This Real Estate Robopreneur Plans to Purchase More than a Million Homes

Back in April of 2019, we put out a piece declaring We’re Living in the Era of the Robopreneur, setting forth a guideline to being an ultra-productive entrepreneur with software and algorithms by one’s side. We told the story through relevant Robopreneurs such as Casey Neistat, Kylie Jenner, and even Casper mattresses.

And then we found out about an incredibly savvy, early Robopreneur that’s still working his magic more than 3 decades under his belt as a Robopreneur.

Early Algorithmic Wins

Sean Dobson is the CEO of Amherst, a real estate investing firm with $20 billion under management. It owns or manages some 16,000 single-family homes, scattered across the Midwest and the Sunbelt.

Back in 1987, before intelligent data-modeling software was just a Google Search away, Sean was pioneering his own models to price home loans. Specifically, he was looking for instances of mispricing.

The seeds of his big score were planted during the housing bubble, when his models predicted a disaster in “Alt-A securities,” packages of loans granted to homeowners who had often refinanced multiple times. “The market was predicting a default rate of 5%, and our models showed it would be 30% [even] if home prices didn’t fall at all,” Dobson recalls. He recruited a group of investors that took short positions in Alt-A, reaping a $10 billion profit—10 times the investment, according to Dobson—when home prices tumbled.

Shawn Tully, Fortune

But Sean hasn’t taken his chips and left the table. Instead, he’s making an even bigger stack and betting big, once again.

The Million Homes Challenge

Main Street Renewal (an arm of Amherst) has set a goal to scoop up 1 million single-family homes, in the $140-200k price range, over the next 15 years. But unlike the TLC shows, Sean has no intention of flipping these homes.

Homeownership, long a bedrock of financial stability, has become unattainable or undesirable for many middle-income workers—for reasons including tighter lending standards, large college-debt loads, and lagging wage growth and savings.

These trends translate into roughly 5 million households that are renting single-family homes rather than taking out mortgages and building equity, and that’s Amherst’s target market.

“We’re catering to a whole new class of Americans—the former buyers who are now either forced renters or renters by choice.” And Dobson is betting that this new class is a permanent one.

Shawn Tully, Fortune

This goes against traditional real estate investing values of multi-billion-dollar funds. Generally, big funds will scoop up apartment complexes because of the density of units and the centralization of their maintenance.

His nontraditional view is really a bet on this massive change in consumer behavior.

What’s particularly fascinating is the process they have for finding homes – specifically the interplay between humans and algorithms:

[They’re] searching for “sweet spot” neighborhoods that combine affordable rents with a strong middle-income employment base. Around 70% of Amherst’s 16,000 homes are in Sunbelt cities.

Amherst depends on humans to find cities, towns, and neighborhoods where fixer-uppers can become profitable, then relies on automation to pick individual homes.

Amherst now targets around 1,000 zip codes in 30 metro areas. Choosing homes there is the job of Amherst’s highly automated purchasing system. In its 19th-floor office on New York City’s Madison Avenue, a dozen buying specialists screen leads on their workstations, delivered by a proprietary program called Explorer, an offshoot of the software Dobson developed to price mortgages. Each morning, the team gets alerts on newly listed homes that meet its price range and geographic criteria—around 1,400 listings a day.

Shawn Tully, Fortune

With listings in hand, they use more machine learning models to predict the ancillary costs and risk of these houses.

For each “first cut” listing, Explorer estimates the costs of renovation. This is machine learning at work: The estimate is based on Amherst’s experience with homes of similar age and size in the same or nearby neighborhoods. In an older home, this might include replacing the HVAC system; for one whose listing photos suggest wear and tear, it might include a new roof. (Team members help the software make that call.) Explorer has become so precise, Negri says, that the actual renovation costs average within 5% of the estimates.

Shawn Tully, Fortune

I want to make it clear that they wouldn’t be anywhere near as successful and accurate without their very rich, proprietary datasets they’ve collected over the years. This is truly the key driver to their algorithm-led approach. (All companies should be thinking “What proprietary data do we own?”)

Explorer also runs a separate calculation, finding three homes being rented within a two-mile radius that are close in age, size, and bed-and-bath specs to the newly listed home. Machine learning helps the software estimate what each house would rent for based on these “comps.” Explorer then churns out an estimated “rental yield”—the net rent after such expenses as taxes and maintenance, divided by all-in cost.

Shawn Tully, Fortune

Right now, their process is bearing about 3 home purchases a day (more than most people own in a lifetime). If they want to reach their 1 million homes in 15 years, they’re going to need to ramp up by more than 60x.

From what I’ve read about Sean Dobson, thus far, I’d put my money behind him. Not because he predicted the 08 housing bubble, but because he’s the pinnacle of what it means to be a Robopreneur.

I admire his ingenuity and foresight, more than 30 years in the making. I think this will jumpstart many other Robopreneurs in real estate. Especially with other real-estate-focused AI software emerging, such as Naborly, Skyline, Enodo, and even Zillow.