The first cut is the deepest – Dublin’s falls and Ireland’s property paradox

This week’s report revealed some intriguing findings in relation to the current state and trajectory of Ireland’s property market. As was discussed yesterday, for example, while east peaked earlier than west, north has fallen further than south since the peak. One of the conclusions of both these findings is that Dublin and its commuter counties have experienced falling prices first and deepest.

This goes somewhat counter to conventional wisdom, although conventional wisdom hasn’t done too well in the last couple of years it must be said! Conventional wisdom would suggest that whatever about the Section 23 wastelands and ‘ghost estates’ of Ireland’s mid-West and elsewhere, the capital – as focal point for Ireland’s public and internationally trading sectors and their upstream and downstream employers – would be alright, at least in relative terms. In an Ireland where prices fell 20% in the crash, Dublin might be 15% or so while “somewhere else” would be worst hit.

While easy to mock, there is something in this from a long-term perspective. I have argued before on this blog – in December and again in February – that the ‘overhang’ of property looks a lot worse, even with just approximate calculations, in the mid-West than in the capital or indeed any of Ireland’s five cities. With stock falling slightly in the last six months, no harm revisiting the ‘overhang per county’ chart again, with stock levels taken from today.

Percentage of property for sale by county, Ireland, April 2009Again, the message is pretty clear – Cavan, Donegal, Leitrim and Roscommon have significant property ‘overhang’ compared to the likes of Monaghan, Kilkenny and Dublin and its commuter counties. The conclusion that I would draw is as follows: as it is home to the vast majority of Ireland’s top earners, to the extent that Dublin’s property market priced in expected future GDP and wage growth – i.e. confidence – it is to be expected that prices will fall most there, as confidence collapses from a high in late 2006 to a low in 2009. (The implication is that prices would be more likely to turn around faster, were confidence to somehow rematerialize.)

Taking a longer term perspective, though, unless prices adjust faster in places like Donegal, they face the prospect of longer peak-to-trough. Indeed already, some on are fretting about the future of places like Roscommon. On a thread entitled “Rents getting very cheap in the west“, mikewest’s message makes glum reading for property holders in Roscommon:

The house prices down here are still utterly crazy because something the developers never noticed is that there is shag all work in Co. Roscommon and if you dont have work then nobody wants to live there. People talk about the ghost estates in Longford and Leitrim but they don’t hold a candle to Roscommon. Every village and town has empty or virtually empty estates and / or apartment blocks…

There is another teeny tiny problem in the west. There are one or two houses too many in some towns right now so asking prices for rents are really more aspirational than actual but not quite as aspirational as asking prices for houses.

Lopping the top half off & Ireland’s property market in a global perspective

On Monday the latest report came out, showing that asking prices had fallen just over 4% in the first three months of the year. Yesterday, I changed focus on the blog a little, as it was Budget day, and tried instead  to put some numbers on what a potential property tax could raise.

Today, I hope to give a little more detail on the findings from the report itself, in particular regional trends, and then give an international perspective also – or at least start to give one, which I think is always instructive. Below is a graph showing the quarter-on-quarter change in asking prices for the last two quarters, i.e. Q4 2008 and Q1 2009, in each county.  The most obvious finding – probably not a surprise to anyone – is that asking prices fell in almost all counties in both quarters. A second clear finding is that there does not appear to have been one or two counties more affected in the last six months than elsewhere (although one could make the argument that Munster has got off relatively unscathed since September).

Quarter-on-quarter changes in house prices, 2008q4-2008q1

Quarter-on-quarter changes in house prices, 2008q4-2008q1

What also jumps out is that the two quarters saw very different patterns. In the final three months of 2008, a few counties – such as Galway, Westmeath and to a lesser extent Donegal and Leitrim – saw the largest downward adjustments in asking prices. Two counties, Mayo and Tipperary actually saw no fall in their asking prices. This quarter, Mayo and Tipperary actually had slightly larger falls than average – perhaps a sign that sellers there had been holding for the start of the year before acceding to the realities of the market. On the flip side, sellers in Galway and Westmeath believed in Q1 that their large adjustments in late 2008 did not need to be followed up with more adjustments straight away.

Sligo has been the worst hit county in terms of falling house prices, with a fall in the region of 10%in three months alone. (Dublin city centre and Waterford city actually saw bigger falls but they are lessened by other parts of their counties.) Aside from that, it seems that Dublin generally and the counties around it were among those with larger adjustments since the start of the year.

