Jump to content
House Price Crash Forum
Sign in to follow this  
BillyShears

Looking Closer At The Figures

Recommended Posts

I'm still going through these land registry figures as on the BBC site. Still not quite sure what they mean.

I grabbed the data for "regions" from the BBC page, and edited them into csv format so that I could load them into a statistics package. I haven't done heavy duty analysis, but this is what I've found so far. I include the full set of data down the bottom so that people can check that my numbers are correct.

FIrst, if I find the average percentage change per region, for the last quarter, it's 3.116514% The median change is 1.8%, much lower, There are a number of regions with anomalously large percentage increases. Or, some regions with unbelievable HPI, such as

place,price,quarter increase,year increase,sales

"Merthyr Tydfil",89920,26,8.6,192

"Blaenau Gwent",82491,22,8.5,209

"City Of Kingston Upon Hull",83231,21.2,17.3,1211

Notably, these are regions with low house prices. In the graphs attached to this message they can be easily seen as anomalously high increases. In particular, I presume that these house prices are after the massive increases for the last quarter, but these are places where there is little work. So it would be interesting to consider what is actually happening in those markets, who is buying the houses, and how they are being financed. But that would be another thread.

Now, the problem here is that we've counted each region as being equally important when calculating the median. A quick thing to check is whether there is a relationship between the rate of HPI in areas and the number of houses. Calculating Pearson's correlation coefficient, we find that the correlation is: -0.07143944. Small. I'm not going to bother calculating whether results are stastically significant, but it's unlikely that this number is significant. So despite the three most positive regions being farily small in terms of sales, there is not a noticable tendancy for high HPI to be in regions with small numbers of sales. The next thing to check is whether cheaper or more expensive regions have higher or lower HPI. The correlation is -0.4147684. With 109 regions, this negative correlation should be very statistically significant, showing that cheaper regions in the UK tend to have higher HPI in the last quarter. (Aside: Spearman's rank based correlation coefficient gives a measure of -0.4454562, not much different). There is a graph attached which shows a straight x-y plot of house price. If we recalcuate the average price of houses in the regions by minusing the HPI from each, then the correlation between price and HPI in the last quarter becomes a much more noticable -0.5235153. (Aside: again using rank-based statistics make little difference). So we definitely have a situation where much of the HPI is due to cheap regions catching up to the more expensive regions.

Looking at the figures for the whole year, the average HPI is 4.484404%, and the median HPI is 3.9%. The graphs attached show that over the whole year, only a very small number of regions are negative. Over the last "boom" quarter, a larger number of regions in the country are negative than in the last year. So, while the headline figures are "boom boom boom", I'm not convinced that this is a real turnaround. It seems to be a continuation of mixed figures across the country, wtih 79 out of 109 regions being positive and 30 out of 109 being negative. This is a much larger proportion of negative regions than over the last year where only 12 out of 109 regions are negative. This is a noticable trend, and means that these land registry figures certainly do not show that the change in market conditions from HPI to HPC is certainly not on hold.

Note that these figures do not include Scotland as all the Scottish figures are N/A on the BBC site.

Looking at the graphs attached, the two barplots show the distribution of changes in regions. The booming regions are clearly visible at the left of the graphs, and that most of the country is much less active can also be seen. The xy plot shows the relationship between average price and HPI. The negative correlation is visible, but it's clearly no straight line.

Finally the last graph shows the actual changes in prices. I calculated these incorrectly, though the differences are marginal. So I include a second graph where I use the proper calculations, both as a scatterplot and a barplot. The scatter plot shows the stronger negative correlation between price and HPI, and the barplot shows how many regions are negative, and how many positive.

What would be very useful is if data could be found showing how many houses are in each of the regions. Where can I get this? Anyone know?

