Creating Confidence in Data & Valuation
HouseCanary Forges Ahead in the World of Automated Real Estate Data and Valuation
By Carole VanSickle Ellis
Real estate investors know all about waiting, but Jeremy Sicklick and Christopher Stroud, co-founders of valuation-focused real estate brokerage HouseCanary, really do not like to wait. That is particularly the case when it comes to waiting on information that could help determine the value of a potential property acquisition in a competitive market.
“As early as 2008, when I was still working with Boston Consulting Group helping real estate investors buy large land banks and deploy capital to buy real estate during the downturn, I knew there had to be a better way to value and underwrite investments,” Sicklick, CEO of HouseCanary, recalled. He knew that data and the ability to engineer effective, accurate data science would be the key to creating this type of product, and he also knew exactly who could help him design that sort of system. By 2013, Sicklick and Christopher Stroud, present-day chief of research at HouseCanary, had founded the company and begun making waves in the world of automated real estate data and valuations.
“When I met Chris [Stroud], he was a doctoral student studying economics,” Sicklick said. Stroud’s doctoral research revolved around dynamic models of financial risk and much of his research involved studying how financial market predictions might be improved using modern technology.
“I assumed at the time that all industries were already highly automated and being driven in some part by engineered data science,” Stroud said. “That was not the case in real estate. It was exciting to realize there was an opportunity to use my particular skillset to make an impact on what was at the time already a $30 trillion market.”
Just a few years later, the company was making headlines as one of the industry’s most interesting and successful technology and data-forward real estate brokerage startups, adding in data from Google Cloud in 2017 and establishing itself as a clear contrast to data providers like Case-Schiller, which, at that time, was using about 20 indices to estimate home pricing compared to HouseCanary’s 20,000.
“What we set out to do – and what we do today – is make an automated valuation model [AVM] available to our clients that is constantly updated, benchmarked, and adjusted to reflect the hundreds of thousands of real estate transactions that happen in the industry every month,” Stroud said. “We not only are able to provide investor clients with the best information available, but we are also able to provide metrics indicating how certain we are of that AVM so that clients can factor certainty and uncertainty into their investment decisions.”
Naturally, another big “selling point” for investors is the speed at which these AVMs are delivered. While certain real estate-data behemoths have been providing branded estimates of home value on demand for years, those estimates are notorious in the industry for being less than reliable. A consistently reliable, industry-agreed-upon standard like an appraisal can take weeks or months to obtain. For investors competing in hot markets or deploying large volumes of capital on tight deadlines, sometimes there simply is not time to wait. At times like these, an AVM can help bridge the gap in the decision-making process and get acquisitions and underwriting underway.
“We highlight to our clients that we identify how confident we feel in a valuation as well as how accurate we believe a valuation to be,” said Sicklick. “This is one of the main reasons many industry leaders are using HouseCanary’s brokerage services and valuation tools to do everything from underwrite properties to fully value them.” He explained that many investor clients have incorporated HouseCanary into their buying process to the extent that they can now “identify a property in a few seconds, fully underwrite that property, and be in a position to make an offer in 10 minutes or less.”
“That is where having trust, the right tools, and the right processes creates huge advantages,” Stroud added, noting seven of the top 10 single-family institutional buyers are using HouseCanary for “part or all” of their process. “We are constantly innovating to continue to improve speed and accuracy,” he said.
Knowing When to Pull the Trigger on Investment Strategies
HouseCanary prides itself on the volume, accuracy, and timeliness of its data, and one thing that both co-founders emphasize when discussing the company is how many of their investor clients use the platform not just to acquire properties but also to monitor their portfolios on an ongoing basis. “Our clients prioritize their ability to make good decisions about whether to keep capital where they have it or make alternative investment decisions,” said Sicklick. “They use our ongoing information stream to get a real-time view of what the markets are doing and identify the right time to ‘pull the trigger’ on changes in investment strategies.”
Clients typically use HouseCanary’s rental evaluation tools, which help investors determine current values of rentals and what they might expect in terms of appreciation and rental rate increases in the near term. “A lot of our clients have portfolios with thousands of properties,” said Stroud. “They come back to us on a regular basis and monitor that entire portfolio by looking at comparables for rents as well as values.”
At the same time, clients can receive information about new properties that have become available, evaluate those properties, and even explore their financing options using HouseCanary’s desktop underwriting system, Property Explorer. “This service provides a contextual property valuation that lets users choose their own comps and valuation methods based on their unique investment strategies,” Stroud said. The companion tool, Rental Explorer, operates in a similar manner, providing in-depth analysis and neighborhood-level market data on rental properties.
These tools not only help HouseCanary clients make good decisions for their portfolios; they are influencing valuation on a broader level as well. “There have been a number of ongoing issues around racial bias in housing valuation for a long time, and Fannie Mae and Freddie Mac released research that, in our view, indicated there is very real, ongoing, systemic bias around how properties are valued or undervalued,” said Sicklick. “Our automated valuation models and tools are able to remove much of that systemic bias as a result of how the system comes up with accurate, trusted, and fair values. The model does not care about ethnicity; all it cares about is the aggregate average price for a property. Because HouseCanary models may factor in tens of thousands of transaction pairs in order to identify a value, much if not all of the racial bias in the valuation is removed in the process.”
