I. Introduction
10.1 Drawing on the definitions and concepts set out previously, this chapter explains how financial soundness indicators (FSIs) for the nonfinancial sectors are to be calculated and interpreted.
10.2 Nonfinancial sectors comprise nonfinancial corporations (NFCs), households, and real estate markets. This chapter covers these FSIs: consolidation basis, data sources, definition, analytical interpretation, and the calculation of their underlying series, as well as potential issues that compilers should be aware of.
10.3 In general, the accounting principles underlying source data calculation are similar to those recommended for deposit takers (DTs) and OFCs. NFCs and households are defined in Chapter 2. This chapter elaborates on real estate markets and prices. The accounting framework and accounting principles for FSIs are discussed in Chapter 4. The sectoral financial statements and memorandum items for NFCs and households, from which underlying series are derived, are covered in Chapter 5. Annex 10.1 summarizes the source data for compilers of recommended FSIs for the nonfinancial sectors.
II. Consolidation Basis
10.4 As described in Chapter 6, data for NFCs and households should be compiled on a resident-based approach, that is, data cover only resident institutional units without intra-group consolidation adjustments. This is because underlying data are mainly obtained from national accounts or other macroeconomic data sets, which are compiled based on the concept of institutional unit and do not consolidate intra-group positions and flows.1
III. Calculation of Financial Soundness Indicators for NFCs
10.5 As with the deposit-taking sector, most FSIs for NFCs are calculated by comparing two underlying series to produce a ratio. For some FSIs, when one or both of the underlying series can be defined in alternative ways, these alternatives are explained.
10.6 Unlike in the case of DTs and OFCs, FSI compilers normally do not have access to accounting records of individual nonfinancial corporations. Therefore, it is often not possible to construct sectoral balance sheets and income statements aggregating the financial statements of all institutional units of the sector, as is the case for DTs. The series needed to calculate these FSIs can be drawn from national accounts-based data; or from specific surveys covering a representative sample of the sector (see Box 10.1). This restriction can pose additional challenges in terms of frequency and timeliness.
10.7 The Guide recommends compiling seven FSIs for the NFCs sector (see Table 1.1). These FSIs focus on NFCs’ solvency, leverage (or gearing), profitability, and debt-servicing capacity. These indicators are useful in predicting corporate distress or failure. NFCs’ poor financial performance will impair their capacity to service their obligations. To the extent that these NFC draw funding from DT’s, NFCs’ distress or failure may negatively impact DTs’ asset quality.
10.8 Unless otherwise stated, all the line references in this section refer to Table 5.5 Nonfinancial Corporations. As already stated, the data to be used to calculate FSIs for this sector are not adjusted to eliminate intra-group positions and flows among NFCs in the reporting populations, but in some countries where these series are compiled from counterpart may be on a consolidated basis, which should be noted in metadata.
European Central Balance Sheet Data Offices
The European Central Balance Sheet Data Offices provides an example of NFCs’ data compilation based on individual accounting records, which eventually could be used to construct FSIs for NFCs.
The national data offices collect, store, and disseminate descriptive and accounting data of NFCs. Data collection is based on a sample of the corporations, which is afterwards expanded for the estimates of the whole population. Two-thirds of the data are collected on a mandatory basis, while the remaining one-third is obtained on a voluntary basis. Periodicity and timeliness vary across countries. The most frequent periodicity (over 60 percent of data sources) is annual, and the rest is collected on a quarterly basis. Almost three-quarters of the products and services are made available within one year of the reference period.
The Bank for Accounts of Companies Harmonized (BACH) database : https://www.bach.banque-france.fr/?lang=en collected by the European Committee of Central Balance Sheet Data Offices contains aggregated and relatively harmonized accounting data of NFCs for 12 European countries. These include 41 balance sheet items, 22 income statements items, and several economic and financial ratios collected at a national level. They complement national accounts data with a detail of subsets of institutional sectors.
Total Debt to Equity
10.9 The FSI for NFCs total debt to equity measures corporate leverage, that is, the extent to which activities are financed through liabilities other than own funds. Given the need to make interest and principal payments on debt, high corporate leverage increases the vulnerability of corporate entities in the event of economic, interest rate, or other financial market shocks and may impair their repayment capacity. More generally, the extent of corporate leverage—together with the volatility of the environment in which corporations operate—could be important indicators of the probability of corporate financial distress, as illustrated in Box 10.2.
10.10 This FSI is calculated by using debt (line 26 in Table 5.5) as the numerator, and capital and reserves (line 29 in Table 5.5) as the denominator. Debt is defined similarly as for DTs as the outstanding amount of those actual current and non-contingent liabilities (paragraph 5.69). Capital and reserves is the accounting concept defined in paragraph 5.144. It is assumed that capital is denominated in domestic currency.
10.11 As discussed, data should be compiled on a resident-based approach. NFCs’ debt and capital and reserves can be drawn from national accounts-based data, more specifically from flow of funds accounts or similar frameworks Alternatively, they can be obtained from data collected from a representative sample of NFCs’ financial statements.
10.12 Equity investments in associates and subsidiaries (and reverse investments) are to be recorded in the investor’s balance sheet on the basis of the investor’s proportionate share in the capital and reserves of the associate and subsidiary, and not using the market value of the traded equity.
External Debt to Equity
10.13 The NFCs external debt to equity FSI is a measure of NFCs’ exposure to nonresident creditors. This indicator is useful for macroprudential analysis and systemic risk monitoring, as there are potential risks associated with a high exposure to nonresidents—usually denominated in foreign currency. As is well-documented, this funding has shown significant volatility, especially for emerging economies.2
10.14 This FSI is the ratio of total debt to nonresidents (line 32 in Table 5.5) to capital and reserves (line 29 in Table 5.5). Data on NFCs’ external debt are readily available from external debt statistics if they are compiled with full-sector breakdown in accordance with the External Debt Statistics: Guide for Compilers and Users (2013). NFCs’ external debt can also be sourced from the International Investment Position (IIP). If sourced from IIP, NFCs’ external debt can be estimated by taking liabilities in the form of (a) direct investment intercompany lending, (b) debt securities under “portfolio investment,” plus all items, except other equity, under “other investment” for “other sectors.” Due to its classification in the IIP, to identify liabilities of NFCs, the items of portfolio investment and other investment should deduct data for OFCs. Where available, NFC’s external debt can be obtained from the sectoral accounts and balance sheet statistics. The extent to which the resulting data would be consistent with the concepts in the Guide would require further consideration.
