8. Financial Statistics
Author: Mr. Jose M Cartas and
• 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

## Abstract

8.1 Financial statistics cover the stocks and flows of financial assets and liabilities between all sectors of the economy, and between the sectors of the economy and the rest of the world. Financial statistics thus have broader sectoral coverage than monetary statistics, which cover the stocks and flows of the assets and liabilities of financial corporations (FCs).

## I. Introduction

8.1 Financial statistics cover the stocks and flows of financial assets and liabilities between all sectors of the economy, and between the sectors of the economy and the rest of the world. Financial statistics thus have broader sectoral coverage than monetary statistics, which cover the stocks and flows of the assets and liabilities of financial corporations (FCs).

8.2 Financial statistics can have various degrees of details. They can cover both stocks and flows, or only stocks, or only flows; and the breakdown of institutional sectors or categories of financial instruments can vary. Further, financial statistics can either be two- or three-dimensional.

8.3 The two-dimensional financial statistics are similar to the financial account and balance sheets, as defined in chapters 11 and 13 of the System of National Accounts 2008 (2008 SNA), respectively. Two-dimensional financial statistics do not provide information on counterpart sectors. For example, it can show the purchase by the nonfinancial corporations (NFCs) sector of debt securities but not identify the sector that issued these securities.

8.4 Three-dimensional financial statistics have from-whom-to-whom information, where the breakdowns of the financial stocks and flows of each sector also include the counterpart sectors—for example, the purchase by NFCs sector of debt securities issued by the central government. Compilers can present financial statistics, or selected components, in different ways, depending on the availability of data sources, level of detail compiled, analytical needs, and other considerations.

8.5 Financial statistics are increasingly used for analytical purposes. The from-whom-to-whom information highlights the role of FCs in financial intermediation. For instance, it can illustrate how the relative importance of various types of financing between sectors/subsectors is changing over time. Macroprudential analysis is another type of analysis, where financial statistics are useful. For instance, the balance sheet approach (BSA)1 uses financial statistics to see how interconnections among sectors impact the whole economy through debtor/creditor relationships.

8.6 From a statistical compilation point of view, financial statistics can reveal inconsistencies or discrepancies in the underlying data across institutional sectors and instruments.

8.7 Financial statistics follow the 2008 SNA framework (see Figure 8.1 and Annex 8.1) and, in principle, use the same institutional sectors, categories of instruments, and valuation and recording principles (covered in Chapters 3 to 5). Financial statistics include the financial account, balance sheets, and the other changes in assets account of the 2008 SNA.

8.8 This Manual recommends compiling financial statistics on a quarterly or annual basis, but encourages quarterly compilation. Compilation on a quarterly basis is applicable to countries that have quarterly data for the current accounts of their national accounts statistics, or are currently working on migration from annual to quarterly national accounts statistics. The degree of detail provided on a quarterly frequency depends on country circumstances with respect to sectoral and financial instrument coverage and the level of detail of the financial statistics being compiled. This Manual does not make a specific recommendation on the timeliness of the financial statistics but encourages a time lag of around one quarter for quarterly data. Data from the FCs’ sectoral balance sheets can be utilized to compile a substantial part of the financial statistics.

8.9 This chapter first provides an overview of the framework for financial statistics, including the 2008 SNA. The next section covers the compilation and presentation of financial statistics, and provides examples of presentational formats, including possible aggregations of subsectors and categories of financial instruments of the 2008 SNA. The last section discusses data sources and related issues. Figure 8.1, Annex 8.1, and Table 8A.1 provide an overview of the 2008 SNA focusing on aspects relevant for financial statistics.

## II. Framework and Scope of Financial Statistics

8.10 As noted in the introduction, financial statistics are a comprehensive set of data on stocks and flows of financial assets and liabilities between all sectors of an economy and with the rest of the world, on a from-whom-to-whom basis where applicable. Financial statistics show: (1) flows in financial instruments between the sectors of an economy, and (2) the corresponding financial asset and liability positions. The framework for financial statistics is the 2008 SNA balance sheets and accumulation accounts, because these accounts provide an internationally recognized set of guidelines for integrating financial stocks and flows into a complete system of accounts. The integration of sectoral flows and stocks also follows the principles of the Guidelines on Integrated Economic Statistics2  which sets out a consistent framework for measuring a country’s economic activity in an increasingly interconnected global economy.

8.11 The most comprehensive financial statistics cover the following SNA accounts: the financial account, balance sheets, and the other changes in assets account, all on a from-whom-to-whom basis (see Figure 8.1 and the section The Structure of the 2008 SNA Accounts of Annex 8.1).

8.12 The sequence of accounts of the 2008 SNA provides the integrated and comprehensive framework for both flows and stocks. The current accounts record the production of goods and services, income generation and distribution, and the use of income for consumption and saving during the period. The current accounts are followed by the accumulation accounts, which record the acquisition and disposal of financial and nonfinancial assets and liabilities and changes in net worth. Finally, the balance sheets show assets and liabilities of each institutional sector at the beginning and end of the period. The opening and closing balance sheets are linked through transactions and other flows, also called horizontal adding-up requirement (see paragraph 5.12). The balance sheets complete the sequence of accounts and show the final result of the entries in the production, distribution and use of income, and accumulation accounts.