An Ongoing Journey of Constant & Dedicated Innovation
When HouseCanary was founded in 2013, the valuation industry was still largely reliant not just on physical experts visiting assets in real time but also on local comparable properties with which to compare a potential acquisition. This remains true – to a degree – today, but HouseCanary has spent the last nine years developing new and innovative options for investors and other players in the real estate space who need faster speeds, higher degrees of accuracy, and more information from their valuation professionals about just how fast and how accurate those professionals believe they can be.
One area of research that HouseCanary’s research department, headed by Stroud, was pioneering long before the terms “remote showing” and “remote appraisal” came into the common lexicon revolves around the use of image recognition and machine learning in the valuation process. “Through photo recognition, we are able to translate photos of properties into accurate analyses about the condition of that property,” said Sicklick. “That helps with the automated angle of things and increases trust and accuracy along the way.”
The photo recognition software uses computer vision models, which many readers may find most familiar in the context of accessing websites that use CAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) software to prevent bots from accessing platforms and prohibit automated attacks. In many cases, after filling in a series of numbers, a user will be asked to review a series of nine photos, usually in a grid formation, and select for a certain object like a traffic light or motorcycle. This process not only confirms that the user in question is not a robot, but it also confirms to the computer vision model that the images in question contain the requested object.
HouseCanary started its own highly customized (and much more elaborate) valuation computer vision model by running a company-wide initiative in which real estate experts in multiple departments tagged large sets of properties for specific issues and items.
“We asked them to tag photos based on the level of damage in the property, whether it was new construction, if there was lumber framing, and whether the image contained highly distressed conditions,” Stroud explained. “Our team worked through thousands and thousands of images to reach critical mass!”
Then, the research team created a set of models that could identify, with a high degree of accuracy, the condition of the asset in the photo and, even more importantly in some cases, what type of materials were present. Stroud noted that this was particularly useful for investor clients using the platform to make decisions about single-family residential acquisitions since the computer models are extremely accurate when it comes to rooms like kitchens, which have a large number of factors that contribute to overall appeal and value.
“Kitchens are so important to residents, and that makes them so important to our clients,” Stroud said. “Our computer vision models can see more than 10 different countertop types and assess the state of that countertop, and they do the same thing for appliances, flooring, etc. They do the same things an individual would do if they were flipping through photos, but they do it better and faster.”
HouseCanary’s computer vision models may sound like a triumph of data analytics – and they are – but for the co-founders, this achievement is just the beginning.
“Our focus is really automating the real estate transaction so that people can buy, sell, and finance in a much faster way by having a trusted valuation system and the tools to help underwrite a property,” said Sicklick. “To do that, we cannot stand still. We are constantly innovating and bringing together the data, the tools, and the technology to further automate the transaction and make it a better experience for our customers.”
SIDEBAR
Responding to the Need for Fast Decision-Making
When the COVID-19 pandemic reached the United States in early 2020, most real estate investors expected the housing market to take a hit, possibly even experiencing another tsunami of foreclosures similar to the one that crashed into the housing market in the mid-2000s. Instead, the market turned hard in the other direction, and, fortunately, HouseCanary was ready to help clients value houses in ways and at speeds they had often never used before.
“There was so much more demand to buy homes and there were so many more questions about how to do that,” Sicklick recalled. “The research team was constantly being asked what was happening in the market.” HouseCanary responded to this onslaught of interest in several ways, including the debut of the HouseCanary “Market Pulse,” which was published weekly for much of the pandemic and shifted to monthly updates in 2021.
“We took a view of housing on a market-by-market basis, just telling people what was happening and breaking down the demand metrics,” said Stroud. “It enabled our clients to make real-time decisions based on real-time data rather than waiting on quarterly reports. It was essential at a time when we had never seen those levels of fear and the need to make decisions so quickly.”
Although much of the initial mystery has disappeared from the equation as the pandemic has shifted toward endemic status, both Sicklick and Stroud believe its effects on housing will be far-reaching. “We remain pretty bullish on 2022 price growth,” Stroud said. “There has been a nationwide shortage of available homes for sale in the wake of COVID. We are down between 60 and 65 percent in terms of listings, and there is really no near-term end in sight to that supply shortage.”
Sicklick agreed, observing that housing supply would not begin to return to normal until sellers begin listing their homes at what would have been considered a “normal” rate prior to the pandemic. “We see affordability itself getting increasingly pressured by the end of 2022 and into the first half of 2023, at least,” he said. “There is going to be really strong demand for housing for the next decade or more driven by the sheer number of households – largely Millennial – that are being created.”
Read the latest Market Pulse report at HouseCanary.com/Resources.
SIDEBAR
HouseCanary Numbers & Recognitions
2013 — Year founded
125+ active employees at HouseCanary
98% transacted properties valued by HouseCanary that fall within Fitch Ratings 80% confidence threshold
Best Midsize Companies to Work for in San Francisco, 2022, Built In
Best Midsize Companies to Work for in Colorado, 2022, Built In
Best Paying Companies in Colorado, 2022, Built In
Tech100 Real Estate Winner, 2022, HousingWire
Tech Trendsetter — Jeremy Sicklick, 2021, Housingwire
Tech100 Real Estate Winner ComeHome by HouseCanary, 2021, HousingWire
Tech100 Mortgage Winner, 2020, HousingWire
Rising Star — Chris Stroud, 2019, HousingWire
Tech 100 Winner, 2019, HousingWire
