The Washington PostDemocracy Dies in Darkness

Opinion Zillow sent its algorithm to take on the housing market. The housing market won.

The Zillow app on a mobile phone in Dobbs Ferry, N.Y., in May. (Tiffany Hagler-Geard/Bloomberg News)
Placeholder while article actions load

Americans who have spent their pandemic living out real estate fantasies through Zillow can take some comfort in the fact that Zillow was doing the corporate version of the same thing. Except that it actually bought a bunch of houses, instead of just looking at them, and lost buckets of money.

Now, the company is cutting its losses and closing “Zillow Offers,” launched in 2018, which made all-cash deals for houses and then fixed them up to sell. Schadenfreude aside, that’s a shame. Zillow, and its fellow competitors in the “iBuying” business — including Opendoor, Offerpad and Redfin — brought speed and efficiency to a slow and capricious housing market. That market is now one competitor down, with lingering questions about whether anyone will ever rationalize it.

Though some people have described iBuying as “house flipping,” what Zillow, etc. have been doing bears little resemblance to the theatrical gut-rehabs you may have seen on HGTV. Generally, their automated, data-driven approach is more common in easy-to-price newer homes in homogeneous developments. They then woo sellers with convenience and speed.

Putting a home on the market generally means making repairs, decluttering and staging for maximum effect, then clearing out for every showing. Even when you find a buyer, they often probably need to sell their own home before they can close. You become one link in a chain of transactions, any one of which can fail and tank your deal. Closing can take months.

Follow Megan McArdle's opinionsFollow

An iBuyer can close in a matter of days. Got a new job in Oregon but own your home in North Carolina? You can be on your way with a few mouse clicks. This has “huge social value,” says Tomasz Piskorski, a Columbia Business School professor who co-wrote a major research paper on iBuying. “There’s tremendous need for intermediation in this market.” It’s not just good for the buyers. Workers who can sell their homes easily can move to more productive jobs.

In return for providing that value, Piskorski’s data shows that iBuyers have been making a roughly 5 percent gross profit on each transaction — buying at a discount for speed, then selling a little above market. They’re still unprofitable once you factor in overhead. But the numbers were headed in the right direction.

So why is Zillow pulling out? The answer to that question will determine whether iBuying is the future of (some of) the real estate market, or a nice idea that never quite works. In solving problems for homeowners, iBuyers can create new ones for themselves.

Housing is a highly idiosyncratic market. Consider a street of Brooklyn brownstones, superficially identical. Inside, some are newly renovated. Others still have ancient plaster and eccentric plumbing. Some have good light, while others are shadowed by tall buildings; one is next to a nightclub, another in the comparatively quiet middle of the block.

Piskorski doubts blocks such as these will ever be good candidates for iBuying. Even with a really good pricing algorithm, iBuying might be able to cover only 20 percent of all home sales.

Or perhaps less. In a relatively new development, there can still be a lot of variation. An algorithm can’t tell you which houses have ugly views or unpleasant neighbors. The seller, however, knows. If they’re willing to sell at a discount, that might be an indication their house isn’t as good as the ones your algorithm is comparing it to. While Piskorski’s research suggests that iBuying is potentially profitable, this will be true only if companies can manage this adverse selection problem.

What if they can’t? Zillow may simply have realized before anyone else that adverse selection is intractable. If so, other iBuyers will eventually fail, too.

Of course, Zillow might also have belatedly realized it was competing with the real estate agents who provide most of the company’s profits, and may be getting mad. Or perhaps Zillow just flubbed the implementation — a haywire pricing model, an inability to get houses upgraded and sold in a timely manner. Companies that execute better might survive.

But Zillow also had some major competitive advantages. Pretty much every seller starts by going to Zillow to look at their “Zestimate,” so Zillow Offers had a built-in customer funnel. And the company had ways to make money even without a sale — channeling those who decline their offer to traditional agents who pay the company for leads, or promoting add-on services such as mortgages and title insurance. In theory, Piskorski told me, Zillow could make money in the iBuying business even if it sold homes at a loss.

With all those advantages, it’s surprising that Zillow couldn’t make a go at iBuying. So maybe there’s a reason the housing market is irrational: Like their owners, houses are too individual to be traded like commodities.