Adelaide Advertiser Real Estate Page Count

[Sim]

Some of you might remember a while back (nearly 18 months now !) I published an update to my analysis of the Adelaide real estate market based on the count of pages in the real estate liftout of the Advertiser newspaper every Saturday.

Well, I decided to tidy up the study this weekend, and thought it might be time to get rid of the pile of newspapers sitting in the corner (several feet high - and that's just the real estate liftouts !).

I've updated my charts with the latest figures and I've even been able to start mapping out some basic trends, which show, unfortunately, nothing terribly surprising.

You can blame Michele_B for the idea - it was her who put me on to the observation that the number of pages in the real estate liftout changed quite significantly over the course of a year and suggested that it might be an indication of the state of the local market.

The theory is that the more pages there are in the liftout, the more houses there are on the market. This could mean that houses are not selling (dead market), but it could also just mean that they are selling, and more people are continuing to sell (boom market). This naturally shows the danger in making naieve analysis of the raw statistics in that you could have the same figure mean two totally different things.

The question is - do people tend not to sell if the market is considered dead ? Obviously there are people who will sell because they have to or decide it is time to trade up or move areas, but in a boom period I assume you would get a lot more people putting their house on the market because "the Smiths down the road got a great price for their house and ours is much nicer than theirs !!". So perhaps in a dead market you would not see anywhere near as many houses in the real estate section, and so the page count would be a lot less than in a boom time when everyone wants to sell.

I guess the only real measure of the state of the market is how long properties take to sell. In a really hot market, they would be selling within days of listing, and so may only ever be in the paper for one weekend advert. In a dead market, they would potentially take weeks or even months to sell, and so might continue to be listed in the paper week after week, which of course starts to mess with any analysis we might choose to make about what the number of pages means.

So what does it really mean ? Not a lot I don't think. It's just some more interesting statistics that you shouldn't really try and read too much in to I suggest.

Anyway, attached are two charts - the first shows the raw count of pages in the real estate sales lift out, plus the count of the number of columns of unit rentals and house rentals as well.

The second graphic shows the same view, but normalised back to a maximum of 1 (being the highest value for the observed period and 0 being the lowest value), and has the three different graphs superimposed on each other so we can attempt to see if there are correlations between the number

[Sim]

My observations from the last three years of figures shows that there is a clear spike in the number of properties being listed each spring. And this peak has actually risen each year for the last three years. This would indicate to me that the boom has been accelerating.

Notice the sharp drop off around Christmas time - this is mostly because of the lack of advertsing done over Christmas. Interestingly, the figures never get back to their peaks post-Christmas, there just seems to be a steady rise until the next peak in mid-Autumn.

Winter time sees a distinct drop-off in the number of properties being advertised.

The only surprising thing I see from these stats is just how high the page count has been this Autumn. Then again, this might be explained by the recent interest rate rises. Spring was a very busy time - the peak of the boom. Now after our Christmas advert drop off, we actually still have a lot of unsold properties out there since people have largely stopped buying given the scare the quick 0.5% hike in rates caused.

I suspect the page count will drop steadily until the bottom in Winter from now, unless we have some more interest rate rises in the next couple of months which may have an even stronger impact on the market.

Of course, I could still be just reading way too much in to all of this :D

[Sim]

From the second graph, you can quite clearly see the strong growth in the number of houses being advertised for rent around the middle of 2003, along with an equally strong drop in the number of houses being advertised for sale during the same period.

To me, this indicates the booming market - winter has fewer listings, so the strong purchasing for investment leads to a sales stock shortage. At the same time, there are a lot of new investment properties hitting the rental market, shown by the high numbers of houses being advertised for rent.

[Rickardo]

Very interesting statistics.

I won't try and read too deeply into them because i really don't follow the Adelaide market, but i'm sure if you graphed the same indicators of every capital city you would find similar trends in peak/trough selling times.

But generally this is a good research complimentary indicator that could help you gauge the way a market is moving, and compliment other research or your experiences within these markets.

Would be nice to see a Melbourne and/or Sydney display of this :)

[Aceyducey]

Sim,

You've clearly put a huge amount of work into this. Is it entirely based on column-inches & page counts? Have ads remained the same size over the period? Variations in size could easily distort the figures.

However your graphs reflect largely what I've seen in other metro markets, in terms of seasonal trends.

I wouldn't go reading too much into the details of your analysis either however :)

Cheers,

Aceyducey

[Sim]

Yes, the work was huge. Every Saturday when I pick up my copy of the Advertiser at my local newsagent I pull out the real estate lift out and the pages with the rental info in it and put them in a pile in my study. Time taken - about 60 seconds per week.

Then once ever 6 - 12 months (took nearly 18 months this time) I sit down, sort them into order and the run through them counting and adding the data to the spreadsheet. Last night it took me about 3 hours.

Total time for the last 18 months is about 4 hours and 18 minutes, or about an average of 30 seconds a day.

I think I can spare that :D

Yes it is a very simple measurement based on page counts and column lengths. Rental advert sizes haven't really changed - although sometimes for rental listings, agencies will put in a block of adverts rather than individual listings in the columns, but that doesn't seem to have much of an impact on the total count.

From what I've seen, the size of the sales advert listings hasn't changed much either. Even if they did, they wouldn't generally change from week to week, so the trends would still show, even if the total number of pages would be different for the same number of actual listings in a previous period when adverts were smaller. Either way, the number itself is less important than the trends anyway, so I don't worry too much about these things.

I am most interested to see what happens at other phases of the property cycle, if the same seasonal trends continue (which I expect they will) and what impact it will have on the size and duration of the peaks and troughs.

And yes, I would expect to see similar patterns show from other cities. Adelaide is relatively easy to track though, because there is only one place that properties are ever listed for sale or rent - that's the Saturday Advertiser (although it will be interesting to see how much the internet has impacted on the number of adverts they get).

[Sim]

mmm... now I have a rather large pile of newspaper that had been sitting happily on my bookshelf, but which my wife now wants removed from the house.

Some excuse about me no longer needing it now I've done my analysis. :rolleyes:

Guess a couple of trips downstairs to the recycling bins is in order. *sigh*

:D