Insolvency Statistics: Change to the method for statistics on ... - Gov.uk
Insolvency Statistics: Change to the method for statistics on company insolvencies Last reviewed: January 2018
Contents Summary ............................................................................................................................................. 2 Background ......................................................................................................................................... 3 Reasons for change ............................................................................................................................. 3 Details of the change .......................................................................................................................... 5 Impact of the change .......................................................................................................................... 5 Communicating the change ................................................................................................................ 6 Detailed impact of the change ............................................................................................................ 7 England & Wales ............................................................................................................................. 7 Scotland......................................................................................................................................... 12 Northern Ireland ........................................................................................................................... 18
Summary In accordance with our revisions policy, we are pre-announcing a change to the methodology used to calculate the number of company insolvencies in a given quarter. This document sets out the reasons for making the change, and a summary of the impact it will have on the statistics. Currently, company insolvencies are counted from the date the insolvency started, with statistics for the most recent quarter being an estimate based on incomplete data. Recent instances of bulk insolvencies of connected companies have had an impact on the quality of these initial estimates. To address this, from April 2018, most company insolvencies will be counted from the date of registration. This means that statistics will no longer be estimated and then revised between quarters, and we will be able more readily to identify and account for bulk insolvencies. This change does not affect statistics on compulsory liquidations in England & Wales and in Northern Ireland, for which data are obtained from administrative records from the Insolvency Service in each jurisdiction. All statistics in this document are not seasonally adjusted (NSA). If you have comments or feedback regarding this change, please contact the Insolvency Service’s statistics team via email at [email protected]. If we do not hear compelling feedback to the contrary, we intend to implement this change beginning with the 2018Q1 edition of the Insolvency Statistics (to be published in April 2018).
Background In 2015, having found weaknesses in the method used to calculate company insolvencies, the Insolvency Service ran a public consultation proposing changes to the method. Three options were presented: 1. No change. 2. Count insolvencies in the quarter they are first registered at Companies House (regardless of when they actually started). 3. Count insolvencies in the quarter they started (regardless of when they were registered). The main advantage of option 2 was that it would provide an accurate record of all insolvencies. The main disadvantage was that, since insolvencies were frequently registered late, the statistics for a given quarter were likely to be an over- or underestimate of the “true” position. The main advantage of option 3 was that statistics for a given quarter would accurately reflect the number of insolvencies. The disadvantage was that, since insolvencies were frequently registered late, statistics for the most recent quarter would be estimated, and revised in subsequent quarters. After considering responses to the consultation, the Insolvency Service selected option 3, and implemented the change in October 2015. Details of the method used were published. Using historic data, the Insolvency Service calculated the accuracy of first estimates of company insolvency, if this method had been in place since 2008. The results were also published. In each quarterly edition of the Insolvency Statistics, the latest quarter’s statistics and the change since previous periods were referred to as “estimated”.
Reasons for change The difference between first estimates of company insolvencies, and subsequent revisions, has been monitored by the Insolvency Service’s statistics team on an ongoing basis. By and large, the differences have been fairly small in absolute terms, with the exception of creditors’ voluntary liquidations (CVLs) in England and Wales (E&W) in 2017Q2.
Estimate vs actual - CVL, E&W Revision
Actual
Estimate
Insolvencies (NSA)
5,000 4,000
3,000 2,000 1,000
0 -1,000 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
This period coincided with an unusually high numbers of CVLs of connected companies, which themselves resulted from a closure of a tax loophole which made the companies unviable. These CVLs were also registered on the Companies House website in the following quarter. A similar set of “bulk insolvencies” was also recorded in 2016Q4, but as these were registered in the same quarter these could be accounted for. The statistics team was made aware of these bulk insolvencies by operational colleagues in the Insolvency Service, and made proactive adjustments to the statistics to account for their expected effects. However, the difference between the estimate and the revised figure for CVLs was larger in nominal terms compared with periods without bulk insolvencies. Moreover, these revisions also affected the percentage difference between a given period and the same quarter the previous year: in 2017Q2, the estimated percentage change compared with 2016Q2 was +36%, which was subsequently revised to +44%.
Estimate vs actual - 4Q change, CVL, E&W Estimate
Actual
80% 70% 60% 50% 40% 30% 20% 10% 0% -10%
2015Q3
2016Q1
2016Q3
2017Q1
The statistics team is aware of a further tranche of bulk insolvencies, which have not all been registered on the same date, which in turn reduces the ability to estimate the effect on the statistics. Further bulk insolvencies cannot be ruled out in the future. The result of this could be that revisions to the data may change the narrative – for example we may initially estimate that numbers of insolvencies have decreased, but revised data may instead show an increase – which would undermine the usefulness of the statistics as a measure of recent trends.
Details of the change Beginning in the 2018Q1 edition of the statistics (to be published in April 2018), we are planning to change to produce statistics using option 2 outlined above. We will also provide statistics on a consistent basis back to 2000Q1. This will give the following benefits to users:
The statistics will not be revised between editions (except in the case that processing errors are identified). We will be able more readily to identify and account for bulk insolvencies. We will be able to publish breakdowns at industry level covering the latest quarter (the current method only allows us to publish these statistics three months in arrears).
Impact of the change The chart below shows the volumes of company insolvencies each quarter since 2000Q1, using both the option 2 and option 3 methods. The time series for the two methods is highly aligned, the largest differences occurring when bulk insolvencies are registered in the quarter after they started. In most cases, the difference in one quarter is cancelled out the following quarter. Similarly, the percentage change in the statistics compared with the same period a year earlier is closely aligned between the two methods, except where bulk insolvencies are registered late. It follows that statistics based on date of registration are a good proxy for the “true” number of insolvencies in a given period.
