FISIM | May 2014 Changes to the UK National Accounts: Financial Intermediation Services Indirectly Measured (FISIM) Authors: Karen Grovell and Daniel Wisniewski, Office for National Statistics Date: 29 May 2014 Executive summary This article covers the changes to financial intermediation services indirectly measured (FISIM) that will be introduced when revised figures for the UK National Accounts, consistent with Blue Book 2014, are published in September 2014. The change to FISIM is as a result of ensuring comparability in measuring Gross National Income (GNI) across the European Union (EU). The ONS implemented the concept of FISIM into the UK National Accounts at Blue Book 2008. The Office for National Statistics (ONS) in collaboration with the Bank of England (BoE) have reviewed FISIM methodology, and will be implementing improvements to the reference rates used in calculating current price FISIM. In addition, interbank FISIM, calculated on services between FISIM producing financial intermediaries, has been removed following clarity in international statistical standards. 1 Introduction Gross National Income (GNI) is an important statistic within the National Accounts. It is used in the calculation of a Member State’s contribution to the EU budget. Due to its operational importance, the EU statistical office (Eurostat) carries out regular audits of the methods and data used to estimate GNI. In 2012, following a comprehensive audit of the methods used across EU countries, a number of areas for improvement were identified which Member States have to address by 2014. One of the improvements is to bring the calculation of FISIM in line with European regulation; this article provides a technical overview of the associated methodology. The UK National Accounts, consistent with Blue Book 2014, will be published in September 2014 and will include improvements to methods and data to address this issue with respect to the UK. This article does not provide a numerical assessment of the impact; this can be found in the article ‘Impact of ESA95 changes on current price GDP estimates’. 1.1 Overview of the key methodological changes The ONS implemented the concept of FISIM into the UK National Accounts at Blue Book 2008. Recently an international task force led by Eurostat, the European Central Bank (ECB) and the Organisation for Economic Co-operation and Development (OECD) reviewed FISIM methodology. The report of that task force was used by Eurostat to redefine the regulation and was used by the ONS and BoE to review UK FISIM estimates. Although European regulations are closely followed both in the calculation of FISIM and in the allocation of FISIM into user sectors, to come fully in line with regulations the UK will be implementing a number of improvements; firstly, the reference rates used in calculating current price FISIM will be changed, resulting in a new internal reference rate and a new external reference rate; and secondly, interbank FISIM will be removed following clarity in international statistical standards. Office for National Statistics | Page 1 FISIM | May 2014 2 National Accounts concepts FISIM is the process whereby a financial institution such as a bank accepts deposits from units (people and institutions) wishing to receive interest on funds, and lends them to units wishing to borrow. The bank acting as an intermediary provides a mechanism to allow the first unit to lend to the second. The unit lending funds accepts a rate of interest lower than that paid by the borrower. A 'reference rate' of interest is the rate at which both lender and borrower would be happy to strike a deal. The difference between interest calculated at the reference rate and interest actually paid to depositors and charged to borrowers is a financial intermediation service charge indirectly measured (FISIM). Total FISIM is the sum of the implicit fees paid by the borrower and the lender. FISIM are produced by the following financial intermediaries (FIs): deposit-taking corporations except the central bank (S.122); and other financial intermediaries, except insurance corporations and pension funds (S.125). Interbank FISIM are the FISIM produced between FIs. By convention, no interbank FISIM has to be calculated between resident FIs, nor between resident FIs and non-resident FIs. FISIM are calculated in respect of non-bank user institutional sectors only. FISIM are allocated into intermediate and final consumption. A full description of the implementation of FISIM in the UK National Accounts can be seen in the article, ‘Improving the measurement of banking services in the UK National Accounts. Economic & Labour Market Review’. 3 3.1 Data sources and methodology Current approach 3.1.1 Current calculation of reference rates According to European regulation, the internal reference rate used to calculate FISIM in relation to resident user sectors should be calculated as the ratio of accrued interest received on interbank loans to the corresponding average stocks. However, in the UK we use official interest rates as the reference rates. The sterling rate applies to sterling business and Euro/US rate applies to foreign currency business, based on the resident sector split of balance sheet holdings. Regulation also specifies that an external reference rate should be used to calculate FISIM exports and imports stemming from business with non-residents. This is calculated as the ratio of accrued interest received on loans and paid on deposits between resident and non-resident FIs to the corresponding average stocks. However, separate data with non-resident FIs were not available in the UK, and the external reference rate is calculated as the mid-rate between the calculated loan and deposit rates from the stocks and interest data (see 3.2.1 below). 3.1.2 Current interbank FISIM Recent guidance clarifies that “by convention, no interbank FISIM have to be calculated between resident FIs, nor between resident FIs and non-resident FIs. FISIM are calculated in respect of non-bank user institutional sectors only.” (European System of Accounts (ESA 2010) section 14.11). Currently the UK excludes the calculation of interbank FISIM between resident deposit-taking corporations except the central bank (S.122); it also excludes interbank FISIM between non-resident deposit-taking corporations except the central bank (S.122) and resident FIs. Office for National Statistics | Page 2 FISIM | May 2014 3.2 New approach to be implemented in September 2014 3.2.1 New calculation of reference rates Internal reference rate Following the regulation, the rate can be calculated using either loans or deposits data, or both, depending on which is more reliable. It was agreed (between ONS and BoE) that the loan data was of better quality than the deposit data, and thus the new reference rate is calculated as follows: Internal reference rate = interest receivable on domestic (intra-UK) interbank loans average stock of domestic (intra-UK) interbank loans The average stock of loans over the