Ronan Lyons | Personal Website
Ronan Lyons | Personal Website

The emergence of regional property markets across Ireland

Last week, released its latest House Price report, which showed that asking prices fell another 4.7% in Q3 2009. Essentially, prices have fallen about five percent on average for each of the past 4 quarters. As I mentioned last week, though, this average hides regional variations. Some parts of Dublin have seen their asking prices fall by twice as much as some parts of Munster.

On the face of it, one could argue that this may just represent different strategies by sellers – in some parts of the country, they price low to get attention, in others they don’t bother and wait for the potential buyers to come in a certain discount below the asking price. There is quite a bit of evidence, though, that differences in asking prices do reflect differences in underlying transactions.

The first evidence is the total stock for sale across the country. The graph below shows the year-on-year change in stock for sale in four high-level regions. A positive number indicating more properties sitting on the market than a year ago – i.e. a bad thing. As you can see, Dublin is the only region with a significant reduction in the number of properties sitting on the market. In Munster, where prices have fallen least, the stock for sale is still rising.

Annual change in total stock for sale, by region
Annual change in total stock for sale, by region

The next graph compares the change in stock for sale and the fall from peak in asking prices to September at a more granular level – and thus more prone to regional data quirks. Nonetheless, the correlation coefficient – 67% – is quite high, suggesting that the relationship between falls in asking prices and stock for sale is not coincidental.

Changes in asking prices and stock for sale, by region, September 2009
Changes in asking prices and stock for sale, by region, September 2009

Economists think in terms of prices and quantities – most of the measures typically reported are prices ones (e.g. PTSB and daft’s own price index), but the two above are all quantity ones. Another quantity-based measure is to see what proportion of properties sell (or at least come down from the website) in a particular period of time. Back in the boom days, one might have looked at a period of 2 months or maybe even 6 weeks, but the more appropriate measure now seems to be six months. The graph below shows the percentage of properties that come down from the site within six months (not quite the same as sales but presumably the trend is representative), across different regions of the country.

Proportion of properties selling within six months, 2007-2009
Proportion of properties selling within six months, 2007-2009

The measure is a three-month moving average to give some stability. What’s interesting is that by this measure, the “least desirable” region to be selling a property is not Munster (as it was by the total stock for sale), but Connacht-Ulster, which has seen some large falls in asking prices recently.

Across both metrics though – and indeed median time to sell – it seems clear that the last six months have seen an improvement in market conditions in Dublin (in transaction, not price terms), but elsewhere there is only the hint that things may turn around soon. And it seems that bigger reductions in asking prices have had something to do with that.

What’s unclear, though, is whether there’s a further round of causality. If some other region had seen the biggest reductions, would it have improving market conditions, like Dublin is now? Or was it always going to be Dublin first, because of the size of its market and/or its dominant role in the Irish economy?

  • A Haythornthwaite ,

    Ronan,thankyou for this.
    I have a problem with the veracity of some of the data from
    If for example you look at co.wexford and draw up a list of all houses for sale on daft , it throws up 4447 choices.
    However,if you interpret the Census data there should be about 40000 house units in the county and a recent Goodbody’s report found that some 8 % or 3200 houses are available, taking both new and second hand into account.(There is a view that equilibrium lies at 5% . Dublin has fallen below this so prices are due to rise)
    I was trying to find where the difference lay and I returned to the listings and looked at how long they were on the market(Clicking twice on the date entered button wiil show you how long the properties have been listed.) It starts with the ones that have been listed for over 1500 days!
    This shows me that the Daft data is skewed to begin with,those properties aren’t actively for sale. Then one has to consider they are analysing prices being quoted rather than prices achieved.
    Marian Finnegan,Economist in Sherry FitzGerald prepares a review of a basket of valuations that we all contribute to from the offices all across the country(I am in Wexford town)and I believe this source gives a truer real time picture of where we are today.We have for example documented a fall in values from peak of 44%-dramatically more than Daft’s data would suggest.
    anyhow I look forward to following your views.Thankyou Adrian Haythornthwaite

    • Ronan Lyons ,

      Thanks for your comments. I guess it’s to be expected that, with you contributing to one dataset (SherryFitz’s) and me analysing another, we may differ on this.
      It’s worth bearing in mind, though, that properties only enter the daft house price index when they’re first posted. It’s not a simple average of all properties on the site at any one time. It’s an econometric method called hedonic price regressions (also used by the ESRI and most national house price indices), and the entries for a particular month are only newly listed properties or those that change their price during the month. Either way, the series captures a register of sellers’ expectations at any one time.

      The huge advantage such a series has is that it is based on the separate deliberations and judgements of thousands of people with their estate agents every month. I would suggest that, when surveyed, estate agents may be more naturally drawn to think of the larger falls, which make more a psychological impression than smaller but equally prevalent falls (hence the even smaller fall from the peak in the PTSB index).

      Thanks for the comment and the opportunity to explain what is and isn’t in the index.


      Leave a comment