Nonfinancial Corporations Debt to Equity in the United States
The graph, sourced from the United States Federal Reserve Bank of St. Louis, shows how a sharp increase in the ratio of NFCs’ debt to equity (measured as NFCs’ credit debt as a percentage of the market value of corporate equity) preceded recession periods in the United States.
Nonfinancial Corporate Business; Credit Market Debt as a Percentage of the Market Value of Corporate Equities
Source: Board of Governors of the US Federal Reserve System; Federal Reserve Bank of St. Louis.Note: Shaded areas indicate US recessions.Nonfinancial Corporate Business; Credit Market Debt as a Percentage of the Market Value of Corporate Equities
Source: Board of Governors of the US Federal Reserve System; Federal Reserve Bank of St. Louis.Note: Shaded areas indicate US recessions.Nonfinancial Corporate Business; Credit Market Debt as a Percentage of the Market Value of Corporate Equities
Source: Board of Governors of the US Federal Reserve System; Federal Reserve Bank of St. Louis.Note: Shaded areas indicate US recessions.10.15 If the FSI and external statistics compilation fall under the purview of different agencies, FSI compilers are encouraged to obtain the data from compilers of external sector statistics through a well-established data sharing arrangement.
10.16 Issues for compilers regarding NFCs’ capital and reserves are discussed in the paragraphs 10.09–10.12. As data are compiled on a resident-based approach, resident parent NFCs’ debt liabilities to any nonresident subsidiaries should be included. Issues for compilers regarding equity investments in associates and subsidiaries (and reverse investments) and goodwill, are also discussed in paragraphs 10.09–10.12.3
Foreign Currency Debt to Equity
10.17 The FSI for NFCs foreign currency debt to equity provides an indication of NFCs’ total debt in foreign currency to both residents and nonresidents, compared with their capital. It is intended to gauge NFCs’ exposure to potential foreign currency risk. High levels of foreign currency debt increase NFCs’ foreign currency risk and, if the corporations’ foreign currency debt is not offset by foreign currency receipts from exports or other sources, they may face difficul-ties in case of a sharp depreciation of the domestic currency. This could be partially or totally ameliorated if the foreign exchange risk is hedged.
10.18 This FSI is calculated by using total debt in foreign currency (line 33 in Table 5.5) as the numerator, and capital and reserves (line 29 in Table 5.5) as the denominator. Debt is defined in paragraph 5.69, capital and reserves is the accounting concept defined in paragraph 5.144, and foreign currency is defined in paragraph 5.37.
10.19 In cases where most debt to nonresidents is in foreign currency, external debt statistics and IIP may provide useful source data. Data on foreign currency debt vis-à-vis the resident central bank, DTs, and OFCs are available from the IMF’s standardized report forms (SRFs) for monetary and financial statistics. SRFs provide data broken down by type of finan-cial instrument, currency of denomination (domestic and foreign), and counterpart sector. If SRFs for the central bank, other depository corporations, and OFCs are available, NFCs’ foreign currency debt to these financial corporations can be approximated as the sum of these financial corporations’ claims on NFCs in the form of loans, debt securities, and other accounts receivable denominated in foreign currency.4
10.20 Issues for compilers regarding NFCs’ capital and reserves are discussed in paragraphs 10.09–10.12.
10.21 Foreign currency debt among NFCs in the reporting population that are part of the same group are included, as FSIs for NFCs are compiled on a resident-based approach. Regarding equity investments in associates and subsidiaries (and reverse investments) as well as goodwill, issues for compilers are discussed in paragraphs 10.09–10.12.
Total Debt to GDP
10.22 The FSI for NFCs total debt to GDP is intended to measure the overall level of NFCs’ indebtedness (both in domestic and foreign currency, to both residents and nonresidents) compared to the size of the economy. It should be analyzed together with other FSIs on NFCs’ debt (see the previous three FSIs for NFCs). A high level of corporate debt in relation to gross domestic product (GDP) is a signal of increased vulnerability of corporations to shocks, which may impair their repayment capacity. This FSI is one of several measures of the NFCs’ level of debt, which is also used to determine NFCs’ debt sustainability (see Box 10.3 for an application of this ratio).5
10.23 This FSI is calculated by using debt (line 26 in Table 5.5) as the numerator, and annual GDP as the denominator. Debt data should be end-period stock and are defined in paragraph 5.69.
10.24 Issues on source data for total debt are discussed in paragraphs 10.09–10.12. If data on total debt are obtained from a sample of financial statements, the results must be extrapolated to estimate the value for the whole sector. GDP data are available from national accounts sources. It should be noted that both underlying data series for this indicator already exist for compiling other FSIs on debt— e.g., “total debt” is the numerator for compiling total debt to equity for NFCs, while GDP is the denominator for compiling several FSIs for OFCs.
Nonfinancial Corporations Debt to GDP by Instrument
This figure is presented in the IMF’s Regional Economic Outlook: Western Hemisphere (April 2016). It illustrates NFCs’ indebtedness using the total debt to GDP ratio split by type of financial instrument, which makes it even more useful for analyzing corporate solvency risks.
As loans account for more than two-thirds of total debt in most of these countries, the focus is on ensuring the adequacy of buffers in the banking system, in terms of both provisions and capital.
Nonfinancial Corporate Debt by Instrument
(Percent of GDP, 2014)
Sources: Bank for International Settlements; Dealogic; IMF, International Financial Statistics database; and IMF staff calculations.Nonfinancial Corporate Debt by Instrument
(Percent of GDP, 2014)
Sources: Bank for International Settlements; Dealogic; IMF, International Financial Statistics database; and IMF staff calculations.Nonfinancial Corporate Debt by Instrument
(Percent of GDP, 2014)
Sources: Bank for International Settlements; Dealogic; IMF, International Financial Statistics database; and IMF staff calculations.10.25 GDP data should be obtained from national accounts source. Regardless of which frequency is used to compile this FSI, the annualized GDP should be used as the denominator.
