MQ5: investment by insurance companies, pensions funds and trusts QMI

• surveys of insurance companies, pension funds and trusts have been in existence in one form or another since the 1960s
• data from the surveys are mainly used in the Financial and Sector Accounts and for creating Gross Domestic Product (GDP) estimates
• all data are reported on a current price basis

MQ5 reports quarterly estimates of net investment from financial transactions (investments) made by insurance companies, self-administered pension funds, investment trusts, unit trusts and property unit trusts.

Also included are quarterly balance sheet data for short-term assets and liabilities, and quarterly income and expenditure data for insurance companies and self-administered pension funds.

Annual balance sheet data for all the institutional groups is contained in the 3rd quarter release, giving information on the market value of assets and liabilities. Annual income and expenditure data for insurance companies are also reported at this time.

The surveys have been devised mainly to meet the needs of the Financial and Sector Accounts and the compilation of Gross Domestic Product (GDP) estimates within the UK National Accounts and UK Balance of Payments.  

Investment by insurance companies, pension funds and trusts (MQ5) is published by Office for National Statistics (ONS) as a quarterly statistical bulletin. The quarterly and annual sample surveys used in the compilation of the publication have been in existence in one form or another since the 1960s, evolving over time.

MQ5 reports quarterly estimates of net investment arising from financial transactions (investments) made by insurance companies, self-administered pension funds, investment trusts, unit trusts and property unit trusts. Also included are quarterly balance sheet data for short-term assets and liabilities, along with quarterly income and expenditure data for insurance companies and self-administered pension funds.

All data are reported on a current price basis. In addition, every third quarter release contains annual balance sheet data for all the institutional groups; providing information on the market value of assets and liabilities. Annual income and expenditure data for insurance companies are also reported at this time.

Information relates to insurance companies, pension funds and trusts who are authorised and/or regulated in the UK (see “How the output is created”).

This report contains the following sections:

• Output quality
• About the output
• How the output is created
• Validation and quality assurance
• Concepts and definitions
• Other information, relating to quality trade-offs and user needs
• Sources for further information or advice

Output quality

This report provides a range of information that describes the quality of the output and details any points that should be noted when using the output.

We have developed Guidelines for Measuring Statistical Quality; these are based upon the five European Statistical System (ESS) Quality Dimensions. This report addresses these quality dimensions and other important quality characteristics, which are:

• relevance
• timeliness and punctuality
• coherence and comparability
• accuracy
• output quality trade-offs
• assessment of user needs and perceptions
• accessibility and clarity

More information is provided about these quality dimensions in the following sections.

Relevance

(The degree to which the statistical outputs meet users’ needs.)

The primary use of data from the surveys is in the financial and sector accounts and the compilation of gross domestic product (GDP) estimates within the UK National Accounts and UK Balance of Payments. The surveys have been devised mainly to meet these needs and they also provide data for the MQ5 statistical bulletin. National accounts users were extensively involved in the re-design of the questionnaires in 2008 and the data aggregations that are produced to meet their needs. Future changes to the national accounts, as a result of the changes to the reporting standards such as the European System of Accounts: ESA 2010, will result in amendments to the MQ5 suite of questionnaires.

We actively seek feedback from external users. Each publication of our data brings queries from users of our statistics. We welcome these as an indication of whether we are providing the data required and to understand better how it is used. Those uses range from finding basic structural data, such as the value of assets held by pension funds, to forming an input for modelling by businesses and consultancies trying to produce forecasts. Such users include:

• Bank of England – data are used for monetary policy and financial stability monitoring

• Department for Work and Pensions – specifically interested in the investment activity of pension funds and any pension business undertaken by insurance companies

• HM Revenue and Customs – data are used to aid taxation analysis of financial institutions

• Association of British Insurers – compares its own data with that of Office for National Statistics (ONS) to ensure both datasets display similar trends

• Debt Management Office – data are used to monitor the investment activity in British government securities (gilts)

