Why are Irish house prices still falling?

Gerard Kennedy and Kieran McQuinn from the Central Bank give their view on where house prices should be in Ireland, and why they haven’t stopped falling in this new paper (.pdf). This paper is a kind of update to Morgan Kelly’s 2007 piece on the likely extent of house price falls (.pdf) as well as other papers.

From the abstract of Kennedy and McQuinn, then:

In this note, the continued fall in Irish house prices is examined. The increased rate of decline in 2011 resulted in Irish prices being almost 50 per cent down from peak levels of mid 2007. Accordingly, in over forty years of house price data, the fall is now one of the most significant across the OECD. We outline the current state of activity in the housing market and, using a suite of models, assess whether the fall in house prices is in line with that suggested by current fundamental factors within the Irish economy. Given that the analysis suggests prices may have overcorrected since 2010 we discuss possible reasons for this continued decline.

By Stephen Kinsella

Senior Lecturer in Economics at the University of Limerick.

47 replies on “Why are Irish house prices still falling?”

Really sad that the money supply must be tied to banks favourite extraction mechanism.
If the state was divorced from these assets (became a national economy again) it would not care what price these “assets” reached.
In the case of Japan I beleive they introduced new base money into the system for public work projects and the like.
While many of these were corrupt Turkeys or roads to nowhere they were far better investments (see Richard Koo) then the consumption sinks known as houses who in many cases have a higher input cost then their rental yield i.e. they have a negative value.

If you want to save a Town such as Youghal you spend new base money on something rational in a +100 $ world.
Otherwise the place rots into the ground.

You also add base tokens to peoples checking accounts and tax waste (think rising home fuel oil tax to road diesel prices)

We need to get back to a Fiat currency….. this will create a far more rational national economy where you issue and tax as opposed becoming bagmen for serial credit hyperinflationists of the Lawson variety.
The problem now is we don’t have a rational economy – its merely a conduit for fluid international capital flows.
This unfortunetly means that rationality is not factored into the economic equations.
Although you could argue this is very rational from a criminal perspective.

I heard reference to this on RTE radio lunchtime news. I was curious about the four models being used. Perhaps what I heard was different to what was said, but I got the impression that it was some international set of models had been used. I was a little surprised as these had somehow escaped my attention til now.

Glancing through this report, it appears these are just ‘in-house’ models developed during the bubble. My bet is that they are about as effective as the Central Bank was when these models were (eh) developed. (Hint: given arrears levels, low interest rates don’t necessarily determine affordability – I could go on but this post would get too long).

Dismissing the content, it is prudent to ask what is the purpose of this report? The CB has a strong interest in seeing house prices stabilise/ rise. Price falls are getting too close to the Blackrock adverse scenario for comfort. I’d guess Gov Honohan would prefer not to put more capital in banks. Similarly taxpayers (many being homeowners) would benefit from higher prices in some ways. But wanting something does not make it happen. Comparing Ireland with other property bubble countries does not work as it ignores the currency union – both in the growth of the bubble and, especially now, in the inability to devalue.

Given “…it emerged that the Central Bank has given the country’s main lenders until the end of next month to outline how they plan to deal with the growing problem of mortgage arrears.” (http://www.irishexaminer.com/breakingnews/ireland/banks-have-no-alternative-to-writing-off-some-mortgage-debts-549582.html#ixzz1tXCcf2qw), will the central bank, based on their own research, advise banks to horde repossessed properties?

@PR Guy: You are being most charitable.

Its mathematical poofery. Irish residential properties increased (in aggregate) 325% – if you believe the data. And it would be hard to refute it.

That’s a bubble in anyone’s understanding, so a correction will occur. But for how long, and for how far (down). The answers are in the realm of astrology. But a few pointers are available.

Housing bubble corrections retrace for about x 1.5 times the time it took for the bubble to form. Lets be generous and say that was from 2002 until 2007 (actually it was from 1995 until 2007 with a two year pause). So, that’s a 7.5 year downturn-time, minimum. We have done (I am being generous again) 5.5 years. So, a bit to go yet. My money is on 2015 – absent a significant energy shock.

The timing of this? Do I need to consult my millner about some metallic foil headgear.? If it were fashioned from tin, it would be a tad expensive. Tin is a truely beautiful, wonderful metal. Aluminium by contrast, is brash and nasty. But I’m poor! 😎

However. Terms and Conditions apply, as they say. We had the Mother and Father of a credit boom. This produces a lot of debt. And we have a nasty predicament with this debt. Unless it clears, no property recovery – except in the fevered imaginations of student statisticians, who steadfastly belive in the construct of The Meaningless Mean. [With sincerest apologies to the late Mr Gosset.]