8.13 The current accounts record transactions relating to production and the generation, distribution, and use of income. Each account records the resources available to the institutional units comprising the sector and uses of these resources by these units. Each account also contains an accounting construct, a balancing item obtained as the difference between the resources or changes in liabilities and the uses or changes in assets. The balancing items have significant analytical value, representing the results of the economic behavior recorded in the specific account. The balancing item for a specific account is recorded as a use in that account and then introduced in the next account as a resource.

8.14 The major balancing items of the current accounts for the total economy are gross domestic product (GDP), gross national income (GNI), gross national disposable income (GNDI), and gross saving. Each of these balancing items may be recorded either as gross or net. The difference between gross and net recording in the accounts is an adjustment for the consumption of fixed capital. This adjustment is undertaken in the production account and carries through all the other balances in the current accounts. Consumption of fixed capital represents the decline in the current value of the fixed assets due to their use in the production process.

8.15 The accumulation accounts cover changes in assets and liabilities, and net worth. They show all changes that occur between the opening and closing balance sheet dates for a subsector, sector, or the entire economy. Accumulation accounts comprise the capital account, the financial account, the other changes in the volume of assets account (OCVA), and the revaluation account. Accumulation accounts thus record transactions as well as other flows. The financial account shows how an institutional sector with a balance on net lending, makes these surplus resources available to other sectors by acquiring net financial assets or reducing net liabilities. Likewise, the financial account shows how a net borrowing sector obtains financial resources by reducing its net holdings of assets or increasing its net liabilities.

8.16 The balance sheets show stocks of nonfinancial and financial assets and liabilities on the balance sheet date for all the sectors of the economy and the counterpart positions for the rest of the world. The difference between total assets and liabilities is net worth. For each category of assets and liabilities, and thus net worth, changes between the opening and closing balance sheets result from transactions and other changes recorded in the accumulation accounts.

### A. Flow Accounts

8.17 Flow accounts include the current accounts and all the accumulation accounts, and from the perspective of financial statistics, flow accounts comprise financial account and other changes in financial assets and liabilities only. Therefore, financial flow accounts comprise the flows of financial assets and liabilities of all the sectors of the economy and with the rest of the world. Various presentational formats can be used with different degrees of cross-classifications and details.

8.18 Following the 2008 SNA, the saving in the last current account (use of disposable income account) should, in principle, match the capital account, which covers the transaction flows in nonfinancial assets and net lending/borrowing (see Figure 8.1 and Table 8.4). Net lending indicates a sector’s net financing of other sectors and net borrowing indicates a sector’s net borrowing from other sectors.

Table 8.1

Two-Dimensional Financial Statistics for Transactions

Source: 2008 SNA, Table 11.1. Values for this table are based on Tables A2.1, A2.2, and A2.3 in Appendix II of this Manual.Note: NPISHs = nonprofit institutions serving households; SDRs = Special Drawing Rights.
Table 8.2

Two-Dimensional Financial Statistics for Stocks (Financial Balance Sheet)

Monetary gold is a financial asset by convention, without a corresponding liability (no asset of the RoW). Therefore, there will be a discrepancy between assets and liabilities.

Note: NPISHs = nonprofit institutions serving households; SDRs = Special Drawing Rights.
Table 8.3

Two-Dimensional Financial Statistics for Stocks, Transactions, and Other Flows

Note: NPISHs = nonprofit institutions serving households; SDRs = Special Drawing Rights.
Table 8.4

Integrated Capital and Financial Account

Note: NPISHs = nonprofit institutions serving households; SDRs = Special Drawing Rights.

8.19 The financial account of the 2008 SNA records the net acquisition of financial assets and net incurrence of liabilities for all institutional sectors by type of financial asset3  and shows how net lending or net borrowing is reflected in transactions in financial assets. The net lending/borrowing from the financial account should, in principle, be equal to the net lending/borrowing from the capital account. In practice, there is always a difference, which is either presented as a statistical discrepancy or distributed to residual sectors for each financial instrument (see below subsection Statistical Discrepancies).

8.20 The entries in the financial account may result from: (1) transactions involving the exchange of financial assets and liabilities; and/or (2) counterpart entries to nonfinancial transactions. The sale of goods, services, or nonfinancial assets may have as its counterpart a change in currency or transferable deposits. Alternatively, the counterpart may be reflected in the financial account in trade credit or other category of accounts receivable or payable. This information is valuable in identifying the financial instruments that net borrowing sectors use to finance their deficits and the instruments that net lending sectors use to allocate their surpluses. In cases where a sector is a net-zero borrower,4  the financial account is still useful because it shows how that sector’s composition of financial assets and liabilities has changed.

