A. Macroeconomic Scenarios⁷

17. The solvency stress test for the SIs includes a baseline and an adverse scenario, covering a 3-year span, from 2019–21. The baseline corresponds to the April 2019 WEO projections, which projects a slowdown in real GDP growth rate relative to recent years and continued elevated sovereign yield levels.⁸ The adverse scenario is simulated using the Global Macrofinancial Model, a structural macroeconometric model of the world economy, disaggregated into forty national economies, documented in Vitek (2018).⁹ The simulation is based on a narrative that captures the risks discussed in the Risk Assessment Matrix (Appendix I), with attention paid to the main vulnerabilities of Italian banks and borrowers. The reference date for the stress test is end-2018.

18. The main feature of the adverse scenario is a reemergence of sovereign stresses in Italy, resulting in a sharp rise in risk premia. The mission’s work builds on the stress test results undertaken in the context of the recent EA FSAP and the EBA stress testing exercises. The two exercises were centered on the main external risks featured in the RAM. Accordingly, the FSAP focused its stress testing scenario on risks emanating domestically. The stresses are assumed to be primarily driven by Italy-specific factors, with limited spillovers to EA periphery countries.¹⁰ The scenario assumes a 180 bps rise in Italy’s long-term risk premium shock (the peak increase relative to the baseline), whereas safe-haven capital inflows reduce term premia by 70 bps in the EA core, and by 40 bps in other advanced economies.

19. The trigger for the scenario is assumed to be a loss of investor confidence. Concerns about rolling back of reforms, less market-friendly policies, and budget proposals which are expansionary relative to Italy’s European commitments could lead investors to leave the Italian bond market. The increase in sovereign risk implies a sharp drop in investment and output in Italy. Furthermore, the recession has an adverse impact on the primary fiscal balance due to lower tax receipts. Lower revenues and higher sovereign borrowing costs force a tightening of the fiscal position, which exacerbates the downturn. In addition, corporates are affected by the resulting higher borrowing rates, and sharply reduce investment expenditure. The resulting drop in aggregate demand exacerbates their balance sheet position through the impact on profits.

20. The scenario leads to a sharp decline in output and worsening of macrofinancial conditions (Tables 2 and 3 and Figure 5). The peak to trough decline in GDP throughout the horizon is 5.1 percent. This represents a fall of 7.2 percent relative to baseline by 2021, reflecting a 6.9 percent fall in consumption and a 22.3 percent fall in investment. The deviation of growth relative to the baseline is equivalent to a 2.1 standard deviations shock to GDP growth in the second year of a three-year horizon (U-shaped profile).¹¹

^{Table 2.}

Italy: Adverse Scenario Calibration Deviation from the baseline (in percentage points; unless specified otherwise)

Source: IMF staff calculations.

^{Table 2.}

Italy: Adverse Scenario Calibration Deviation from the baseline (in percentage points; unless specified otherwise)

	Adverse Scenario
	2019	2020	2021
Real GDP	-3.4	-6.9	-7.2
Policy interest rates	0.1	0.1	0.0
Short-term money market rate	0.8	1.1	0.9
Spread of short-term money market rate	0.7	1.0	0.9
Long-term government bond yield	1.3	1.8	1.5
Real effective exchange rate appreciation
(-)/depreciation (+)	8.0	12.7	12.3
Nominal exchange rate appreciation
(-)/depreciation (+)	7.7	11.5	10.4
Inflation rate (CPI)	-0.3	-1.2	-0.9
Unemployment rate	1.1	3.4	4.5
Memo:
Baseline Real GDP growth (in percent)	0.0	0.9	0.7
(Adverse) Real GDP growth (in percent)	-3.3	-2.7	0.3
Cumulative real GDP growth (from 2018)	-3.3	-6.0	-5.7
Severity: deviation of growth from baseline in multiples of standard deviation of historical growth volatility	1.9	2.1	1.6

Source: IMF staff calculations.