Comparison of methods - Total, E&W Difference
Reg date (option 2)
Start date (option 3)
6,000
4,000 2,000 0
-2,000 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
8,000
Comparison of methods (4Q % change) - Total, E&W Reg date (option 2)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2004Q1
2003Q1
2002Q1
2001Q1
100% 80% 60% 40% 20% 0% -20% -40%
Communicating the change We will:
Pre-announce the change in the 2017Q4 edition of the Insolvency Statistics. Describe the reasons for the change and the impact it will have (this document). Notify known users in advance of the change. Publicise the change using other channels e.g. the Insolvency Service’s Twitter account. Draw attention to the change in the 2018Q1 edition of the Insolvency Statistics.
If you have comments or feedback regarding this change, please contact the Insolvency Service’s statistics team via email at [email protected].
Detailed impact of the change England & Wales
Revision
Estimate
Estimate
5,000 4,000
3,000 2,000
1,000 0
-1,000 2016Q1
2016Q3
2017Q1
2015Q3
2017Q3
Comparison of methods - Total, E&W Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - Total, E&W
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
100% 80% 60% 40% 20% 0% -20% -40%
8,000 6,000
4,000 2,000 0
-2,000 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Actual
70% 60% 50% 40% 30% 20% 10% 0% -10% -20%
6,000
2015Q3
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned except where there have been bulk insolvency events that have been registered in the following quarter.
Actual
2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
There was one occasion where the first estimate was that insolvencies had increased since the same period the previous year, but when the figures were revised this was changed to a decrease.
Estimate vs actual - 4Q change, Total, E&W
Estimate vs actual - Total, E&W
Insolvencies (NSA)
Total company insolvencies The difference between the first estimate of the total number of company insolvencies in England and Wales, and the actual number, tended to be fairly small in absolute terms with the exception of in 2017Q2 when a bulk insolvency event was registered late with Companies House.
Revision
Actual
Estimate
Estimate
4,000
3,000 2,000 1,000
0 -1,000 2015Q3
2016Q1
2016Q3
2017Q1
2015Q3
2017Q3
Comparison of methods - CVL, E&W Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVL, E&W
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
Compulsory liquidations Statistics on compulsory liquidations in England and Wales are obtained from administrative records of the Insolvency Service, and the method for producing quarterly statistics is not changing as part of this proposal.
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2004Q1
2002Q1
2001Q1
100% 80% 60% 40% 20% 0% -20% -40%
5,000 4,000 3,000 2,000 1,000 0 -1,000 -2,000 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Actual
80% 70% 60% 50% 40% 30% 20% 10% 0% -10%
5,000
2003Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned except where there have been bulk insolvency events that have been registered in the following quarter.
Estimate vs actual - 4Q change, CVL, E&W
Estimate vs actual - CVL, E&W
Insolvencies (NSA)
Creditors’ voluntary liquidations (CVLS) The difference between the first estimate of the number of CVLs in England and Wales, and the actual number, tended to be fairly small in absolute terms with the exception of in 2017Q2 when a bulk insolvency event was registered late with Companies House.
Revision
Estimate
Estimate
400 350 300 250 200 150 100 50 0 -50
10%
5% 0% -5% -10%
-15% 2015Q3
2017Q3
Comparison of methods - Administration, E&W Reg date (option 2)
Reg date (option 2)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2,500 2,000 1,500 1,000 500 0 -500 -1,000
2017Q1
300% 250% 200% 150% 100% 50% 0% -50% -100%
2009Q1
Start date (option 3)
2016Q3
Comparison of methods (4Q % change) Administration, E&W
2001Q1
Difference
2016Q1
2008Q1
2017Q1
2007Q1
2016Q3
2006Q1
2016Q1
2005Q1
2015Q3
Insolvencies (NSA)
Actual
15%
2004Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned except where there have been bulk insolvency events that have been registered in the following quarter.
Actual
2003Q1
Insolvencies (NSA)
There was one occasion where the first estimate was that insolvencies had decreased since the same period the previous year, but when the figures were revised this was changed to an increase.
Estimate vs actual - 4Q change, Administration, E&W
Estimate vs actual - Administration, E&W
2002Q1
Administrations The difference between the first estimate of the number of administrations in England and Wales, and the actual number, tended to be fairly small in absolute terms.
Revision
Actual
Estimate vs actual - 4Q change, CVA, E&W
Estimate
Estimate
140
100 80 60 40 20
0 -20 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - CVA, E&W Difference
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVA, E&W
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2007Q1
0
-100
2006Q1
100
2005Q1
200
2002Q1
300
2001Q1
100% 80% 60% 40% 20% 0% -20% -40% -60% -80%
400 Insolvencies (NSA)
2016Q1
2004Q1
The time series since 2000 are closely aligned for both option 2 and option 3.
Actual
50% 40% 30% 20% 10% 0% -10% -20% -30% -40% -50%
120
Insolvencies (NSA)
The difference between the first estimate of the number of CVAs in England and Wales, and the actual number, tended to be fairly small in absolute terms – usually fewer than ten companies.
Estimate vs actual - CVA, E&W
2003Q1
Company voluntary arrangements (CVAs) Volumes of CVAs each quarter have been fairly small – usually fewer than 100 companies each quarter.