I hope this report gets the shredding it deserves.

It is hard to gauge how much prices have really fallen because most properties cannot be sold given the limited supply of buyers with cash piles.

However, I have recently come across properties selling for more than their asking price and have heard anecdotes of others being hard to buy.

Accordingly, I don’t think prices are falling. We appear to be bouncing along the bottom in a lumpy market. (Note – “lumpy” is a current favourite with lingo lovers.)

With that said, further falls cannot be ruled out as the property tax situation evolves.

There was a report in this week’s Economist which I read which suggested that Irish house prices have another 20% fall to go. Only then would they be at the correct value. Given that markets tend to over-correct as much as over-hype on the way up, we could see a 30% fall in Irish property prices over the next 18 months before we truly hit the bottom and can begin to see an upward trend resuming.

In any case, I would say that the recovery will be regional – Dublin has come the furthest through this crash and will be the first place to benefit from a recovery. I think Dublin prices may bottom this year, stay flat for another two or so years before starting to rise again.

The authors should put the estimation results for the four pricing models on an accessible file somewhere. It is impossible to get the full picture from this CBI paper/note without the model specifications and model estimation details. Describing the current-period model residuals without showing the model specifications and sample-period estimation results is too limiting.

‘using a suite of models’….

I can’t think of any combination of models that can describe the mess we’re in. The only piece of data that I see encouraging about property investment is rents holding up despite Joan Burtons rent allowance cuts. Though we’re only 4 months in to the year.

Who among those people in mortgage worthy employment at the moment have enough faith in this country to consider getting a mortgage now? After seeing what mortgages can and have done to friends, family and fellow citizens.

For me it’s 2 problems. No/low levels of credit and a fear of credit developing in society.

I can’t believe we will see either supply (of mortgages) or demand (from purchasers) until we see evidence of sustained and real economic growth and unemployment steadily declining in Ireland. If we ever get that (seriously – I’m still thinking the West might be finished), it’s going to be a long time coming.

It’s already clear that what was once 400k (at peak) is still not attracting buyers at 150k (I’m thinking of some specific 2-bed houses/cases I know). When people on 30-40k p.a. can afford modest 2 or 3-bed houses at 3 x salary, maybe that will be bottom? It doesn’t seem a ridiculous aspiration in a society that hasn’t gone out of control.

@Brian Woods Senior

“PR Guy You are being most charitable”

Yes, I’m all heart 🙂

But it really is wishful thinking be be saying we have overshot at this stage of the game. There is so much more to play out – maybe not originating here but the waves will wash up on these shores.

I’d second what Gregory Connor says above and further to the report stating that the econometric data was available from the authors requested same from the Bank this afternoon.

It is not even clear what the source is for actual prices, CSO, Allsop Space, Myhome, DAFT – there’s quite a difference between an actual decline of 43% (Myhome) and Allsop Space (70%).

It is hard to square the 26% undervaluation based on affordability with The Economist last November which said Ireland was still above “fair value” based on income and rents.

It’s difficult to understand how the authors of this report can reach the conclusion that house prices in Ireland are below fundamental value. Based on almost any reasonable measure, house prices in Ireland are not particularly low.

Actually, I would have thought that the continuing high property prices in Ireland are a testament to the extraordinary amount of taxpayers’ money the government is willing to spend propping up prices through NAMA and the banks’ policy of refusing to repossess property almost regardless of how bad the loan.

I don’t agree that the property market has no floor until the economy improves or the backlog of empty properties is sold. The government is willing to spend an almost unlimited sum on keeping property off the market and by making sure that those with big property losses are under no pressure to sell.

Given the low supply of property available at a market clearing price we can expect that the market will reach an equilibrium price that is far higher than might happen in a free market. In fact, we’re seeing this already in Dublin where the shortage of family homes has already caused prices to stabilise at a level that is high for the average citizen.

I know this might sound silly, given that there are four of them and all and that they probably have some very nice sums and datasets and all, but could it be, perchance, that their models are pants? You know, total and utter rubbish? Wrong on the way up and now wrong on the way down?

Just a thought…

PS In Figure 4, are we really to believe that actual prices are below 2000 levels? How is this assumed in the absence of transaction data going back to then?

Property of all sorts is being sold for less than the cost of rebuilding.

Building materials are not dropping in price, if anything they are rising due to increased energy costs and demand elsewhere for raw materials.

Petroleum products are steadily marching upwards. Tractor oil that cost 68 euro two weeks ago cost 71 euro today. Trivial but it has had at least four price rises over the past year to my recall.

@ What Goes Up: “Japlanic!”

Brilliant! Good as Manama!

