PPR Versus CSO Statistics - which is more accurate and why?

alastair

The_Conductor wrote: »

There are distinct trends emerging in CSO and (lagging the CSO statistics) PPR statistics. They do not support a continuation of what has been happening for the past 30 months.

The distinct trend evident in the CSO figures is of continued growth in property prices. Slower this year than the rapid burst last year, but growth nonetheless, and growth evident across both Dublin and the rest of the country.

The_Conductor

alastair wrote: »

The distinct trend evident in the CSO figures is of continued growth in property prices. Slower this year than the rapid burst last year, but growth nonetheless, and growth evident across both Dublin and the rest of the country.

In the Dublin area- that growth has trended downwards in almost a straight line- and month on month has been negative more often than positive since January- earlier large increases in the last 10-11 months ago- are the main drivers of any advised year-on-year growth.

We actually discussed this pretty definitively earlier on in this thread.

The distinct trend in CSO figures- has been a reduction in annual growth rates in the Dublin area- in a startling manner- with an increase in regional prices counteracting the Dublin decreases.

PPR prices- give a better indication (they include the now dwindling cash sales- the CSO figures are based on mortgage drawdowns)- however, the PPR figures are only retrospectively (as-in several month later) on a parr with the confidence intervals you can ascribe to the CSO figures.

alastair

The_Conductor wrote: »

In the Dublin area- that growth has trended downwards in almost a straight line- and month on month has been negative more often than positive since January- earlier large increases in the last 10-11 months ago- are the main drivers of any advised year-on-year growth.

We actually discussed this pretty definitively earlier on in this thread.

The distinct trend in CSO figures- has been a reduction in annual growth rates in the Dublin area- in a startling manner- with an increase in regional prices counteracting the Dublin decreases.

PPR prices- give a better indication (they include the now dwindling cash sales- the CSO figures are based on mortgage drawdowns)- however, the PPR figures are only retrospectively (as-in several month later) on a parr with the confidence intervals you can ascribe to the CSO figures.

Dublin figures are safely in the positive growth for 2015 territory - irregardless of any drop from higher growth last year - growth is still growth. That's an ongoing trend of growth in prices. It's even stronger outside Dublin. That's the reality of the trend. PPR prices are not a price index, and can't be interpreted as such. You've been doing your best to deny the reality of the CSO stats, but they're the reality nonetheless.

In Dublin residential property prices rose by 2.8% in August. Dublin residential property prices were 8.2% higher than in August 2014. See table 6.

...

Outside of Dublin residential property prices rose by 1.9% in August. Prices were up 10.8% compared with August 2014. See Table 4.

The_Conductor

alastair wrote: »

Dublin figures are safely in the positive growth for 2015 territory - irregardless of any drop from higher growth last year - growth is still growth. That's an ongoing trend of growth in prices. It's even stronger outside Dublin. That's the reality of the trend. PPR prices are not a price index, and can't be interpreted as such. You've been doing your best to deny the reality of the CSO stats, but they're the reality nonetheless.

CSO Stats are one reality- the reality of mortgage drawdowns.
PPR Stats- are a different reality- they are the prices actually obtained inclusive of cash buyers (40% of the market) that the CSO stats do not figure- however, they are subject to significant revision as they are reported in arrears (CSO stats are uptodate- but only feature mortgage drawdowns).

You can argue this as much as you like- neither sets of figures are ideal- one dismisses 40% of the market- versus the other that is late.........

As for the Dublin CSO stats- they are positive on an annualised basis- however, if you do them for calendar 2015- they are in fact negative in the calendar year of 2015 to date (this is the point I was making).

Its really a case of making of the statistics as you will- I'm not buying or selling at the moment- and indeed- other than an interest in keeping my property tax as low as possible- don't have any specific interest in the actual prices (my property- other than rental property- is a roof over my head).

Time will tell. You can make stats mean whatever you want- I'm trying to take as broad a view as possible.

alastair

The_Conductor wrote: »

CSO Stats are one reality- the reality of mortgage drawdowns.
PPR Stats- are a different reality- they are the prices actually obtained inclusive of cash buyers (40% of the market) that the CSO stats do not figure- however, they are subject to significant revision as they are reported in arrears (CSO stats are uptodate- but only feature mortgage drawdowns).

You can argue this as much as you like- neither sets of figures are ideal- one dismisses 40% of the market- versus the other that is late.........

As for the Dublin CSO stats- they are positive on an annualised basis- however, if you do them for calendar 2015- they are in fact negative in the calendar year of 2015 to date (this is the point I was making).

Its really a case of making of the statistics as you will- I'm not buying or selling at the moment- and indeed- other than an interest in keeping my property tax as low as possible- don't have any specific interest in the actual prices (my property- other than rental property- is a roof over my head).

Time will tell. You can make stats mean whatever you want- I'm trying to take as broad a view as possible.

The PPR figures are not a price index. It says so on the tin. It's not because they're late that they shouldn't be used as a proxy for an actual price index like the CSO produce - it's because they aren't intended, and shouldn't be interpreted as an index - at any time.

If you take the CSO figures for Dublin on a calendar year to date basis, they're up, not down. Not sure how you're seeing otherwise. Positive months account for 5.6%, negative months account for only 3.1%. There's no real chance of them falling into negative YoY territory by the end of the year either, unless some rapid sea change occurs.

The_Conductor

alastair wrote: »

The PPR figures are not a price index. It says so on the tin. It's not because they're late that they shouldn't be used as a proxy for an actual price index like the CSO produce - it's because they aren't intended, and shouldn't be interpreted as an index - at any time.

