In this State, we have enough Irish jokes – what, with the 2008 bank guarantee and paying bondholders in utterly bust banks whilst putting at risk the credit worthiness of the State. But we haven’t had as good a joke as the report published by the Central Bank of Ireland (CBI) this morning. To be fair, the author Eoin O’Brien says at the outset that he is expressing his own opinion and it is entirely his work, but when something is published under the CBI brand, the CBI as a body corporate should bear some responsibility for its content.
So why the conniption?
The report sets out to examine prices of residential property achieved at recent Allsop Space Irish property auctions and to compare the current auction prices with peak asking prices and current asking prices. It’s not that he’s even comparing apples and oranges – or actual prices with asking prices – because at the peak where there was a sellers’ market, the asking price should intuitively be pitched considerably in excess of actual prices but today in a buyers’ market, you would expect asking prices to more closely reflect actual prices. So, 2007 oranges may well have a different quality to 2012 oranges. But what will really amuse you is the fact that he uses the results from the Allsop Space auctions to work out the actual price of residential property today, and the way he does it, he takes the total hammer prices and divides that by the number of properties sold, to get his average – or rather averages, mean and median (see * below for a refresher on the two). There is no attempt to match two bedroom apartments in Ballsbridge with other local two bedroom apartments, or even with two bedroom apartments in Ballyfermot. The statisticians refer to this matching of like with like as “hedonic regression”, but there is no attempt to do this with the auction results.
The report then goes on to calculate yields – simplistically the annual rent divided by the value of a home – but wait! For historical average prices, it uses the Department of the Environment, Community and Local Government average house price. Remember that’s the one that take a €58m sale like Walford on Shewsbury Road and a studio sale in Ballyfermot for €50,000 and says the average price of an Irish home is €29,025,000! No seriously, and by the way, even though the CSO index has become the premier index in the country, junior minister at the Department, Jan O’Sullivan has said that she will continue to fund the “marginal cost” of maintaining this utterly flawed series (which by the way excludes cash-only transactions!)
One of the regular commenters on here, WGU will shortly be on to say that auctions themselves are skewed towards cheaper property, and indeed the report’s author today does confirm this with his “Kernal density estimate” which shows the overwhelming majority of auction sales are in the €75-175,000 range.
So what are the conclusions anyway from the report? Here we go (the ranges are there to show mean and median)
(1) Dublin property is down 60-67% from peak asking prices
(2) Dublin houses are down 55-62%
(3) Dublin apartments are down 66-68%
(4) Non-Dublin property is down 72-77% from peak asking prices
(5) Non-Dublin houses are down 69-73%
(6) Non-Dublin apartments are down 80-83%
(7) Dublin property actual today is down 21-32% from current asking prices
(8) Dublin houses are down 3-23%
(9) Dublin apartments are down 28-33%
(10) Non- Dublin property actual today is down 46-57% from current asking prices
(11) Non-Dublin houses are down 39-48%
(12) Non-Dublin apartments are down 63-66%
If you were to extract messages from the above
(1) Dublin is down less from peak asking than non-Dublin, which contradicts the CSO index which suggests prices in Dublin are down is the other way around – to the end of August 2012, Dublin was down 57% and non-Dublin was down just 46%.
(2) Apartments are down more from peak asking than houses everywhere.
(3) Non-Dublin sellers are being completely unrealistic today with asking prices double auction actual prices.
Lastly it is worth saying that auctions generally tend to achieve lower prices than private tender sales through estate agents. There are apparent exceptions like the Northumberland Road property which sold for €685,000 last week at the Allsop Space auction after being purchased three months earlier for €550,000 but the consensus “on the street” was that the property was undersold three months ago. Buyers at auction are taking more risks and tend to pay in cash. So actual auction prices may tend to overestimate the decline compared with what would be regarded as normal mark prices. But since we’re comparing apples and oranges, does it really matter?
Thanks to the Property Price Register we now have actual sales prices. It may be some weeks before the data is analysed so that some third party provides an index, and the data only goes back to January 2010, but the age of the asking price is coming to a close, and the prediction on here is that if there are two headlines next January 2013 – “asking prices down 10% in quarter” or “actual prices down 5%”, I be willing to bet you will click on the latter headline and ignore the former.
*Now it might have been some years since some of you did you junior cert, so to recap on mean and median take the following numbers
13, 18, 13, 14, 13, 16, 14, 21, 13
The average (or “mean”) is the sum of these numbers divided by number of numbers or ((13, 18, 13, 14, 13, 16, 14, 21, 13)/9 or 15
The median is the middle number when these numbers are sorted in ascending order so
(13, 13, 13, 13, 14, 14, 16, 18, 21) or 14
What exactly is actually disgraceful about the report?
The author says explicitly that there implications for the wider market should be taken with caution due to the sample size.
Its a very limited piece of work in scope and the only way I could possibly agree with your headline was if the author or CBI were somehow misleading with the conclusions.
Disgraceful?
Blusterbuster, mollified to “flawed”
@NWL
I tend to agree with ‘Rob S’ and ‘blusterbuster’ the author outlines that the sample size is small and drawing too many conclusions from the numbers would not necessarily be overly wise. So taking that health warning into account the report is a useful exercise however I do have a few issues namely:
The Daft.ie surveys are correctly stated as being based on asking prices however Ronan Lyons recently wrote that following Daft.ie own internal investigations they have calculated that the overall discount to actual contracted prices from asking using their most recent surveys is about 7%. I have previously posted workings incorporating this further adjustment into my own peak to trough assessment of house prices and therefore for the Central Bank not to adjust Daft.ie asking prices to contracted prices is a serious gap in this reports PTT analysis. This 7% adjustment hold true for rents as well as house prices.