Seriously, those so-called econs in the Central bank should be lacerated. The purchase of a home is THE most significant and costly financial transaction anyone will take on. Worse, your in a recourse situation – so if property values decline, as they are now doing (its present tense, with a vengence), your bangjaxed!

Look at the legal contortions that nurse had to endure to sort out her negative equity situation. How many hundreds more have to endure this? And these clueless CB bozos are asserting that residential property is probably OVERVALUED! Its rank incompetence. Have they no moral fibre whatsoever? I suppose they’re too young to understand.

@CB: Yeah Colm. Those abandoned McMansions have no floor. Its been pilfered! Detroit may the first of many.

I think they should call a halt to this scrapping of NIR class 450s DMUs, at least until we see where we are going with this oil thingy. (they are not that old really -1985ish)
We have continual declines of our CIE Bus fleet and now this…. put them in storage for a year please.
We might feel the need to use them down south next year when we are using Punts again.

Update: My rage at those CB ninkin-poops: Over- should of course read of under – sorry for that forced error.

One of the key questions from the point of view of employment, is have the cost of houses fallen below the current cost of building the house, even with a zero value assigned to the site. I suspect prices have fallen below the current cost of building and therefore the prospects of any upturn from an employment point of view are non-existant.

IMHO, apart from a feel good or a fell bad wealth indicator, the concentration on how far house prices have fallen is somewhat misplaced in terms of a future effect on employment.
There is still a hankering back to the misconceived notion that the price at which we sell houses to each other is a significant indicator of our economic wealth or of our economic prospects.

We only have housing bubble data going back to the 70s because before that a high proportion of the money supply existed as cash. It’s the digitization of money which has allowed such bubbles.

It’s important to note that mortgages repayments have hit the natural limit of taking two concurrent careers to repay. So all bets are off regarding what might happen to house prices.

We could reduce the effects of asset bubbles by allowing Central Banks to carefully create a debt free source of digital money for their Governments.

The trick to reading this paper is to imagine a few devout Harland & Wolff engineers studying the mathematical models they used and scratching their heads at the “theory versus reality” conundrum.

They then go off and write a paper – “Why is the Titanic still underwater?”

The world looks on and is embarrassed for them.

Like Gregory Connor id like to see the econometric models behind these figures but just a few points.

I think the models are dependent on income levels and interest rates as the main determinants of house prices. This may explain the results as interest rates have stayed very low. One thing that struck me as odd is that according to the diagram 2 of the models claimed that house prices were undervalued in 2003 which i find it difficult to accept.

On the rent to house prices model it seems that if you look at the figures one way then house prices are still overvalued.

But id like to see the econometric models which arrived at the results. Might be interesting to compare the models with the ESRI and Bacon econometric models to see what they would arrive at.

Perhaps some members of the society of chartered surveyors,the property professionals, might like to contribute to the non professionals?

How about a Paper on how to stop Irelands property obsession.

We’d need a section on ending bubbles, increasing tenants rights and ending subsidies.

I suggest we also inform people that avoiding ownership does not mean they’ll:

1. End up in a famine pit.
2. End up homeless. (rent/social housing are alternatives)
3. Go to hell.
4. Be hated by their children.
5. Be poor.

My own thoughts, published on my blog last week after the publication of the latest CSO property price data: http://vicduggan.wordpress.com/2012/04/26/rock-bottom/

Much play has been made of the latest CSO data on property prices, which showed prices avoiding a fall in March for only the second month since their peak in September 2007. Just as one swallow doesn’t make a summer, however, one month’s data doesn’t make a trend.

Overall, average house prices nationwide have fallen by49% since their peak according to the CSO, whose data is compiled on the basis of mortgage drawdowns. This likely understates the extent of the collapse in house prices, as it discounts cash transactions which have become increasingly important in recent years.

Flat prices in March must be viewed in the context of their accelerating decline through the Autumn and Winter. Moreover, there is clear divergence between Dublin and the rest of Ireland, the former seeing a 0.7% increase month-on-month while the latter saw a 0.6% drop. This may be explained by the fact that Dublin property saw the biggest appreciation up to 2007, and has since seen a greater correction (57% compared to 45% in the Rest of Ireland and 49% on average). Over-supply is also less of an issue in Dublin then elsewhere, in general.

There is no doubt that houses are more affordable now than they have been in years, when benchmarked against disposable incomes, rental yields etc., but there are two reasons why a robust recovery is unlikely in the near future: credit & confidence.

It is well known that the Irish banking system is undergoing massive deleveraging at the behest of the troika, but also because the banks’ loans are not backed by sufficient deposits. This process is certain to constrain credit supply for years to come.