• The Investment Association – compares its own data with that of ONS to ensure both datasets display similar trends; it also uses the data to provide an overall view of the UK savings and pensions markets and the components that make it up

• European Union’s Statistical Office (Eurostat) – uses data to compile statistics at a European level to enable comparisons between countries and to support the development of European fiscal policy

• Organisation for Economic Co-operation and Development (OECD) – analyses investment activity to help formulate economic growth and financial stability recommendations for member countries

• Trade associations, city analysts, institutional investors and fund managers use these data for modelling or forecasting purposes and also to track asset allocation trends; academics and journalists also use the data for research purposes

In March 2015, we announced the likelihood of wide-ranging changes to the surveys used to collect the data presented in MQ5. This would be necessary in order to become compliant with the new European System of Accounts 2010: ESA 2010. At the time we indicated that changes would be rolled out over a 3-year period (that is, through to 2018). However, due to a number of factors, changes to the MQ5 surveys (which would impact upon this statistical bulletin) have been deferred until further notice. Users will be kept informed of developments through the bulletin.

Estimates of net investment for the following asset types are included in the bulletin:

• short-term assets (those maturing within 1 year of their originating date)
• UK government sterling securities (gilts – fixed income or index-linked bonds issued by the UK government)
• UK corporate securities and overseas securities (comprising ordinary shares, corporate bonds and preference shares)
• other assets (for example, mutual fund investments, insurance-managed funds, UK government securities denominated in foreign currency)

Further detailed asset breakdowns are shown in the associated datasets.

Timeliness and punctuality

(Timeliness refers to the lapse of time between publication and the period to which the data refer. Punctuality refers to the gap between planned and actual publication dates.)

The validation of individual respondent data and the process undertaken to produce a set of aggregated results is usually completed approximately 8 weeks after the end of a quarter (3-month period). A further 3 to 4 weeks are set aside to prepare the data for publishing and compile the statistical bulletin. This time lag allows some contingency for taking on late respondent returns and resolving complex queries that might minimise data revisions. The statistical bulletin is released on a quarterly basis. It is published approximately 3 months after the period to which the data refer.

For more details on related releases, the GOV.UK release calendar provides 12 months’ advanced notice of release dates. If there are any changes to the pre-announced release schedule, public attention will be drawn to the change and the reasons for the change will be explained fully at the same time, as set out in the Code of Practice for Official Statistics.

The MQ5 output is compiled from data collected from insurance companies, pension funds and trusts. This section details how the output is created and contains information on the target population, sampling frame, sample size and design for each of the surveys. Information is also provided on the outlier, imputation, weighting, estimation and statistical disclosure processes.

Target population and sampling frame

Insurance companies: The target population is made up of all companies that undertake insurance business in the UK, regulated by the Prudential Regulation Authority (PRA) and authorised by the Financial Conduct Authority (FCA). There are approximately 450 companies in the population.

Pension funds: The target population is obtained from Pension Funds Online (PFO), which produces a list of self- administered pension funds. From this, UK pension funds are selected (excluding those that are fully insured). There are approximately 1,400 pension funds in the population. The PFO list does not cover the entire population. In order to account for this under-coverage of the population, a methodology review concluded that the estimates from the survey (for private sector funds) should be scaled up by a factor of 22%.

Investment trusts: The target population is obtained from the Association of Investment Companies (AIC) and is made up of investment trust fund managers in the UK. There are approximately 73 trusts in the population.

Unit trusts: The target population is obtained from the Investment Association (IA) and is made up of UK unit trust fund managers (authorised by the FCA). There are approximately 102 unit trusts in the population.

Property unit trusts: The target population is obtained from the Association of Real Estate Funds (AREF) and is made up of UK property unit trust fund managers authorised under the terms of the Financial Services and Markets Act 2000. There are approximately 25 trusts in the population.

Sample size (approximately)

Sample design

Insurance companies: A stratified random sample is used, with the value of premium income as the stratification variable.