Revision
Actual
Estimate
Estimate
6
500%
5
400% 200%
3
100%
2
0%
1
-100%
-200% 2015Q3
2017Q3
Comparison of methods - Administrative Receiver, E&W Reg date (option 2)
2017Q1
2016Q1
2015Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2014Q1
0
-100
2013Q1
100
2012Q1
200
500% 400% 300% 200% 100% 0% -100% -200%
2011Q1
300
2017Q1
Start date (option 3)
2010Q1
400
Reg date (option 2)
2009Q1
Start date (option 3)
2016Q3
Comparison of methods (4Q % change) Administrative receiver, E&W
2001Q1
Difference
2016Q1
2008Q1
2017Q1
2007Q1
2016Q3
2006Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
2016Q1
2005Q1
2015Q3
2004Q1
0
Insolvencies (NSA)
Actual
300%
4
2003Q1
Insolvencies (NSA)
There tended to be no difference between the first estimate of the number of administrative receiverships in England and Wales, and the actual number – the exception being in 2017Q1 when two cases were registered in the following quarter.
Estimate vs actual - 4Q change, Administrative Receiver, E&W
Estimate vs actual - Administrative Receiver, E&W
2002Q1
Administrative receiverships The number of administrative receiverships has been very low since 2010 – and in recent quarters there have been a maximum of five cases. Because of this, no adjustment is made to account for late registration of administrative receiverships.
Scotland
Actual
Estimate vs actual - 4Q change, Total, Scot
Estimate
Estimate
350
40%
300
30%
250
20%
200
10%
150
0%
100
-10%
50
-20%
0
-30%
-50
Actual
-40% 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Total, Scot Difference
2016Q3
2017Q1
Comparison of methods (4Q % change) - Total, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3) 500
150%
400
100%
300
2016Q1
Start date (option 3)
50%
200
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
-100% 2002Q1
-100 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
0
-50% 2001Q1
0%
100
2004Q1
Insolvencies (NSA)
Revision
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - Total, Scot
2003Q1
Total company insolvencies The difference between the first estimate of the total number of company insolvencies in Scotland, and the actual number, tended to be fairly small in absolute terms.
Revision
Actual
Estimate vs actual - 4Q change, CVL, Scot
Estimate
Estimate
Actual
60% 50% 40% 30% 20% 10% 0% -10% -20% -30%
80
60 40 20
0 -20 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - CVL, Scot Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVL, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
400%
100 80 60 40 20 0 -20 -40
300% 200%
100% 0% 2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2002Q1
2001Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
-100% 2004Q1
Insolvencies (NSA)
100
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - CVL, Scot
2003Q1
Creditors’ voluntary liquidations (CVLS) The difference between the first estimate of the number of CVLs in Scotland, and the actual number, tended to be fairly small in absolute terms.
Actual
Estimate vs actual - 4Q change, Compulsory Liquidation, Scot
Estimate
Estimate
200
60%
150
40%
Actual
20%
100
0% 50
-20%
0
-40%
-50
-60% 2017Q3
2015Q3
Comparison of methods - Compulsory Liquidation, Scot Reg date (option 2)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2012Q1
0
-100
2011Q1
100
250% 200% 150% 100% 50% 0% -50% -100%
2010Q1
200
2017Q1
Start date (option 3)
2009Q1
300
Reg date (option 2)
2008Q1
Start date (option 3)
400
2016Q3
Comparison of methods (4Q % change) Compulsory liquidation, Scot
2001Q1
Difference
2016Q1
2007Q1
2017Q1
2006Q1
2016Q3
2005Q1
2016Q1
2004Q1
2015Q3
2003Q1
Insolvencies (NSA)
Revision
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - Compulsory Liquidation, Scot
2002Q1
Compulsory liquidations The difference between the first estimate of the number of compulsory liquidations in Scotland, and the actual number, tended to be fairly small in absolute terms.
Revision
Actual
Estimate vs actual - 4Q change, Administration, Scot
Estimate
Estimate
60 100%
50
80%
40
60%
30
40%
20%
20
0%
10
-20%
0
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
-40% 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Administration, Scot Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) Administration, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
1200% 1000% 800% 600% 400% 200% 0% -200%
100 80 60 40 20 0 -20 -40 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Actual
2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
The difference between the first estimate of the number of administrations in Scotland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administration, Scot
Insolvencies (NSA)
Administrations The number of administrations in Scotland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrations, and all revisions are upwards.
Actual
Estimate vs actual - 4Q change, CVA, NI
Estimate
Estimate
7
250%
6
200%
5
150%
Actual
100%
4
50%
3
0%
2
-50%
1
-100%
0
-150% 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - CVA, Scot Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVA, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2007Q1
0
-5
2006Q1
5
2005Q1
10
2002Q1
15
2001Q1
600% 500% 400% 300% 200% 100% 0% -100% -200%
20
2004Q1
Insolvencies (NSA)
Revision
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - CVA, Scot
2003Q1
Company voluntary arrangements (CVAs) The number of CVAs in Scotland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of CVAs, and there have been no revisions to date.
Revision
Actual
Estimate vs actual - 4Q change, Administrative Receiver, E&W
Estimate
Estimate
6
100%
5
50%
4
0%
-50%
2
-100%
1
0
-150% 2017Q3
2015Q3
Comparison of methods - Administrative Receiver, Scot Reg date (option 2)
Reg date (option 2)
2017Q1
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
600% 500% 400% 300% 200% 100% 0% -100% -200%
2009Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Start date (option 3)
50 40 30 20 10 0 -10 -20
2016Q3
Comparison of methods (4Q % change) Administrative receiver, Scot
2001Q1
Difference
2016Q1
2008Q1
2017Q1
2007Q1
2016Q3
2006Q1
2016Q1
2005Q1
2015Q3
2004Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Actual
3
2003Q1
Insolvencies (NSA)
The difference between the first estimate of the number of administrations in Scotland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administrative Receiver, Scot
2002Q1
Administrative receiverships The number of administrative receiverships in Scotland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
Northern Ireland
Revision
Actual
Estimate vs actual - 4Q change, Total, NI
Estimate
Estimate
140
Actual
60%
120
40%
100 20%
80 60
0%
40
-20%
20 -40%
0 -20
-60% 2017Q3
2015Q3
Comparison of methods - Total, NI Reg date (option 2)
Reg date (option 2)
200
60%
150
40%
Start date (option 3)
20%
0%
2017Q2
2016Q4
-60% 2016Q2
-40%
2015Q4
0 -50
2015Q2
-20%
2011Q4
50
2017Q1
Comparison of methods (4Q % change) - Total, NI
Start date (option 3)
100
2016Q3
2014Q4
Difference
2016Q1
2014Q2
2017Q1
2013Q4
2016Q3
2013Q2
2016Q1
2012Q4
2015Q3
2012Q2
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - Total, NI
Insolvencies (NSA)
Total company insolvencies The difference between the first estimate of the total number of company insolvencies in Northern Ireland, and the actual number, tended to be fairly small in absolute terms. Because most statistics are not adjusted to take account of late registrations, most revisions tend to be upwards.