Billy Shears

"place","price","quarter","year","sold",

"Merthyr Tydfil",89920,26,8.6,192,

"Blaenau Gwent",82491,22,8.5,209,

"City Of Kingston Upon Hull",83231,21.2,17.3,1211,

"Hartlepool",98770,17.2,32.3,415,

"Neath Port Talbot",102936,17.1,7.1,503,

"Middlesbrough",108982,16.2,22.7,485,

"Gwynedd",153558,14.9,8.5,348,

"Ceredigion",178402,10.3,7.9,216,

"Pembrokeshire",166873,10.3,5.1,388,

"Poole",261746,10.2,4.1,736,

"Blackburn With Darwen",98762,9.8,8.5,677,

"Denbighshire",140981,9.2,1.3,336,

"Stoke-On-Trent",89910,9.1,4.4,1089,

"Blackpool",119112,8.7,9.6,750,

"Powys",169929,8.5,4.3,335,

"City Of Peterborough",146481,7.9,8.5,852,

"Caerphilly",115784,7.9,7.2,542,

"Halton",127289,7.7,2,464,

"Bridgend",133572,7.6,12.8,404,

"Bracknell Forest",239397,7.3,3.6,530,

"City Of Plymouth",151301,7.3,1.7,1154,

"West Yorkshire",138242,7.2,7,9129,

"Wrexham",147293,6.9,7.1,427,

"Torfaen",120546,6.8,5.4,253,21645

"Greater London",306664,6.5,5.9,29242,

"Durham",114329,6.4,6.5,1956,

"Lancashire",128806,5.8,8.8,5417,

"Cumbria",143851,5.6,7.7,2006,

"North East Lincolnshire",102179,5.4,4.8,786,

"The Vale Of Glamorgan",178523,5.3,16.4,495,

"Monmouthshire",199073,5.2,7,294,

"South Yorkshire",126326,5.2,7.8,4577,

"Merseyside",132854,4.7,8.2,4379,

"Cardiff",171619,4.6,2,1186,

"Staffordshire",161232,4.4,3.9,2886,

"Rhondda Cynon Taff",93453,4.2,2.7,775,

"Carmarthenshire",137252,4.1,7.2,537,

"Hertfordshire",257311,3.8,3.2,4520,

"Tyne And Wear",132526,3.8,9.1,4023,

"Redcar And Cleveland",109289,3.8,5.2,502,

"Torbay",174208,3.7,4.1,642,

"Newport",144790,3.7,4.8,530,

"Greater Manchester",134935,3.6,8.5,10155,

"York",184466,3.1,4.7,728,97281

"Bedfordshire",193746,3,3.9,1804,

"Flintshire",142754,3,6.8,406,

"Brighton And Hove",222241,2.9,1.7,1328,

"Bournemouth",202293,2.9,5.7,992

"Thurrock",169523,2.6,-1.5,619,

"Buckinghamshire",307451,2.5,8,1733,

"North Yorkshire",200881,2.5,5.4,2183

"Swansea",139153,2.4,7.8,876

"Surrey",318980,2,8,5037,112259

"West Midlands",146903,1.9,8,8775,

"West Berkshire",255346,1.8,2.6,721,121755

"Essex",212460,1.5,3.4,6001,

"Cambridgeshire",197085,1.5,1.6,2373,

"Southend-On-Sea",182346,1.4,5.5,948,

"Reading",196094,1.3,2.6,743,

"Isle Of Wight",181138,1.3,3.9,683,

"Slough",187772,1.2,2.9,449,

"Milton Keynes",175337,1.2,1.1,1224,

"Luton",150880,1.2,4.6,782,

"Oxfordshire",257528,1.1,1.6,2417,

"Conwy",151565,1,-6.6,405,

"City Of Derby",137082,1,3.3,935,

"Darlington",125556,1,2.3,530,

"Leicester",134130,0.9,6.6,954,

"Bath And North East Somerset",249488,0.7,3.6,646,

"Cornwall",203006,0.6,-0.8,2131,

"East Riding Of Yorkshire",154431,0.6,7.6,1475,

"Nottinghamshire",146871,0.5,-0.9,3059,

"Norfolk",166062,0.3,2.9,3495,

"Stockton-On-Tees",128643,0.3,-2.9,682,

"Lincolnshire",147659,0.2,1.8,3091,

"North Lincolnshire",122869,0.1,2.6,592

"Northumberland",149225,0,5.9,1129,

"Herefordshire",208672,-0.2,6.2,658,

"North Somerset",185780,-0.3,2,928,

"Gloucestershire",201799,-0.4,1.1,2512,

"Portsmouth",154803,-0.4,1.2,867,

"Swindon",155868,-0.5,0.5,994,

"Windsor And Maidenhead",335761,-0.7,1.1,496,

"Wokingham",272890,-0.8,3.8,697,

"Medway",154231,-0.8,0.6,1125,

"Kent",206628,-1.1,3.9,6353,

"Northamptonshire",158537,-1.1,-0.1,3322,

"Wrekin",140819,-1.2,4.3,679,

"Cheshire",185036,-1.4,1.6,2618,

"Devon",212600,-1.5,1.3,3123,

"Southampton",161716,-1.5,1.8,950,

"Derbyshire",147537,-1.5,1.6,2901,

"Shropshire",184664,-1.6,2.8,1127,

"Leicestershire",170145,-1.6,-1.5,2555,

"Hampshire",230858,-2.2,3.7,5307,

"South Gloucestershire",182959,-2.3,4.6,1058,

"City Of Bristol",178116,-2.3,-0.9,1767,

"Rutland",228672,-2.4,7.1,148,

"Suffolk",176077,-2.7,0.2,3238,

"Isle Of Anglesey",145033,-2.8,-0.7,213,

"East Sussex",198541,-3.5,-1,2544,

"City Of Nottingham",119797,-4,1.2,1035,

"Dorset",226010,-4.1,0.2,2051,

"West Sussex",221415,-4.3,-0.4,3827,

"Wiltshire",209024,-5,1.2,1813,

"Somerset",183850,-5.3,-2,2296,

"Worcestershire",184936,-5.9,1.1,2102,

"Warrington",156929,-6.3,1.7,713,

"Warwickshire",184860,-7.2,-2.4,2255

graphs.png

graph2.png

post-3706-1147207914.png

post-3706-1147207929.png

Share this post


Link to post
Share on other sites

:unsure:

I think I need to brush up on my stats.

Well done Billy.

I'm not sure that congratulations are in order. I've only had a first crack at these numbers. If this were my day job I might spend a week full time analysing the data. And I'm not sure that the page I copied them from was complete, as I'm sure that there should be an "East Midlands" in there at -3.X%. Or, is that not a region.

In any case, the figures are nowhere near as bullish as people seem to have been claiming. With the noticable loosening of credit and with a confidence boost for buyers given that no interest rate rises have happened for a while, and it being spring, some gain in the market is understandable. But the loosening of credit is a once only change, and the summer and then the winter months are coming. More regions being negative in the last quarter versus the last year is also a "not quite the boom its made out to be" moment.

Edit: and there are even typos in the title of one of the graphs. Grrrrr.....

Billy Shears

Edited by BillyShears

Share this post


Link to post
Share on other sites

I'm not sure that congratulations are in order. I've only had a first crack at these numbers. If this were my day job I might spend a week full time analysing the data. And I'm not sure that the page I copied them from was complete, as I'm sure that there should be an "East Midlands" in there at -3.X%. Or, is that not a region.

In any case, the figures are nowhere near as bullish as people seem to have been claiming. With the noticable loosening of credit and with a confidence boost for buyers given that no interest rate rises have happened for a while, and it being spring, some gain in the market is understandable. But the loosening of credit is a once only change, and the summer and then the winter months are coming. More regions being negative in the last quarter versus the last year is also a "not quite the boom its made out to be" moment.

Edit: and there are even typos in the title of one of the graphs. Grrrrr.....

Billy Shears

Even so, very interesting. Can you recommend any reading for those of us who are statistically challenged?

Share this post


Link to post
Share on other sites

Even so, very interesting. Can you recommend any reading for those of us who are statistically challenged?

Schaum's Business Statistics will give you a good grounding.

SW

Edited by SHERWICK

Share this post


Link to post
Share on other sites

What would be very useful is if data could be found showing how many houses are in each of the regions. Where can I get this? Anyone know?

Billy

I came across this report about the affordability of intermediate housing (presumably 2nd & 3rd TB'ers) from pages 24 to 35 it gives the number of working households in the areas - this will give an idea but in some areas such as Bournemouth and parts of Devon and Cornwall where there are many non working (ie retired) household the figures will be less reliable

Even if it doesn't help with the stats it is interesting reading albeit 2 years out of date

http://www.jrf.org.uk/bookshop/eBooks/185935243X.pdf

Hope this helps we lecturers must stick together

Regards

CS

Share this post


Link to post
Share on other sites

Number of Households from 2001 Census

Billy

extract from 2001 census about households

There are a total of 22,539,000 households in England and Wales. 21,660,000 of these are occupied (20,451,000 in England and 1,209,000 in Wales), 727,000 are vacant (676,000 in England and 51,000 in Wales) and 151,000 are second homes or holiday accommodation (135,000 in England and 16,000 in Wales).

Vacancy rates are above average (3.9 per cent or more) in the North West, North East, Wales and Yorkshire & the Humber while in London, South East, East and South West vacancy rates are 2.8 per cent or less.

Among local authorities South Holland in the East Midlands reports the highest proportion of whole houses or bungalows, followed by Derwentside and Staffordshire Moorlands. However, almost half of London's households are flats, maisonettes or apartments. Of these 33 per cent are purpose-built flats and 14 per cent are part of a converted or shared house.

Of the 20 districts with the highest proportion of flats, maisonettes or apartments, 18 are in London. The other two districts are the south coast resorts of Brighton & Hove and Bournemouth.

In both England and Wales only 0.4 per cent of households are a caravan or other mobile or temporary structure.

Good luck with the calcs

CS

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
Sign in to follow this  

  • Recently Browsing   0 members

    No registered users viewing this page.

  • 301 Brexit, House prices and Summer 2020

    1. 1. Including the effects Brexit, where do you think average UK house prices will be relative to now in June 2020?


      • down 5% +
      • down 2.5%
      • Even
      • up 2.5%
      • up 5%



×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.