Hacker News Reader

Comment by michaericalribo | Hacker News Reader

No doubt Zillow’s data must be fascinating and deep and extensive.

In general, I’m curious about how an individual could use this type of data to their own benefit—to “game the system,” as it were, by surfacing / finding “outliers”, by which I mean properties that are renting / selling far below their “market-calibrated” price.

Then again, I guess developing that into a product is (ostensibly) the entire job of an analytics PM somewhere like Zillow...

Still, we (collectively) are close to it being straightforward to scrape this data, train a model, and score new listings as they come in. The tools are all there, the tooling to orchestrate it off-the-cuff for personal use is still a bit far off.

View on HN

One reality of liquid markets is that, whenever someone discovers an edge, it is (generally) only viable for a limited time. Either the original observer arbitrages the advantage into oblivion or multiple observers notice the same thing and do so together.

A more-reliable way to go is to determine whether a particular property is selling for less than it is worth to you. If it is, and there isn't a better option, hit it.

This is good advice. In economic terms, you always have an edge for your own personal purchases since you have superior access to information about your own preferences.

Real estate market's liquidity is far less than the stock market.

From what I see, as a compulsive zillow-with-24hr-filter-scout, the market is a lot less liquid than one might believe. There are a lot of economists' $20 bills laying around. (Especially during COVID, but also generally). Even some $100 bills. And they almost always sit on the market for at least 12, if not 24 hours.

My explanatory theory is that people who are very interested in getting great deals on apartments are also the same kind of people who sign long-term leases.

IMO, the biggest barrier to arbitrage, that allows these good deals to float on the market for several days, is that most people are not interested in pouncing on an apartment that's come on the market <8 hours ago. Breaking your lease (the rental market equivalent of "selling", a requirement for "selling up") is just way too expensive.

Brokers (the people who are ostensibly paid to arbitrage) are also not super interested in getting their client the best arbitrage deal, because their commission will usually come out lower.

All this to say: next time you're moving, consider hounding zillow in the mornings and evenings every day, with a 24hr filter, in the 5-6 weeks prior to your move. Your patience will usually be rewarded.

>Brokers (the people who are ostensibly paid to arbitrage) are also not super interested in getting their client the best arbitrage deal, because their commission will usually come out lower.

AFAIK it works the other way around. The commissions might be say... 10% lower, but if you save time (by not having to maintain the listing and attract buyers), you can more than make up for it in volume. This was covered by freakomics a while ago https://youtube.com/watch?v=aFYlgqv3T-w

I believe that GP was referring to buyer's brokers. In their case, if the client pays more, their commission grows. Furthermore, if they can close quickly, by making a higher offer or purchasing something that isn't quite the right deal, the buyer's agent increases volume, leading to the same effect.

I don't want to be the economist who doesn't pick up the $20 bill but the entire real estate market is based on trying to identify dislocations in price. Plus even if you write a python script that finds underpriced houses on average how many losers can you afford to finance and how long can you hold out for a buyer?

Zillow does house flipping on the side with their "Zillow Offers" product. Apparently it is now a significant portion of their revenue.

In another life when I was a consultant, I did work for a startup that scraped public data for exactly this sort of use.

They still exist and use nonpublic sources now, but built a nice subscription business scraping lis pendens notices and such. No reason someone else can't do the same.

>> "lis pendens"

For others who thought this was a typo: https://en.wikipedia.org/wiki/Lis_pendens

>> "In United States law, a lis pendens is a written notice that a lawsuit has been filed concerning real estate, involving either the title to the property or a claimed ownership interest in it. The notice is usually filed in the county land records office. Recording a lis pendens against a piece of property alerts a potential purchaser or lender that the property’s title is in question, which makes the property less attractive to a buyer or lender. Once the notice is filed, the legal title of anyone who nevertheless purchases the land or property described in the notice is subject to the ultimate decision of the lawsuit."

Reply on news.ycombinator.com