Distressed loan trading is expected to see a substantial uptick in activity during 2021, with tech-forward traders looking to instant property condition assessments enabled by geospatial imagery and AI to gain a competitive edge.
In this post, we’ll look at how geospatial imagery and artificial intelligence are powering the rise of instant property condition assessments, the technologies behind them, and how traders can use these solutions to make buy and sell decisions quickly while choosing the right disposition path for each loan. This includes:
First, let’s look at market conditions that could spark increased loan trading in the months ahead.
Why 2021 May See a Surge in Distressed Loan Trading
The realities of vaccine rollouts to fight COVID-19, high unemployment, and the inevitable end to mortgage relief measures has the market preparing for what could be an unprecedented spike in distressed loan trading.
Indeed, if you think 2020 was a year of uncertainty, there’s little reason to feel better about 2021. Mortgage forbearance programs that have provided a financial lifeline for homeowners amid a pandemic-ravaged economy are rapidly approaching their 12-month expiration dates. According to the Wall Street Journal, more than half of 2.7 million active forbearance plans are set to end after the March expiration date.
The Biden Administration has already directed federal agencies to extend protections for their mortgage borrowers through June, and further extensions are sure to follow. But at some point this year, it’s likely these relief measures will start to be phased out. And that has the market holding its breath to see which path borrowers follow after expiration.
Will they extend forbearance, move to current, or fall into some stage of delinquency? The answers to these questions will significantly impact the timing and magnitude of what could be an unprecedented rise in distressed loan trading. So it’s little wonder that traders are looking for ways to optimize their earnings—including new technologies that leverage geospatial analytics to provide instant property condition assessments at scale.
What Are Instant Property Condition Assessments?
Instant property condition assessments provide the property intelligence needed to make real-time transaction decisions with trading-quality accuracy. But what makes this so game-changing to distressed loan trading?
It’s fairly straightforward: the greater the credit risk of a mortgage, the more critical the collateral value behind a loan. In looking at non-performing assets, property condition is a primary consideration traders should take into account. Unfortunately, getting a high-fidelity understanding of property conditions quickly, easily, and cost-effectively has always been next to impossible.
But a new breed of solutions that provide instant, accurate property condition assessments is changing all that.
Powered by AI, these solutions leverage high-resolution geospatial imagery and computer vision to extract structured data from the visuals. Think size of the property, type of roof construction and its condition, the presence of solar panels, whether there are overhanging trees, and whether the property has a pool, yard debris, and more.
The most robust of these solutions can provide instant access to tens of millions of properties nationwide. Our own solution, for instance, offers property condition data on more than 100 million properties across the US—delivered with the speed and coverage of traditional property record data.
But Aren’t Google Maps Good Enough?
It’s true that the pandemic has led to the growing use of geospatial imagery to better understand property condition. Analysts preparing to bid on loan portfolios will often review satellite images from Google Maps and retrieve “Zestimates” from Zillow, with its accompanying property photos.
But the basic assumption of most automated valuation models (AVMs)—including Zestimates—is that the property is in average condition, so analysts spend a significant amount of time looking at imagery in an attempt to discover the outliers.
This continues after having won the bid, too. Investors’ due diligence teams may rely on valuation providers to deliver images from various street-level angles, which are often incorporated into a broker price opinion (BPO). Additionally, diligence analysts scour the images these reports contain to glean new information to help further assess a property’s condition.
This can be a helpful exercise, but it’s nowhere close to the level of actionable intelligence provided by very high-resolution imagery analyzed with modern machine learning. To understand why let’s start by examining different types of geospatial imagery used for various kinds of analysis.
As you’re about to see, different image capture methods come with distinct advantages and disadvantages. And a manual review of these images by analysts creates its own set of challenges.
Geospatial Imagery Sources for Distressed Mortgage Trading
Not all geospatial imagery is created equal. On one end of the spectrum, there are high-coverage, lower-resolution imagery sources such as satellites. On the other, lower-coverage, or low-latency, high-resolution imagery offered in drive-by inspections. Let’s take a closer look at the pros and cons of each.
Satellites offer great coverage, providing overhead views of the entire United States and in fact, the globe. In the distressed mortgage market, we often hear people using tools such as Google, Bing, or Zillow to get eyes on a property. These channels utilize multiple imagery sources, including satellite imagery. Analysts who access commercially available satellite imagery are able to understand the layout of a neighborhood and parcel. They may also be able to key into other critical variables—is the property next to a landfill? Or a factory?
But it comes with a serious trade-off: Image recency. Satellite imagery can be wildly out of date. And update schedules? That’s anyone’s guess. Take the images below. The image on the left is over four years old. The age of the image on the right is just 18 months. The difference is striking: In the first image, the 16 homes in the new development are yet to be built.
Move the slider to see the difference:
Commercial satellite providers are reasonably priced and offer a top resolution of one square foot per pixel (with 30 cm per side per pixel, which is known as 30 cm Ground Sampling Distance (GSD)—but is more colloquially referred to as 30 cm resolution). Typical resolution ranges from 50 to 80 cm resolution.