8.21 The financial account does not identify counterpart sectors for financial transactions and, therefore, the question of who is financing whom is not answered. For a full understanding of financial flows and the role they play in the economy, it is important to identify counterpart sectors. Adding counterpart sectors for financial transactions enables from-whom-to-whom type analysis, which allows tracking how surpluses by one sector are allocated amongst domestic sectors and cross-border by type of financial instrument, and how sectors with deficits meet their financing needs in terms of financial instruments used and financing sectors.

8.22 This type of from-whom-to-whom approach is particularly important in analyzing the role of FCs in financial intermediation; that is, mobilizing financial resources and making them available to other sectors in forms suitable to these sectors, including through the transformation of different characteristics of the financial instruments, such as maturity. The FCs sector and its subsectors play a critical role in directing financing flows from net lending sectors to net borrowing sectors and in supplying the instruments used.

8.23 In addition to financial transactions, flow accounts may also be compiled that identify counterparts to changes in financial assets and liabilities due to revaluations or OCVA.

### B. Stock Accounts

8.24 Stock accounts comprise the balance sheets5  of all the sectors of the economy and with the rest of the world. Various presentational formats can be used for financial balance sheets with different degrees of cross classifications and details.

8.25 The SNA balance sheets show the opening and closing balance sheets of all sectors of the economy and the counterpart assets and liabilities of the rest of the world for each financial instrument category, as well as changes during the reference period. Like the financial account, the SNA balance sheets do not identify counterpart sectors for financial assets and liabilities. Adding counterpart sectors enables from-whom-to-whom analysis, highlighting the financial intermediation role of the FCs sector and its subsectors.

## III. Compilation and Presentation of Financial Statistics

8.26 The complex and detailed information of financial statistics offers both presentational opportunities and challenges. Striking a balance between providing sufficiently detailed data and conveying information efficiently to users is challenging, particularly because different users may have different needs. Tables with too much detail may hinder a user’s ability to extract the required information or to uncover trends and relationships of interest.

8.27 Possible different presentations range from either sector-specific and/or instrument-specific tables for stocks and/or flows, or specific flow components, to detailed data for all sectors on a from-whom-to-whom basis. In addition to data for a specific period, compilers and users can focus on sets of templates for several time periods to trace the changes over time. This approach is relevant particularly for presentations containing more detailed information, such as covering counterpart sectors for all instruments and stocks/flows covering multiple time periods.

8.28 A fully articulated set of standard templates covering stocks were developed in the context of the G-20 Data Gaps Initiative (DGI). These templates are a set of internationally comparable sectoral accounts and balance sheets that provide a minimum and encouraged classification for sectors, financial assets and liabilities, and nonfinancial assets.6  These standard templates, however, do not identify counterpart sectors, and so do not provide from-whom-to-whom information.

### A. Two-Dimensional Financial Statistics

8.29 Two-dimensional financial statistics can be prepared for: (1) transactions; (2) stocks; and (3) other flows, including separately for revaluations and OCVA; or (4) all flows together. Two-dimensional financial statistics can be constructed for one period or several time periods.

8.30 Two-dimensional financial statistics for transactions are similar to the financial account of the 2008 SNA (see Table 27.3 of the 2008 SNA). Each sector has two columns—one for net acquisition of financial assets and one for net incurrence of liabilities. To emphasize that transactions in financial instruments can be directly measured, net financial investment is used (instead of net lending/borrowing) and is calculated as net acquisition of financial assets less net incurrence of liabilities.

8.31 Conceptually, for each sector the sum of all columns for transactions in financial assets equals the sum of all columns for transactions in liabilities and net worth if all the transactions are covered. A financial instrument may appear both as a change in financial asset and a change in liability for a sector, for example, when a sector both issues and buys equity shares. For each financial instrument category, the sum of changes in assets across all the sectors equals the sum of changes in liabilities. If the identities do not hold, the table would show discrepancies between assets and liabilities for sectors and instruments. Table 8.1 presents an example of two-dimensional financial statistics for transactions for a single period. In the case of debt securities, for example, total change in financial assets (5.1 = 40.8 + (−35.7)) is equal to total changes in liabilities across sectors (5.1 = 8.4 + (−3.3)).

8.32 Two-dimensional financial statistics for stocks are similar to the 2008 SNA balance sheets (see Table 13.1 of the 2008 SNA). For each sector, assets and liabilities are shown separately. The net financial position equals financial assets less liabilities (see Table 8.2).

8.33 The other flows table can be presented in the same format as transactions, which bridges the gap between transactions and period-to-period changes in stocks. For each sector, the other flows table has columns for changes in assets and liabilities, respectively. Further, instead of net financial investment, the other flows presentation shows net other flows. For more detailed information, separate tables for revaluations and OCVA may be constructed, if the data sources allow.

8.34 A combined presentation of balance sheet opening stocks, transactions, other flows, and closing stocks can be used as shown in Table 8.3. This combined presentation highlights the adding-up requirements that must be satisfied for each sector and for each instrument.