Return on Equity (ROE)
10.26 The FSI for NFCs return on equity is commonly used to capture NFCs’ efficiency in using capital. It also indicates NFC’s ability to internally generate capital through retained earnings and to potentially attract new equity investment. Profitability is a critical determinant of corporate strength, affecting capital growth, the ability to withstand adverse events and, ultimately, repayment capacity. Sharp declines in corporate sector profitability, for example, as a result of economic deceleration, may serve as a leading indicator of NFCs’ financial difficulties and a potential credit risk exposure that will affect the financial corporations’ asset quality. However, account should be taken of cyclical movements in corporate sector profitability and of market structure—that is, industry characteristics, competitive environment, and pricing flexibility. The diversified types of businesses within the NFC sector mean that the actual performance of subsectors is likely to vary widely from the overall NFCs’ ROE. It is useful to examine the components of ROE to determine whether the change in NFC’s positions is driven by leverage or net income.
10.27 The FSI is calculated by using net income after taxes (line 9 in Table 5.5) as the numerator and the average value of capital and reserves (line 29 in Table 5.5) over the same period as the denominator. As with DTs, net income after taxes is used in the calculation of this FSI because, in addition to be an indicator of profitability, ROE is a measure of return on shareholders’ investments in NFCs—that is, shareholders’ interest is on income after taxes. Net income is described in paragraphs 5.136–5.138. Capital and reserves is the accounting concept defined in paragraph 5.144.
10.28 Data can be drawn from national accounts-based data or, if available, from central balance sheet offices. For the large entities, data might be drawn from published corporate financial statements and aggregated to get both the numerator and the denominator for this FSI. However, the extent to which the resulting data would be consistent with the concepts in the Guide would require further consideration.
10.29 Regarding capital, issues for compilers— including the definitions of capital—are discussed in paragraphs 10.09–10.12. As data are collected on a resident-based approach, transactions and positions among NFCs in the reporting population that are part of the same group are not eliminated.
10.30 Being a ratio of a flow to a stock, the same considerations as for the case of DTs and insurance corporations for a similar indicator apply here. That is, net income should be annualized and compilers should report the income annualization method in the metadata. At a minimum, the denominator can be calculated by taking the average of the beginning and end-period positions (e.g., average of the beginning and the end of the reference quarter if this FSI is compiled on a quarterly basis), but compilers are encouraged to use the most frequent observations available in averaging the capital stocks.
Earnings to Interest and Principal Expenses
10.31 The FSI for NFCs earnings to interest and principal expenses measures NFCs’ capacity to cover their debt-service payments (interest and principal). It serves as an indicator of the risk that NFCs may not be able to make the required payments on their debts. The NFCs’ default on debt obligations will negatively affect the creditors’ asset quality and profitability. This FSI thus is potentially a leading indicator of deterioration in the DT sector as NPLs may increase in future if the NFC sector has a low ratio of earnings to interest and principle expenses.
10.32 This FSI is calculated by using earnings (net income) before interest and tax (EBIT) (line 31 in Table 5.5) as the numerator, and debt-service payments (line 34 in Table 5.5) over the same period as the denominator. EBIT is a commonly used measure of earnings for the calculation of debt-service coverage. EBIT and interest receivable from other NFCs are defined in paragraph 5.146, and debt-service payments are defined in paragraph 5.149.
10.33 Data on earnings and debt-service payments may not be available from national accounts and, therefore, they should be obtained from other sources. Potential data sources include external debt statistics, which requires collection of data on debt service payments on external debt. Data on domestic debt-service payments need to be additionally collected. For the larger entities, data might be drawn from published corporate financial statements and aggregated to calculate both the numerator and the denominator for this FSI. Another source might be data stored by central balance sheet data offices, which will usually have a flow-off-funds-type framework. Specific survey data may be required.
10.34 Debt-service coverage, and particularly interest coverage, is a concept used in the analysis of corporate accounts. However, the extent to which the resulting data would be consistent with the concepts in the Guide would require consideration. Debt-service payments among NFCs in the reporting population, regardless of whether they are part of the same group or not, are included in the denominator. The numerator includes interest receivable (including those among NFCs in the reporting population that are part of the same group) from other NFCs. Therefore, the numerator and denominator have the same coverage.
10.35 The underlying flow data used to calculate this FSI should be reported on a cumulative basis—that is, data should be accumulated from the beginning of the reference year until the end of the reporting period.
Earnings to Interest Expenses
10.36 The FSI for NFCs earnings to interest expenses measures NFCs’ capacity to cover interest payments, providing insights into the risk that NFCs may not be able to make the required interest payments. Lack of capacity to pay interest may constitute an early warning that NFCs might fail to pay overall debt obligations. As mentioned earlier, NFCs’ default on debt services will lead to a deterioration of the lending financial corporations’ asset quality and profitability (see Box 10.4).
10.37 In some cases, it is difficult to collect data on principal payments, while data on interest payments are generally available from accounting records.6 For this reason, the FSI earnings to interest expenses is an alternative to report a debt-service ratio in case data on principal payments are not available. If data for both principal and interest payments are available, both FSIs earnings to interest expenses and earnings to interest and principal expenses should be compiled and disseminated.
10.38 This FSI is calculated by using EBIT (line 31 in Table 5.5) as the numerator and interest expenses (line 5 in Table 5.5) over the same period as the denominator.
10.39 Sources of data for both numerator and denominator are discussed in paragraphs 10.33–10.34.
10.40 Issues on reporting flow data and data definitions are the same as earnings to interest and principal expenses discussed earlier.
Earning to Interest Expenses
This figure is presented in the Regional Economic Outlook: Western Hemisphere (April 2014). It illustrates the capacity of Latin American NFCs to cover interest payments, using the median of the ratio of EBIT to interest expenses. As EBIT were three to four times higher than interest payments in most of these countries during the reference period, the rise in leverage did not appear to have compromised the debt-servicing capacity of the corporations in the sample. However, these ratios are prone to marked declines in the event of a pronounced economic downturn or rise in interest rates. Moreover, statistics for the median firm conceal vulnerabilities in the weaker tail of the sample.