In addition despite my many posts here and elswhere suggesting that the only true and reliable long run driver of house prices is long run rental yields it eventually seems the penny is finally beginning to drop with the CBI as the analysis presented, although somewhat simplistic, is suggesting that the CBI is turning its attentions towards the effect this metric is playing in the market. I had previously written to the CBI over the years imploring them to the dangers of not taking this metric into account and the likely effect it would have on house prices when the mean reversion process ensued as it inevitably would and did. Unfortunately the Central Bank told me not to worry as rental yields were not an important driver to owner occupied house prices. Oh dear. How they now wished they had studied the long run mean reversion process in asset prices a bit more closely.
For the record; using the mean reverting process will see house prices eventually settle to throw off a net yield of 7% (gross 8.4%) and this will see house prices exhibit an average PTT fall of c75%. This will happen. I said it would happen 5 year ago, 4 years ago, 3 years ago, 2 years ago and up until recently I repeated the dose. ALL other metrics will prove inferior in placing a value on a house regardless of its use. Who would bet against this logic today?
The numbers are as follows:
Market asking gross yields per Daft.ie in July 2007 (market peak) 3.12%
Adjustment to bring asking to contracted rent (7% discount market average)Therefore Contracted gross yield at peak was in fact 2.90%
Falls of -24.8% in nominal rent to date from peak gives a peak yield at todays rent of 2.18%
Gross Yields for market with significant over supply will see prices revert to throw off gross yields of 8.4% (being 7% net adjusted upwards for 10.5 paying months)
Minimum peak to trough market average fall to revert a 2.18% yield to 8.4% is a required house price fall of 74.0%
Bottom line we have a long way to go until prices bottom out as the current (q2 2012) Daft.ie rental market average yield is still 5.3% which is still way to low meaning house prices are still way to high – in fact still about 35% too high.
“Minimum peak to trough market average fall to revert a 2.18% yield to 8.4% is a required house price fall of 74.0%
Bottom line we have a long way to go until prices bottom out as the current (q2 2012) Daft.ie rental market average yield is still 5.3% which is still way to low meaning house prices are still way to high – in fact still about 35% too high.”
Why do you make the assumption that the rental yield should remain the same (i.e. at ~8%) and the house prices should adjust to suit? Why can’t the rental amounts and yield change to suit the house prices?
Chicken – egg, cart – horse and all that.
New house prices will not fall below the cost of materials+ or house building will stop completely until builders can recoup material costs (including land) and a ‘reasonable’ profit.
Rental values have rarely been directly in line with, i.e. a driver for, house prices. As often as not the rentals have been driven by mortgage costs rather than house prices.
I accept that I may be a little out of touch as it has been many years since I was ‘in’ the estate agency business, but for the time I was this was certainly my perception.
@PunterMan
You raise an issue that has troubled the very best of market technicians for many years i.e. the chicken and egg scenerio with regards to mean reversion in asset pricing i.e. does the earnings adjust the prices or do the prices adjust the earnings. This is not a simple circle to square.
Over many years or as market heads like to call it ‘the long term’ the mean reverting phenomenon has indicated that despite booms and busts in asset prices long term average valuation trends bizzarly continue to hold true. Remember price is what you pay but value it what you get over the long term. Over the long term risky assets in the RoI have commanded a premium over ‘risk free’ assets by a margin of about 4%. Credit Suisse produce analysis on this on a country by country basis stretching back as far as prices and data are reliable for each country under review. The long run analysis for the RoI is based on data for about 120 years and the numbers suggest that over the long term risky assets yield c6.5% to c7.5% in nominal terms.
Now risky assets in this context is normally assumed to be equity assets however we also know that the illiquidity of property tends to ensure a higher risk premium over and above normal equity investments and hence why over the long terms RoI property gross rental yields have been higher than yields on pure equity investments.
The issue of course is what drives house prices and this is where the property market and normal equity pricing gets skewed. For the vast majority of the past 50 years Ireland has had access to mortgage related products for the purchase of houses. The availability of credit/leverage is the single most important driver in the housing market bar none. In simple terms those in control of the leverage control the market – and that’s the banks and lenders generally. The control of leverage by the banks basically means that they have the power to control the prices of property. This is not the same with the equity makets for instance where leverage represents a significantly smaller proportion of the cash going into the marketplace. In the normal property market – from about the early 1960s to about the middle of 2008 between 95% to 98% of all transactions in the residential market happened only because a 3rd party lender allowed it so. From the middle of 2008 to date this proportion has fallen dramatically and with it house prices. So make no mistake about it house prices can, do and will in many instances fall below the cost of production because the supply demand dynamic needs to be considered at all times. Hence the reason why in many parts of the country houses are selling for significantly lower than production costs.
The argument I have been making for years is that rents should in fact be the ONLY driver of house prices. We know that housing is not a ‘cash’ market it is a leverage market and as such pricing can be controlled by those in control of the leverage. The mistake the banks made is lending money into the market not on the earnings ability of the house in question or its proxy equivalent to generate a known rent but to lend on the earnings capacity of the individual to repay a loan. This is a deeply flawed lending model and hence the reason why the banks are bust. Had the banks looked at what the cash market was paying for accomodation i.e. rent and price housing off the asset that it was taking as security rather than the individual we would not be having this conversation. Cash based earnings are always a better guide to asset values than artifical leverage induced pricing. Hundreds of thousands now in negative equity are learning this the hard way.
Reblogged this on Machholz's Blog and commented:
Nothing from the central bank assures me, they continue to give a clean bill of health to the two biggest banks in the country even though they are both toxic and corrupt and stuffed to the gills with x corrupt politicians on the make
Auctions themselves are skewed towards cheaper property!
That is all :-)
13, 18, 13, 14, 13, 16, 14, 21, 13
Now, for the PhD in statistics, what is the mode?
13