8.35 Financial statistics are developed within the integrated framework of the 2008 SNA. Conceptually and analytically, the inter-linkage between the capital account and the financial account allows linking the nonfinancial accounts and the financial accounts through the net lending/net borrowing balancing item, which can be derived both from the capital account and from the financial account. Table 8.4 presents an example of an integrated capital and financial account for one quarter, showing how net lending or borrowing in the financial account is equal to that in the capital account. It draws on the net incurrence of liabilities (95 for the total economy), net acquisition of financial assets (120 for the total economy), net acquisition of nonfinancial assets (192), and changes in net worth due to savings and capital transfers in the capital account (217) all from Table 8A.1. Because it is an example, the table does not show any statistical discrepancy between saving and capital transfers, and total net investment, although in practice they might occur.

### B. Three-Dimensional Financial Statistics

8.36 Three-dimensional financial statistics add the counterpart sector to the two-dimensional financial statistics, allowing for from-whom-to-whom analysis. Figure 8.2 presents the concept of the three-dimensional financial statistics, which provides a framework to present flows and stocks for all categories of instruments for all the sectors or subsectors by counterpart sector. Within this framework, it is possible to trace who finances whom, by which category of financial instrument, and the financing amount. For transactions, the three-dimensional presentation shows both parties to the transaction as well as the financial instrument being used—this is also sometimes called flow-of-funds statistics. For stocks, the similar three-dimensional presentation shows the creditor and debtor for each financial instrument category—this is also sometimes called BSA.

8.37 For analyzing bilateral cross-border stocks and flows, the three-dimensional presentation may be expanded further by breaking down the rest of the world by country, and even sector (i.e., both domestic and cross-border information are on a from-whom-to-whom basis, by country and sector). Ultimately, such an expansion across many countries would allow constructing bilateral financial statistics at the global level. Such “global flow of funds” data have high analytical value, like analysis of interconnectedness, global liquidity flows, and global financial networks. Individual economies could also benefit by being able to identify possible spill-over channels of external shocks into the domestic economy and its sectors.

8.38 A three-dimensional presentation for stocks is shown in Table 8.5. It shows the creditor and debtor sector for each financial instrument category. Assets and liabilities are shown for each sector and, for convenience, the net financial position for each financial instrument category, similar to the net financial position for the sectors in Tables 8.2 and 8.3.

Table 8.5

Three-Dimensional Financial Statistics for Stocks

Note: A = financial assets; L = liabilities; NP = net position (financial assets less liabilities); NPISHs = nonprofit institutions serving households; SDRs = Special Drawing Rights.
Table 8.5a

A Simplified Example of Three-Dimensional Financial Statistics for Two Sectors

Note: A = assets; L = liabilities; NP = net financial position (financial assets less liabilities); SDR = Special Drawing Rights.Source: Values for this table sourced from Tables A2.1, A2.2, and A2.3 in Appendix II of this Manual.

8.39 The presentation shown in Table 8.5 can also be used for transactions and other flows (where relevant), on a from-whom-to-whom basis. Instead of a net financial position, the transactions presentation shows net financial investment, and the other flows table, net other flows. In principle, separate from-whom-to-whom tables for revaluations and OCVA may also be constructed for financial assets, if the data sources allow.

8.40 This Manual recommends compiling the three-dimensional financial statistics on an unconsolidated basis, which means that if a financial instrument is both an asset and a liability of the same sector, both the asset and liability positions are shown (in the shaded cells in Table 8.5). Furthermore, in unconsolidated data, the reciprocal asset/liability positions could be compared. Consolidated data could result in the loss of analytical value. When data are consolidated, the diagonal cells (shaded boxes) representing intra-sector stock positions are empty.

8.41 Table 8.5a shows an extract of cross-cells for two sectors—FCs and general government. The net financial position of the FCs with the general government is 394.1 (see the total cell under net financial position in the upper-right cell, showing a negative position of the general government vis-à-vis FCs), which represents the net claim on the general government. The entries for debt securities of 610.7 under liabilities in the top-right cell and under assets in the lower-left cell show debt securities issued by central, state, and local governments and held by FCs.

8.42 As the number of financial instrument categories and sectors increases, compilers may present three-dimensional financial statistics separately for each sector. For example, complete sectoral balance sheets may be presented for each sector similar to those in Appendix II showing opening stocks, transactions, valuation changes, and OCVA for all financial and nonfinancial assets with counterpart sectors where relevant. Similarly, from-whom-to-whom tables for positions (by issuer and holder sectors) and transactions (by debtor and creditor sectors) may be presented for a specific financial instrument, for example debt securities.7

8.43 Using standardized monetary statistics sources, such as the SRFs, allows further breakdowns of all the instruments by currency (into domestic and foreign currency).

### C. Balance Sheet Approach

8.44 A common use of three-dimensional financial statistics is the BSA analysis. The BSA can be used to analyze vulnerabilities arising from balance sheet positions and mismatches at a point in time, as well as of the buildup of such vulnerabilities over time. Simulations can be conducted to analyze the spillover of possible shocks from one sector into another, such as a sudden withdrawal of bank deposits or an inability to “rollover” maturing external debt.

