Photographs Courtesy of Amherst Holdings

Meet the A.I. Landlord That’s Building a Single-Family-Home Empire

Shawn Tully (Jun 21, 2019 11:34 AM EDT)

Erin Burrus has endured some misfortune in recent years: After a cancer diagnosis, she lost her home to foreclosure. Today she’s healthy again, and a stable job in sales has helped her mend her finances. “I’m climbing my way back up,” says Burrus. One symbol of her stability is the two-­bedroom home she shares with her husband and their children in Greenwood, a solidly middle-class suburb of Indianapolis. The family rents the place rather than owning their home. But it was important to Burrus that they not be in an apartment. “I wanted to get a house with a yard for the kids, for that family atmosphere,” she says.

Burrus’s landlord is a company called Main Street Renewal; she found out about it from her mother, who rents a nearby home from the same outfit (and runs a thriving dress-alteration business with Burrus). And each is now playing a small part in an ambitious experiment.

Main Street Renewal is an arm of Amherst Holdings, 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. That portfolio makes Amherst one of the biggest, fastest-growing players in institutionally owned rental homes, a $45 billion subsector of the real estate industry that barely existed before the Great Recession.

Sean Dobson, Amherst’s CEO, is an imposing Texan data savant who dropped out of college to get into mortgage trading. A decade ago, he made a killing shorting shaky debt during the housing crash. Today he’s adding 1,000 homes a month to his empire with the help of artificial intelligence, using data modeling to make dozens of offers a day on potentially profitable houses. The Main Street homes are a $3.2 billion investment that generates around $300 million in annual rental income, but Dobson harbors far bigger ambitions: “We want to get to 1 million homes in the next 15 years or so,” he says. While that figure reflects as much bravado as realism—it’s more than 60 times the number of homes Amherst owns today—the fact that it’s conceivable shows how much the housing market has changed, and how technology is helping investors profit from those changes.

The rise of the single-family-rental industry reflects profound shifts in the finances and attitudes of America’s families. 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. According to ­Yar­deni Research, slightly more than one in three households that would have been buying first homes before the financial crisis is now either renting or still living with their parents.

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. Its specialty is grabbing run-down properties in nice, middle-class subdivisions—guided by algorithms that help it avoid bidding wars and money pits—which it then spruces up for the new rental generation. Amherst’s typical customers are couples in their early forties with one or two kids and household incomes around $60,000. They’re paying an average rent of $1,450 a month. “That’s almost exactly what they’d pay on a mortgage and other expenses if they owned the house,” says Dobson. “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.

Single-family-home rentals have long been dominated by local entrepreneurs—mom-and-pop investors or groups of businesspeople who own and manage no more than a couple of dozen properties (and often as few as one). Historically, when bigger fish, such as hedge funds and real estate investment trusts (REITs), invested in rental housing, they focused on apartment buildings—larger assets whose bunched-together density made them more cost-effective to manage.

The housing crash of the 2000s changed the math. As hard-pressed households gave up on ownership, and demand for rentals increased, investors realized single-family houses could be a more stable income source than apartments. An empty unit is a money loser, and houses were empty less often. Tom Barrack, head of real estate investment firm Colony Capital, explains that in single-family homes, “families stayed for two or three years, versus six months to a year in apartments.” Demand has stayed high, he adds, in part because consumers who used to see homes as investments are no longer confident that prices will rise.

The business remains highly fragmented: Institutional investors own only about 2% of America’s 15 million single-family rental homes. But over the past seven years, those investors have amassed a substantial portfolio—some 300,000 houses in all. The biggest players include Invitation Homes, a REIT that’s the product of a merger of rental divisions of several investment firms, including Blackstone, Starwood Capital, and Colony Capital; American Homes 4 Rent; and Amherst. All these landlords use automated house-hunting to fuel their growth. But Amherst differs from its rivals in focusing its computer models—and its business model—on affordable suburbs in the solid middle of the U.S. housing sector.

Dobson spent his childhood far from those burbs, in a trailer in an East Texas state park where his family owned a campground concession. “My mom and dad rented cabins and sold gas,” recalls Dobson. “Then oil prices spiked, people couldn’t afford vacations, and that was the end of the redneck paradise.”