More intangible is the ‘confidence fairy’. While credit may be in short supply, this is only a problem if demand for credit is significantly greater. Regardless of affordability, many prospective purchasers are clearly staying on the sidelines simply because they expect prices to fall further. If all buyers do the same, price falls become a self-fulfilling phenomenon divorced from economic fundamentals. This is what economists call ‘adaptive expectations’, i.e. people adapt and adjust their behaviour to everyday experience. Once asset price deflation has taken hold, the market will usually overshoot ‘true’ valuations.

Trying to call the bottom of the market may feel like trying to catch a falling knife, particularly if you are planning to put your money where your mouth is. When house prices do bottom out, however, and it is unlikley that they have yet reached their nadir, they are more likely to experience L-shaped stagnation than V-shaped recovery.


Celebrating the Work of an Accounting Great: Professor W.T. Baxter A Symposium was held at LSE on Saturday 15 July 2006 to celebrate the work of Professor W.T. Baxter, CA, Emeritus Professor of Accounting at LSE. Professor Baxter, the first full-time Professor of Accounting in Great Britain, sadly died on 8 June 2006, just short of his 100th birthday. He was a highly respected and greatly loved teacher, colleague and friend, and the Symposium was a fitting tribute to his achievements as an academic and as an individual. Around 100 former students, colleagues, and members of Professor Baxter’s family spent the day listening to presentations by academics and practitioners, who celebrated his significant contributions to accounting theory in areas such as deprival value and the measurement standards applied to accounting, which have led to current developments in accounting standards.

The Symposium opened with a keynote address by Professor Christopher Napier (University of Southampton) who provided an engaging insight into the ‘History of Accounting at LSE’ over the last hundred years. Professor Baxter’s contribution to the School and the development of accounting as a discipline was situated in the context of the School’s emerging focus on economic approaches to analysing business issues. Professor Baxter was central to developing undergraduate and postgraduate teaching in this newly-emerging academic discipline that recognised the business needs of the profession, but provided students with a rigorous and theoretically sound basis for understanding how businesses work. This is a tradition that continues at LSE today, with the Department of Accounting focusing on providing the intellectual training and theoretical understanding of accounting as a discipline, leaving the professional accounting firms to train graduates on the practical, number-crunching exercises that form part of a young accountant’s daily life.

Following Professor Napier’s keynote address, the first session of the Symposium focused on Professor Baxter’s theory of deprival value. The practitioner perspective on this concept was provided by Sir Ian Byatt (Past Director General of the Office of Water Services) and Peter Holgate (Senior Technical Partner, PricewaterhouseCoopers). Sir Ian illustrated its use in nationalised and privatised utilities in the UK, and showed how it was critical to setting an initial regulatory capital value. Peter Holgate examined deprival value and its use in current cost accounting and the setting of accounting standards. Professor Michael Bromwich (LSE) then gave an academic view of deprival value, its relationship to other prominent academic accounting theories of current value, and the importance of this concept to the Sandilands Committee and the Accounting Standards Committee.

The afternoon session examined the research importance of Professor Baxter’s work, looking at how it had been used in the past, and its relevance today. Professor Shyam Sunder (Yale University and President of the American Accounting Association) looked at Professor Baxter’s contribution to the accounting standards debate, and discussed how Professor Baxter was one of the first academics to recognise and articulate the negative consequences of authoritative measurement standards on the accounting profession. Dr Joanne Horton (LSE) described how the concepts of deprival value are still central to the teaching of financial reporting by pointing to the relevance of this work to standard setters obsessed with ‘fair value’. The session concluded with a presentation by Professor Ken Peasnell (Lancaster University) who examined where deprival value is today and its role in contemporary financial reporting and potentially again in inflation accounting. Professor Richard Macve (LSE) gave the concluding overview.

Leading figures from both sides of the Atlantic gave oral reminiscences of working with Will. One of the highlights of the day was a presentation made to Mrs Leena Baxter by Emeritus Professor of Economics, Basil Yamey, one of Will’s oldest colleagues and friends, who he met in South Africa, and who subsequently became an LSE colleague. Mrs Baxter was presented with a bound book of reminiscences, contributed by people who were Will’s students, colleagues and friends. The reminiscences showed the warmth with which former students remembered their teacher and the respect in which he was held by his students and former colleagues. The second highlight was the informal reminiscences provided by a number of alumni during the lunchtime reception. The memories shared by these former students reminded all of the participants of Will Baxter’s unique nature, his sense of humour and his keen intellectual engagement with accounting history, theory and practice.

The Symposium was jointly and generously sponsored by LSE, the British Accounting Association, the Institute of Chartered Accountants in England and Wales, and the Institute of Chartered Accountants of Scotland (where Professor Baxter trained as a Chartered Accountant).