8.45 The BSA framework considers four main types of balance sheet mismatches that can point to vulnerabilities for all sectors of the economy. These are: (1) currency mismatches, which arise when borrowers’ liabilities denominated in a foreign currency are larger than their assets denominated in that currency, or vice versa; (2) remaining maturity mismatches between liabilities and assets, such as short-term liabilities versus longer-term assets that create funding and interest rate risks; (3) capital structure problems, including an excessive reliance on debt rather than on equity (or high leverage); and (4) solvency or counterpart risk if the assets of a debtor are not sufficient to cover its liabilities, including contingent liabilities. Thus, the BSA identifies inter-sector linkages and possible balance sheet mismatches by instrument, counterpart sector, currency, and possibly maturity.

8.46 For completeness, the BSA may include memorandum items to allow the calculation of control totals. These could include, for example, nonfinancial assets and other assets/liabilities for which counterpart sectors are not identified, as well as certain off-balance-sheet items taken from other sources (for example, contingent liabilities taken on by certain sectors such as credit commitments).

### D. Statistical Discrepancies

8.47 Statistical discrepancies arise when two data sets provide different numerical values for the same data category. In financial statistics, there is typically a statistical discrepancy between net lending/borrowing (NL/NB) derived from the capital account and net financial investment (NFI) derived from the financial account, which conceptually should equal (see paragraph 8.35). From the capital account, the identity for net lending/borrowing is:

$NL/NB\equiv Net\text{\hspace{0.17em}}Saving+Net\text{\hspace{0.17em}}Capital\text{\hspace{0.17em}}Transfers-Net\text{\hspace{0.17em}}Capital\text{\hspace{0.17em}}Formation$

From the financial account, the identity for net financial investment is:

$NFI\equiv Net\text{\hspace{0.17em}}Acquisition\text{\hspace{0.17em}}ofFinancial\text{\hspace{0.17em}}Assets-Net\text{\hspace{0.17em}}Insurrence\text{\hspace{0.17em}}of\text{\hspace{0.17em}}Liabilities$

8.48 No clear international consensus on the treatment of this discrepancy exists. One approach is to keep the discrepancy through the use of a residual (referred to as a balancing item). This approach assists users in gauging the magnitude of errors and the overall quality of the data. Alternatively, the discrepancy can be removed by distributing the discrepancy across one or more items in the capital account, the financial account, or both—treating the discrepancy as a transaction or valuation change or, more likely, as OCVA. The advantage of removing discrepancies is that it provides “balanced” accounts, and ambiguity is eliminated for the users. Consistent with the 2008 SNA, it is recommended to provide users with the recorded data for NL/NB and NFI, and not distribute the discrepancy, while indicating which data set is more reliable.

8.49 For compilers, data on the discrepancies provide valuable information for identifying areas that need improvement in data collection, estimation, and compilation system.

## IV. Data Sources for Financial Statistics

8.50 Monetary statistics are often the main data source for financial statistics. This is because of the financial intermediation role of FCs in an economy and the good coverage of monetary statistics. The FCs sector is often the counterpart to financial transactions either as the creditor/holder or the issuer/debtor of a financial instrument.

8.51 Conceptually, data sources for financial statistics can be obtained from both parties to each financial transaction/position. In practice, compilers may need to rely on the data reported by only one party to a financial transaction/position. For instance, the household sector, which consists of many small units, channels financial transactions through the FCs sector, which has much fewer institutional units and typically reliable data recording and reporting framework. It may therefore be feasible to derive the data on financial assets and liabilities of the household sector from FCs’ records, rather than directly from the households, although financial assets and liabilities of the household sector vis-à-vis sectors other than FCs, such as the nonresident sector, would need to be captured through alternative data sources.

8.52 Furthermore, standard sources for monetary statistics, such as sectoral balance sheets/SRFs presented in Appendix II, provide detailed counterpart information for almost all types of financial assets and liabilities of the FCs sector and its subsectors, which enables developing from-whom-to-whom tables for stocks.

8.53 If monetary statistics are compiled for the components of flows, as shown in Appendix II, these data can be used for compiling flows within the framework of the three-dimensional financial statistics. Otherwise, compilers must use different direct or indirect data sources and/or methods to estimate transactions and other flows involving FCs’ assets and liabilities, as explained in Chapter 5 (paragraphs 5.69–5.85). The remaining stock and flow data to complete the three-dimensional financial statistics can be taken from other data sources such as government finance statistics, external sector statistics, and securities statistics, compiled with counterpart sector detail.

8.54 Table 8.6 describes the relative reliability that is typical of various data collected or estimated for financial statistics. Highly reliable data are those that can be directly obtained from FCs’ reported data, and the IIP and balance of payments statistics. The moderately reliable data are those that involve estimation, but for which some data source are available on an annual or less frequent basis, or are contained in surveys. The less reliable data are those that are difficult to obtain or for which data sources do not usually exist. Many estimates of data in this category are based on residual calculations, as explained later in this section. The nonshaded cells are those for which liabilities do not usually exist. For example, households and NPISHs do not issue currency nor deposits, and only the government issues government securities.