LA5 (Brazil, Chile, Colombia, Mexico, and Peru ): Median Ratio of Earnings before Interest and Taxes to Interest Expenditure, 2003–13
Source: IMF, Regional Economic Outlook: Western Hemisphere, April 2014.LA5 (Brazil, Chile, Colombia, Mexico, and Peru ): Median Ratio of Earnings before Interest and Taxes to Interest Expenditure, 2003–13
Source: IMF, Regional Economic Outlook: Western Hemisphere, April 2014.LA5 (Brazil, Chile, Colombia, Mexico, and Peru ): Median Ratio of Earnings before Interest and Taxes to Interest Expenditure, 2003–13
Source: IMF, Regional Economic Outlook: Western Hemisphere, April 2014.Household Debt to GDP
The graph shows the steep increase in the ratio of household debt to GDP in the United States in the years prior to the financial crisis; as well as its steady decline afterwards, as households reversed their consumption pattern from previous years.
Household Debt to GDP
Source: IMF, Financial Soundness Indicators website.Household Debt to GDP
Source: IMF, Financial Soundness Indicators website.Household Debt to GDP
Source: IMF, Financial Soundness Indicators website.10.41 The denominator includes interest payments to other NFCs (including payments among NFCs in the reporting population that are part of the same group). The numerator includes interest receivable from other NFCs (including those among NFCs in the reporting population that are part of the same group). Therefore, the numerator and denominator have the same coverage. Regarding the calculation of earnings and data consolidation, issues for compilers are discussed in paragraphs 10.33–10.35.
IV. Calculation of Financial Soundness Indicators for Households
10.42 The analysis of the household sector balance sheet is also key for financial stability considerations. Sharply rising household debt, for example, could lead to distress in DTs with considerable household exposure. Economic activity and interest rate shocks may impact the ability of households to service their debt, as well as the value of their collateral. As with FSIs for NFCs, FSIs for households serve as leading indicators of the expected evolution of DTs’ asset quality.
10.43 The vulnerability of households may be assessed through the use of sectoral accounts, flow of funds, and other macroeconomic data. Indicators include the ratios of household debt to GDP, household debt to income, household debt service and principal payments to income, household debt to assets, and household debt to the value of collateral pledged. Household vulnerability on the asset side includes households’ exposure to equity and real estate price movements.
10.44 Unless otherwise stated, all the line references in this section are to Table 5.6 Households. Data for households are compiled on a resident-based approach and no consolidation adjustments are required as they are not applicable to the household sector. The simplified presentation of Table 5.6 is based on information sourced from national accounts, which is derived from sample surveys subject to response and reporting errors. Obtaining data on the household sector is difficult and therefore coordination with the agency compiling national accounts statistics is essential.
Household Debt to GDP
10.45 The FSI for household debt to GDP measures the overall level of household indebtedness (usually related to consumer loans and mortgages) as a share of GDP. As with the NFC sector, a high rate of growth and level of borrowing increases the vulnerability of households to economic and financial market shocks and may impair their repayment capacity (see Box 10.5).
10.46 This FSI is calculated by using household debt (line 19 in Table 5.6) as the numerator, and GDP as the denominator. Debt data should be end-period stock. Household debt is defined in paragraph 5.156.
10.47 Both the numerator and the denominator should be compiled using national accounts data, which provide a broader coverage of household debt and GDP. If data on household debt are not available from national accounts sources, data from the financial sector sources can be used—although in this case, it would cover only household debt to resident financial corporations. Data sources should be documented in the metadata.
10.48 Data for household debt comprise debt incurred by resident households of an economy only. Regardless of which frequency is used to compile this FSI, the annualized GDP should be used as the denominator.
Household Debt-Service and Principal Payments to Income
10.49 The FSI for household debt-service and principal payments to income measures the capacity of households to cover their debt payments (interest and principal). It can also be used as a leading indicator of consumer spending growth: a high debt-service ratio over a period of time might be a sign of slow growth of personal consumption in the period ahead.
10.50 This FSI is calculated by using household debt-service and principal payments (line 22 in Table 5.6) as the numerator, and gross disposable income (line 6 in Table 5.6) over the same period as the denominator. Household debt-service payments are defined in paragraph 5.158, and gross disposable income is defined in paragraph 5.153.
10.51 Information on household disposable income should be available from national accounts sources. However, data on debt-service payments might not be available from national accounts sources and so additional data may need to be separately requested (see paragraph 5.150). Most likely, the household sector borrows from resident financial corporations, although some borrowing from abroad might exist, in which case there may be a need to capture cross-border borrowing activity. Additionally, households might obtain commercial or retail credit directly from NFCs, which in some economies could constitute an important part of household debt. The required data series on debt service could be included in household surveys. Alternatively, data from the resident financial sector can be used together with some assumptions about repayment schedules to estimate household debt service.7
10.52 Both the numerator and the denominator are flow data, which should be reported on a cumulative basis—that is, data should be accumulated from the beginning of the reference year until the end of the reporting period.
Household Debt to Income
10.53 The FSI for household debt to income is intended to assess the debt sustainability of the household sector, with a high or growing ratio signaling sector’s vulnerabilities. A high level of household debt coupled with inadequate capacity to service could cause a shock to the country’s financial sector. In this regard, this indicator should be analyzed together with the previous two FSIs for households.
10.54 This FSI is calculated using household debt (line 19 in Table 5.6) as the numerator and gross disposable income of households (line 6 in Table 5.6) over the same period as the denominator.
10.55 Information on household debt and disposable income should be available from national accounts sources (see paragraphs 5.153 and 5.156). If data on household debt are not available from national accounts sources, data from the financial sector source can be used—in this case covering only household debt to resident financial corporations.
10.56 Issues on both underlying series are the same as the previous two FSIs for households (paragraph 10.51).
10.57 Household debt should be measured as outstanding stock at the end of the reporting period, whereas the denominator is the households’ annual-ized gross disposable income. Compilers should report the income annualization choice in the metadata.
V. Real Estate Markets
10.58 For macroprudential analysis, it is highly desirable to have indexes of real estate prices because deposit takers (DTs) may have large exposures (both direct and indirect) to real estate and may be affected by volatile price movements. Moreover, real estate assets are a major component of private sector wealth, a determinant of private consumption and, consequently, of economic activity.