The family moved to Houston when Sean was starting high school, and his father bought him the toy that would change his life, a TRS-80 computer from Radio Shack. The device generated so much static, Dobson says, that the family’s TV picture dissolved when the computer was running. But he became an expert programmer, and the summer after his high school graduation in 1987, he got an IT job on a mortgage-­trading desk. He became a pioneer in building sophisticated models to price home loans—and in using those models to find instances when investors were mispricing mortgage-backed securities (MBSs) based on faulty projections of their risks.

In 1994, Dobson founded the forerunner to Amherst, and by the early 2000s, Amherst was selling $25 billion a year in MBSs to pension funds and insurers. 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.

Dobson’s front-row seat at the housing collapse helped him recognize the opportunity in rentals. By 2011 he had begun a campaign to persuade investors to finance a new venture—a fund to buy and rent out single-family homes on an industrial scale. Some of his former partners saw the potential. “Single-family rentals are basically a big information game,” says Curtis Arledge, head of Mariner Investment Group. “You collect all kinds of information if you buy at scale. That data gives him a competitive advantage.”

Most were far more skeptical. To bolster his campaign, Dobson had purchased 215 houses in Phoenix and Dallas. “The portfolio wasn’t ideal,” he concedes. “We had graffiti-scarred houses in the inner city and houses in the suburbs six miles from the nearest house [we owned]. Did I mention that at least one dwelling was a former bordello?” Many investors saw the motley collection as epitomizing everything wrong with being a landlord—the deterioration of the properties, the hassles of maintaining a far-flung portfolio. “They said I was nuts, that this was an impossible business that would suffer ‘death by a thousand cuts,’ ” Dobson says.

It took a year of hard selling for Dobson to raise $200 million. But that seed money was enough to prove his concept. His first properties yielded enough profit to persuade investors to finance future rounds. Since 2011, Amherst has raised eight rental-­housing funds totaling $5 billion. In most cases, it has partnered with a single big investor—among them, private equity giants like TPG. The funds have produced average annual percentage returns in the mid-teens on their cash stakes, according to investors, including income from rent and price appreciation. (Amherst occasionally sells packages of homes when prices rise sharply, including to other investors.) And those returns are bigger than they would otherwise be, thanks to the firm’s digitally driven bargain hunting.

On a drive through Arlington and DeSoto, two Dallas suburbs, Amherst managing director Joe Negri is quick to point out the fixer-uppers. About one house in five qualifies. Negri shows me the classic signs: bedsheets stuffed in the windows, rusting AC units in the side yards. On the inside, he says, we are likely to find glued-down vinyl tiles peeling off the concrete floors.

Finding shabby abodes like these and making them respectable is the load-bearing wall of Amherst’s strategy. Amherst depends on humans to find cities, towns, and neighborhoods where fixer-uppers can become profitable, then relies on automation to pick individual homes. Negri, 31, heads the human team. He spends 150 days a year on the road overseeing Main Street ­Renewal’s operations from Atlanta to Denver, 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: Atlanta and Dallas, combined, account for about 5,300; Houston, Charlotte, and Jacksonville are also big markets. Amherst also favors Rust Belt “metros” with a sturdy foundation of jobs, including Indianapolis, Louisville, and St. Louis. These markets are all shaped by forces that keep housing costs in check. In Sunbelt cities, new construction plays that role; in the Rust Belt, relatively modest economic and workforce growth keep housing cheap. Each is an antithesis to coastal markets such as Los Angeles and Boston, where a dearth of new building and ­su­perheated local economies inflate prices. Focusing on ­fixer-uppers in modestly priced markets helps keep Amherst’s all-in costs for each home, including repairs, remarkably low, ranging from an average of $140,000 in Memphis to $208,000 in Dallas. (The median existing-home price nationwide is $267,300.) They’re almost always priced below the average in those markets too.

Making sure low prices aren’t a sign of economic zombiehood is Negri’s job. “The No. 1 criteria is diversity of employment,” he says, especially in blue-collar and middle-class jobs. Before Amherst chooses a new metro, Negri explores its neighborhoods firsthand. “I’ll live in a hotel for a month straight, driving around with an iPad,” he says. “I was driving the Florissant area of St. Louis early in the morning, and one out of every two or three people are dressed in Boeing uniforms. That gave me a lot of confidence.” A dealbreaker: cars sitting in the driveway in mid-morning, a sign that a lot of residents aren’t getting paychecks.

Based on research like Negri’s, 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.

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.

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.