Revision
Actual
Estimate vs actual - 4Q change, CVL, NI
Estimate
Estimate
35
40%
25 20%
20 15
0%
10
-20%
5 -40%
0 -5
-60% 2017Q3
2015Q3
Comparison of methods - CVL, NI Reg date (option 2)
Reg date (option 2)
2017Q2
2011Q4
-10
2016Q4
0
2016Q2
10
80% 60% 40% 20% 0% -20% -40% -60%
2015Q4
20
Start date (option 3)
2015Q2
30
2017Q1
Comparison of methods (4Q % change) - CVL, NI
Start date (option 3) 40
2016Q3
2014Q4
Difference
2016Q1
2014Q2
2017Q1
2013Q4
2016Q3
2013Q2
2016Q1
2012Q4
2015Q3
Insolvencies (NSA)
Actual
60%
30
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - CVL, NI
2012Q2
Creditors’ voluntary liquidations (CVLS) The difference between the first estimate of the number of CVLs in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Compulsory liquidations Statistics on compulsory liquidations in Northern Ireland are obtained from administrative records of the Insolvency Service in Northern Ireland, and the method for producing quarterly statistics is not changing as part of this proposal.
Estimate
Estimate
14
2015Q3
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Administration, NI Reg date (option 2)
Reg date (option 2)
Start date (option 3)
30 20 10 0 -10
2017Q1
Start date (option 3)
300% 250% 200% 150% 100% 50% 0% -50% -100% -150%
2012Q4
Insolvencies (NSA)
40
2016Q3
Comparison of methods (4Q % change) Administration, NI
2012Q2
Difference
2016Q1
2017Q2
0
2016Q4
2
2016Q2
4
2015Q4
6
2014Q2
8
2013Q4
10
250% 200% 150% 100% 50% 0% -50% -100% -150%
2013Q2
12
Actual
2015Q2
Actual
2014Q4
Revision
Estimate vs actual - 4Q change, Administration, NI
2011Q4
The difference between the first estimate of the number of administrations in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administration, NI
Insolvencies (NSA)
Administrations The number of administrations in Northern Ireland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
Revision
Actual
Estimate vs actual - 4Q change, CVA, NI
Estimate
Estimate
14
250%
12
200%
10
150%
50%
6
0%
4
-50%
2
-100%
0
-150% 2017Q3
2015Q3
Comparison of methods - CVA, NI Reg date (option 2)
Reg date (option 2)
Start date (option 3)
2017Q2
2016Q4
2016Q2
2015Q4
300% 250% 200% 150% 100% 50% 0% -50% -100%
2015Q2
Insolvencies (NSA)
2017Q1
Comparison of methods (4Q % change) - CVA, NI
Start date (option 3) 25 20 15 10 5 0 -5 -10
2016Q3
2014Q4
Difference
2016Q1
2014Q2
2017Q1
2013Q4
2016Q3
2013Q2
2016Q1
2012Q4
2015Q3
2012Q2
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Actual
100%
8
2011Q4
The difference between the first estimate of the number of CVAs in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - CVA, NI
Insolvencies (NSA)
Company voluntary arrangements (CVAs) The number of CVAs in Northern Ireland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
The difference between the first estimate of the number of administrative receiverships in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administrative Receiver, NI Revision
Actual
Estimate vs actual - 4Q change, Administrative Receiver, NI
Estimate
Estimate
4
Insolvencies (NSA)
Administrative receiverships The number of administrative receiverships in Northern Ireland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
250% 200% 150% 100% 50% 0% -50% -100% -150%
3 2 1
0
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Administrative Receiver, NI Difference
Insolvencies (NSA)
Start date (option 3)
8 6 4 2 0 -2 -4
Actual
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) Administrative receiver, NI
Reg date (option 2)
Reg date (option 2) 250% 200% 150% 100% 50% 0% -50% -100% -150%
Start date (option 3)
© Crown copyright 2018 You may re-use this document/publication (not including logos) free of charge in any format or medium, under the terms of the Open Government Licence v3.0. To view this licence visit http://www.nationalarchives.gov.uk/doc/open-government-licence; or write to the Information Policy Team, The National Archives, Kew, Richmond, Surrey, TW9 3DU; or email: [email protected]. This document/publication is also available on our website at https://www.gov.uk/government/collections/insolvency-service-official-statistics. Any enquiries regarding this document/publication should be sent to us at [email protected].