If you take the CSO figures for Dublin on a calendar year to date basis, they're up, not down. Not sure how you're seeing otherwise. Positive months account for 5.6%, negative months account for only 3.1%. There's no real chance of them falling into negative YoY territory by the end of the year either, unless some rapid sea change occurs.

Period Percentage changes
Dublin - all residential properties 1 month 2015 2014 3 months 2015 3 Months 2014 Annual Price Increase 2015 Annual Price Increase 2014
Jan -1.9 -1.3 -1.8 0.3 21.4 13.6
February -0.7 -0.6 -2.4 -1.6 22.8 13.3
March 1.1 0 -1.6 -1.9 20.2 14.3
April 1 3.1 1.3 2.5 15.2 17.7
May -0.1 4.2 2 7.4 11.1 22
June -0.4 3.3 0.5 11 9 23.9
July 0.7 2.7 0.2 10.5 8.2 23.2
August 2.8 3.5 3.1 9.8

Cumulative year to-date -0.3 11.4

The_Conductor

alastair wrote: »

The PPR figures are not a price index. It says so on the tin. It's not because they're late that they shouldn't be used as a proxy for an actual price index like the CSO produce - it's because they aren't intended, and shouldn't be interpreted as an index - at any time.

You can interpret any dataset you choose- as an index.
Whether or not the person who compiled the dataset intended it to be used as an index or not- is moot. The CSO normalise their data- the PPR index is raw data- not normalised. It doesn't mean you can't present it as such (and indeed- the CSO use the PPR raw data to check their own stats against). It just means extra work for the person using the PPR data- that the CSO factor into their index.

alastair

The_Conductor wrote: »

Period Percentage changes
Dublin - all residential properties 1 month 2015 2014 3 months 2015 3 Months 2014 Annual Price Increase 2015 Annual Price Increase 2014
Jan -1.9 -1.3 -1.8 0.3 21.4 13.6
February -0.7 -0.6 -2.4 -1.6 22.8 13.3
March 1.1 0 -1.6 -1.9 20.2 14.3
April 1 3.1 1.3 2.5 15.2 17.7
May -0.1 4.2 2 7.4 11.1 22
June -0.4 3.3 0.5 11 9 23.9
July 0.7 2.7 0.2 10.5 8.2 23.2
August 2.8 3.5 3.1 9.8 25.1

Cumulative year to-date -0.3 11.4 -1.8 28.2 107.9 128

Dunno where you're getting a cumulative year to date figure of -0.3?
If the % increase is 5.6 and the % drop is 3.1 that reads as a cumulative +2.5% to me (2.46% to be pedantic and take the changes in their correct order).

alastair

I don't know why you felt the need to pull this out of the property prices thread btw - leaving your factually incorrect claim that the CSO figures show anything other than a trend for price increases without an appropriate counter?

alastair

The_Conductor wrote: »

You can interpret any dataset you choose- as an index.
Whether or not the person who compiled the dataset intended it to be used as an index or not- is moot. The CSO normalise their data- the PPR index is raw data- not normalised. It doesn't mean you can't present it as such (and indeed- the CSO use the PPR raw data to check their own stats against). It just means extra work for the person using the PPR data- that the CSO factor into their index.

Two points worth highlighting

1. The point of being clearly told that the PPR date shouldn't be used as a price index is certainly not moot. It's clearly flagged for a reason. That you opt to ignore that, and plough ahead claiming the data as a proxy price index is clearly ill-advised (literally)
2. Let's be honest here - you don't have the information to normalise the PPR data into anything approaching a legitimate price index. You've no idea if one month's set of stock has any relationship to the previous, or any other month's set of stock, or what the useful comparators are.

The only actual price index worth a damn is the CSO one - for all it's limitations. The Daft asking price index is really only for entertainment purposes, and the PPR data isn't an index, nor do you have the means to turn your interpretation of those tea leaves into anything resembling a worthwhile index.

alastair

mid wrote: »

By your calculations, if there was a 20% decrease and then a 25% increase, then that means a cumulative 5% increase, which doesn't seem correct

Eg.
100,000 with 20% decrease = 80,000

80,000 with 20% increase = 96,000

So you need a larger increase to get back to the original amount

This example has a 20% decrease, then a 20% increase, but the cumulative was a 4% drop

My calculations are based on a set of sequential changes with a set starting value. So they take account of the varying values the percentages are applied to. It's still a cumulative increase of 2.46%.

slowjoe17

The_Conductor wrote: »

Period Percentage changes
Dublin - all residential properties 1 month 2015 2014 3 months 2015 3 Months 2014 Annual Price Increase 2015 Annual Price Increase 2014
Jan -1.9 -1.3 -1.8 0.3 21.4 13.6
February -0.7 -0.6 -2.4 -1.6 22.8 13.3
March 1.1 0 -1.6 -1.9 20.2 14.3
April 1 3.1 1.3 2.5 15.2 17.7
May -0.1 4.2 2 7.4 11.1 22
June -0.4 3.3 0.5 11 9 23.9
July 0.7 2.7 0.2 10.5 8.2 23.2
August 2.8 3.5 3.1 9.8

Cumulative year to-date -0.3 11.4

Where do these numbers come from?

They don't make sense to me.

I Am The Law

From the PPR website?

slowjoe17

I Am The Law wrote: »

From the PPR website?

Maybe. But their "news" section ends in 2013 and the website is badly organised, so a link would be helpful.

alastair

I Am The Law wrote: »

From the PPR website?

They're using August CSO figures, but then some voodoo calculations attached to them - and yes - they make no sense.

http://www.cso.ie/en/releasesandpublications/er/rppi/residentialpropertypriceindexaugust2015/