It’s not everyday a business leader stands up and says: You know that huge business opportunity we’ve been ploughing cash into and thought was the bee’s knees? Well, we have to stop doing that thing PDQ, otherwise it might bankrupt the company.

That’s essentially the message Rich Barton, co-founder and executive of Zillow Group Inc., was forced to communicate Nov. 2.

His $22 billion company is shutting its Zillow Offers iBuying business — digitally-enabled home flipping, to you and I — after its pricing algorithms failed to forecast home prices accurately, causing large losses.

During a conference call explaining the decision to shut the unit, which sadly entails significant layoffs, Barton said the pricing errors had made Zillow look like a “leveraged housing trader” and its models and assumptions had “a high likelihood, at some point, of putting the whole company at risk.”

If that’s indeed a serious risk, then the only possible conclusion from Zillow’s humiliating about-turn is: It’s absolutely the right thing to do. Zillow still has a wildly popular real-estate listings website. That might not have the whizzy growth and total addressable market (TAM) of iBuying, but its margins are significantly better (the Internet, Media & Technology segment’s adjusted ebitda margin was 43% in the third quarter). There’s no shame in focusing on that, along with adjacent activities that don’t involve betting the company.

I’m sure this episode, and Barton’s response, will one day become a case study in business schools. (By its own admission, Zillow got into iBuying partly for defensive reasons. In hindsight, that wasn’t a good reason. But should companies stick to what they do best? Sure: Providing they’re willing to risk getting disrupted.)

At this point one important question remains unanswered: Is iBuying terribly risky, or was Zillow just really bad at it? This sounds like a cop-out, but my guess is it’s a bit of both.

The still relatively neophyte business of iBuying first grabbed my attention a year ago when Zillow’s big rival Opendoor Technologies Inc. said it would go public via a Chamath Palihapitiya SPAC, and giddy retail investors were referring to it as the “Amazon of housing.” (Palihapitiya called it his next 10x idea.)

iBuyers purchase houses, refurbish them and sell them again a few months later. There’s no question Opendoor, Zillow Offers and their ilk provide a convenient service — they make cash offers and much of the process happens digitally — but the rich valuations seemed to ignore how it’s low margin and enormously capital intensive.

These businesses need huge scale for the investments to pay off. But with scale comes even more risk, a point Barton emphasized this week. There are just so many houses held in inventory. What happens when the housing market suddenly cools and an iBuyer is left with a fast depreciating asset plus large inventory holding costs such as financing? The $300 million inventory write-downs Zillow has taken suggest things can go wrong pretty quickly.

I’m sure Opendoor will also be asked some version of the same question next week when it’s set to unveil third quarter results. It’s sure to put up a spirited defense. (A spokesperson told Bloomberg News that it was “well-positioned to meet consumer demand” and “open for business.”)

iBuyers stress they’re not to trying to take a massive punt on the housing market. They aim to make money from flat fees they charge sellers and buyers, not from rising house prices.

Right now, of course, appreciating house prices do create additional upside — because by the time the iBuyer comes to sell them, they’re worth more. That’s one reason Opendoor made a small adjusted profit in its latest quarter. Offerpad Solutions Inc. is also making decent money, considering it’s still pretty small. Just because — to use Barton’s words — Zillow took a “big swing” and missed, doesn’t mean these others will too.

Though the U.S. housing market has cooled a little recently, conditions remain relatively benign. That suggests Zillow fumbled its golden opportunity, due to a combination of useless algorithms and an inability to keep up with demand. In contrast, when the pandemic caused the housing market to temporarily shut down last year, Opendoor shed its inventory impressively fast (though this handicapped sales when the market quickly rebounded). With Zillow now leaving the scene, it has one less competitor to worry about.

Even so, after Zillow’s humbling, there’s sure to be a reevaluation of the risks of iBuying. That’s no bad thing. It’s better to have any vulnerabilities in this business model become apparent now. These companies are still selling just a few thousand homes a year. It’s scary to imagine what it would have been like if the weaknesses showed up after they’ve taken over the whole housing market.

Chris Bryant is a Bloomberg Opinion columnist covering industrial companies.