comments 2

Zillow pauses algorithmic homebuying business

Zillow just announced that it has paused its (algorithmic) US homebuying business for the remainder of this year. The company acquired some 3,800 homes in Q2 of this year and, apparently, it now has a backlog of repairs and sales to work through. As a reminder, this business model, which is sometimes referred to as iBuying, is based on using algorithms to quickly value and buy homes (mostly online). The homes are then renovated and flipped for a profit. The problem, as most of you know, is that this pandemic has, among other things, disrupted construction supply chains and made it difficult to hire people. That has hurt the renovation component of this model.

Today’s news was bad for Zillow’s stock, but good for Opendoor’s stock, which is their main competitor. Opendoor subsequently came out and announced that they remain open for business. (Disclosure: I am long $OPEN). But this announcement is perhaps a good reminder that buying and selling real estate remains a different animal than, say, buying and selling stocks. And so there are some perfectly understandable reasons for why real estate hasn’t been disrupted by the internet in the same way that other industries have. Matt Levine does a great job explaining this in his recent column, “Sorry, Zillow’s Computer Can’t Buy Your House Right Now.”

Here’s an excerpt:

“I’ll pay you $350,000 for your house as long as a human can go out there, look around, and make sure that price isn’t wildly off” is an interesting model but it’s not quite the same as “push this button to sell your house for $350,000.” And “I’ll pay $350,000 for a house and then send out a crew to replace the carpets” is not quite the same as “I’ll pay $350,000 for a house and flip it 20 minutes later for $355,000, collecting a small spread for providing liquidity.” Computerization has come into the housing market, but it hasn’t taken it over yet.

One of the challenges is that the supply of homes is heterogeneous, even in a suburban community or in a multi-family building where you might have the same set of floor plans that repeat. Because maybe the home has been renovated and fit out entirely in gold. Or maybe it’s the opposite and it has been poorly maintained. There are variables to contend with that have historically necessitated more rather than less human involvement. Homes are also something that don’t trade all that frequently, which is less than optimal when it comes to online marketplaces.

But what if buying and selling a home was dramatically cheaper and easier to do? How often would people actually do it? Presumably more often. I agree with Matt that “computerization” hasn’t taken over the real estate industry just yet. But algorithmic homebuying still appears to be one of the more promising approaches.


  1. There’s also too much friction for the buyer: closing costs, interest rate variations, tearing oneself from one community to live in another or, if just flipping, contending with local laws that discourage absentee ownership (understandably) or things like Airbnb which reduce stock available to buy and compete with hotels.
    Rhove and others are attempting to make crowdsourcing available to the masses, but it’ll never be like buying & selling a stock.


  2. Pingback: Zillow exits algorithmic home-flipping business – BRANDON DONNELLY

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s