10.59 Sharp drops in real estate prices affect DTs negatively due mainly to the impact they have on the value of collateral, the increase in the real estate loan to value ratio, the negative wealth effect on debtors, and therefore on the quality of DTs’ loan portfolios. There is a well-documented relationship between real estate cycles and economic cycles, with rapid increases in real estate prices (bubbles) and excessive lending being early indicators of an impending financial crisis.8 During an upswing in real estate prices, real estate may be used as collateral for extension of credit for further purchases. However, once conditions begin to reverse, such exposure could lead to a mutually reinforcing downward spiral.
10.60 DTs’ exposure to real estate prices can arise through many channels: (1) ownership of real estate; (2) loans collateralized by real estate; (3) risk of prepayment; (4) holding of pass-through (or asset-backed) securities9 backed by real estate (mortgage) loans; or (5) exposure to households and corporations that can be affected by changes in the servicing costs of real estate related borrowing or price movements in real estate markets.
10.61 The reasons why real estate prices are potentially volatile are varied. Real estate markets are illiquid, with final prices negotiated individually between the contracting parties and with high transactions costs. Supply is inelastic in the short-term owing to the time needed to plan projects and complete construction, making real estate markets cyclical.10 Development is often subject to many legal or other restrictions, such as a shortage of urban land that can be developed. Under these conditions, the impact on prices of changes in demand is exacerbated. While international capital flows into or out of real estate can rapidly and unpredictably affect market sales and prices, price volatility is also endogenously induced through the provision and cost of domestic credit.
Measuring Real Estate Prices
10.62 International guidance in constructing representative real estate price indices is relatively limited. The first comprehensive overview of conceptual and practical issues related to the compilation of price indices for residential properties is available in the Handbook on Residential Property Prices Indices (RPPIs), published in 2013.11 And in 2019, the IMF will be issuing a Practical Guide on the Compilation of the RPPI. Methodological guidance on commercial property prices was at a developmental stage at the time the Guide was published.
10.63 Contrary to a general price index, where prices for identical goods and services can be observed over time, real estate markets are highly heterogeneous (both within and across countries), with properties having unique locations and structural characteristics. Furthermore, prices can only be observed sporadically—when properties are transacted. Consequently, the construction of a real estate price index is substantially more difficult than the construction of other price indexes based on a matched model methodology because:
Because dwellings are not homogenous, there is normally no uniform market price for real estate.
Diversity and lack of standardization result in the need to gather a wide range of data to compile indices to represent various market segments, with associated challenges to securing access to suitable data and high technical sophistication requirements.
Representative real estate prices in residential and commercial markets can be hard to measure accurately given that there may be disparate prices for apparently similar properties, and prices may be volatile.
Transactions of the same dwelling are infrequent.
Experience has shown that there can be particular difficulties in acquiring representative source data for measuring commercial real estate prices across the economy.
10.64 When developing real estate price indices, compilers should be aware of a number of factors: (1) the wide range of differences among properties, leading to difficulties in identifying “a standard real estate unit”; (2) the mix of transactions by type, complicating the construction of weights to use in indices; and (3) different methods of compiling real estate price indices.
10.65 To capture changes in real estate price trends, the Guide advocates, at a minimum, quarterly compilation of data. Metadata describing in detail the content and coverage of—and the conceptual approach underlying—any price index disseminated is essential.
Residential property price indices
10.66 In the case of residential property, the objective for compilers is to construct a constant quality residential property price index (RPPI) that can control for differences in the characteristics of the properties sold over time. The goal is for changes in the RPPI to measure only price changes in the real estate market. Typically, the most important characteristics that need to be accounted for include (1) the location of the property; (2) the property type (e.g., detached house or apartment); (3) the size of the property (structure or plot); (4) the age of the structure; (5) the materials used in the construction; and (6) any other price determining characteristics.12
10.67 There are several methods to calculate RPPIs, all described in detail in Eurostat’s Handbook on Residential Property Prices Indices, namely: (1) simple mean or median indices; (2) stratification or mix adjustment methods; (3) hedonic regression methods; (4) repeat sales methods; and (5) appraisal-based methods. Since this is an area beyond the scope of this Guide, it is sufficient to enumerate them and to make FSI compilers aware of the complexities involved in the calculation of RPPIs.
10.68 FSI compilers rely on other agencies or data providers for the source data used in producing the FSI measuring residential property prices. The quality, coverage, and detail of data will, to a very large extent, determine what RPPI might be used and, ultimately, the quality of the FSI. Ideally, the index should cover a large number of transactions nationally rather than just a subset (say for the capital city or only mortgage funded transactions); reflect actual transaction prices; and be timely, accurate, and continuously available over time.
Commercial property price indices
10.69 The principles described for RPPIs also apply to commercial real estate, but with additional complexities. Commercial real estate comprises four very different types of properties: offices, retail, industrial, and residential (if developed for commercial purposes). Within these four categories, properties are heterogeneous and transactions irregular, hindering comparisons of average transaction prices for a fixed-quality bundle of properties over time. Even where repeat transactions can be used, the population of properties sold more than once in the period of the index can be very limited and unrepresentative of the total population of commercial properties.
10.70 For retail property, value depends heavily on the profits of the occupant’s business, and therefore it will fluctuate with the economic cycle. Another complicating factor in compiling a CPPI is that statistical reporting systems often do not effectively pick up the relatively small number of commercial transactions— as they may involve privately negotiated sales—and the changing patterns of new construction. Rather, experience suggests that commercial real estate indices tend to be based on localities, such as big cities, where there are specific concentrations of properties available commercially. Consequently, the compiled CPPI may be unrepresentative of the whole economy.
10.71 Facilitating the process of compiling price indexes for commercial real estate is the fact that commercial real estate can be characterized as a commodity consisting of square meters of commercial space for which rental or use values can be estimated. Rental rates are often expressed in terms of the annual cost per unit of space, most commonly per square meter. Such measures can also be used for purposes of international comparisons of rental costs.