Insurance companies engaged in long-term business, whose premium income is £400 million or more are always included in the quarterly and annual samples. Similarly, insurance companies engaged in general business are always included in the quarterly and annual samples if their premium income is £600 million or more. Stratified random sampling is used below these thresholds.

It is possible for the two types of insurance companies to have negative premium income. Insurance companies may report negative premiums for a number of reasons but common reasons include the transfer of a pension fund and a reinsurance deal set-up with another insurer. Both of these result in a transfer out of premiums previously taken on. If this amount exceeds the premiums received during business as usual, the company will report a negative premium. Any companies in this category are typically always included in the quarterly and annual samples. The stratification bands (in pounds million) used in the results processing are:

Pension funds: A stratified random sample is used, with fund membership as the stratification variable.

All private and public corporation pension funds with a membership of 20,000 or more are always included in the quarterly and annual samples. Funds with fewer than 20,000 members are sampled on a random basis. All local authority pension funds are included in the quarterly and annual samples. Funds covered by a crown guarantee have their own stratum and are all included in the quarterly and annual samples. The stratification bands (by number of members) used in the results processing are:

Investment trusts: A stratified random sample is used, with market value of funds under management as the stratification variable.

Investment trusts management companies with trusts under management with a market value of £500 million or more are always included in the quarterly and annual samples. Stratified random sampling is used below these thresholds. The stratification bands (in pounds million) used in the results processing are:

Zero up to 99 100 up to 499
500 up to 1,999
2,000 or more

Unit trusts: A stratified random sample is used, with market value of funds under management as the stratification variable.

Investment trusts management companies with trusts under management with a market value of £500 million or more are always included in the quarterly and annual samples. Stratified random sampling is used below these thresholds, but no unit trusts with funds under management of less than £100 million are selected. The stratification bands (in pounds million) used in the results processing are:

Zero up to 99 100 up 499
500 up to 999
1,000 up to 2,999
3,000 or more

Property unit trusts: The property unit trusts sample is not stratified due to the total population being small. All in-scope trusts are sampled quarterly and annually.

Outliers

Businesses supplying atypical or extreme returns are automatically identified and treated as outliers using a one-sided Winsorisation method. As some questions on the survey questionnaires can have both positive and negative returns, separate cut-off values for the Winsorisation process are calculated for each sign. Additional manual treatment of outliers may also occur if appropriate.

Imputation

Period-on-period movements are used to impute returns for non-responding businesses that have returned data for the previous period. If newly-selected businesses fail to return data they are treated as non-selected and taken into account in the estimation procedure. In the case of the annual surveys, imputation takes into account equivalent quarterly data that might be available from individual businesses.

Weighting and estimation

Each sampled business is used to represent a number of similar businesses or funds based on premium income (insurance companies), membership (pension funds) or funds under management (trusts). After imputation for non-responders, ratio estimation is applied to produce estimates for the entire population by individual data variable.

The data are not subject to deflation or seasonal adjustment.

Statistical disclosure

Data disclosure rules mean that not all data collected on the survey questionnaires can be published individually. Where individual data items are identified as being disclosive, higher level aggregates, combining the disclosive data with other data items, are created for publication purposes. Even though unavoidable, non-ONS users looking to analyse data at a more detailed level could consider this to be a limitation of the output.

Accuracy

(The degree of closeness between an estimate and the true value.)

Questionnaires for the quarterly surveys used to compile the statistical bulletin are despatched at the end of the period to which they relate. Questionnaires for the annual insurance companies and pension funds surveys are despatched in the March after the end of the reference year, whereas the annual trusts survey questionnaires are despatched in the January after the end of the reference year.

To try to minimise non-response bias, all quarterly and annual surveys issue one paper reminder letter followed by telephone or email response-chasing. There is also the possibility of using the legal powers of the Statistics of Trade Act 1947 to force response, though we prefer to work together with businesses to produce the necessary information.