Contents Summary ............................................................................................................................................. 2 Background ......................................................................................................................................... 3 Reasons for change ............................................................................................................................. 3 Details of the change .......................................................................................................................... 5 Impact of the change .......................................................................................................................... 5 Communicating the change ................................................................................................................ 6 Detailed impact of the change ............................................................................................................ 7 England & Wales ............................................................................................................................. 7 Scotland......................................................................................................................................... 12 Northern Ireland ........................................................................................................................... 18
Summary In accordance with our revisions policy, we are pre-announcing a change to the methodology used to calculate the number of company insolvencies in a given quarter. This document sets out the reasons for making the change, and a summary of the impact it will have on the statistics. Currently, company insolvencies are counted from the date the insolvency started, with statistics for the most recent quarter being an estimate based on incomplete data. Recent instances of bulk insolvencies of connected companies have had an impact on the quality of these initial estimates. To address this, from April 2018, most company insolvencies will be counted from the date of registration. This means that statistics will no longer be estimated and then revised between quarters, and we will be able more readily to identify and account for bulk insolvencies. This change does not affect statistics on compulsory liquidations in England & Wales and in Northern Ireland, for which data are obtained from administrative records from the Insolvency Service in each jurisdiction. All statistics in this document are not seasonally adjusted (NSA). If you have comments or feedback regarding this change, please contact the Insolvency Service’s statistics team via email at [email protected]. If we do not hear compelling feedback to the contrary, we intend to implement this change beginning with the 2018Q1 edition of the Insolvency Statistics (to be published in April 2018).
Background In 2015, having found weaknesses in the method used to calculate company insolvencies, the Insolvency Service ran a public consultation proposing changes to the method. Three options were presented: 1. No change. 2. Count insolvencies in the quarter they are first registered at Companies House (regardless of when they actually started). 3. Count insolvencies in the quarter they started (regardless of when they were registered). The main advantage of option 2 was that it would provide an accurate record of all insolvencies. The main disadvantage was that, since insolvencies were frequently registered late, the statistics for a given quarter were likely to be an over- or underestimate of the “true” position. The main advantage of option 3 was that statistics for a given quarter would accurately reflect the number of insolvencies. The disadvantage was that, since insolvencies were frequently registered late, statistics for the most recent quarter would be estimated, and revised in subsequent quarters. After considering responses to the consultation, the Insolvency Service selected option 3, and implemented the change in October 2015. Details of the method used were published. Using historic data, the Insolvency Service calculated the accuracy of first estimates of company insolvency, if this method had been in place since 2008. The results were also published. In each quarterly edition of the Insolvency Statistics, the latest quarter’s statistics and the change since previous periods were referred to as “estimated”.
Reasons for change The difference between first estimates of company insolvencies, and subsequent revisions, has been monitored by the Insolvency Service’s statistics team on an ongoing basis. By and large, the differences have been fairly small in absolute terms, with the exception of creditors’ voluntary liquidations (CVLs) in England and Wales (E&W) in 2017Q2.
Estimate vs actual - CVL, E&W Revision
Actual
Estimate
Insolvencies (NSA)
5,000 4,000
3,000 2,000 1,000
0 -1,000 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
This period coincided with an unusually high numbers of CVLs of connected companies, which themselves resulted from a closure of a tax loophole which made the companies unviable. These CVLs were also registered on the Companies House website in the following quarter. A similar set of “bulk insolvencies” was also recorded in 2016Q4, but as these were registered in the same quarter these could be accounted for. The statistics team was made aware of these bulk insolvencies by operational colleagues in the Insolvency Service, and made proactive adjustments to the statistics to account for their expected effects. However, the difference between the estimate and the revised figure for CVLs was larger in nominal terms compared with periods without bulk insolvencies. Moreover, these revisions also affected the percentage difference between a given period and the same quarter the previous year: in 2017Q2, the estimated percentage change compared with 2016Q2 was +36%, which was subsequently revised to +44%.
Estimate vs actual - 4Q change, CVL, E&W Estimate
Actual
80% 70% 60% 50% 40% 30% 20% 10% 0% -10%
2015Q3
2016Q1
2016Q3
2017Q1
The statistics team is aware of a further tranche of bulk insolvencies, which have not all been registered on the same date, which in turn reduces the ability to estimate the effect on the statistics. Further bulk insolvencies cannot be ruled out in the future. The result of this could be that revisions to the data may change the narrative – for example we may initially estimate that numbers of insolvencies have decreased, but revised data may instead show an increase – which would undermine the usefulness of the statistics as a measure of recent trends.
Details of the change Beginning in the 2018Q1 edition of the statistics (to be published in April 2018), we are planning to change to produce statistics using option 2 outlined above. We will also provide statistics on a consistent basis back to 2000Q1. This will give the following benefits to users:
The statistics will not be revised between editions (except in the case that processing errors are identified). We will be able more readily to identify and account for bulk insolvencies. We will be able to publish breakdowns at industry level covering the latest quarter (the current method only allows us to publish these statistics three months in arrears).
Impact of the change The chart below shows the volumes of company insolvencies each quarter since 2000Q1, using both the option 2 and option 3 methods. The time series for the two methods is highly aligned, the largest differences occurring when bulk insolvencies are registered in the quarter after they started. In most cases, the difference in one quarter is cancelled out the following quarter. Similarly, the percentage change in the statistics compared with the same period a year earlier is closely aligned between the two methods, except where bulk insolvencies are registered late. It follows that statistics based on date of registration are a good proxy for the “true” number of insolvencies in a given period.
Comparison of methods - Total, E&W Difference
Reg date (option 2)
Start date (option 3)
6,000
4,000 2,000 0
-2,000 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
8,000
Comparison of methods (4Q % change) - Total, E&W Reg date (option 2)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2004Q1
2003Q1
2002Q1
2001Q1
100% 80% 60% 40% 20% 0% -20% -40%
Communicating the change We will:
Pre-announce the change in the 2017Q4 edition of the Insolvency Statistics. Describe the reasons for the change and the impact it will have (this document). Notify known users in advance of the change. Publicise the change using other channels e.g. the Insolvency Service’s Twitter account. Draw attention to the change in the 2018Q1 edition of the Insolvency Statistics.
