I’m generally skeptical of the prediction market philosophy, but current real-time housing data isn’t always indicative of the market. Using these markets to replace lagging indicators like Case-Shiller with daily price discovery might fill the gap.
While the 'betting' aspect of housing prices feels problematic to me, the resulting data might be one of the few ways to get a picture of pricing in real-time.
"If traded heavily enough by market participants and experts, a prediction market for home prices could provide valuable signals on where home prices are going," says Berner. "An efficient market is the best way to set prices and to quantify the wisdom of the crowd, so if this tool gained enough traction it could be an effective prediction tool."
The issue, at least for now, is that there probably won't be enough transaction volume for sufficiently fine housing markets.
One of the issues with prediction markets is that they work best on things people want to bet on. Prediction markets would need much higher participation to be as useful as they might end up being.
On top of that real estate is highly localized. These aggregates of nationwide data are interesting but in the context of people living in an area they have little utility.
The OP says predictable. What do they mean? A specific house in a specific market? Probably not but why would one want the macro to be predictable??
You’re absolutely right that real estate is hyper local. The market in NYC is very different from the market in Austin.
In this case, I see “more predictable” as real-time data availability on a market level. Having information on how “hot” (or “cool”) the market is would help inform pricing for sellers and offers for buyers. It's effectively moving from delayed snapshots to real-time market sentiment.
If participating in prediction markets was a normal part of daily life, then we might expect robust markets around average local housing prices, maybe even markets around average prices for certain types of houses.
I find that people do tend to be very interested in what homes in their neighborhoods will sell for, so maybe there could be functional markets on individual houses. I do have more neighbors than Predyx has users, after all.
That’s a great point. If the prediction market was part of the real estate app there might be more participation, especially among those who frequently listings. But it would be tough to get people to participate in prediction markets if they have to go off a platform where they’re already in the mindset (for example, from Zillow/Realtor/Redfin to Polymarkets).
Being able to embed external prediction markets into relevant apps could bridge that gap
I'm skeptical. You average user of Zillow today has no clue how off it is on price very often.
These apps are useful but in most cases my friends in real estate will tell you that it is wildly off. A good agent will run comps manually and even those estimates are often off. It's all guess work until someone actually puts down an offer.
And so much of what a house will go for is not data based. Presentation, irrational thinking, and fear play in. If you all are talking about betting markets on individual houses sale price I am skeptical. Not that someone would bet, but that it would have utility.
Agree that so much of real estate is non-rational and qualitative. People get very emotional when selling or buying a house.
I would also be skeptical of a prediction market if it were to be specific to individual houses. As you mention, comps absolutely matter because they take those hyper-local factors into account. Each house is unique, and the price is ultimately whatever a buyer is willing to pay.
Having a real-time macro signal for a whole city could potentially help people better understand market dynamics, but I’d agree that every house needs to be priced individually.
Redfin estimate is a good starting point for negotiations in my experience