The intellectual contributions of the day have now been collected and edited by Professor Pauline Weetman (Strathclyde University) and will appear in the September 2007 edition of Accounting and Business Research (Vol. 37, No. 2) under the title ‘Comments on deprival value and standard setting in measurement: from a symposium to celebrate the work of Professor William T. Baxter’.

I was fortunate in the 1970’s to study under Professor Will Baxter and Professor Basil Yamey at the LSE and was honoured to attend the above symposium.
Professor Baxter spent all his long life trying to develop theories of asset valuations and developed the concept of deprival value.

Professor Basil Yamey was an expert on the economics of industry and was a member of the UK Monoplies and Mergers Commission. He lectured on cartels, monopolies, etc etc.

The Irish commercial property merket is an organised cartel. This will be proved in the Irish courts very soon.

It’s very sad with all the Irish Professors of Economics none were capable of noticing the greatest property bubble in the history of mankind or the commercial property cartel

Your six year old could have spotted both.

Again, this is a case where we could have definitive data is the different departments and bodies would work together instead of staking out their own patch.

AFAIK there is an obligation under EU Law to have a proper index in place im the very near future. The CSO plan of building an price index based on geo-locations is pants of the highest order, especially with apartment blocks. It is no wonder they have little visible progress to date.

The sooner the Property Registration Authority, The Valuations Office and Ordnance Survey Ireland are merged the better. They should then be charged with building the house price index and the Central Statistics Office should only have a consultative role in designing systems for gathering and storing the information instead of being in charge of the creation of the index.


Folks, I’ve read the paper from the CBI and those of you who have been reading posts of this nature over the past number of years will recognise that the CBI have a dismal record in forecasting anything, let alone an asset class which they had to employ 3rd party external consultants to assess for them last year. And the small matter of paying €30m for that advice. Lets not go there.

One of the better critiques of where the Irish property markets is headed was provided by Cormac Lucy back in January –


on a previous post I copied in the link – quite obviously the CBI didn’t review same because on almost all topics their view is at variance with Cormac Lucys except for one. And that one has been something of an obsession of mine for a long long time and thats mainly the importance of rental yield analysis in property valuation. In its lastest paper the CBI specifically notes:

.”..If, on the other hand, one regards the
relationship between 1982 and 1995 as being more
reflective of what the long run relationship between
rents and house prices should be, then house prices
still have some way to fall.

Notwithstanding the implications of the rent
price analysis, the empirical evidence, in general,
would appear to suggest that Irish house prices
have fallen by more than what they should have
over the past couple of years…”

Now allow me to tell you a story. I wrote to the CBI following the publication of their PCAR document and suggested their house price falls estimate made absolutely no sense at any level but importantly at the rental yield level. If one was to follow their logic their adverse house price fall assumption would have seen net rental yields bottom at about 5% (in fact marginally below but we’ll use 5% as a close yardstick). No response. In fact after about the 7th time of asking they finally wrote to me and suggested that using a yield analysis was not their preferred methodology and pretty much dismissed what I had set out in detail to them.

Move on c1 year to today. Its very clear to me that the CBI simply don’t get this. The quote from the paper above is indicative of a form of arrogance of those who believe they know better than the rest of us. They don’t. The simple truth is that the CBI have bought into the ‘location, location, location’ property industry spin and dismiss the fundamentals of any asset class valuation in order to give the air of the all seeing guru where they seek to pick the model that fits their view of the world. This is nonsense. The fact is they have been proven to have paid for advice and can’t now logically be seen to put the hands up and say it as it is – ‘we was wrong’.

Instead they have the nerve to produce this claptrap and dismiss the fundamentals on the ground.

The most important number that’s produced in the Irish property market comes from the Daft.ie survey of rentals every quarter. The latest quarterly average asking gross rental yield number for Dec 2011 was 5.1%. Netting this back using the long term standard metrics and those recently provided to us by Ronan Lyons in adjusting from asking to contracted numbes on Daft (7% discount) suggested that at the end of Dec 2011 the average net yield in the Daft survey was 4.35% (5.1%*11/12*93%) which equates roughly to an asset trading at about a p/e of 23 times its most recent earnings. This is far from cheap or indeed ‘undervalued’ as the CBI would have you believe.

Data produced by Credit Suisse and the LSE suggests that the historical review of Irish asset returns over the past 100 years or so would lead one to the conclusion that the average long run equity premium for assets defined as ‘risky’ has Ireland throwing off a premium of 4.5% over and above risk free equivalents.