quarter is calculated as an average of the monthly data reported in the four months across the quarter External reference rate The regulation specifies that the calculation of the external reference rate should follow a similar approach using interest flows and comparable stocks for business with non-resident banks. However, using this approach in the UK raises several concerns. The biggest of which is the effect that large intrabank flows have on the reference rate. As London is a major financial centre, there are sizeable intrabank loans and deposits, which are often just used for liquidity/cash flow management purposes, rather than reflecting true funding. The associated interest does not reflect the cost of funding or a realistic price of risk, so there are often large balance sheets with little corresponding interest resulting in an unrealistically low reference rate. The low reference rate then leads to sustained and implausible negative deposit FISIM when applied to non-bank sectors. This issue was discussed with Eurostat, and they supported the UK’s efforts to bring methodology closer in line with the guidance but understood the UK’s concerns with the proposed approach to calculating the external reference rate. Considering that the use of a separate external reference rate seems inappropriate for the UK, Eurostat agreed that the resident reference rate can be applied in the UK for calculating exports and imports of FISIM. This more accurately reflects the pure cost of funding for banks, creating a more realistic time series. 3.2.2 New interbank FISIM In the new approach to the calculation of FISIM the convention of interbank FISIM has been comprehensively adhered to. In addition to the current practice of on interbank FISIM (see 3.1.2), the UK will exclude interbank FISIM between resident deposit-taking corporations except the central bank (S.122) and resident other financial intermediaries (S.125); it will also exclude interbank FISIM between resident deposit-taking corporations except the central bank (S.122) and non-resident FIs. 4 4.1 Impact of the changes Impact overview This article does not provide a numerical assessment of the impact; this can be found in an associated article ‘Impact of ESA95 changes on current price GDP estimates’. Office for National Statistics | Page 3 FISIM | May 2014 In the new methodology the reference rates are generally higher. This has the impact of reducing the margins receivable by FIs on loans and increasing the margins receivable by FIs on deposits. GDP and final consumption FISIM Revisions to GDP are positive in most periods. The overall change to final consumption FISIM is mostly from large upwards revisions to consumer deposit FISIM across the period. These are then partly offset by downwards revisions to both export loan and deposit FISIM. Resident FISIM With respect to the changes to resident FISIM, revisions are driven by the change in reference rate. The new methodology more realistically reflects interbank lending rates, which are higher than official interest rates in all periods, and significantly so during the UK financial crisis. The reference rate is thus higher in all periods; the result being higher deposit FISIM and lower loan FISIM than we see in current data. Given that the deposit stocks that feed into final consumption FISIM are larger than the corresponding loan stocks, the increase in deposit FISIM exceeds the fall in loan FISIM. The exception to this is when the margin on deposits turns negative. The largest revisions are within consumer deposit FISIM. These occur because of the high proportion of stocks here relative to other sectors. Any increase in margin caused by the increase in the reference rate therefore impacts consumer deposit FISIM most significantly. This impact is at its largest during 2008-2009 where the difference between current and new reference rates is greatest; during periods when there is less stress on interbank markets these revisions are reduced significantly. Under the current methodology the low reference rate resulted in persistent negative consumer deposit FISIM from Q3 2008. While there are still some periods of negative FISIM under the new methodology, these are much smaller than those seen previously and intuitively the concept of negative consumer deposit FISIM on a small scale is credible. The change in methodology has therefore led to a much more plausible time series. Non-resident FISIM Revisions to non-resident FISIM will be predominantly downwards across the period. These are for the most part driven by the exclusion of interbank FISIM FI business. As outlined in section 3.2.1, export FISIM is currently calculated on business with all non-resident FIs, while the new methodology will calculate export FISIM only on business with non-resident non-FIs. The removal of interbank FISIM FI business will therefore lead to a fall in both export loan and deposit FISIM. The largest revisions will occur over 2008 where in current data there is a peak. This was caused by a sharp increase in both non-resident loan and deposit stocks over the year, the majority of which was business with non-resident FIs. This large increase in stocks led to elevated FISIM exports, when in reality FISIM should not have been calculated on the majority of this activity. By removing interbank FISIM FI business we get large downwards revisions to both loan and deposit FISIM exports, but this is a more accurate depiction of the real time series. The change in the reference rate also impacts non-resident FISIM, but works in the opposition direction to the removal of interbank FISIM FI business. The change in the reference rate alone increases total export FISIM, as the rate increases across all periods reducing loan FISIM but increasing deposit FISIM. As on the resident side, the deposit stocks are larger, and thus the increase here more than offsets the fall in loan FISIM. The removal of interbank FISIM FI business, however, exceeds this effect. Around 75-80% of all non-resident loan and deposit stocks are with FIs; by removing this non-resident business, FISIM exports fall significantly. It is therefore this methodological change that is driving the revisions to non-resident FISIM. Office for National Statistics | Page 4 FISIM | May 2014 5 Conclusions The changes that have been made bring the UK National Accounts in line with international statistical standards. The new reference rates better reflect the cost of funding for FIs. This combined with the removal of interbank FISIM greatly improves the UK’s FISIM time series. 6 Contact Karen Grovell email: Karen.l.grovell@ons.gov.uk 7 Acknowledgements ONS would like to thank their BoE colleagues for their contribution to this paper: Katrina Farrell, Lauren Bowers, Perry Francis and Tristan Broderick. 8 References Akritidis, L. (2007): Improving the measurement of banking services in the UK National Accounts. Economic & Labour Market Review European System of Accounts (ESA 2010) Office for National Statistics | Page 5