10.72 Two main types approaches have been developed for constructing CPPIs: (1) appraisal-based and (2) transaction-based indices.13 Beyond methodological limitations, compilers of CPPIs face data availability problems. Data on commercial real estate are sparse and sometimes not available for some types of properties, especially for industrial property, and the mix of transactions can differ greatly over time. Currently, most price index series for commercial real estate are provided by private sector organizations. Indices disseminated by private sources may not disclose the methodology used for their calculation, hampering comparison between data sources. There is also possible bias if the private sector organization only covers certain segments of the market.
Financial Soundness Indicators for Real Estate Markets
10.73 The four FSIs for real estate markets are (1) residential real estate prices (a core FSI), (2) commercial real estate prices, (3) residential real estate loans to total gross loans, and (4) commercial real estate loans to total gross loans.
Residential real estate prices
10.74 The FSI residential real estate prices provides a metric to gauge the exposure of DTs in case of rapid increases in residential real estate prices, which can be followed by a sharp decline when credit conditions deteriorate (see Box 10.6).
10.75 This core FSI, which covers residential real estate price indices, is calculated as the percentage change in the index during the 12 months prior to the reporting period.
10.76 FSI compilers must rely on source data from third parties for this indicator. They usually do not determine the way RPPIs are estimated, as they use indices produced by other agencies and available to the public. Ideally, the index should have a broad coverage in terms of geography (country-wide, or the largest cities in the country), property type (detached homes, townhomes, apartments), and price-range coverage.
10.77 FSI compilers should be aware of the advantages and disadvantages of the four main methods for calculating RPPIs;14 and ensure that comparable data are collected, stored, and compiled.
10.78 If more than one RPPI is disseminated, compilers of FSIs should acknowledge possible trade-offs between frequency, timeliness, accuracy, and coverage of the selected real estate price index. The coverage of the index should be as broad as possible, and its frequency should be at least quarterly. If a general aggregated price index with a comprehensive geographical coverage is not available, then FSI compilers should decide which of the narrower indices is the most representative of the residential real estate market and use it for the calculation of this core indicator. Metadata on this indicator must be also disseminated, clearly explaining the data sources and compilation methods used.
Commercial real estate prices
10.79 The FSI commercial real estate prices provides a metric to gauge the exposure of DTs in case of rapid increases in commercial real estate prices (often fueled by expansionary monetary policy and capital inflows), which can be followed by a sharp decline in case of an economic downturn or when credit conditions deteriorate.
The Housing Bubble in the United States
The graph shows the evolution of the S&P Case-Shiller U.S. National Home Price Index, a generally accepted RPPI, in the run-up to the financial crisis of 2008.
Partly fuelled by a “loose” monetary policy, housing prices peaked in mid-2006 and reached levels 120 percent higher than 10 years prior. When the real estate bubble busted, prices dropped by almost 30 percent in the next six years.
The collapse of the housing market triggered a financial crisis in the United States that spread to the rest of the world. The U.S. government had to bail out the banking system through special loans and rescue packages.
S&P Case-Shiller U.S. National Home Price Index evolution
Source: S&P/Case-Shiller U.S. National Home Price Index.S&P Case-Shiller U.S. National Home Price Index evolution
Source: S&P/Case-Shiller U.S. National Home Price Index.S&P Case-Shiller U.S. National Home Price Index evolution
Source: S&P/Case-Shiller U.S. National Home Price Index.10.80 This FSI covers commercial real estate price indices. It is calculated as the percentage change in the commercial real estate price index during the 12 months prior to the reporting period.
10.81 For this indicator, FSI compilers must rely on indices produced by other public or private agencies. Shortcomings regarding geographical coverage and types of properties surveyed may negatively affect the quality of the index used. Contrary to the case of RPPIs, the main sources of data for CPPIs are often private sector organizations. This raises the issue about possible bias on the available data. Another source of information for CPPIs may be financial institutions active in lending to the commercial real estate market.
10.82 As explained earlier, the calculation of CPPIs involves the same difficulties as the estimation of RPPIs, and FSI compilers face similar issues as for the indicator on residential property prices. However, the differences are compounded by the fact that commercial real estate comprises four very different types of properties: offices, retail, industrial, and residential (if developed for commercial purposes), making it difficult to compare the prices.
10.83 The same considerations discussed for RPPIs regarding frequency, timeliness, and coverage of the indices apply also to CPPIs, and hence the need for extensive metadata on the CPPI used when calculating this indicator.
Residential real estate loans to total loans
10.84 The FSI residential real estate loans to total loans provides a metric to gauge the DTs’ exposure to the residential real estate market. Experience has shown that, in many instances, a real estate boom characterized by a rapid rise in real estate prices has been preceded or accompanied by a boom in mortgage lending.15 Following a subsequent tightening of these policies, and a collapse in market prices, there have been episodes of financial sector problems when debtors face difficulties meeting their payments. The drop in value of the residential real estate collateral worsens the situation.
10.85 This FSI is calculated by using residential real estate loans as the numerator (line 50 in Table 5.1), and gross loans (line 18.i in Table 5.1) as the denominator. Residential real estate loans are defined in paragraph 5.97 as all loans collateralized by real estate, while loans are defined in paragraphs 5.41–5.43. Household debt collateralized by real estate can be used alternatively as the numerator (line 23 in Table 5.6). While not all real estate lending to households is collateralized by residential real estate, such collateralized debt predominates.
10.86 The definition of this FSI requires not only data on residential real estate (mortgage) loans, but also data on all loans collateralized by residential real estate, regardless of the purpose of those loans. In many countries, loans collateralized by real estate may comprise a significant portion of credit to the household sector. National practices may differ on how these loans are classified. The required series are not available from the consolidated balance sheet of DTs but will need to be provided by DTs as supplementary memorandum series. Total loans can be sourced from the consolidated balance sheet of the DTs.
10.87 For cross-border consolidated data, data on residential real estate loans by subsidiaries abroad may need to be additionally requested, if not available from supervisory sources. The available information may need to be aggregated.
10.88 For data compiled on a domestic location consolidation basis, residential real estate loans may be available from monetary and financial statistics sources that provide an industrial classification of lending by type of economic activity. Otherwise, additional data may need to be separately requested.