At the end of Dec the net property returns in Ireland per Daft were less than the premium required forgetting at all about the additional risk free returns. One could argue as to what now constitutes a ‘risk free return’ but even assuming it somewhere between Germany, France, Finland, Luxembourg etc one would work on the basis of at least 2% (forgetting inflationary effects). So here we have it, Irish data analysed by Credit Suisse and the LSE suggesting that ‘risky’ assset classes have exhibited returns of at least 6.5% (notwithstanding the property over supply issues in the RoI or the lack of credit or the lack of confidence or the liquidity issues) for the best part of 100 years and the CBI are suggesting that yields at about 4.35% are now about 25% undervalued i.e. net yields should really be 3.26% i.e. fair value in their world is an asset valued at 31x current earnings. I’m sorry folks but this is lunacy.

There is a footnote in the paper and it mentions that if you contact the authors they will send the estimation details. I sent a note and got the code the next day. The estimation code is easy to read (you need to know RATS but most people who do time series empirical work can read RATS).

A big difficulty is how to treat the bubble period in the estimation problem – the authors include all available data. This likely affects the forecast results since property prices during the bubble are optimally fitted by the models. The high observed prices during the bubble will tend to raise the forecasts for the post-bubble period. Censoring the bubble period from the estimation sample (or censoring any subsample) is also very problematic of course. Perhaps we need to develop new models which capture the special features of the bubble period while also fitting the “normal” periods.

@ gregory

is the answer to the problem of smoothing out for the bubble to include dummy variables to account for shock/irrational periods? i suppose if such periods are the norm you have a problem….there is no question that normalised bubble periods which arguably make up a large portion of the data is going to generate biased upward results….but is this not all a case of the CB doing the bidding of Nama in much the same way a broker pushes stock he has a vested interest in?

@ YoB: Thanks for that useful update.

The McWilliams chap spent a few mins on Morning Ireland wittering on about yields and all. Not a fan of the chap, but he did get the point across. You must pay heed to yields. Pity it is so logical and rational. Folk are only interested in sentiment. Hence they will believe the CBI pumpers. Requires less cognitive effort.

“The quote from the paper above is indicative of a form of arrogance of those who believe they know better than the rest of us. ”

What does this tell you? Dismiss any critique out of hand. If that does not work, use – “Its the only game in town” sort of dismissal. If that fails deploy fear – “If there is a No vote, I will bring in an even worse budget”. If that fails to quell dissenters: slime them.

Trouble is these behaviours work. Folk are sheeple.

I’ll write to Honahan about that disgraceful commentary. Not that I expect anything positive. But if you do not complain … … “garde l’eau!”

@V Barret — The same dummy-variable approach occurred to me so I went ahead and stuck a bubble period dummy into the estimation code and re-ran their RATS program – adding this dummy lowers the end of sample mispricing a bit giving 10-15% underpricing instead of 10-26% underpricing. So it does not change the sign just narrows the range of findings across the models. The model uses log prices and the bubble dummy explains about 20% overpricing during the bubble period 2002-2007.

@ Gregory Connor,

Are you saying the CB models are trying to predict property price (market prices) using other variables? This is quite different to deriving a fundamental price.

What type of model are they using? regression? and what’s the fit?

Do they indicate the time series used in determining the weightings? What would the models predict if you cut the time series off at 1997?

@Ahura Mazda — They use four standard macro-type property price models estimated with monthly frequency log-linear regressions, using affordability, demographic, interest rate variables as explanatory variables. See the CBI note.

@ gregory.

if what is being captured by the dummy is the unusual level of credit flowing in the market at the time given other fundamentals, then applying the same dummy to this period might (assuming there is unusually low credit) then one wonders would we hit on a price that was just what we have in the market now 🙂

@ Gregory Connor,

I hope my previous comment didn’t appear rude. My abrupt series of questions reflects more my stream of consciousness than anything else.

I’m boxing in the dark a little as I don’t have the data, but I’d note the following:
1. Their models attempt to predict market prices not a fundamental price.
2. FIG4 suggests the models are too sensitive to short term variations to be considered as predicting a long term value.
3. Removal of outliers – in this case I’d argue removing 1998 + data. It’s a whole chunk, but it would be interesting to see the outcome. If you consider rental price from 2000 to 2008, it would appear to have little value in predicting house prices. You’d probably get an inverse relationship between rent and property prices between 2009 to 2012. Though intuitively, there should be a positive correlation.

I won’t address the variables being used as I don’t know their exact nature. Regarding a dummy variable, is this a constant value?

