I’m thinking of doing a price/sq.ft. analysis of all properties on myhome.ie … or at least those for which the “beds-type” field on their search results have a format like this: “4 Bed Detached House For Sale - 154m²”.
I know there are folks here who have run analyses against some of the property sites. While I can easily rustle up the tools to do this, maybe someone could tell me: (1) has someone else done this already, (2) is there some easy way to get at a property database other than hit the website for some thousands of pages, 10 results per page (and will this have any nasty consequences), (3) any other tips/advice, (4) anything else I could usefully analyse while I’m at it?
Why? I am becoming increasingly concerned that we will enter a cash market in 2011 where the only relevant metric is the necessity to sell. The Irish banks are in seriously bad shape, the winter snow will have stressed many businesses - how can you talk of averages in the worlds most distressed and yet irrational market?
I wouldn’t waste all your talent and time and effort. This place is a powder-keg, and something’s about to give. The complexities that have caused our economic downfall go far far beyond just property and bad governance. If these concerns are adressed and clarified I think you will see some pretty hefty changes coming into place. Everything will change, while everything will remain the same. I know that probably sounds a bit conflictive but its not. But anyways, your idea is good, greater clarity and ease, right? That has to extend much farther than just property websites. One thing that has to be very clear, the days of property speculation have come to their final dead end.
I think this would be valuable as the market is going to become very utilitarian in 2011; by that I mean most buyers will only be buying if they have to and will focus on the question:
“Now that I have decided not to emigrate and not to rent, how do I house my family and make the most of the situation that prevails in this country”
So you are not going to see people “trading down” (because they won’t want to sell in the first place; they’ll wait for their kids to inherit), or people moving for aspirational reasons (they’ll emigrate first or sell to rent). Most of those smart enough to dodge the crash are also smart enough not to buy in 2011. The buyers pool should be overwhelmingly be those who dodged the crash but absolutely have to buy (new baby; already raising 2 kids in a 2 bed).
I suggest you add into the analysis:
Last known renovation date or which decade it was decorated in, judging from photos.
Garage present and / or converted
Number of photos posted (1 or less usually indicates a wreck).
Garden commentary. Whether “south facing” or “west facing” omitted.
CO and V4V, I take your points and of course there is a bigger picture, but I still think an additional evidence-based metric would be useful to have. It’s one that could be of value even when the powers-that-are-soon-not-to-be grace us with the fabled property price database.
FbyR and YRG, interesting suggestions. I’m kind of limited to what a piece of software can extract from the property listings. Floor area is one that is available in a structured way, whereas renovation date, garage and garden info, may be impossible to parse in a reliable way. (Although if anyone out there has written a parser for EA-speak and survived with their sanity intact, let me know ). Number of photos probably could be done – I’ll keep it in mind.
SM - was there any breakdown of build costs by region? Might be interesting to see if second hand asking prices bear any correlation.
I think this would be very very useful for apartments, where you’re not getting any external area.
For houses you’d need to work out some way of weighting in the size of the site, in fact I’d go so far as to say you shouldn’t be looking at price/sqft for houses, but at:
(Price of property - (Sqft of house * build cost per sqft))/the area of the site itself.
Basically a cost per sqft for the site without the house on it.
Unfortunately the size of the site itself isn’t usually given in the ads.
Ok, here – with major caveats – are some preliminary results. This is all the properties on myhome.ie for which area, price and property type were all given. (Only about one third of properties have areas). I’ve converted from sq.mtr. to sq.ft.
The myhome data was horribly buggy in many ways. They’ve entered sq.ft as sq.mtr. in many cases, and even some acres. To combat this I omitted everything under 50 sq.mtr and above 500 sq.mtr, and everything with a price per sq.ft. outside the range 40-1000. (A quick visual check suggests that this gets rid of many/most anomalies).
Sample size shows how many properties in each category – rows with low sample sizes should be treated as (even more) suspect. Anyway, FWIW:
If anyone wants the raw data I can provide a csv or (maybe) xml file. Data captured were: county/postcode, #beds, type of house/apt/site, sale/auction, tax type (section 23, pre '63 etc.), floor area sq.mtr, price, num photos on myhome, homogenised price/sq.ft, hyperlink to ad on myhome.ie. There are about 60,000 entries (including overlaps of counties/postcodes) and about a third have floor areas.
[EDIT: show houses and apartments in a single table]
[EDIT: better outlier filtering]
That’s fantastic work. I’m a big fan of the price per sq. ft., it’s the only proper way to look at the value of a place IMO. Although I prefer price sq metre, i’m more used to it
I lived in Holland for a while, and the main site for there is www.funda.nl. Houses are listed with size and you can compare easily. Best parts of Amsterdam around the Vondelpark area (Holland has unemployment rate 4%, no IMF, no ECB…) at around the same level per square metre as Dublin 4 or 6.
It would be interesting to do a comparison with other countries, if people have knowledge of the market.
With the same health warning as last time, here are prices per square foot scraped from myhome.ie for the beginning of April. I didn’t bother with apartments this time – I may do that for cities later. Q/Q change is given to the nearest percent. As before, treat small sample sizes as suspect. For instance, the 5% drop in Dublin South with over 1100 houses is more reliable than the 8% rise in Dublin 1 with only 14.
EDIT: I commented last time that “seems like in other counties you get more bang for your buck the more rooms you buy, whereas in Dublin you pay a premium for larger houses”. One thing I just happened to notice this time around is the big difference in overall size between 3 and 4 bed houses in Dublin. 4-beds are on average over 50% bigger, which I guess goes to show what a load of shitbox 3-beds were built as the standard Dublin house in recent times.