10.89 The consistent application by DTs of a definition of residential real estate is central. This should include houses, apartments, and other dwellings (e.g., houseboats and mobile homes)—and any associated land—intended for occupancy by individual households. Furthermore, it is very important that all DTs follow the definition of residential real estate loans recommended by the Guide, namely not only residential real estate loans but also any other loan collateralized by residential real estate regardless of the purpose of those loans. Regarding total loans, issues for compilers are the same as for other core and additional FSIs for DTs where they are used as denominator, as discussed in Chapter 7.
Commercial real estate loans to total loans
10.90 The FSI commercial real estate loans to total loans provides a metric to gauge the DTs’ exposure to the commercial real estate market. Many of the same considerations described earlier for residential real estate apply for commercial real estate.
10.91 This FSI is calculated by using loans collateralized by commercial real estate, loans to construction companies, and loans to companies active in the development of real estate (line 51 of Table 5.1) as the numerator; and gross loans (line 18.i) as the denominator. Commercial real estate loans are defined in paragraph 5.98 and loans are defined in paragraphs 5.41–5.43.
10.92 The definition of this FSI requires not only data on loans for commercial real estate but also data on all loans collateralized by commercial real estate, plus loans to construction companies and corporations active in the development of real estate. These series are not available from the consolidated balance sheet of DTs but will need to be provided by DTs as supplementary memorandum series.
10.93 For cross-border consolidated data, data on commercial real estate loans by subsidiaries abroad may need to be additionally requested, if not available from supervisory sources. The available information may need to be aggregated.
10.94 For data compiled on a domestic location consolidation basis, commercial real estate loans may be available from monetary and financial statistics sources that provide an industrial classification of lending by type of economic activity. If so, lending among resident DTs that are part of the same group should be deducted. Otherwise, additional data may need to be separately requested.
10.95 As with residential real estate loans, the consistent application by DTs of a definition of what constitutes commercial real estate lending is central. Commercial real estate lending among DTs in the reporting population that are part of the same group is deducted. Regarding total loans, issues for compilers are the same as for other core and additional FSIs for DTs where they are used as denominator, as discussed in paragraph 7.38.
ANNEX 10.1. Summary of Financial Soundness Indicators for Nonfinancial Sectors
| Definition | Source Data | Issues for Compilers |
|---|---|---|
| Nonfinancial Corporations | ||
| Total Debt to Equity | ||
| Nonfinancial corporations’ (NFCs’) total debt as a percentage of capital and reserves. | - National accounts. | - Goodwill should not be deducted from capital and reserves. |
| External Debt to Equity | ||
| NFCs’ debt liabilities to nonresidents as a percentage of capital and reserves. | - External debt statistics. - International investment position. - National accounts. | - NFCs’ debt liabilities to nonresident subsidiaries are included in external debt. |
| Foreign Currency Debt to Equity | ||
| NFCs’ debt in foreign currency as a percentage of capital and reserves. | - External debt statistics. - International investment position. - National accounts. - Standardized report forms (foreigncurrency-denominated domestic debt). | - Issues on capital and reserves are discussed in Total debt to equity. |
| Total Debt to GDP | ||
| NFCs’ total debt as a percentage of GDP. | - National accounts. - For large corporations, data might be sourced from published financial statements. | - Denominator is annualized GDP. |
| Return on Equity | ||
| NFCs’ net income after taxes as a percentage of average capital and reserves. | - National accounts. - For large corporations, data might be sourced from published financial statements. | - Indicator is a ratio of a flow divided by a stock. Numerator is annualized net income after taxes. Denominator is average capital and reserves over the same period. |
| Earnings to Interest and Principal Expenses | ||
| NFCs’ earnings before interest and taxes as a percentage of debt service payments. | - Data might need to be obtained from sources other than national accounts. - For large corporations, data might be sourced from published financial statements. | - Interest payments among NFCs are included in both the numerator and denominator. |
| Earnings to Interest Expenses | ||
| NFCs’ earnings before interest and taxes as a percentage of interest service payments. | - Data might need to be obtained from sources other than national accounts. - For large corporations, data might be sourced from published financial statements. | - Interest payments among NFCs are included in both the numerator and denominator. |
| Households | ||
| Debt to GDP | ||
| Total household debt as a percentage of GDP. | - National accounts. - Monetary statistics (debt to resident financial corporations only) | - It covers only debt of resident households. - Annualized GDP should be used. |
| Debt-service and Principal Payments to Household Gross Disposable Income | ||
| Household debt-service and principal payments as a percentage of household disposable income. | - National accounts. - Financial corporations or their regulatory agencies. | - Coordination with national statistical offces and financial corporation regulatory agencies is essential to obtain source data. |
| Debt to Household Gross Disposable Income | ||
| Total household debt as a percentage of household gross disposable income. | - National accounts. -Financial corporations or their regulatory agencies. | - Annual gross disposable income covering the last 12 months ending in the reporting period. - Coordination with national statistical offces and financial corporations’ regulatory agencies is essential to obtain source data. |
| Real Estate | ||
| Residential Real Estate Prices | ||
| Twelve-month percentage change in residential property price index. | - Offcial statistics. - Real estate agents. - Financial institutions active in lending to real estate market. | - Compilers usually rely on indices produced by other agencies. - Coverage of the index might not be suffciently broad. |
| Commercial Real Estate Prices | ||
| Twelve-month percentage change in commercial property price index. | - Private sector organizations. - Financial institutions active in lending to commercial real estate market | - Data availability. - Heterogeneity of types of commercial real estate properties hindering the comparability of prices. |
| Residential Real Estate Loans to Total Loans | ||
| Deposit takers’ (DTs’) residential real estate loans as a percentage of their total loans. | - Supervisory data, when available. - DTs’ internal records. | - Consistent application by DTs of a definition of residential real estate loans. |
| Commercial Real Estate Loans to Total Loans | ||
| The sum of DTs’ loans collateralized by commercial real estate plus loans to construction companies and to companies active in the development of real estate, as a percentage of DTs’ total loans. | - Supervisory data, when available. DTs’ internal records. | - Consistent application by DTs of a definition of commercial real estate loans. |