If you have comments or feedback regarding this change, please contact the Insolvency Service’s statistics team via email at [email protected].
Detailed impact of the change England & Wales
Revision
Estimate
Estimate
5,000 4,000
3,000 2,000
1,000 0
-1,000 2016Q1
2016Q3
2017Q1
2015Q3
2017Q3
Comparison of methods - Total, E&W Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - Total, E&W
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
100% 80% 60% 40% 20% 0% -20% -40%
8,000 6,000
4,000 2,000 0
-2,000 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Actual
70% 60% 50% 40% 30% 20% 10% 0% -10% -20%
6,000
2015Q3
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned except where there have been bulk insolvency events that have been registered in the following quarter.
Actual
2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
There was one occasion where the first estimate was that insolvencies had increased since the same period the previous year, but when the figures were revised this was changed to a decrease.
Estimate vs actual - 4Q change, Total, E&W
Estimate vs actual - Total, E&W
Insolvencies (NSA)
Total company insolvencies The difference between the first estimate of the total number of company insolvencies in England and Wales, and the actual number, tended to be fairly small in absolute terms with the exception of in 2017Q2 when a bulk insolvency event was registered late with Companies House.
Revision
Actual
Estimate
Estimate
4,000
3,000 2,000 1,000
0 -1,000 2015Q3
2016Q1
2016Q3
2017Q1
2015Q3
2017Q3
Comparison of methods - CVL, E&W Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVL, E&W
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
Compulsory liquidations Statistics on compulsory liquidations in England and Wales are obtained from administrative records of the Insolvency Service, and the method for producing quarterly statistics is not changing as part of this proposal.
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2004Q1
2002Q1
2001Q1
100% 80% 60% 40% 20% 0% -20% -40%
5,000 4,000 3,000 2,000 1,000 0 -1,000 -2,000 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Actual
80% 70% 60% 50% 40% 30% 20% 10% 0% -10%
5,000
2003Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned except where there have been bulk insolvency events that have been registered in the following quarter.
Estimate vs actual - 4Q change, CVL, E&W
Estimate vs actual - CVL, E&W
Insolvencies (NSA)
Creditors’ voluntary liquidations (CVLS) The difference between the first estimate of the number of CVLs in England and Wales, and the actual number, tended to be fairly small in absolute terms with the exception of in 2017Q2 when a bulk insolvency event was registered late with Companies House.
Revision
Estimate
Estimate
400 350 300 250 200 150 100 50 0 -50
10%
5% 0% -5% -10%
-15% 2015Q3
2017Q3
Comparison of methods - Administration, E&W Reg date (option 2)
Reg date (option 2)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2,500 2,000 1,500 1,000 500 0 -500 -1,000
2017Q1
300% 250% 200% 150% 100% 50% 0% -50% -100%
2009Q1
Start date (option 3)
2016Q3
Comparison of methods (4Q % change) Administration, E&W
2001Q1
Difference
2016Q1
2008Q1
2017Q1
2007Q1
2016Q3
2006Q1
2016Q1
2005Q1
2015Q3
Insolvencies (NSA)
Actual
15%
2004Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned except where there have been bulk insolvency events that have been registered in the following quarter.
Actual
2003Q1
Insolvencies (NSA)
There was one occasion where the first estimate was that insolvencies had decreased since the same period the previous year, but when the figures were revised this was changed to an increase.
Estimate vs actual - 4Q change, Administration, E&W
Estimate vs actual - Administration, E&W
2002Q1
Administrations The difference between the first estimate of the number of administrations in England and Wales, and the actual number, tended to be fairly small in absolute terms.
Revision
Actual
Estimate vs actual - 4Q change, CVA, E&W
Estimate
Estimate
140
100 80 60 40 20
0 -20 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - CVA, E&W Difference
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVA, E&W
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2007Q1
0
-100
2006Q1
100
2005Q1
200
2002Q1
300
2001Q1
100% 80% 60% 40% 20% 0% -20% -40% -60% -80%
400 Insolvencies (NSA)
2016Q1
2004Q1
The time series since 2000 are closely aligned for both option 2 and option 3.
Actual
50% 40% 30% 20% 10% 0% -10% -20% -30% -40% -50%
120
Insolvencies (NSA)
The difference between the first estimate of the number of CVAs in England and Wales, and the actual number, tended to be fairly small in absolute terms – usually fewer than ten companies.
Estimate vs actual - CVA, E&W
2003Q1
Company voluntary arrangements (CVAs) Volumes of CVAs each quarter have been fairly small – usually fewer than 100 companies each quarter.
Revision
Actual
Estimate
Estimate
6
500%
5
400% 200%
3
100%
2
0%
1
-100%
-200% 2015Q3
2017Q3
Comparison of methods - Administrative Receiver, E&W Reg date (option 2)
2017Q1
2016Q1
2015Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2014Q1
0
-100
2013Q1
100
2012Q1
200
500% 400% 300% 200% 100% 0% -100% -200%
2011Q1
300
2017Q1
Start date (option 3)
2010Q1
400
Reg date (option 2)
2009Q1
Start date (option 3)
2016Q3
Comparison of methods (4Q % change) Administrative receiver, E&W
2001Q1
Difference
2016Q1
2008Q1
2017Q1
2007Q1
2016Q3
2006Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
2016Q1
2005Q1
2015Q3
2004Q1
0
Insolvencies (NSA)
Actual
300%
4
2003Q1
Insolvencies (NSA)
There tended to be no difference between the first estimate of the number of administrative receiverships in England and Wales, and the actual number – the exception being in 2017Q1 when two cases were registered in the following quarter.