Table 8.6

Reliability of Data Sources

Note: DC = depository corporation; OFC = other financial corporation; NFC = nonfinancial corporation; NPISH = nonprofit institutions serving households.

### A. Main Data Sources

8.55 Compilers of financial statistics will need to use a variety of data sources. The compilation usually starts by determining the stocks for each sector and the rest of the world. Table 8.7 presents the most common sources for stock and flow data.

Table 8.7

Main Data Sources

Note: SRF = Standardized Report Form; FC = financial corporation.

### B. Supplementary Data Sources

8.56 Supplementary data sources include administrative records, survey data, and market data outside the regular compilation of macroeconomic datasets. These sources can be used to improve the data estimates and to fill gaps in the main data sources. Some types of supplementary data are described in Table 8.8.

Table 8.8

Supplementary Data Sources

Note: OFC = other financial corporation; OCVA = other changes in the volume of assets.

### C. Application of Data Sources

8.57 The main and supplementary data sources provide most of the necessary data for financial statistics, but often do not properly cover flows and stocks for which both counterparts are NFCs, or households and NPISHs. Further, for negotiable financial instruments full disaggregation by sector is often not available, especially for holders. Therefore, compilers need to allocate the aggregated data among issuers and holders, using available data sources and ancillary information.

8.58 In such circumstances, compilers usually apply either of two techniques—counterpart data or residual data. Both techniques apply the principle that every financial claim (other than gold bullion included in monetary gold) is a financial instrument that has a counterpart liability, or that for each category of financial instrument the sum of the net acquisitions (including those of the rest of the world) must equal the sum of the net incurrence of liabilities. A residual sector may need to be designated to ensure that the total value of the transactions or positions in a financial asset/liability category is fully allocated, if data are available for all but one sector.

8.59 To fill data gaps for nonfinancial sectors and negotiable instruments, compilers can rely on control totals for financial assets and liabilities. A control total refers to the total amount (stocks and flows) of a certain financial instrument issued and held. The compilers can obtain the data for control totals from the following sources presented in Tables 8.7 and 8.8:

• a. Balance sheet data of FCs (mainly for nonnegotiable instruments such as deposits and loans)

• b. Government records for government debt

• c. Other sources such as custodians/security registration offices (mainly for negotiable instruments such as debt securities and shares).

8.60 For flow data, in addition to the main and supplementary source data shown in Tables 8.7 and 8.8, compilers may need to apply procedures for estimating the transactions, revaluations, and OCVA data, similar to those for monetary statistics described in Chapter 5.8  Securities price indices are often used to estimate valuation changes for securities, when market values of the securities are unavailable.

#### Multiple data sources

8.61 Compilers often deal with a lack of data sources, but sometimes multiple or overlapping sources for a sector or financial instrument may be available containing different numbers. In these cases, compilers need to assess the relative quality of the data sources and try to find out the possible reasons for differences in order to use the most reliable data aligned with statistical principles. For example, the government authorities may report data on most of their lending to public nonfinancial corporations (PNFCs). Liability data for the same loans may be reported by the PNFCs themselves, as borrowings from the government. The two sets of data may contain different numbers, because one side of the reporting may not include data for all PNFCs. In this case, compilers should use the most comprehensive and reliable data to estimate the asset and liability positions for loans extended by government to PNFCs.

8.62 Multiple data sources used for compiling financial statistics are not always based on the definitions, classifications, and accounting principles of the 2008 SNA. For monetary statistics, there are at least two such cases. First, equity on the liability side of the sectoral balance sheets is book-valued, but needs to be market-valued for financial statistics. Second, provisions for expected losses on financial assets are included in other accounts payable in monetary statistics, but are not recognized in the 2008 SNA and thus in financial statistics. Figure 2.2 in Chapter 2 shows how monetary statistics are reconciled with those of SNA for these two cases.

8.63 To the extent possible, compilers need to adjust the data sources to meet the requirements and quality standards for compiling financial statistics. Examples of such cases in data sources include:

• a. Sectoring—the disaggregation with respect to counterpart sectors may differ in the data source from the sectoring used for financial statistics.9

• b. Classification of financial instruments—the categories of financial assets and liabilities may differ in the data source from the instrument classifications required for financial statistics.

• c. Valuation—negotiable (or foreign currency denominated) financial instruments may not be valued at current market price (or market exchange rate) or fair value.

• d. Time of recording—transactions may not be recorded on an accrual basis and the time of recording may differ for the two parties to the transaction or position.

• e. Coverage—macro data for a sector may cover units that do not belong to that sector or may exclude some units that belong to the sector; for example, a nonresident branch of a financial corporation may be included in the accounts of the parent but should be excluded as it is a nonresident institutional unit.

8.64 Compilers of financial statistics often need to deal with data reporting that is inconsistent, partial, or indirect, as well as with the absence of data reporting for certain units, subsectors, or even sectors. When both parties report a transaction and a position, several types of errors may occur; this introduces a data inconsistency. For example, the parties may use different valuations of transactions or positions, different timing for recording transactions, and different classifications by instrument or counterpart sector. These kind of inconsistencies need to be adjusted based on the more reliable source.