View Table

^{Table 2.}

Italy: Adverse Scenario Calibration Deviation from the baseline (in percentage points; unless specified otherwise)

	Adverse Scenario
	2019	2020	2021
Real GDP	-3.4	-6.9	-7.2
Policy interest rates	0.1	0.1	0.0
Short-term money market rate	0.8	1.1	0.9
Spread of short-term money market rate	0.7	1.0	0.9
Long-term government bond yield	1.3	1.8	1.5
Real effective exchange rate appreciation
(-)/depreciation (+)	8.0	12.7	12.3
Nominal exchange rate appreciation
(-)/depreciation (+)	7.7	11.5	10.4
Inflation rate (CPI)	-0.3	-1.2	-0.9
Unemployment rate	1.1	3.4	4.5
Memo:
Baseline Real GDP growth (in percent)	0.0	0.9	0.7
(Adverse) Real GDP growth (in percent)	-3.3	-2.7	0.3
Cumulative real GDP growth (from 2018)	-3.3	-6.0	-5.7
Severity: deviation of growth from baseline in multiples of standard deviation of historical growth volatility	1.9	2.1	1.6

Source: IMF staff calculations.

^{Table 3.}

Italy: FSAP ST Baseline and Adverse Scenario (in percent; unless specified otherwise)

Source: IMF staff calculations.

^{Table 3.}

Italy: FSAP ST Baseline and Adverse Scenario (in percent; unless specified otherwise)

	Baseline			Adverse
	2019	2020	2021	2019	2020	2021
Real GDP Growth	0.0	0.9	0.7	-3.4	-2.7	0.3
Policy interest rate	0.0	0.2	0.3	0.1	0.3	0.3
Short-term money market spread	0.0	0.0	0.0	0.8	1.1	0.9
Long-term government bond yield	3.5	3.8	4.0	4.8	5.5	5.5
Nominal exchange rate app (-)/dep(+)	0.0	0.0	0.0	7.7	11.5	10.4
Inflation rate (CPI)	0.8	1.4	1.6	0.5	0.2	0.7
Unemployment rate	10.8	10.6	10.4	11.9	14.0	14.9

Source: IMF staff calculations.

View Table

^{Table 3.}

Italy: FSAP ST Baseline and Adverse Scenario (in percent; unless specified otherwise)

	Baseline			Adverse
	2019	2020	2021	2019	2020	2021
Real GDP Growth	0.0	0.9	0.7	-3.4	-2.7	0.3
Policy interest rate	0.0	0.2	0.3	0.1	0.3	0.3
Short-term money market spread	0.0	0.0	0.0	0.8	1.1	0.9
Long-term government bond yield	3.5	3.8	4.0	4.8	5.5	5.5
Nominal exchange rate app (-)/dep(+)	0.0	0.0	0.0	7.7	11.5	10.4
Inflation rate (CPI)	0.8	1.4	1.6	0.5	0.2	0.7
Unemployment rate	10.8	10.6	10.4	11.9	14.0	14.9

Source: IMF staff calculations.

View Full Size

Adverse and Baseline Scenario

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Source: IMF staff calculations.Note: ^*) The EBA stress test scenario covers the period 2017–2020.

• Download Figure
• Download figure as PowerPoint slide

21. The financial implications of the scenario are equally severe. The long-term yields increase by 180 bps in the second year (relative to the baseline). Spreads relative to German long-term Bunds (a measure of risk premia) reach 580 bps in the second year of the scenario before narrowing to 560 bps, partly due to the declining German Bund yield under the scenario.¹² Furthermore, heightened risk aversion reduces equity prices by 45 percent in Italy over two years, relative to the baseline.¹³

B. Methodology

22. The TD exercise for the banks is based on the IMF’s internally developed solvency stress testing framework. This stress test includes a comprehensive set of risks, including credit risk associated with all exposures, market risks (equity, exchange and interest rate risks), sovereign risk, and interest rate risk on the banking book. By contrast, the derivatives book was not considered, due to lack of access to granular enough information to stress the derivatives portfolio in a meaningful way (see the Stress Testing Matrix (STeM) in Appendix II for more details).