This is another important source of overhead images. Multiple commercial aerial imagery providers rotate their fleets of airplanes around the United States throughout the year. Aerial imagery providers visit highly populated areas every few months, so the images offer good data currency at a similarly reasonable price.
A key advantage of aerial imagery is resolution. While the leading satellite imagery providers offer 30 cm resolution, aerial providers deliver resolutions of 7 cm or better. The images above showcase the difference in resolution between satellite and aerial. An analyst can identify debris in both images, but the aerial photo allows for the identification of critical details—debris type, driveway cracks, landscaping, roof condition—and more.
BPOs and PCRs (Property Condition Reports) typically include imagery captured in drive-by inspections. Both analysts and brokers use these images to determine the curb appeal of a given property. Often, they are used to influence the broker’s opinion of value. The resolution of these images is extremely high, typically below 1 cm square per pixel.
Here, the key distinction is cost. Drive-by inspections require sending a person onsite. Inspection providers have extensive networks, so most properties are in coverage. But while other sources of imagery are stored and instantly available, inspections have turn-times that vary by provider and geography. Generally, results are returned within a week, although less populated areas can have extreme outliers. The requirement that a qualified individual visit the property has a deleterious impact on the economics of this option.
Another critical challenge is the inherent limits the term “drive-by” implies—these images typically reflect only the view from the street. As such, drive-by images do not capture the entire parcel and include more human variability in capture quality than any of the other imagery collection sources.
Computer Vision: Assessing Opportunities at Scale
Regardless of how it’s collected, geospatial imagery is valuable to traders of distressed whole loans. But relying on human analysts to manually evaluate this imagery presents significant challenges—most notably speed and scalability.
Even for a highly-trained analyst, it can take at least a couple of minutes to evaluate a set of property images. But with modern computer vision technology, it takes just a couple of seconds to provide equally accurate and detailed information.
For those new to the term, computer vision automates the extraction, analysis, and understanding of useful information from images. Because computer vision offers the ability to analyze properties in seconds instead of days, evaluating images on large pools of mortgages becomes quite straightforward.
To put it into perspective: The 19,800 loans contained in a reperforming loan sale recently announced by Fannie Mae would take an analyst more than 300 hours to assess, even if it took just a minute per property. Computer vision would allow investors to evaluate these properties in under 19 minutes.
This kind of fast, reliable analysis makes more powerful and predictive models available, and frees analysts to catch outliers they would otherwise have missed and thus make more profitable decisions.
Turning Imagery into Actionable Insight
That’s not to say any of this is easy—transforming imagery into structured data is intricate and requires expertise.
Using our own approach as an example, we begin by stitching together a coast-to-coast view of the United States from imagery provided by our imagery partners. Then, we deploy highly-trained computer vision algorithms to identify key elements in the imagery relevant to risk and valuation. Next, additional data, ranging from parcel vectors to area hazards, are layered in to create a comprehensive view of the collateral/property. Finally, the outputs are returned as structured data linked to an address.
Impact of Geospatial AI on Risk
These property characteristics can greatly impact risk. As an example, let’s use the Roof Condition Rating (RCR) created by Cape Analytics. Cape’s RCR solution analyzes every roof on a five-point scale ranging from severe to excellent, as shown in the figure below.
RCR has proven useful to insurance providers, where policies on properties with a bad roof have had loss ratios three times higher than those on properties with good and excellent roofs. RCR has also proven useful in predicting mortgage risk. In one study of 443,274 properties, mortgages on properties with roofs in severe condition were twice as likely to default as those on homes with roofs in excellent condition.
Other variables have been advantageous as well. For example, using computer vision to identify trash littering a parcel using overhead imagery has helped analysts flag risk on properties that drive-by inspections failed to identify. The detection of negative condition elements early in the process allows analysts to make more informed bids and focus diligence efforts where they are most impactful. Clearly, catching these issues before close improves returns.
Further, these risks change over time. Transforming images to structured data allows portfolio managers to more easily monitor those changes. Understanding this change retrospectively allows analysts to trace the path of millions of properties, useful in model fitting and market analysis. Understanding this change prospectively enables easy-to-manage change detection, flagging portfolio risks early.
Deploying Property Condition Data into Analyst Workflows
This conversion of images into structured data allows risk analysis at a massive scale. With scalable cloud infrastructure, properties are analyzed quickly, and results can be integrated into customer workflows with ease. Data delivery is possible via API, batch files, or a customer web portal.
Often, combinations of delivery methods are used, where large portfolios are analyzed via API, and a web portal is used to hone in on an individual property and more fully comprehend the risk of previously unknown characteristics that could hurt investment returns.
With a potential surge in distressed loan trading on the horizon, the kind of instant property condition assessments enabled through geospatial imagery and computer vision could make all the difference in avoiding such risks and maximizing returns amid unprecedented opportunity.
To learn more about instant property condition assessment technologies and other trends in distressed loan trading watch our on-demand webinar, Deploying Geospatial AI in Distressed Loan Trading