8.65 The types of adjustments needed to align specific sets of source data with the requirements of financial statistics or to be internally consistent may vary depending on particular circumstances.

#### Estimation of missing data

8.66 Compilers of financial statistics may need to rely on various techniques to estimate incomplete data. For example, to maintain a timely release calendar for financial statistics, compilers may need to estimate missing data that will become available later, or use annual data to estimate quarterly data.

8.67 For estimating missing data for which historical series are available but the next observation becomes available with a delay, compilers may rely on known estimation techniques such as repeating the previous observation or extending the trend of historical series. For estimating higher-frequency data (e.g., quarterly) using lower-frequency data (e.g., annual), different interpolation methods may be used such as constant ratios, data smoothing, and regression methods.10

#### Editing and checking data

8.68 Compilers apply different validation and plausibility checks for rows and columns of financial statistics to identify data problems. Construction of charts and tables of time series data may also be useful in revealing outliers to be verified or corrected.

8.69 For plausibility testing of the aggregate data, unexpected movements in the data should be explained in terms of economic behavior, if not attributable to data collection, estimation, or compilation errors. For this purpose, basic underlying relationships among macroeconomic data need to be applied. Compilers’ knowledge may need to be complemented by consultation with different experts and analysts.

8.70 Data problems are more apparent in time series than for a single period. The presentational format for a period or end of period enables checking that the data meet adding-up requirements and broad plausibility tests, but time series presented in tables or charts—in differenced or percentage-change form, or as ratios—are useful for identification of outliers that need to be investigated.

## V. Systematic Development of Financial Statistics

8.71 Countries need a wide range of data sources and methods to compile the full range of financial statistics. As these data sources may be collected by different agencies, close cooperation among statistical agencies within an economy is crucial. Given differences in availability of data sources across countries, financial statistics can be developed with varying degrees of detail depending on data availability and analytical needs. Table 8.9 presents different levels of compiling financial statistics in terms of covered details.

Table 8.9

Different Levels of Financial Statistics

Note: OCVA = other changes in the volume of assets; 2008 SNA = System of National Accounts 2008; OFC = other financial corporation.

8.72 An example of a possible presentation of financial statistics at the intermediate level is set out in the IMF’s Special Data Dissemination Standard Plus: Guide for Adherents and Users, Table 2.1.11  This table is based on the requirement of the Special Data Dissemination Standard Plus for its adherents and presents “minimum” classifications, by sector and instrument. Financial statistics provide most of the necessary input for compiling SNA balance sheets.

8.73 The most comprehensive financial statistics with full details can include additional subsectors and subcategories of financial instruments. Table 8.10 illustrates possible additional subsectors and financial instruments.

Table 8.10

Examples of Disaggregated Sectors and Instruments

Note: ICPF = insurance corporations and pension funds; SDR = Special Drawing Right.

## Annex 8.1 Overview of the 2008 System of National Accounts

### Introduction

8.74 This annex provides an overview of the 2008 System of National Accounts (2008 SNA) and illustrates how it presents the overarching framework for the development of financial statistics. Financial statistics are developed within the framework of the 2008 SNA, which provides for comprehensive coverage of stocks and flows for the total economy as well as for each of its sectors.

8.75 The 2008 SNA serves as a coordinating framework for macroeconomic statistics and provides a consistent conceptual and accounting framework for integrated macroeconomic aggregates. It provides a presentation of macroeconomic aggregates relating to all institutional sectors of the economy and the economic relationships of an economy with the rest of the world. These macroeconomic and sectoral data allow for key macroeconomic indicators to be derived on corporate borrowing, profitability, household wealth and savings, leverage ratios and estimates of debt service burden. As such, the 2008 SNA provides consistency and coherence with other internationally accepted standards for macroeconomic statistics.

8.76 The integrated economic accounts of the 2008 SNA comprise the full sequence of accounts by institutional sector, the accounts for the rest of the world, and the accounts for the total economy. The key conceptual elements underlying the integrated economic accounts are as follows:

• a. Institutional units that are further grouped into sectors

• b. Transactions and other flows

• c. Assets and liabilities.

8.77 Chapters 3 and 4 of the 2008 SNA and Chapters 3, 4, and 5 of this Manual elaborate on the conceptual framework for macroeconomic statistics. The 2008 SNA is a closed system in that it is designed to record all flows undertaken by all units resident in the domestic economy and the assets and liabilities of these units. There are two aspects of the 2008 SNA that facilitate this recording: (1) quadruple-entry accounting, and (2) the rest-of-the-world account.

8.78 Quadruple-entry accounting is the result of the simultaneous application of vertical and horizontal entries, where each transaction results in four entries in the system. Vertical double-entry accounting—also simply referred to as double-entry bookkeeping—results in two entries, commonly referred to as a credit entry and a debit entry. It is the accounting system commonly used in business accounting. Horizontal double-entry bookkeeping reflects mutual economic relationships between two units.