Hurdle Rate

25. The hurdle rates for banks were differentiated between systemic and non-systemic banks (Table 4). Hurdle rates under the baseline consist of Basel III regulatory minima on CET1 (4.5 percent) and include other systemically important institution (O-SII) buffers for each G-SIB/O-SII. The O-SII buffer differs for each O-SII bank, ranging from 0.19 percent to 1.0 in 2021 and the phase-in period of the O-SII buffers during 2018 to 2021 was taken into account.¹⁶ The final capital level is calculated on a fully loaded basis, including for IFRS 9 implementation. In addition to CET1, we evaluate the banks’ T1 ratio and CAR. The leverage ratio (Tier 1 capital to non-risk weighted total assets) during the stress test horizon was also compared against the 3 percent Basel III minimum requirement. Banks that end the stress test horizon with a capital level or a leverage ratio below the relevant hurdle rates are considered to have failed the test.

^{Table 4.}

Italy: Hurdle Rates for Solvency Stress Tests

(in percent)

Source: IMF staff calculations.

^{Table 4.}

Italy: Hurdle Rates for Solvency Stress Tests

(in percent)

Minimum capital ratios	Hurdle rate (Baseline and Adverse scenario)
	OSII	Other banks (Non-OSII)
Total Capital ratio (total capital to RWAs)	8 + OSII buffer	8
Tier I Capital ratio (Tier 1 capital to RWAs)	6 + OSII buffer	6
Common Equity Tier I Capital ratio (CET1 capital to RWAs)	4.5 + OSII buffer	4.5
Leverage ratio (Tier 1 capital to total assets)	3	3

Source: IMF staff calculations.

View Table

^{Table 4.}

Italy: Hurdle Rates for Solvency Stress Tests

(in percent)

Minimum capital ratios	Hurdle rate (Baseline and Adverse scenario)
	OSII	Other banks (Non-OSII)
Total Capital ratio (total capital to RWAs)	8 + OSII buffer	8
Tier I Capital ratio (Tier 1 capital to RWAs)	6 + OSII buffer	6
Common Equity Tier I Capital ratio (CET1 capital to RWAs)	4.5 + OSII buffer	4.5
Leverage ratio (Tier 1 capital to total assets)	3	3

Source: IMF staff calculations.

Credit Risk Analysis

26. Credit risk constituted the largest risk factor for the banking system (Figure 6). RWAs of credit risk accounted for 88 percent of total RWAs in the sample banks, in line with the banking system’s asset composition. The largest portion of assets was loans, representing 71 percent, followed by debt securities. By sector, loans were mostly to large firms (27 percent), followed by mortgages (19 percent), small- and medium-sized enterprises (SMEs) (17 percent), and other financial corporations (11 percent).

View Full Size

Sample Banks’ Balance Sheet Composition

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; European Central Bank; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

27. Most of the sample banks apply partial IRB approach. Out of the nine banks, two exclusively apply the standardized approach to credit risk, while the remaining banks apply the IRB approach partially. Based on the RWAs, the IRB approach accounts for 50 percent (the median) of total credit exposure of the seven sample banks, with a range of 24 percent to 71 percent.

28. By geographic distribution, most of the sample banks’ loan exposures are in Italy. Out of the total loans, 60 percent are distributed in Italy, with most of the remaining portfolio in EA countries (Figure 7). For customer loans, which consist of loans to non-financial corporations and households, 65 percent of the loan exposure is distributed in Italy, while the rest is in other countries, such as Germany, Austria and Turkey. Three of the SIs in the sample have a significant portion (larger than 30 percent) of their loan exposures outside Italy.