Returned respondent data are run through a series of validation tests to try to identify potential errors. The tests are designed to check whether all of the data required have been provided, whether the period of returns meet certain criteria and that the returned data are credible. The tests also compare returned data with previous data for each respondent. Data that fail the tests are identified for further action. Where failures are judged potentially significant to survey results, they will be queried with respondents, with the aim of correcting (editing) or confirming the original data.

The introduction of annual survey results with the third quarter figures each year leads to revisions of the published quarterly estimates, both to income and expenditure, and to transactions data. Revisions to transactions data are usually caused by problems with quarterly misreporting of data by businesses, which are identified as part of the quality assurance of the corresponding annual survey returns made by the businesses.

For income and expenditure, the revisions are due to the incorporation of the annual insurance survey results, which are based on larger samples and also, generally reflect audited accounts. It is important to note that for both the pension funds and trusts sectors an annual income and expenditure survey is not undertaken.

For each institutional group (for example, pension funds) there is a common sample for all their surveys (for example, quarterly transactions and quarterly income and expenditure), but each survey is conducted independently, which can result in different response rates.

In some instances individual survey questionnaires are completed by different people within the business, and with limited linkage within existing systems between the surveys at the individual respondent level. Therefore, there can be discrepancies at an aggregate level between the numbers emerging from the transactions and income and expenditure surveys.

The set of annual surveys includes balance sheet data from the insurance companies and pension funds sectors. This allows the data to be ”aligned” so that transactions, income and expenditure and the balance sheet are consistent. The alignment process assumes that the transactions data are the weakest of the three strands of information and therefore take the necessary revision. This assumption has been confirmed by contact with respondents when data have been queried. It is important to note that no alignment process is currently undertaken for the trusts sector.

The accuracy and quality of the statistical bulletin are checked using a series of extensive consistency checks within the survey aggregation systems.

With revisions being the main quality indicator for these data, a revisions policy is available to assist users with their understanding of the cycle and frequency of data revisions. Users that make use of the data for modelling or forecasting purposes are clearly told to pay careful attention to this policy. Given present practices within the processing area, more extensive revisions can be expected with the publication of the figures for the third quarter each year. Data revisions and survey response rates are published in each edition of the statistical bulletin. There is also a table displaying the average values and revisions, over the last 5 years, arising from the take-on of the annual survey results.

Coherence and comparability

(Coherence is the degree to which data that are derived from different sources or methods, but refer to the same topic, are similar. Comparability is the degree to which data can be compared over time and domain for example, geographic level.)

From the first quarter of 1998, in order to comply with the requirements of the European System of Accounts, all the financial institutions’ survey questionnaires were changed to collect additional asset breakdowns.

From Quarter 1 (January to March) 2004, there were changes to the long-term insurance funds income and expenditure questionnaires. Total pension contributions made to funded schemes cannot be derived by summing pension premiums reported by insurance companies and contributions reported by pension funds. To do so would result in double-counting since pension business premiums from insurance companies include any premiums (including transfers) received from pension funds and any transfers within the long-term insurance sector. The changes from 2004 were:

• new questions on premium income and claims from insurance managed funds business
• separate data on minimum contributions received from the National Insurance Contributions Office and bulk buyouts
• more detailed guidance on what should be reported under personal and occupational pensions business

The revised questions and guidance provide a fuller and more accurate breakdown of income. In addition to the changes to the quarterly questionnaires, new questions were included on the annual survey of long-term insurance funds’ income and expenditure questionnaire. The new questions were:

a breakdown of:

• personal and stakeholder pension premiums by type of pension: personal pensions; stakeholder pensions; group personal pensions and other pensions
• occupational pension premiums by type of pension: defined benefit; defined contribution and hybrid

transfers in of:

• personal and stakeholder pension business included in premiums figures
• occupational pension business included in premiums figures

transfers out of:

• personal and stakeholder pension business included in claims figures
• occupational pension business included in claims figures

Estimates on private pension contributions, having adjusted for the double counting issue are incorporated in the release Pensions in the National Accounts, A Fuller Picture of the UK's Funded and Unfunded Pension Obligations and detailed in the article Improved methods for calculating private pension contributions.

