2021 Irish Property Market chat - mod warnings post 1

hometruths · 05-02-2021 12:28AM

Graham wrote: »

do tell

I’ll use fliballs source:

fliball123 wrote: »

I replied to Props with this link that shows his figures are completely inaccurate with regard to houses coming on stream of course when you prove him wrong he simply ignores the post..Here it is again for you props

https://www.housing.eolasmagazine.ie/the-challenge-of-housing-obsolescence/#:~:text=According%20to%20Sirr's%20figures%2C%20a,19%2C000%20is%20due%20to%20obsolescence.

These are figures from 2016 but the method of calculation is the same:

According to Sirr’s figures, a total of 123 houses become obsolete in Ireland every week, or 6,394 houses a year.  The difference between the headline figure of nearly 51,000 houses being built in Ireland annually and the actual net gain of just short of 19,000 is due to obsolescence.

Apply to that to 2020. The difference between 21k houses being built and the net gain of 28k is due to obsolescence.

I.e the amount of people renovating/rebuilding properties is greater than that letting them fall into dereliction.

We don’t have an obsolescence problem and Props was right to disregard it in his assumptions.

Timing belt · 05-02-2021 12:31AM

[/HTML]

schmittel wrote: »

I’ll use fliballs source:

These are figures from 2016 but the method of calculation is the same:

According to Sirr’s figures, a total of 123 houses become obsolete in Ireland every week, or 6,394 houses a year.  The difference between the headline figure of nearly 51,000 houses being built in Ireland annually and the actual net gain of just short of 19,000 is due to obsolescence.

Apply to that to 2020. The difference between 21k houses being built and the net gain of 28k is due to obsolescence.

I.e the amount of people renovating/rebuilding properties is greater than that letting them fall into dereliction.

We don’t have an obsolescence problem and Props was right to disregard it in his assumptions.

I am confused as the housing stock you are getting your 28k from includes vacant properties....Are you saying that the additional housing is derelict properties being brought back to life?

hometruths · 05-02-2021 12:33AM

Timing belt wrote: »

[/HTML]

I am confused as the housing stock you are getting your 28k from includes vacant properties....Are you saying that the additional housing is derelict properties being brought back to life?

I can see you are confused. But I cannot help you as I don’t understand why you are confused.

hometruths · 05-02-2021 12:35AM

Timing belt wrote: »

I know a housing report that does not have a category for apartments....

I meant who’d have thought your figures didn’t add up.

Graham · 05-02-2021 12:36AM

schmittel wrote: »

We don’t have an obsolescence problem and Props was right to disregard it in his assumptions.

I don't think obsolescence is a problem. Obsolescence is a necessary part of things changing/improving.

At the same time, it would be silly to suggest the loss of 5000 - 8000 properties per year is of no consequence and shouldn't be taken into account when considering new build requirements.

Timing belt · 05-02-2021 12:40AM

schmittel wrote: »

I meant who’d have thought your figures didn’t add up.

Its your logic that does not add up and as I said before if you want to argue go find a mirror I am sure there must be one in one of the vacants

hometruths · 05-02-2021 12:41AM

Graham wrote: »

I don't think obsolescence is a problem. Obsolescence is a necessary part of things changing/improving.

At the same time, it would be silly to suggest the loss of 5000 - 8000 properties per year is of no consequence and shouldn't be taken into account when considering new build requirements.

But if we’re seeing a net gain of we are not losing 5-8k properties a year. That’s the point. Older properties are being maintained better or brought back into use.

It would be silly to make new build need projections based on losing 8k houses a year when we are actually gaining 7k a year. That’s a 15k swing.

hometruths · 05-02-2021 12:43AM

Timing belt wrote: »

Its your logic that does not add up and as I said before if you want to argue go find a mirror I am sure there must be one in one of the vacants

I’m not trying to argue, your figures clearly were incorrect, and I was just trying to clarify what I meant. No offense meant.

Graham · 05-02-2021 12:47AM

schmittel wrote: »

But if we’re seeing a net gain of we are not losing 5-8k properties a year. That’s the point.

We're still losing 5-8k properties a year and replacing them with 5-8k new properties.

If I have 4 apples, lose them and buy 8 new apples, It doesn't change the fact I've lost 4 apples.

hometruths · 05-02-2021 12:50AM

Graham wrote: »

We're still losing 5-8k properties a year and replacing them with 5-8k new properties.