This leads on to perhaps a more interesting question – how have counties fared since their property prices peaked? To do that, I’ve set up another Manyeyes dataset (which anyone can access) with the percentage gap between house prices in a given quarter and the peak, for each county. Where a county is sandy coloured, that means it has peaked. The deeper the blue, the bigger the fall. (One little trick with these figures is that for a county’s earlier “blues”, prices are still going up. By the second row, that’s no longer an issue.)

Change in asking prices from the peak, 2007-2009

Change in asking prices from the peak, 2007-2009

A couple of findings emerge, based interestingly on alternate axes of the country:

  • East peaked before west, on average, and by almost six months. If you draw a line from Cavan down to Wexford, 10 of the 13 counties peaked in the first half of 2008, more than half the country in population terms, including all of Dublin and its offshoots. Cork, Galway, Limerick and a few other counties actually peaked in the second half of 2007, while a couple of stragglers – Tipperary and Westmeath to be precise – only peaked in early 2008. (Interesting to note, in passing, their sellers’ totally different reactions to conditions in late 2008, as per the first chart above.)
  • North is falling faster than south, on average. If you draw a line from Dublin over to Galway, 9 of the 10 worst affected counties so far come from that half of the island. The top half of the property market – literally! – has been lopped off more than the bottom half. This means that the north-east – essentially Dublin-plus – fell first and is falling hardest, while the south-west – Munster – was last to fall and has fallen least so far. It will be interesting to compare these emerging trends, two years into the property crash, with the final statistics on Ireland’s property readjustment/crash/Armageddon/return to sanity/fill in name here.

Speaking of writing the history books, perhaps it’s no harm to have a quick look to our left and our right and see how other property markets are faring. Below is a chart of about 20 countries (with two different measures in there for the US, the first is the OFHEO measure, while US* is the Case-Shiller national index). I’ve based this on data posted on the Economist’s website, but have surreptitiously replaced the 2007/2008 ESRI data, about which there is a lot of scepticism currently, with data. The bars show the annual rate of change in house prices, including a 1997-2008 average, and figures for 2007 and 2008. (As per the Economist website, some of the Q4 08 figures are actually Q3 08 while a couple, including Ireland, are Q1 09.)

International comparison of property markets, 1997-2009

International comparison of property markets, 1997-2009

Replacing the ESRI data with the had the effect of moving Ireland from the “Club of Moderates” such as Denmark and the Netherlands, to the “Bleeding Edge” group with Hong Kong, the UK and the US (at least one measure for the US at any rate). I will do my best to try and track down the original data for this series so that a change-from-peak measure can be contructed as again that may be more instructive than a year-on-year change, particularly in six months time.

In the meantime, though, I’ll leave this up here and ask for any insights, comments or queries, as per usual! Fire away…

Ireland’s property overhang: Homeowners in Roscommon, Cavan and Leitrim beware

Last week, I posted a visualization of changes in asking prices for Irish property, since 2006, using the IBM Manyeyes tool. It’s proved very popular, not least with the crowd on I’ve been happy to take suggestions on what’s the most important thing to be mapping and one suggestion – which ties in nicely with some ideas I’d been working on – was to measure the number of properties for sale by county, per capita.

What I’ve done for today’s visualization is take the number of permanent households from the 2006 Census – and taken it as fixed. I then plugged in county-level figures for the stock of property for sale from the database, and used the two to calculate an approximate percentage of the total property stock in a county that’s currently for sale. There are a range of potential data issues, from taking the Census figures as fixed to how to capture the size of new developments – it’s my hope that while all those issues are valid ones, the overall story should be relatively clear.

As before, the results are available for all to see and download on Manyeyes. It’s probably pretty clear what the overall message is, though, from the preview below (click on the picture to go through to Manyeyes):

Percentage of properties for sale, by county, Jan-Oct 2008

Percentage of properties for sale, by county, Jan-Oct 2008

Some initial thoughts:

  • First off, Dublin seems among the least affected areas.
  • “Holiday home land”, i.e. counties like Wexford, Kerry, Cork, Donegal and Galway, have seen their overhang increase over the course of the year, but again, they are not the worst affected areas.
  • A trio of counties, Roscommon, Cavan and Leitrim, however, steal the show. Those three, as of start-October, had more than 10% of their properties for sale.

Data on how many properties have churned through the market since the start of 2007 would probably confirm that the hysteresis which has gripped the Irish property market is worst in some of the areas where the property boom probably reached its most irrational. As before, all comments, questions and thoughts welcome.