During 2008, a review of the quarterly and annual insurance companies and pension funds surveys was completed. The aim was to be more compatible with European requirements and new international accounting standards, while also taking the opportunity to improve aspects of the survey processing. Following the review, improved quarterly questionnaires were introduced in respect of the first quarter of 2009. The new annual questionnaires came into force in respect of the 2009 survey year.

The main changes from this review were:

• cash liquidity funds being included in “any other liquid deposits”
• where separate information used to be asked for banks and building societies, they are now merged into single questions
• the annual derivatives questions split between UK and overseas counterparties enhanced with a product breakdown (assets and liabilities), and the quarterly product questions replaced with a UK and overseas counterparty split
• change in the maturity breakdown for British government sterling securities
• additional category for other UK mutual funds
• removal of the “of which loans to financial institutions” question
• changes to the guidance notes for the fixed asset questions
• the revised income and expenditure questionnaires take a more logical approach, which is more comparable to accountancy standard reporting

The introduction of the new questionnaires caused some data discontinuities. Data adjustments to remove those discontinuities were made for the years 2000 to 2008 so that the data from 2000 now form a consistent time series. Table 1 displays the percentage effect the adjustments had on the net investment and holdings data for total assets for all the institutional groups.

Other sources

There are various external organisations that collect data (through surveys or for administrative purposes) for the insurance companies, pension and trusts sectors, such as the Association of British Insurers (ABI), the Pensions and Lifetime Savings Association (PLSA), the Investment Association (IA) and the Prudential Regulation Authority (PRA). The information collected by these organisations differs to that collected by Office for National Statistics (ONS). The ONS surveys for these sectors are the only sources of data that collect the level of financial detail required to satisfy the requirements of the national accounts.

Ownership of UK quoted shares is a biennial ONS publication reporting on the size of the UK stock market and the institutions that own the shares traded on it. Conceptually, data should be coherent with that reported in MQ5. However, the underlying methodology of the surveys feeding into both publications differs greatly. A review of the share ownership methodology was published in January 2012. This review draws attention to the differences between the two sets of data and provides some insight into the reasons for any discrepancies.

The Annual Survey of Hours and Earnings (ASHE) collects information about earnings and pensions for a 1% sample of employee jobs from the Pay As You Earn (PAYE) system. The survey identifies which of these jobs has pension provision and, within the overall category of workplace pensions, it separately identifies occupational pensions and personal pensions. Results on pension participation and contribution rate bands are published annually (around February or March), with breakdowns by sex, age, industry sector, earnings and firm size. This differs from MQ5, which focuses on financial investments undertaken by pension funds. As such, there is no direct overlap between the two outputs.

In addition, the Occupational Pension Schemes Survey (OPSS) collects and publishes information on scheme membership, benefit structure and contribution rates. OPSS covers both private and public sector occupational pension schemes registered in the UK. Results from OPSS provide a detailed view of the nature of occupational pension provision in the UK. In contrast, MQ5 reports on financial investment activity by pension funds. The two data collections therefore have minimal overlap.

MQ5 data are also used to help compile the UK National Accounts and associated publications. It should be noted that MQ5 data are purely based on survey returns from insurance companies, pension funds and trusts. However, during the national accounts compilation process, combinations of data sources are used to produce a balanced set of accounts. This not only includes the survey data that underpins the MQ5 release, but also data from administrative sources such as central government and the Bank of England. The process of balancing the data from different sources can sometimes lead to discrepancies in the figures published within the MQ5 and national accounts publications, even though the reporting items may appear to be coherent.