If I have 4 apples, lose them and buy 8 new apples, It doesn't change the fact I've lost 4 apples.

Why are you so certain we’re losing this many properties a year? Where are you getting 5-8k from?

Marius34 · 05-02-2021 12:53AM

schmittel wrote: »

Why are you so certain we’re losing this many properties a year? Where are you getting 5-8k from?

Central Bank estimates around 5K, similarly other sources i've seen is around 5k-7K

Graham · 05-02-2021 12:54AM

schmittel wrote: »

Why are you so certain we’re losing this many properties a year? Where are you getting 5-8k from?

There are a range of estimates out there which is way I used a fairly broad range.

Comparing against other countries our assorted figures appear to be broadly aligned.

Have you any reason to suspect we have a significantly lower amount of obsolescence than anywhere else?

hometruths · 05-02-2021 01:05AM

Graham wrote: »

There are a range of estimates out there which is way I used a fairly broad range.

Comparing against other countries our assorted figures appear to be broadly aligned.

Have another look at those estimates and see how they arrive at the estimate. You'll probably find it relates in some way to the difference between the number of new builds and the net gain in the total housing stock.

Graham wrote: »

Have you any reason to suspect we have a significantly lower amount of obsolescence than anywhere else?

Yes. We had a gain of 28k in the housing stock after building 21k new units - a difference of 7k. It's mathematically impossible to have lost 8k houses in that scenario.

Timing belt · 05-02-2021 01:06AM

schmittel wrote: »

I can see you are confused. But I cannot help you as I don’t understand why you are confused.

Because it one year there was a drop of 6.7k decrease in detached properties

hometruths · 05-02-2021 01:09AM

Marius34 wrote: »

Central Bank estimates around 5K, similarly other sources i've seen is around 5k-7K

Do you have a link to these?

hometruths · 05-02-2021 01:11AM

Timing belt wrote: »

Because it one year there was a drop of 6.7k decrease in detached properties

So what? We’re fond of blaming Leitrim for this sort of thing. It could farmer pats cottage.

The important figure is the total stock increased by 28k.

PropQueries · 05-02-2021 01:19AM

Timing belt wrote: »

Because it one year there was a drop of 6.7k decrease in detached properties

Haven't read the report, but given that was the big outlier, did they explain it?

If they didn’t explain or at least refer to it, as it’s a fairly big elephant sized outlier, there may be some miscalculation in the gathering of original data collection IMO

Graham · 05-02-2021 01:19AM

schmittel wrote: »

Yes. We had a gain of 28k in the housing stock after building 21k new units - a difference of 7k. It's mathematically impossible to have lost 8k houses in that scenario.

If the 21k/28k figures are correct it means one of two things:

a) we've discovered the secret to everlasting property so housing is never obsolete.
b) something else.

I'd look at 'b'

Marius34 · 05-02-2021 01:23AM

schmittel wrote: »

Do you have a link to these?

"Our scenario assumes that 5,000 new dwellings are required each year to replace existing housing stock lost due to obsolescence"

https://www.centralbank.ie/docs/default-source/publications/economic-letters/vol-2019-no-14-population-change-and-housing-demand-in-ireland-(conefrey-and-staunton).pdf

hometruths · 05-02-2021 01:24AM

Graham wrote: »

If the 21k/28k figures are correct it means one of two things:

a) we've discovered the secret to everlasting property so housing is never obsolete.
b) something else.

I'd look at 'b'

I’d look at ‘B’ too - Might be related to the reports that it is impossible to get builders for renovation work because they are all flat out renovating properties. I don’t think this is a secret.

Timing belt · 05-02-2021 01:25AM

schmittel wrote: »

So what? We’re fond of blaming Leitrim for this sort of thing. It could farmer pats cottage.

The important figure is the total stock increased by 28k.

Its an important figure as because I would have thought that if derelict building s were being brought back to life the most profitable ones would be detached properties.

But if we go back an concentrate on the 28k and if you are saying that this is due to derelict properties being brought back to life it doesn't add up unless people are living without Electricity.

Browney7 · 05-02-2021 01:26AM

Graham wrote: »

If the 21k/28k figures are correct it means one of two things:

a) we've discovered the secret to everlasting property so housing is never obsolete.
b) something else.