@Ahura Mazda — I thought your comment was right-on, no problem there. They use log-linear regression and I just fudged a bit of RATS code sticking in a fixed zero-one dummy for 2002-2007 and got a coefficient of .16 to .21 (there are four models). Since it is log price as the dependent variable this means that predicted prices in those years are given an approx 21% uplift by inclusion of the dummy. It was a bit simple-minded to use a fixed dummy variable in this way, but it did convince me that with this adjustment these standard models still find current prices are below model-implied. Of course this does not take account of all the other contemporary factors such as banks’ desire to shrink their balance sheets, Euro area vulnerability, prospects for a renewed financial crises in Ireland, the Eurozone, or even globally. Back in the old days when I got my PhD in financial economics we had to promise never to claim we could forecast asset prices. Not sure if property prices were covered by that promise. Anyway it was a long time ago, and times have changed.

The surveyors red book may have to be revisited. My suspicion is, I havn’t yet proof, but the society of chartered surveyors, the property professionals, may be an organised cult,who are totally brainwashed and highly dangerous.

Perhaps some of their spokespeople might like to comment.


Don’t follow “You’d probably get an inverse relationship between rent and property prices between 2009 to 2012. Though intuitively, there should be a positive correlation.”

Rent would currently be more constant than price. Are you observing the fact that rents have stabilised / gone up a bit while prices still fell?

“2. FIG4 suggests the models are too sensitive to short term variations to be considered as predicting a long term value.”

When you try to model an asset price you are either trying to derive some ‘fundamental’ value as a time series or trying to model investor behavior – or some combination.

The short term variations – to the extent they correlate with actual prices are modelling behaviour. Behaviour and value often part company.

@YOB has a yield preference & the CBI guys’ compare with dividend yield.

Equities strategists still sometimes just print off a long term yield chart and draw some lines on it. It is a bit unsophisticated in a way, but has the advantage of being unadulterated and honest.

The problem with it is that every form of investment competes with all the others for money. You then compare dividend yield with fixed income yields, index linked yields and on and on.

Where Property fits into the bond, pref, equity spectrum is never really certain and in 2006 or so Irish property was almost being rated as high-tech start-up companies. It was ridiculous.

@Gregory, I can’t do the digging right now, but have you got a simple list up could put up, of the variables that went into each of these models? Usually, if you have the list and stare at the graphs, you can discern which component has led to which diversion – which could be informative


“..Where Property fits into the bond, pref, equity spectrum is never really certain and in 2006 or so Irish property was almost being rated as high-tech start-up companies. It was ridiculous…”

Fair point and you are correct there is no definitive box you can place property into in terms of valuation methodology – however there are a few known knowns such as its a risky asset (just ask the banks) , has a tendency to mean revert over time in terms of relative valuation to its long run trend but most importantly and nearly always considered an after thought for the most part its a credit dependant asset class. In that regards its almost unique.

Being a credit dependant assest class changes life a great deal. The riskometer moves up many notches which in fact places it further out the risk spectrum (for most owners) than its close cousin the humble equity which is normally bought for cash (options/cfds/spread betting/derivatives etc aside). The risk premium to which I refered above relates to equities not property. Given the added credit risk any risk premium, all other things being equal, should on a fundamental basis be higher – further negating what the CBI have suggested is an over sold asset class.

Is Real estate an effective hedge against inflation?
Recent research has seriously questioned the notion that real estate is an effective hedge against inflation. Joe Valente says such arguments are missing the point

The inflation hedging quality of real estate is one of its most attractive and enduring investment characteristics. Indeed, it is the reason often given to invest in real estate, particularly by those investors who need to match long-term assets to liabilities. A number of major institutional investors have increased their allocation to real estate recently as a result of their respective house views that assume a significantly higher rate of inflation in the future. It is not just institutions. The continuing attraction of the asset class to an increasing number of high net-worth Individuals with a strategy underpinned by wealth preservation also implies a belief that property can and does act as a suitable hedge against rising inflation.

Many practitioners have long asserted that property can be used as a hedge. Most investors tend to be of the view that property is, or can be, an inflation hedge, particularly over the long term. In the same way, most will recognise that, in the short term, local market fluctuations will tend to prevail and confuse the debate somewhat. It is clear, however, that the debate over the merits of real estate as a hedging tool has long been raging but that the evidence is remains inconclusive.

This debate will undoubtedly gather further momentum given the growing concerns over a higher inflationary environment in the years ahead across Europe. No doubt this will trigger a wave of new research papers on the subject which, if anything like recent ones, will shed next to no light on the issue, and succeed only in adding to the general level of confusion surrounding the subject, or at least glossing over some of the most important characteristics of the asset class and its ability to perform successfully as an inflation hedge.

The definition of what exactly constitutes a hedge is the first, and possibly the most important, source of confusion. An inflation hedge is often taken to mean an investment whose value is directly related to the level of inflation. In other words that there is a direct relationship between property values and some measure of inflation. So if over, say, a five or 10-year period, property values keep pace with inflation, or even outperform a particular benchmark, this is not in itself viewed as sufficient evidence that real estate is capable of performing a hedging function. For this to be the case, so the argument goes, property values have to move with, and react, to a changing inflationary environment.