International comparisons

It is difficult to meaningfully compare investment by UK insurance companies, pension funds and trusts with that of other countries. This is largely due to different rules and regulations surrounding insurance and pension provision, and also because other countries don’t combine data for these specific institutional groups into a single detailed publication. The focus for other countries is on collecting the data for national accounts purposes, not on producing a separate publication for these institutional groups.

Many countries around the world use different sources to collect these data. In some cases the data collection is split between the national statistical office and the central bank (Belgium) or the industry regulator (Finland). The periodicity of data collection also varies between countries; some collect data quarterly (Sweden), others on an annual basis (New Zealand). In addition, some countries use a transactions approach to data collection (UK), while others prefer a balance sheet style (Ireland).

International bodies such as the Organisation for Economic Co-operation and Development (OECD) compare institutional investment data across countries to help formulate economic growth and financial stability recommendations

(Concepts and definitions describe the legislation governing the output, and a description of the classifications used in the output.)

Survey data are collected under the statutory powers of the UK Statistics of Trade Act 1947. The surveys are conducted to meet the requirements specified in the European System of Accounts and Eurostat’s Structural Business Statistics Regulation for these sectors.

Output quality trade-offs

(Trade-offs are the extent to which different dimensions of quality are balanced against each other.)

It is recognised that the production of quarterly estimates, to comply with National Accounts timetable deadlines, can be subject to significant revisions. The results of the annual surveys are considered to be more accurate since they are based on more robust information used by businesses to prepare their annual financial reports. Therefore, more reliable estimates are available once the quarterly data have been aligned to the annual survey results.

Assessment of user needs and perceptions

(The processes for finding out about uses and users, and their views on the statistical products.)

Meetings have been held at least quarterly for some years with the national accounts compilers to discuss the quality of the latest sets of data and any emerging issues.

A range of users of the data were consulted in May 2012 to gauge their views on an Office for National Statistics (ONS) proposal to change from the current method of collecting quarterly transactions data for these surveys, moving instead to collecting quarterly balance sheet data. This will also make the surveys compliant with the revised version of the European System of Accounts, 2010: ESA 2010 and will be implemented around 2016 to 2017. Users will continue to be consulted as the ESA 2010 work progresses.

A user engagement plan is in place, along with a contact log and a Financial Inquiries inbox, used to record all user queries relating to the data. Also, a business statistics community has been set u p on

StatsUserNet to encourage dialogue and information-sharing between users and producers of business statistics. The future intention is to use this forum as another means of engaging with users about the financial surveys that underpin the MQ5 release.

Under the provisions of the Statistics and Registration Services Act 2007, the UK Statistics Authority has a statutory function to assess sets of statistics against the Code of Practice for Official Statistics with a view to determining whether it is appropriate for the statistics to be designated, or to retain their designation, as National Statistics. MQ5 was assessed by the UK Statistics Authority to ensure it was meeting the standards set out in the Code of Practice for Official Statistics. This involved consulting with the users and respondents to find out their opinions as well as reviewing all aspects of the survey output. The assessment report on financial investment (number 224) was published on 28 June 2012.

Accessibility and clarity

(Accessibility is the ease with which users are able to access the data, also reflecting the format in which the data are available and the availability of supporting information. Clarity refers to the quality and sufficiency of the release details, illustrations and accompanying advice.)

Our recommended format for accessible content is a combination of HTML web pages for narrative, charts and graphs, with data being provided in usable formats such as CSV and Excel. Our website also offers users the option to download the narrative in PDF format. In some instances other software may be used, or may be available on request. Available formats for content published on our website but not produced by us, or referenced on our website but stored elsewhere, may vary. For further information please refer to the contact details at the beginning of this report.

For information regarding conditions of access to data, please refer to the following links:

• Terms and conditions (for data on the website)
• Accessibility

In addition to this Quality and Methodology Information, quality information relevant to each release is available in the quality and methodology section of the relevant statistical bulletin.

Useful links

• Background information on the institutions covered in the MQ5 release
• Investment by insurance companies, pension funds and trusts (MQ5) glossary