that. Or they could, someday in the future, be 30-year consumer mortgages.”
“We’re just getting started; we feel there are endless opportunities to improve on current industry gold standard systems and processes,” Stavinsky said. “We can envision disruption in nearly every facet of the business including appraisal, closing, draw disbursement, insur- ance, just to name a few.”
PLAYING OFFENSE IN THE NEXT CYCLE
We are trying to very thoughtfully apply machine learning, AI … we are trying to artfully use some of these analytic tools to understand collateral change.
Investing in data and technology to drive decision-making for servicers is key, according to Lilly. “Assembling a complete dataset in combination with third-party tools can help clearly identify the right path to meet these objectives,” he said,
Daren Blomquist is vice president of market economics at Auction. com, the nation’s leading online
marketplace for residential bank-owned and foreclosure properties. Learn more at www.auction.com.
KEVIN MARSHALL
12 think realty housing news report
february 2019 13
Made with FlippingBook Online newsletter