Dissecting the Mortgage Approval numbers

Posted on 1 August 2017 by Ray Boulger

Be the first to comment


A good example of how different headlines can be accurate but nevertheless convey a hugely conflicting, and possibly misleading, message was apparent after publication of yesterday’s June mortgage approval figures from the Bank of England. Here is an example of two contrasting headlines:

“UK mortgage approvals drop to nine-month low in June.”

“BoE: June mortgage approvals reach highest level since 2008.”

However, a clue to why both these headlines could be justified was provided by a headline in another publication: “Approvals dip but loan values continue to rise.” 

So, as average property values have risen significantly since 2008 the average size of individual mortgage approvals has also increased, even though the number of transactions is still well below 2008 levels.

But these headlines beg other questions as well. Do they refer to total mortgage approvals or just those for purchases? Are they real figures or manipulated (or to use the technical term “seasonally adjusted”) ones? 

As only a tiny minority of people who saw one of these headlines will have been prompted to look at the full data on The Bank of England web site hopefully most will have at least gone behind the headlines and read the accompanying article to get some context.

However, a major problem is that most articles are based on manipulated figures, even though The Bank publishes both these and real figures alongside each other. The actual figures are much more relevant for any business trying to compare its performance with the market as a whole, and relevant businesses will be well aware of any seasonal influences. But other factors influence the market as well, some more predictable than others.

Furthermore, as, for dramatic effect, journalists normally like to quote as large a number as possible (e.g. the technically meaningless “up to 50 people affected” rather than the factual “at least 30….”) the fact that the actual numbers are normally much more volatile, and hence more newsworthy, as can be seen from the table below, should be an added reason to use them as the main basis for any article.

If the main focus of reporting was on the actual figures the monthly analysis could reflect on the reasons for an increase or decrease, month on month and year on year, one of which might be a seasonal influence.  

Another problem with seasonal adjustments is the different ways of calculating the adjustment, depending on who devised the system. It is an art, and a rather dubious one at that, testified by the need for frequent retrospective adjustments, which of course never get reported, rather than a science. This is best demonstrated by the Nationwide and Halifax House Price Indices, where some months Nationwide adjusts the real figure upwards and Halifax downwards, or vice versa! 

The following table compares the actual and seasonally adjusted monthly mortgage approval numbers for the last 12 months. Most of the media focus is on purchases and, as you can see, there are 5 months where the seasonally adjusted change is less than 1% and the maximum change is 6.2%, whereas in the real world there were 6 months where the percentage change was in double figures, with 4 of those being in excess of 20%. 

No doubt economists who are fans of seasonal adjustments would argue that the fact that the seasonally adjusted changes are small means they are taking the volatility out of the system, which is what they are designed to do. On that premise perhaps we should adjust the inflation figures to eliminate the effect of currency changes and wage increases with the intention of keeping the CPI and RPI changes as close to zero as possible!    

 

Purchases

Month

Actual

Seasonally Adjusted

 

Number

% Change

Number

% Change

July 16

66,820

- 13.9

61,956

- 4.7

Aug 16

63,685

- 4.7

61,626

- 0.5

Sept 16

66,139

+ 3.9

63,532

+ 3.1

Oct 16

67,470

+ 2.0

67,475

+ 6.2

Nov 16

70,577

+ 4.6

67,090

- 0.6

Dec 16

49,809

- 29.4

67,187

+ 0.1

Jan 17

47,981

- 3.7

68,654

+ 2.2

Feb 17

60,506

+ 26.1

67,785

- 1.3

Mar 17

79,525

+ 31.4

66,187

- 2.4

Apr 17

62,457

- 21.4

65,012

- 1.8

May 17

74,176

+ 18.8

65,109

+ 0.1

Jun 17

77,836

+ 4.9

64,684

- 0.7

 

 

 

 

 

 

Remortgages

July 16

43,730

- 4.5

43,012

- 0.3

Aug 16

41,347

- 5.4

40,207

- 6.5

Sept 16

44,719

+ 8.2

42,668

+ 6.0

Oct 16

45,211

+ 1.1

43,637

+ 2.3

Nov 16

49,637

+ 9.8

45,421

+ 4.1

Dec 16

39,447

- 20.5

46,508

+ 2.4

Jan 17

38,756

- 1.8

45,845

- 1.4

Feb 17

41,701

+ 7.6

43,836

- 4.4

Mar 17

49,412

+ 18.5

42,674

- 2.7

Apr 17

36,373

- 26.4

40,508

- 5.1

May 17

44,354

+ 21.9

43,145

+ 6.5

Jun 17

47,645

+ 7.4

44,548

+ 3.2 


One interesting take out from these figures is that January is the month with both the lowest number of actual purchases (47,981) and the month with the highest number of seasonally adjusted purchases (68,654)!!

Nationwide’s July house price index was announced today and as can be seen from the actual figures below, taken from its press release, there have only been two months in the last year when monthly prices have changed by more than 0.5%. Even very small increases are seen by some estate agents as a good enough excuse to produce over exuberant comments to try to talk the market up but it is worth bearing in mind that if these figures were a political poll they would all be within the margin of error. With monthly changes being so small it definitely makes more sense to focus on longer term trends.

Month – 2016/17

Average Price (£)

Change

Jul

205,715

+ 0.4

Aug

206,145

+ 0.2

Sept

206,015

- 0.1

Oct

205,904

- 0.1

Nov

204,947

- 0.5

Dec

205,898

+ 0.5

Jan

205,240

- 0.3

Feb

205,846

+ 0.3

Mar

207,308

+ 0.7

Apr

207,699

+ 0.2

May

208,711

+ 0.5

Jun

211,301

+ 1.2

Jul

211,671

+ 0.2


The current annual change is + 2.9% and as can be seen from the table net price changes in the last 5 months of 2016 were negligible. There seems little reason to expect the rest of this year to be much different.

Categories: Property market, Bank of England, House and home

Comments

Post a Comment

Please keep your comments relevant. Charcol reserves the right to edit or delete comments.

The blog postings on this site solely reflect the personal views of the authors and do not necessarily represent the views, positions, strategies or opinions of John Charcol. All comments are made in good faith, and John Charcol will not accept liability for them. We may contact you in response to your comment – by submitting your comment, you are consenting to this.

To find out more about how we collect, use and protect your data, please read our privacy policy.

You are currently offline. Some pages or content may fail to load.