| Definition | Source Data | Issues for Compilers |
|---|---|---|
| Nonfinancial Corporations | ||
| Total Debt to Equity | ||
| Nonfinancial corporations’ (NFCs’) total debt as a percentage of capital and reserves. | - National accounts. | - Goodwill should not be deducted from capital and reserves. |
| External Debt to Equity | ||
| NFCs’ debt liabilities to nonresidents as a percentage of capital and reserves. | - External debt statistics. - International investment position. - National accounts. | - NFCs’ debt liabilities to nonresident subsidiaries are included in external debt. |
| Foreign Currency Debt to Equity | ||
| NFCs’ debt in foreign currency as a percentage of capital and reserves. | - External debt statistics. - International investment position. - National accounts. - Standardized report forms (foreigncurrency-denominated domestic debt). | - Issues on capital and reserves are discussed in Total debt to equity. |
| Total Debt to GDP | ||
| NFCs’ total debt as a percentage of GDP. | - National accounts. - For large corporations, data might be sourced from published financial statements. | - Denominator is annualized GDP. |
| Return on Equity | ||
| NFCs’ net income after taxes as a percentage of average capital and reserves. | - National accounts. - For large corporations, data might be sourced from published financial statements. | - Indicator is a ratio of a flow divided by a stock. Numerator is annualized net income after taxes. Denominator is average capital and reserves over the same period. |
| Earnings to Interest and Principal Expenses | ||
| NFCs’ earnings before interest and taxes as a percentage of debt service payments. | - Data might need to be obtained from sources other than national accounts. - For large corporations, data might be sourced from published financial statements. | - Interest payments among NFCs are included in both the numerator and denominator. |
| Earnings to Interest Expenses | ||
| NFCs’ earnings before interest and taxes as a percentage of interest service payments. | - Data might need to be obtained from sources other than national accounts. - For large corporations, data might be sourced from published financial statements. | - Interest payments among NFCs are included in both the numerator and denominator. |
| Households | ||
| Debt to GDP | ||
| Total household debt as a percentage of GDP. | - National accounts. - Monetary statistics (debt to resident financial corporations only) | - It covers only debt of resident households. - Annualized GDP should be used. |
| Debt-service and Principal Payments to Household Gross Disposable Income | ||
| Household debt-service and principal payments as a percentage of household disposable income. | - National accounts. - Financial corporations or their regulatory agencies. | - Coordination with national statistical offces and financial corporation regulatory agencies is essential to obtain source data. |
| Debt to Household Gross Disposable Income | ||
| Total household debt as a percentage of household gross disposable income. | - National accounts. -Financial corporations or their regulatory agencies. | - Annual gross disposable income covering the last 12 months ending in the reporting period. - Coordination with national statistical offces and financial corporations’ regulatory agencies is essential to obtain source data. |
| Real Estate | ||
| Residential Real Estate Prices | ||
| Twelve-month percentage change in residential property price index. | - Offcial statistics. - Real estate agents. - Financial institutions active in lending to real estate market. | - Compilers usually rely on indices produced by other agencies. - Coverage of the index might not be suffciently broad. |
| Commercial Real Estate Prices | ||
| Twelve-month percentage change in commercial property price index. | - Private sector organizations. - Financial institutions active in lending to commercial real estate market | - Data availability. - Heterogeneity of types of commercial real estate properties hindering the comparability of prices. |
| Residential Real Estate Loans to Total Loans | ||
| Deposit takers’ (DTs’) residential real estate loans as a percentage of their total loans. | - Supervisory data, when available. - DTs’ internal records. | - Consistent application by DTs of a definition of residential real estate loans. |
| Commercial Real Estate Loans to Total Loans | ||
| The sum of DTs’ loans collateralized by commercial real estate plus loans to construction companies and to companies active in the development of real estate, as a percentage of DTs’ total loans. | - Supervisory data, when available. DTs’ internal records. | - Consistent application by DTs of a definition of commercial real estate loans. |
National compilation practices vary. In some jurisdictions, data on financial positions of nonfinancial sectors may be drawn from counterpart data, such as information on bank deposits and loans.
See, for instance, Bluedorn, John, Duttagupta, R., Guajardo, J. and Topalova, P. , 2013, Capital Flows Are Fickle: Anytime, Anywhere, IMF Working Paper 13/183, August.
If data on NFCs’ external debt are not available, but NFCs’ total debt and domestic debt (debt to residents) are available, NFCs’ external debt can be calculated as the difference between total debt and domestic debt.
If complete SRFs for some countries are not accessible from the IMF’s monetary statistics database, FSI compilers might obtain the data from monetary statistics compilers in their respective countries through a well-established data sharing arrangement. Another source of NFCs’ foreign currency debt is the already described balance sheet databases.
2008 SNA, paragraph 2.138 for a definition of GDP.
Interest payments are recorded as a separate item in the income statement as they accrue and, therefore, are usually available. Principal payments are recorded on the balance sheet as a reduction of outstanding liabilities not separately identified. Without a more detailed record keeping, this information may not be readily available to FSI compilers.
For instance, information can be obtained on possible repayment schedules based on remaining maturity data for loan debt and the pattern of credit card debt repayment, providing some rough estimates for debt service.
See, for instance, Gorton, Gary (2008), The Panic of 2007, paper prepared for the Federal Reserve Bank of Kansas City, Jackson Hole Conference, August.
Pass-through securities are securities backed by a pool of loans (prominently, mortgage loans), where the interest and principal payments on the loans are directly passed through to the holders of the securities. Defaults on the interest or principal of the loans, or prepayments of the loans in the pool, are absorbed by the holders of the securities.
See Mueller, Glenn (2002), “What Will the Next Real Estate Cycle Look Like?,” in Journal of Real Estate Portfolio Management, January, pp. 115–125.
The Handbook is a joint publication by Eurostat, the International Labour Organization (ILO), the IMF, the Organisation for Economic Co-operation and Development (OECD), the United Nations Economic Commission for Europe (UNECE), and the World Bank through the Inter-Secretariat Working Group on Price Statistics (IWGPS).
See Eurostat (2013), p. 25.
For a detailed treatment of these approaches, as well as the difficulties of compiling CPPIs, see Silver, Mick, 2013, “Understanding Commercial Property Price Indexes,” Worl d Economics, Volume 14, Number 3, July–September.
See Eurostat (2013).
See Gorton (2008).