I'd look at 'b'

https://www.cso.ie/en/releasesandpublications/er/ndc/newdwellingcompletionsq42020/

It could be a multitude of things. The above link explains CSO is adjusted ESB connections (as the CSO believe raw ESB data over eggs the number). Student accom is likely a decent sized contributor - it's very likely An Post will recognise and capture "Units 1-20 College block X" whereas CSO won't.

In rural areas, derelict or small cottages are often desirable as it's easier to get planning and knock the place to rebuild.

Timing belt · 05-02-2021 01:27AM

PropQueries wrote: »

Haven't read the report, but given that was the big outlier, did they explain it?

If they didn’t explain or at least refer to it, as it’s a fairly big elephant sized outlier, there may be some miscalculation in the gathering of original data collection IMO

They made no reference to it....

Also in relation to your theory on unfinished buildings the ESB data would not add support to your view

Graham · 05-02-2021 01:28AM

That would be my guess Browney7, a multitude of figures all based on data that's not entirely accurate or measuring slightly different things.

Marius34 · 05-02-2021 01:34AM

schmittel wrote: »

We had a gain of 28k in the housing stock after building 21k new units - a difference of 7k. It's mathematically impossible to have lost 8k houses in that scenario.

Honestly do you think we had a gain of 28K in a year of 2020 of housing stocks? Or it's likely those numbers means something else?
Would you the same as an annual gain, if in Geodirectory total stocks numbers would appear only 5K difference? I think you would reject those numbers to call as an annual gain.

Timing belt · 05-02-2021 01:34AM

ESB connection data

Row Labels All connections New dwelling completion Non-Dwelling Reconnection Unfinished
2011 10,289 6,994 827 1,090 1,378
2012 8,345 4,911 723 1,260 1,451
2013 8,207 4,575 653 1,473 1,506
2014 10,929 5,518 753 2,290 2,368
2015 12,623 7,219 845 2,888 1,671
2016 14,923 9,879 843 2,341 1,860
2017 19,185 14,355 956 2,512 1,362
2018 22,385 17,916 908 2,673 888
2019 25,259 21,087 857 2,591 724
2020 24,543 20,676 975 2,399 493

hometruths · 05-02-2021 01:43AM

Marius34 wrote: »

"Our scenario assumes that 5,000 new dwellings are required each year to replace existing housing stock lost due to obsolescence”

https://www.centralbank.ie/docs/default-source/publications/economic-letters/vol-2019-no-14-population-change-and-housing-demand-in-ireland-(conefrey-and-staunton).pdf

Strangely you did not quote the full sentence:

"Our scenario assumes that 5,000 new dwellings are required each year to replace existing housing stock lost due to obsolescence based on Duffy et al.(2014)”

So they’re basing their projections of obsolescence in 2019 on 2014 figures, when we’d just finished up knocking ghost estates.

hometruths · 05-02-2021 01:44AM

Marius34 wrote: »

Honestly do you think we had a gain of 28K in a year of 2020 of housing stocks? Or it's likely those numbers means something else?
Would you the same as an annual gain, if in Geodirectory total stocks numbers would appear only 5K difference? I think you would reject those numbers to call as an annual gain.

Yes honestly I do. I think total housing stock means total housing stock. Not sure what else it could mean.

hometruths · 05-02-2021 02:00AM

Timing belt wrote: »

ESB connection data

Row Labels All connections New dwelling completion Non-Dwelling Reconnection Unfinished
2011 10,289 6,994 827 1,090 1,378
2012 8,345 4,911 723 1,260 1,451
2013 8,207 4,575 653 1,473 1,506
2014 10,929 5,518 753 2,290 2,368
2015 12,623 7,219 845 2,888 1,671
2016 14,923 9,879 843 2,341 1,860
2017 19,185 14,355 956 2,512 1,362
2018 22,385 17,916 908 2,673 888
2019 25,259 21,087 857 2,591 724
2020 24,543 20,676 975 2,399 493

If you were to assume all reconnections and unfinished were renovations of uninhabitable houses then it still leaves 4K difference of unexplained new stock.

Timing belt · 05-02-2021 07:51AM

schmittel wrote: »

If you were to assume all reconnections and unfinished were renovations of uninhabitable houses then it still leaves 4K difference of unexplained new stock.