View Full Size

Geographical Distribution of Loans

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; European Central Bank; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

29. Default rates for exposures in Italy were estimated separately for five different portfolios. Historical default rates at the aggregate level were provided by the BdI for five portfolios: corporate large firms, corporate SMEs, mortgage, consumer, and financial institution.¹⁷ Point-in-time (PiT) PDs are projected using regression models with macro variables as independent variables. Details of the estimations are included in Appendix III.

30. Default rates were estimated using a Bayesian Model Averaging (BMA) approach. Under this methodology, a subset of all possible models is first chosen where all explanatory variables (macrofinancial variables and lags) are statistically significant in explaining changes in PDs. The coefficients are then obtained using a weighted average of default rate estimates across multiple models, with the weights corresponding to the posterior probability of each specification (see Appendix III). The result is then attached to the starting point PDs of each sample bank to build the PD path of each bank (Figure 8).¹⁸

View Full Size

Projected Default Rates

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’ltalia; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

31. The projection of default rates for exposures outside Italy was estimated using Moody’s expected default frequency (EDF) for geographical areas that were significant for the sample banks. Given that the sample banks have loan exposures in countries outside Italy, the PDs were calculated for significant exposures in some countries, including Germany, Austria, Turkey, Russia and the United States. The following three categories were used: corporate consumer durables, consumer nondurables, and services. These categories were mapped to major COREP portfolios; i.e., corporate (including SME and specialized lending), mortgage, and consumer loans. The default rates were projected using similar BMA approach for exposures in Italy. The PiT shifts were then applied to regulatory PDs for non-defaulted exposures.

32. Credit risk associated with fixed-income instruments is differentiated between fair value (FV) and amortized cost (AC) holdings. The credit risk associated with FV holdings is embedded in the market risk methodology: the change in a security’s price (and the resulting capital impact) reflects changes due to risk-free rate movement or changes in credit risk premia. For AC securities, changes in risk-free rates are immaterial for the calculation of capital requirements. Nevertheless, in the case of AC securities, provisions are made according to changes in credit spreads.

33. Conditional PD forecasts were generated based on the estimated model parameters. Given a weak macroeconomic outlook in the baseline, the PDs in most segments are projected to gradually increase in the baseline scenario and to worsen further in the adverse scenario. The impact under the adverse scenario displays idiosyncrasies across segments, with the impact on large firms and SMEs more sizable than those on mortgages and financial institutions. The magnitude of the projected PD shock under the adverse scenario for those severely impacted segments is consistent with historical stress episodes.

34. Real domestic output, unemployment rate, inflation, and short-term and long-term interest rates proved to be relevant for the buildup of credit risk. This is reflected in the higher than prior posterior inclusion probability and sizable long run multiplier estimate (i.e., coefficients for both the contemporaneous and lagged terms of the independent variables) for the sectoral PDs. The type and number of significant variables varies distinctly across segments, as manifested by the individual characteristics of their historical PDs.

35. Bank-by-bank LGD rates were used in the exercise. The FSAP’s analysis on the recovery rate of NPLs, which consist of unlikely-to-pay (UTP) and bad loans, provided the reference of loss given default for the IRB portfolios and provisioning rate for the standardized exposure of the new and the existing NPLs. ¹⁹ The provisioning rate is calculated on a bank-by-bank basis differentiating between secured and unsecured loans for both UTP and bad loans. The secured loss rate was used for mortgage loans, while the weighted average of secured and unsecured loss rates was used for other type of loans. For secured loans, the median loss rate was at 42 percent of total NPLs, with a range between 38 percent and 46 percent. For unsecured loans, the median loss rate was at 65 percent of total NPLs, ranging from 54 percent to 72 percent in the sample banks. The same LGD rates are used for both the baseline and adverse scenario.

36. PIT parameters were used to compute loan loss provisioning. Loan impairments were calculated on all exposures including on-balance and off-balance sheet exposures, considering the migration of off-balance sheet items (credit line and guarantees) to on-balance sheet. Coverage included all asset classes for IRB and STA exposures reported in CRR.