Far from semantics, the use of such a definition will inevitably lead to the conclusion that commodities, or even equities, offers a much better hedging mechanism than that provided by real estate. After all, real estate is lumpy, there is a lack of frequent real-time pricing – quarterly data in a few markets, annual indexes in most – and that’s not to say anything of the constraints imposed upon it by the landlord/tenant relationship which will vary not just between markets but over the course of the real estate cycle as well.

Issues that make real estate different and which will inevitably mean that values won’t be able to ‘react’ sufficiently quickly to changes in inflation. But this ignores the fact that growth in property values may well exceed inflation over a set period. In essence, the conclusion that real estate is a poor hedge against inflation is often solely the result of applying a definition that is ill-suited to the sector or doing so with little understanding of how the property market actually works in the real world.

The key point is that real estate provides a long-term hedge against inflation but, in order for it to do so, certain other criteria have to be met.

First, while the real asset nature of property underpins its inflation hedging quality over the long term, buying tangible assets as a protection against inflation is sensible only if done at the right price. It’s an obvious point but one that is overlooked a little too often.

Second, properties with inflation-linked leases are increasingly attractive to a broad range of investors as they are often seen as defensive in nature. However, while such leases can offer many attractive qualities to the investor, rent indexation does not equal inflation protection – it is not a guarantee even if some investors tend to regard them as such.

The reality is somewhat different – rent indexation in the absence of pro-active management results only in over-rented assets. Such leases should not be viewed as a guarantee of an outcome but rather as a starting point in the quest for inflation protection.

Third, income growth is the key to solving the inflationary puzzle. The ability to retain a tenant in the building, maintain and grow cash flow lies at the very centre of an inflation-hedging strategy. And that applies across all markets and points in the cycle.

And finally, the importance of income growth brings us to the missing link in much of this debate, which is simply the ability and expertise of the asset management team. The situation is the same for all real estate markets: in the UK with its longer leases and upward-only rent reviews; in the euro-zone with its propensity for shorter leases and rent indexation; in the US with year-long leases and no indexation, or in markets such as Turkey where the length and value of a lease can be somewhat debatable. In all these instances it is the ability of pro-active asset management to deliver value that consistently shines through time and again.

The bottom line is very straightforward: a bad asset manager will underperform inflation no matter what the inflationary environment might be, the point in the economic cycle, the holding period or the relationship between landlord and tenant.

Author: Joe Valente

@ grumpy and ahura.

Iv a paper from 2003 outlining the various models and variables to determine fundamental house prices. Im having a bit of difficulty putting it up here but theres a copy of it on the Irish Mortgage Brokers website.

Its in the historical blog there dated 4th august 2010

@ Grumpy,

Re ‘Are you observing the fact that rents have stabilised / gone up a bit while prices still fell?’ Yes. If you were to chart avg. house price to avg. rent over the last 30yr, my guess is you would see three patterns 1. up to late1990s/2000 a positive correlation (i.e. prices tending to increase in tandem) 2. 1990s/2000 to peak little/no correlation 3. Post peak a negative correlation. I’m just observing that the regression model is run over this whole period where we’ve some odds trends.

Re your second point: I’m saying the CB models are predicting historical market prices using a series of variables. I don’t think this is the same as a fundamental price. Although a slight tangent, we should bear in mind that only 30ish% of properties have mortgages and only a small portion of housing stock is for sale at any point in time so it is worth questioning how national level variables relate to the population of buyers. Then we need to consider soft issues which are not captured by data – people on deckchairs queuing to buy off plan (maybe we need to use a poisson – bet you weren’t expecting stats humour :D). Though ‘soft’ factors might help explain historical prices, I’d argue they shouldn’t alter fundamental price.

Also, one unfortunate aspect of the CB models is that they seem to be backward looking. We know that there is a huge whole in the state coffers to be filled, emigration and a backlog of foreclosures to work through etc. From what I can tell these aren’t addressed in the CB model.

@ Gregory,

Is it fair to say that the dummy variable will be constant each year (2002-2007)? It’s explanatory power is limited by forcing it to have the same effect in each of these years. You could create a dummy for each year, but that’s going in to spurious territory.

@ Frank Quinn,

Do you have a link?

BTW there are default models that penalise mortgages originated via intermediaries 😉

@ Ahura

Its on the Irish Mortgage Brokers website (historic blog) 10th August 2010.

The diagrams didnt scan too well but think there better in the introductory piece.

Cant post the link up here (think its a size issue)

I didnt include a mortgage intermediary variable at the time which may explain my results, I knew i was missing something !!

Comments are closed.