If that yield meets Amherst’s target (which Fortune estimates is between 5% and 6%), the team will make an offer. About 20% of each day’s listings qualify; Amherst bids on those candidates no more than 12 hours after they’re first listed, making all-cash offers. Around 10% of its offers—on roughly 30 homes a day—get accepted and go to contract. Amherst dispatches inspectors to assess each home’s condition during the grace period. Unless they find fatal surprises—such as a cracked foundation—the houses pass muster and join the Main Street Renewal portfolio.

Once you own a fixer-upper, of course, you need to fix ’er up. Amherst spends an average $28,000 per home, roughly 20% of the purchase price, on renovations. Many of the middle-class families in Amherst’s customer base could amass the down payment to buy the same low-priced homes, but few would have the savings to also fund big improvements.

Touring a dozen Main Street ­Renewal houses in Dallas and Atlanta, I was impressed by how closely the homes, especially the interiors, resembled new construction. The houses all had different floor plans, but within each varied box, Amherst installed the same features: the kinds of fixtures and brands you’d find in a new middle-to-higher-end subdivision. In a six-year-old, 3,100-square-foot home in Douglasville, a suburb of Atlanta, Amherst had installed four gleaming new GE appliances: stove, dishwasher, fridge, and microwave. The countertops were thick quartz; the downstairs floors were sturdy ceramic tile; Hampton Bay ceiling fans whirred in the living room and master bedroom. The rent: $1,850 a month.

In nearby Austell, a smaller, cheaper, and older Main Street home—1,850 square feet, built in 1997—was undergoing a gut renovation. The carpeting was ripped and stained, and the vinyl ceiling in the kitchen sagged. But workers were installing the same appliances, flooring, and other features as in the Douglasville house. The Austell renovations would eventually cost twice what the Douglasville ones did. But that house would rent for $1,695—enough to reap the yield Amherst seeks.

Economies of scale help these renovations pay off. The improvements that cost Amherst $28,000 would, by the company’s estimates, cost a regular buyer at least $44,000. Because Amherst purchases in such high volumes, it can buy fixtures on heavily discounted national contracts. Its cost for the four GE appliances combined, for example, is $1,850 per home; a do-it-yourselfer would pay around $3,000 at Home Depot. The renovations are handled by outside contractors, but many rely on Amherst for most of their business, so costs are predictable and overruns are rare.

Amherst’s tenants also benefit from a time-honored privilege of renting: not being on the hook for repairs. In-house crews in each market handle most of that. In Dallas and its suburbs, a crew of 28 maintenance workers pilots a fleet of 10 white repair vans, each bearing the Main Street Renewal logo and each stocked with spare tiles, trendy “moth gray” paint, and ceiling fans.

Amherst has figured out how to serve a fast-growing new cohort of renters. The question facing Dobson is whether that cohort will keep growing.

Some experts think the downturn in ownership is temporary and that more millennials and families are on the verge of buying. That might not doom Amherst’s business model, but it would put the brakes on investor enthusiasm, says Ed Pinto, an economist at the American Enterprise Institute and former chief credit officer of mortgage agency Fannie Mae. “That Wall Street money is hot, not patient money,” Pinto says. “They will head for the exits or cut back on acquisitions.”

Dobson acknowledges that a surge in demand could trip up his strategy. If prices spike in the Bargain Belts, Amherst’s acquisition costs would go up. And since single-family rents tend to track home prices, its customers might choose to rent apartments rather than homes. “If home prices outpace income growth, we can’t get the rents to be profitable and grow our portfolios,” he says.

Still, Dobson doesn’t see such threats on the horizon, and he thinks most trends are breaking in his favor. If the economy slows, Amherst could benefit in two ways: Home prices would slacken, creating buying opportunities for investors, and rental demand would rise. Whatever the economy does, he argues, his industry will benefit as it scales up. He’s convinced that the pool of homes available to Amherst will grow by millions, as aging landlords whose kids have no interest in fixing toilets and dunning for rents opt to sell to the big guys. “I have $5 billion to $6 billion from outside investors knocking on the door,” says Dobson. “In the end, we’ll get to 1 million houses.”

However big the empire becomes, it’s unlikely to ever include Dobson’s own home. He and his wife and two kids share a baronial brick manse of more than 7,500 square feet, complete with wine cellar, in trendy Austin. It may not be huge by Texas standards, but it’s the kind of home that would never clear Amherst’s algorithms, in the kind of market Dobson the landlord wouldn’t touch.

A version of this article appears in the July 2019 issue of Fortune with the headline “America’s A.I. Landlord.”