I had a look back at previous Geo-directory reports to compare the housing stock movement year on year with number of new address they added each year to see was there something that would explain the difference.
Geo new address added Difference
2020 2,042,420 28,063 21,851 6,212
2019 2,014,357 20,685 20,359 326
2018 1,993,672 19,323 21,207 -1,884
2017 1,974,349 -46,174 36,218 -82,392
2016 2,020,523 10,627 13,842 -3,215
2015 2,009,896

The movement in 2017 looked strange but when I looked into the report it would appear that the movement was due to no longer including derelict dwellings in their housing stock from 2017 onwards.

This then had me questioning how the geo-directory housing stock claimed to be similar to the 2016 Census considering the census does not include derelict building. So I went to validate the Geo-directory data to the 2016 Census data and found that the geo-directory housing stock is circa 60-70k lower than the CSO figure on a like for like basis
i.e.
Housing Stock Derlict Dwellings Temporary Dwellings Student Accomadation Adjusted Housing Stock Housing Stock per Census 2016 New housing units per CSO CSO Housing stock (Census 2016 + new ) Difference
2020 2,042,420 9,381 10,740 2,022,299 20,676 2,089,271 -66,972
2019 2,014,357 9,239 8,592 1,996,526 21,087 2,068,595 -72,069
2018 1,993,672 9,149 6,444 1,978,079 17,916 2,047,508 -69,429
2017 1,974,349 9,047 4,296 1,961,006 14,355 2,029,592 -68,586
2016 2,020,523 58,855 8,959 2,148 1,950,561 2,003,645 11,592 2,015,237 -64,676

Geo-directory includes student accommodation, Temporary-Dwellings and derelict housing. The CSO does not.

The Temporary dwelling data comes direct from the Geo-directory report
The student accommodation figure comes from https://www.irishexaminer.com/business/economy/arid-40220573.html
The CSO said its figures do not include student accommodation, though it may in time include the units in its release.
Nonetheless, it said over 1,300 “bed spaces” were created as part of student accommodation schemes in the fourth quarter alone and means that 10,740 student units have been built since the summer of 2016.
The Derelict data is derived from the Geo-directory reports

2017 Q2 1,967,698
2016 Q4 2,020,523
-52,825
New housing units per CSO 6030
Difference Assumed to be Derelict Dwellings -58,855

The student accommodation accounts for circa. 2k of the difference and maybe more as I just distributed the 10k of student accommodation evenly over the period when it may have actually come in waves.
In order to try and see if this is the case I looked at the movement in housing stock from 2020 Q4 to 2017 Q2 and compared it to the new completions data from the CSO
2020 2017
Q4 Q2 Difference

DETACHED 647,505 697,672 -50,167
SEMI-DETACHED 501,535 475,151 26,384
TERRACED 569,423 538,728 30,695
BUNGALOW 289,356 223,003 66,353
DUPLEX 25,220 24,135 1,085
TEMPORARY-DWELLING 9,381 9,009 372

2,042,420 1,967,698 74,722
New Completions per CSO 68004
Difference 6,718

This difference could be accounted for by the student accommodation
i.e.
10,740 x 60 months = 179 per month
179 x 42 months = 7518

The 6k difference appeared in 2020 and I don't think all the student accommodation came online then... maybe it was only reported in 2020 who knows.

What is also interesting is that in 2017 geo-directory added 36k new address but the no of completions that year was only 14k giving a difference of 22k and no explanation was provided in the report. It is also the same time that they excluded derelict dwellings from their housing stock and started reporting vacant properties which may just be a coincidence but I think they did a tidy up exercise when they had the CSO data to compare against IMO

This then got me think could it be something similar are they tidying up vacant/derelict dwellings ahead of the census or that they have shared the data with the census enumerators who have undertaken some reviews of the vacant properties and feed it back to geo-directory. At first I thought this would not be possible with Covid but when I looked into the difference of the 6k it seemed to appear between the 2019 Q4 report and the 2020 Q2 Report so there is a good chance it happened pre-covid. That is just my guess based on what I have seen on the data IMO

The Final observation unrelated to the 6k difference is the 60-70k difference between the overall housing stock when you compare geo-directory to the CSO data..it must account for some of the difference on the vacant property figures reported by the CSO and Geo-directory as I can't see the number of occupied properties being understated by geo-directory otherwise people wouldn't be getting post etc..

2021 Irish Property Market chat - *mod warnings post 1*

Comments

2021 Irish Property Market chat - mod warnings post 1