37. The FSAP estimated further provisioning of about €4.9 billion (6.4 percent of existing provisions or 0.44 percent of RWAs) for all SIs and LSIs, largely in relation to the UTP portfolio. The estimates were based on assumptions used for loss rates as indicated above (UTP and bad loans; secured and unsecured), which were based on data on banks’ internal workouts of NPLs from the BdI credit registry and loan servicing companies. The FSAP assumes that, on top of the provisioning needs for new defaulted loans, further loan losses will materialize due to the seasoning of the existing NPL portfolio. In particular, it assumes that, over the three-year horizon, 50 percent of the UTP portfolio will become bad loans, 20 percent will return to performing status, and the rest (30 percent) will be closed in banks’ books as UTP. The required additional provisions were incorporated both in the baseline and stress scenarios.²⁰

38. Interest payments were assumed to accrue only on performing exposures under both the baseline and adverse scenarios. The interest revenue on performing exposures was calculated on the gross carrying amount. While accounting rules allow banks to accrue interest income on non-performing exposures with provisioning required on the more delinquent and uncollectible assets, the stress test exercise took a more conservative approach which does not allow banks to project income on non-performing exposures.

39. The exercise considered potential additional costs arising from NPL disposals. The results without NPL disposal plans include provisioning needs for new defaulted loans and, if applicable, gradual provisioning increases over the three-year stress test horizon to raise the coverage ratio of existing NPLs to the loss rates used by the FSAP on various portfolios (based on internal workout of NPLs assumption) (see above). However, loss rates from market-based NPL disposals are higher than loss rates from internal workouts. Considering that, the exercise included an additional scenario where banks internalize additional costs arising from a market-based disposal of NPLs that would allow them to reach the NPL ratio target for 2021 disclosed in their annual reports.²¹ The additional provisions are calculated as the difference between loss rates of NPL sales and the loss rate from internal workout. The NPL disposal loss rate is, on average, 14 percentage points higher. Both scenarios use fully loaded IFRS 9 capital ratios.²²

Market Risk Analysis

40. Own-sovereign securities constitute the largest share of bank holdings of debt securities (Figure 9). As of December 2018, the share of own-sovereign securities was 44 percent of total debt securities (including FV and AC categories), followed by foreign-sovereign securities (34 percent) and securities issued by credit and financial institutions (19 percent). The own-sovereign securities were mostly booked at FV, either at fair value through profit or loss (FVTPL) or fair value through other comprehensive income (FVOCI), at 61 percent of total own-sovereign securities. The total share of securities booked at FV accounted for 69 percent of total. This classification maximizes the capital impact of sovereign yield changes as the gains or losses would be absorbed directly through net profit and equity.

View Full Size

Sample Banks’ Debt Holdings

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; European Central Bank; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

41. Duration varies across types of debt securities; own-sovereign securities at FV have relatively short duration. The average duration of FV own-sovereign securities is at around 4 years, much lower than the same securities at AC category, at 6 years on average. Foreign sovereign securities have average durations of around 5 and 6 years, for FV and AC categories. The relative shorter duration of sovereign securities at FV reduced the impact of sovereign yield changes on regulatory capital.²³

42. Stress tests assessed the resilience of banks when facing different sources of market risks. In addition to credit risks, banks also faced risks from changes in market variables, such as interest rates, exchange rates and equity prices. These losses would be generated through the net open position in foreign currencies and market valuation losses for debt securities due to changes in market yield. The scope excludes amortized cost positions held in a hedge-accounting relationship, as well as hedge-accounting derivatives.

43. Market risk is treated as an add-on component, with a separate calibration that is consistent with the macroeconomic scenario. The market risk stress scenario will have an impact on both capital resources—either via profit and loss or via other comprehensive income—and capital requirements (RWAs). The impact on capital resources will comprise of positions in the trading book as well as other fair valued items in the banking book. The impact on RWAs for market risk evolve with balance sheet assumptions.

