TR-HNR-March-2019

MY TAKE

Talented Tech: Alternative Valuation Solutions Give Lenders More Flexibility in 2019

ically find the best path to an ac- curate valuation for any particular scenario by training itself on Clear Capital’s unique data set.” An AVM built by economists or statisticians without much valu- ation expertise — the “stats-only approach” — may result in a model that is less accurate and much less scalable, since the intricacies in each market must be identified,

built into the AVM, and updated manually, often involving years of research and fine tuning. An AVM built by computer scientists with valuation expertise — the “ma- chine-learning approach” — results in an incredibly accurate model that is also quickly scalable, since the AVM can teach itself the intricacies from the market level down to the property level, build them into the AVM, and keep them updated auto- matically, in near-real time. “Our ClearAVM customers bene- fit from unprecedented quality, cov- erage, and confidence,” said Clear Capital CEO Duane Andrews. “The real beauty of the machine-learn- ing approach is that our model can automatically look outside of a typ- ical search radius and train itself to consider variables that may add the most value, like a property’s view. It’s constantly adjusting, even as we speak. The future is very exciting.” Clear Capital was able to build

Our ClearAVM customers benefit from unprecedented quality, coverage, and confidence."

DUANE ANDREWS

BY KENON CHEN

A utomated valuation models (AVMs) are effective property valuation tools. But they’re kind of like magic — enter an address, and an estimated value appears. For consumers, that simplicity is perfect. However, for financial in- stitutions that use AVMs for lending decisions, quality control, or other situations that carry risk, the me- chanics behind that simplicity must be well understood. Although lenders and regulators still approach them with (justifi-

able) caution, AVMs are gaining popularity as an effective and trustworthy solution for more and more situations. An AVM, especially when combined with a property inspection, can be an effective val- uation tool for home equity lending. An AVM can provide a quick and cost-effective way to value a pool of loans in a portfolio. An AVM can serve as an effective quality control tool or provide a quick, pre-valua- tion “gut check.” And, an AVM can make collateral review and under-

writing processes more efficient. Clear Capital recently announced a major addition to its analytics solutions with ClearAVM, a lend- ing-grade automated valuation model (AVM) powered by an ad- vanced machine-learning engine. “With rapid advancements in machine-learning capabilities, we saw an opportunity to bring AVM accuracy to a new level,” said Clear Capital President Kevin Marshall. “By embracing machine-learning, ClearAVM can quickly and automat-

a lending-grade AVM from the ground up because of its ad- vancements in data processing, machine learning, and nearly 20 years of property valuation lead- ership. ClearAVM rounds out Clear Capital’s ability to offer solutions that best fit the valuation scenar- io and property, especially when combined with its other analytics solutions and network of apprais- ers and real estate professionals who offer in-the-field valuation and

With rapid advancements in machine- learning capabilities, we saw an opportunity to bring AVM accuracy to a new level.”

KEVIN MARSHALL

16 think realty housing news report

march 2019 17

Made with FlippingBook Online newsletter