Estimate vs actual - 4Q change, Administrative Receiver, E&W
Estimate vs actual - Administrative Receiver, E&W
2002Q1
Administrative receiverships The number of administrative receiverships has been very low since 2010 – and in recent quarters there have been a maximum of five cases. Because of this, no adjustment is made to account for late registration of administrative receiverships.
Scotland
Actual
Estimate vs actual - 4Q change, Total, Scot
Estimate
Estimate
350
40%
300
30%
250
20%
200
10%
150
0%
100
-10%
50
-20%
0
-30%
-50
Actual
-40% 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Total, Scot Difference
2016Q3
2017Q1
Comparison of methods (4Q % change) - Total, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3) 500
150%
400
100%
300
2016Q1
Start date (option 3)
50%
200
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
-100% 2002Q1
-100 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
0
-50% 2001Q1
0%
100
2004Q1
Insolvencies (NSA)
Revision
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - Total, Scot
2003Q1
Total company insolvencies The difference between the first estimate of the total number of company insolvencies in Scotland, and the actual number, tended to be fairly small in absolute terms.
Revision
Actual
Estimate vs actual - 4Q change, CVL, Scot
Estimate
Estimate
Actual
60% 50% 40% 30% 20% 10% 0% -10% -20% -30%
80
60 40 20
0 -20 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - CVL, Scot Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVL, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
400%
100 80 60 40 20 0 -20 -40
300% 200%
100% 0% 2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2007Q1
2006Q1
2005Q1
2002Q1
2001Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
-100% 2004Q1
Insolvencies (NSA)
100
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - CVL, Scot
2003Q1
Creditors’ voluntary liquidations (CVLS) The difference between the first estimate of the number of CVLs in Scotland, and the actual number, tended to be fairly small in absolute terms.
Actual
Estimate vs actual - 4Q change, Compulsory Liquidation, Scot
Estimate
Estimate
200
60%
150
40%
Actual
20%
100
0% 50
-20%
0
-40%
-50
-60% 2017Q3
2015Q3
Comparison of methods - Compulsory Liquidation, Scot Reg date (option 2)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2012Q1
0
-100
2011Q1
100
250% 200% 150% 100% 50% 0% -50% -100%
2010Q1
200
2017Q1
Start date (option 3)
2009Q1
300
Reg date (option 2)
2008Q1
Start date (option 3)
400
2016Q3
Comparison of methods (4Q % change) Compulsory liquidation, Scot
2001Q1
Difference
2016Q1
2007Q1
2017Q1
2006Q1
2016Q3
2005Q1
2016Q1
2004Q1
2015Q3
2003Q1
Insolvencies (NSA)
Revision
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - Compulsory Liquidation, Scot
2002Q1
Compulsory liquidations The difference between the first estimate of the number of compulsory liquidations in Scotland, and the actual number, tended to be fairly small in absolute terms.
Revision
Actual
Estimate vs actual - 4Q change, Administration, Scot
Estimate
Estimate
60 100%
50
80%
40
60%
30
40%
20%
20
0%
10
-20%
0
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
-40% 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Administration, Scot Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) Administration, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
1200% 1000% 800% 600% 400% 200% 0% -200%
100 80 60 40 20 0 -20 -40 2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Actual
2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
The difference between the first estimate of the number of administrations in Scotland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administration, Scot
Insolvencies (NSA)
Administrations The number of administrations in Scotland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrations, and all revisions are upwards.
Actual
Estimate vs actual - 4Q change, CVA, NI
Estimate
Estimate
7
250%
6
200%
5
150%
Actual
100%
4
50%
3
0%
2
-50%
1
-100%
0
-150% 2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - CVA, Scot Difference
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) - CVA, Scot
Reg date (option 2)
Reg date (option 2)
Start date (option 3)
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
2009Q1
2008Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
2007Q1
0
-5
2006Q1
5
2005Q1
10
2002Q1
15
2001Q1
600% 500% 400% 300% 200% 100% 0% -100% -200%
20
2004Q1
Insolvencies (NSA)
Revision
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - CVA, Scot
2003Q1
Company voluntary arrangements (CVAs) The number of CVAs in Scotland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of CVAs, and there have been no revisions to date.
Revision
Actual
Estimate vs actual - 4Q change, Administrative Receiver, E&W
Estimate
Estimate
6
100%
5
50%
4
0%
-50%
2
-100%
1
0
-150% 2017Q3
2015Q3
Comparison of methods - Administrative Receiver, Scot Reg date (option 2)
Reg date (option 2)
2017Q1
Start date (option 3)
2017Q1
2016Q1
2015Q1
2014Q1
2013Q1
2012Q1
2011Q1
2010Q1
600% 500% 400% 300% 200% 100% 0% -100% -200%
2009Q1
2000Q1 2001Q1 2002Q1 2003Q1 2004Q1 2005Q1 2006Q1 2007Q1 2008Q1 2009Q1 2010Q1 2011Q1 2012Q1 2013Q1 2014Q1 2015Q1 2016Q1 2017Q1
Insolvencies (NSA)
Start date (option 3)
50 40 30 20 10 0 -10 -20
2016Q3
Comparison of methods (4Q % change) Administrative receiver, Scot
2001Q1
Difference
2016Q1
2008Q1
2017Q1
2007Q1
2016Q3
2006Q1
2016Q1
2005Q1
2015Q3
2004Q1
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Actual
3
2003Q1
Insolvencies (NSA)
The difference between the first estimate of the number of administrations in Scotland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administrative Receiver, Scot
2002Q1
Administrative receiverships The number of administrative receiverships in Scotland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
Northern Ireland
Revision
Actual
Estimate vs actual - 4Q change, Total, NI
Estimate
Estimate
140
Actual
60%
120
40%
100 20%
80 60
0%
40
-20%
20 -40%
0 -20
-60% 2017Q3
2015Q3
Comparison of methods - Total, NI Reg date (option 2)
Reg date (option 2)
200
60%
150
40%
Start date (option 3)
20%
0%
2017Q2
2016Q4
-60% 2016Q2
-40%
2015Q4
0 -50
2015Q2
-20%
2011Q4
50
2017Q1
Comparison of methods (4Q % change) - Total, NI
Start date (option 3)
100
2016Q3
2014Q4
Difference
2016Q1
2014Q2
2017Q1
2013Q4
2016Q3
2013Q2
2016Q1
2012Q4
2015Q3
2012Q2
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - Total, NI
Insolvencies (NSA)
Total company insolvencies The difference between the first estimate of the total number of company insolvencies in Northern Ireland, and the actual number, tended to be fairly small in absolute terms. Because most statistics are not adjusted to take account of late registrations, most revisions tend to be upwards.