44. Market valuation losses corresponding to holdings of debt securities were estimated using the modified duration approach. For every country (in which the sample banks have significant holdings) and year, sovereign yield curves were constructed by linear interpolation of short- and long-term interest rates, as specified in the macroeconomic scenarios. Then, the average duration of debt portfolio was calculated for each bank based on the supervisory data (COREP). Losses were then calculated as the product of the size of each bond portfolio, average duration, and the changes in the yields. For non-sovereign securities, the yield moved in line with sovereign yield with a credit spread along the three-year horizon. The following formula represents the modified duration approach in the stress test:

$\frac{\mathrm{\Delta }Valuation}{Valuation}=-MD\mathrm{.}\mathrm{\Delta }{y}_{MD}$

where MD is the modified duration of the portfolio and ∆yMD is the change in the yield caused by the shift in the yield curve (vis-à-vis the value prevailing in the previous year) and measured at a point in time that matches the modified duration of the portfolio.

45. For equities held with trading intent, the FV impact was subject to a floor using an approach similar to EBA 2018 stress test methodology for exposures held in the trading portfolio. The market impact from full revaluation of equity holdings was subject to a floor using the following constraint:

$\mathrm{\Delta }Eq=15*(-0.20\%(E{q}^{Long}+E{q}^{Short}))$

where the VaR scaling factor has been set to the upper bound of 1.5, and the trading position includes the FV of equity instruments (assets) and the short positions in equity instruments (liabilities).

Interest Rate Risk in the Banking Book

52. Bank interest rates on new business were estimated and used as the input for interest rate risk in the banking book (IRRBB). Using BMA methodology, the satellite models estimate aggregate funding and lending rates on the portfolio level, which include interest rates on retail and wholesale deposits (both term and overnight), debt securities as well as household and corporate loans. Subsequently, the model outputs were used to project bank-specific interest rate paths by attaching the period changes of the aggregate rates in the forecasting horizon to the bank-specific starting point.

53. The projection of funding and lending rates was mapped to banks’ financial assets and liabilities by product and counterparty using the short-term exercise (STE) IRRBB template. The IRRBB template provide bank-specific maturity ladder for fixed rate instruments and repricing date for floating rate instruments on portfolio level for both assets and liabilities. The template includes the following categories: (i) the asset side of the banking book comprises generic products related to debt securities and loans and advances underwritten by the banks; and (ii) the liabilities side comprises retail and wholesale overnight and term deposits, repos as well as debt securities.

54. The input for interest rate models is very similar to that of credit risk models. Most of the inputs for the credit risk model were used in the interest rate models, such as GDP growth, inflation, exchange rate, unemployment rate, and short- and long-term interest rates. As interest rates were received in a blended form and reflect both domestic and foreign exposures (mainly from the EA and the United States), most explanatory variables came under the form of Italy-specific and EA- and U.S.-based indicators to account for country-specific interest risk associated with both the domestic and foreign creditors/borrowers.

55. The relationship between lending rates and funding cost is incorporated in the model. To simulate bank-specific risk behavior and allow for a partial pass through of the rising funding cost to the lending rate, banks’ funding cost was included as an additional explanatory factor in the projection of the lending rate. Therefore, the model was performed sequentially by first estimating the funding rate, which was then used as input for the projection of the lending rates.

56. The model also allows for the inclusion of PDs to incorporate credit risk in banks’ interest rate determination. Specifically, household and corporate PDs estimated by the credit risk satellite models were included as an add-on risk factor and mapped into respective lending rate categories to capture banks’ premium charges on borrowers with high default risk.

57. The projected interest rate paths were broadly in line with banks’ portfolio characteristics. On the liability side, this is reflected by a more severe impact on the long term and unsecured debt portfolios as opposed to highly liquid and short-term funding. On the asset side, the increase in the lending rates also incorporates the increase in PDs associated with loan portfolios. However, to be conservative, a lower bound (5th percentile for the adverse and 25th percentile for the baseline) within the lending rate forecasted confidence band were selected to factor in the constrain faced by the banks in increasing lending rates. The resulting average decline on the net interest margin amounts to 0.07 percentage points under the baseline and 0.18 percentage points under the adverse scenario, respectively.

