Are active San Francisco listings distorting my 109-day sample?

S

SmallEmber

First-time buyer
I’m looking at San Francisco properties between $612,000 and $918,000, and the country-home listings in my sample appear to need roughly 109 days to find a buyer. Most of the obvious outliers seem to be vacant properties.

Am I over-weighting listings that are still online? Completed-sale data is much harder to find than asking data, so I’m trying to work out whether anything actually changed this month or whether the active stock is giving me a distorted picture.
 
Yes, active stock can distort this badly. Homes that sell quickly disappear from the active pool, while slow or overpriced ones remain and become an increasing share of what you see. Completed sales will also reflect agreements made earlier, so they won’t give you a perfectly current answer either. I’d compare recent completions with new-listing volume and withdrawn listings rather than relying on one snapshot.
 
I have checked the advertised dates, but it is still unclear whether the location groups are genuinely comparable. Moving the boundary by a few streets at this price level could bring in properties with different condition, layouts or buyer demand.

The other missing fact is the full listing history. A property may have been withdrawn and then returned under a fresh date, making its visible time online shorter than the real marketing period. I would give each property a narrow neighbourhood label and record any earlier appearance before deciding whether the 109-day figure reflects local demand or merely the way the sample was drawn.
 
I’m not convinced vacancy is the main explanation. It may just be easier to notice in the listings that have lingered. Condition, awkward layouts or properties that are harder for buyers to finance could all overlap with vacancy.

The useful comparison would be vacant versus occupied properties with similar condition and location, not vacant versus the whole sample.
 
I’d build three small groups: completed, still active and withdrawn. Record original ask, latest ask, first listing date, any visible price-cut date, condition and a narrow location label. Even without a large sample, that should show whether 109 days is being driven by stale stock, relisting, or sellers waiting too long before reducing.
 
The survivor-bias point is probably the obvious part I was missing. I had been treating the listings still online as the main signal because they’re easier to observe, when they may mostly describe what has not worked.

I’ll separate active, completed and withdrawn properties, tighten the neighbourhood groupings, and note price-cut timing. That should also stop vacancy from becoming a catch-all explanation.
 
One more caveat: days alone won’t tell you whether sellers became more realistic this month. A motivated seller who cuts early may complete before an older listing whose owner is willing to wait. Compare asking-price changes and seller behaviour alongside time on market. Buyer financing can also delay a deal without indicating weaker demand for the property itself.
 
You must log in or register to reply here.
Back
Top