Revision
Actual
Estimate vs actual - 4Q change, CVL, NI
Estimate
Estimate
35
40%
25 20%
20 15
0%
10
-20%
5 -40%
0 -5
-60% 2017Q3
2015Q3
Comparison of methods - CVL, NI Reg date (option 2)
Reg date (option 2)
2017Q2
2011Q4
-10
2016Q4
0
2016Q2
10
80% 60% 40% 20% 0% -20% -40% -60%
2015Q4
20
Start date (option 3)
2015Q2
30
2017Q1
Comparison of methods (4Q % change) - CVL, NI
Start date (option 3) 40
2016Q3
2014Q4
Difference
2016Q1
2014Q2
2017Q1
2013Q4
2016Q3
2013Q2
2016Q1
2012Q4
2015Q3
Insolvencies (NSA)
Actual
60%
30
Insolvencies (NSA)
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Estimate vs actual - CVL, NI
2012Q2
Creditors’ voluntary liquidations (CVLS) The difference between the first estimate of the number of CVLs in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Compulsory liquidations Statistics on compulsory liquidations in Northern Ireland are obtained from administrative records of the Insolvency Service in Northern Ireland, and the method for producing quarterly statistics is not changing as part of this proposal.
Estimate
Estimate
14
2015Q3
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Administration, NI Reg date (option 2)
Reg date (option 2)
Start date (option 3)
30 20 10 0 -10
2017Q1
Start date (option 3)
300% 250% 200% 150% 100% 50% 0% -50% -100% -150%
2012Q4
Insolvencies (NSA)
40
2016Q3
Comparison of methods (4Q % change) Administration, NI
2012Q2
Difference
2016Q1
2017Q2
0
2016Q4
2
2016Q2
4
2015Q4
6
2014Q2
8
2013Q4
10
250% 200% 150% 100% 50% 0% -50% -100% -150%
2013Q2
12
Actual
2015Q2
Actual
2014Q4
Revision
Estimate vs actual - 4Q change, Administration, NI
2011Q4
The difference between the first estimate of the number of administrations in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administration, NI
Insolvencies (NSA)
Administrations The number of administrations in Northern Ireland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
Revision
Actual
Estimate vs actual - 4Q change, CVA, NI
Estimate
Estimate
14
250%
12
200%
10
150%
50%
6
0%
4
-50%
2
-100%
0
-150% 2017Q3
2015Q3
Comparison of methods - CVA, NI Reg date (option 2)
Reg date (option 2)
Start date (option 3)
2017Q2
2016Q4
2016Q2
2015Q4
300% 250% 200% 150% 100% 50% 0% -50% -100%
2015Q2
Insolvencies (NSA)
2017Q1
Comparison of methods (4Q % change) - CVA, NI
Start date (option 3) 25 20 15 10 5 0 -5 -10
2016Q3
2014Q4
Difference
2016Q1
2014Q2
2017Q1
2013Q4
2016Q3
2013Q2
2016Q1
2012Q4
2015Q3
2012Q2
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
Actual
100%
8
2011Q4
The difference between the first estimate of the number of CVAs in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - CVA, NI
Insolvencies (NSA)
Company voluntary arrangements (CVAs) The number of CVAs in Northern Ireland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
The difference between the first estimate of the number of administrative receiverships in Northern Ireland, and the actual number, tended to be fairly small in absolute terms.
Estimate vs actual - Administrative Receiver, NI Revision
Actual
Estimate vs actual - 4Q change, Administrative Receiver, NI
Estimate
Estimate
4
Insolvencies (NSA)
Administrative receiverships The number of administrative receiverships in Northern Ireland each quarter has been very low since 2010. Because of this, no adjustment is made to account for late registration of administrative receiverships, and all revisions are upwards.
250% 200% 150% 100% 50% 0% -50% -100% -150%
3 2 1
0
Comparing option 3 (the current method) with option 2 (the proposed change), the time series are very closely aligned.
2015Q3
2016Q1
2016Q3
2017Q1
2017Q3
2015Q3
Comparison of methods - Administrative Receiver, NI Difference
Insolvencies (NSA)
Start date (option 3)
8 6 4 2 0 -2 -4
Actual
2016Q1
2016Q3
2017Q1
Comparison of methods (4Q % change) Administrative receiver, NI
Reg date (option 2)
Reg date (option 2) 250% 200% 150% 100% 50% 0% -50% -100% -150%
Start date (option 3)
© Crown copyright 2018 You may re-use this document/publication (not including logos) free of charge in any format or medium, under the terms of the Open Government Licence v3.0. To view this licence visit http://www.nationalarchives.gov.uk/doc/open-government-licence; or write to the Information Policy Team, The National Archives, Kew, Richmond, Surrey, TW9 3DU; or email: [email protected]. This document/publication is also available on our website at https://www.gov.uk/government/collections/insolvency-service-official-statistics. Any enquiries regarding this document/publication should be sent to us at [email protected].