8.79 When a transaction is undertaken, two pairs of entries are recorded. The first pair of entries records the provision of a good, service, or asset to a unit and the acquisition of the good, service, or incurrence of liability by the other. The second pair of entries usually appears in the financial account and records one party supplying the means of payment (by incurring liabilities or reducing assets) and the other party receiving payment. The second pair of transactions may appear in the current accounts or capital account in cases of unrequited transfers or transfers in kind. Further, in cases where the transaction is based on changing the composition of the portfolio of financial assets or liabilities, both pairs of entries would appear in the financial account.

8.80 The use of a quadruple-accounting system means that accounts add up horizontally and vertically, making the system of national accounts a closed system. All transactions have counterpart entries both horizontally and vertically. Conceptually, all transactions of an individual sector and subsector sum to zero, as do transactions across all sectors; and all liability positions have counterpart financial asset positions, except for gold bullion held as a reserve asset. Table 8A.1 provides a numerical example of the production of the “top-to-bottom” accounts. Such “top-to-bottom” accounts can be produced for each sector of the economy including the financial corporations sector.

#### Rest-of-the-World Account

8.81 The total economy consists of units that are resident in the domestic economy. These units are grouped in five institutional sectors, which may undertake transactions with nonresidents. The 2008 SNA records transactions between residents and nonresidents as if these nonresidents constitute a separate, single institutional sector, which is referred to as the rest of the world.

8.82 Flows in all the accounts may have entries in the rest-of-the-world account, with the exception of the production and generation of income accounts. The production account records domestic production and the generation of income account records how the income generated from production is distributed among domestic sectors. Flows to the rest of the world are shown as uses of the rest of the world and flows from the rest of the world are shown as resources.

8.83 The rest-of-the-world account records the same transactions as the balance of payments statement; it is a mirror of the balance of payments. The balance of payments records transactions from the perspective of residents, while the rest-of-the-world account records these transactions from the perspective of the rest of the world.

### The Structure of the 2008 SNA Accounts

8.84 The 2008 SNA accounts allow for the recording of all flows and stocks relating to economic activity. The accounts provide a consistent and integrated framework that covers all institutional sectors of the economy and the economic relationships of an economy with the rest of the world.

8.85 The accounts comprising the sequence of accounts are grouped into three main categories: current accounts, accumulation accounts, and balance sheets. Figure 8.1 in the main text of this chapter presents an illustration of the inter-relationships of the accounts.

8.86 The transactions recorded in the current accounts often entail counterpart entries in the capital and financial accounts. Every resource in the current account corresponds to an increase in economic value available to the owning unit, and every use corresponds to a decrease. Thus, resources increase the net worth of a unit, and uses decrease net worth. The final balancing item of the current accounts—saving—is that part of income that is not used for final consumption expenditure. Saving is the starting point for the accumulation accounts and, together with net capital transfers, represents resources available for financing capital formation, adding to financial assets, and reducing liabilities. Accumulation accounts cover changes in stocks of nonfinancial and financial assets, in liabilities, and in net worth caused by transactions and other events. Balance sheets cover the stock of nonfinancial and financial assets and liabilities, as well as the net worth of institutional sectors and the economy.

#### Current Accounts

8.87 The goods and services account is not an account in the sequence of accounts, but reflects transactions in goods and services that are undertaken by institutional units. This account shows how total supply of goods and services from output (domestic production) and imports is used for capital formation, intermediate consumption, final consumption, and exports. It is possible to rearrange the account to show how GDP can also be calculated from the expenditure side as the sum of final consumption expenditures, capital formation, and net exports (exports less imports).

8.88 The current accounts comprise the production account, the distribution of income account (primary and secondary), and the use of income account. These are described below.

8.89 Production account. This account records the results of production (output) and the use of goods and services in that production (intermediate consumption). The wear and tear of capital that is used in that production is not recorded as intermediate consumption, but is presented as consumption of fixed capital. The balancing item for the production account of a given sector or unit is value added. The balancing item for the total economy is gross domestic product or net domestic product. Thus, the balancing item for the total economy is derived as the sum of value added plus the sum of taxes on products less subsidies on products. Taxes on products are not allocated to individual units or sectors, but to the economy as a whole.

Table 8A.1

Full Sequence of Accounts and Balance Sheets for the Total Economy1

Abridged version of Tables, 2.13 and 2.14 of the 2008 SNA, with financial assets, liabilities and their changes replaced by numbers based on Tables 8.1, 8.2, and 8.4.

This includes compensation of employees (net, from rest of the world), which is equal to 4 and could be obtained as the difference between compensation of employees in the allocation of primary income account (1,154) and generation of income account (1,150).

Property income (net, from rest of the world) is equal to 6 and could be obtained as the difference of property income on the resources side (397) and uses side (391) of the allocation of primary income account.

Current transfers (net, from rest of the world) is equal to −38 and could be obtained as the difference of current transfers on the resources side (1,174) and uses side (1,212) of the secondary distribution of income account.

Final consumption expenditure for the total economy (1,399) is the sum of final consumption expenditures of households (1,015), nonprofit institutions serving households (32), and general government (352).