58. Variables related to money market rates and long-term sovereign yields are the main contributors in the projections of bank interest income and funding costs (Appendix IV). The long-run multiplier for variables associated with short- and long-term interest rates turn out to be sizable in the determination of both the lending and funding rates. Specifically, on the funding side, the 3-month money market rate and 10-year domestic sovereign bond yield spread explain most of the movement in the interest expense. On the lending side, an almost identical set of variables play similarly significant roles, with the addition of the U.S. long-term sovereign bond yield and the funding cost associated with retail term deposits. The long-run pass-through from Italy’s sovereign bond yield and short-term interest rate on funding rates appears to be large, particularly from the government bond yield to retail term deposit rate and from the short-term rate to the wholesale term deposit rate and retail and wholesale overnight deposit rates.

59. The funding cost was measured using effective funding costs for the whole range of liabilities and incorporate and a repricing structure (Figure 10):

• The pricing of repos followed changes to the short-term money market rates (EURIBOR). The interest rate projections were linked to both the baseline and adverse scenario and therefore were expected to slightly increase under the baseline and considerably tighten under the adverse scenario (maximum 1.1 percentage points deviation from the baseline scenario).

• Funding from retail and wholesale deposits are repriced according to the output from the satellite models. Four deposit categories (i.e., wholesale overnight, wholesale term, household overnight, household term) were estimated by the satellite models. Rates evolve with economic conditions and benchmark rates.

• The pricing of debt securities is estimated by satellite models for new bond with contractual maturities above one year.

View Full Size

Funding Rate Estimation by Portfolio

(in percent)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; European Central Bank; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

60. The interest income is measured using effective lending rates for the whole range of interest-bearing assets and considering the repricing structure (Figure 11):

• The weighted average of lending rates was used to project interest rates on loans and advances, given limited decomposition in the IRRBB template. Specifically, lending rates on consumer loans, mortgages and non-financial corporates were weighted by their respective notional amount as of December 2018 to produce the forward paths.

• Rates on debt securities followed projections on the long-term sovereign bond yield, which was linked to the macroeconomic scenario.

View Full Size

Lending Rate Estimation by Portfolio

(in percent)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’ltalia; European Central Bank; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

C. Solvency Stress Test Results

61. Under the baseline scenario, the unfavorable macroeconomic outlook and the FSAP assumptions used on LGD, which included the transition of 50 per cent of UTPs to bad loans, increased the provisions needed during the three-years horizon.²⁵ The results are as follows (Figure 12):

• With the assumption of “no NPL disposal”, the aggregate CET1 ratio of the sample banks would decline by 56 bps from 11.9 percent to 11.4 percent; the decline was mostly attributable to the increase in RWAs. With the assumption of “NPL disposal”, the CET1 ratio would decline by 102 bps to 10.9 percent.

• Considering the hurdle rate of capital ratio minima plus OSII, with “no NPL disposal” assumption, one bank falls below the T1 threshold and another bank falls below the CAR threshold. When including NPL disposal targets, the first bank would fall short of the three capital thresholds. The resulting shortfall in capital is about 0.05 percent of GDP. The two banks would also see their leverage ratio fall below the minimum threshold of 3 percent.

• The main contributor to the drop in the capitalization ratios under the “NPL disposal” scenario is credit risk (the need for additional loan loss provisioning). Credit risk accounted for 4.5 percentage points of CET1 followed by an increase in RWAs of 0.5 percentage point, which mostly comes from the positive asset growth of the sample banks. Besides credit risk, the increase in the sovereign yields also has an important impact through valuation losses and NII reduction related to higher funding costs (0.4 percentage points each). Meanwhile, the pre-provision revenue, including mainly aggregate NII, non-interest income, and non-interest expenses, increases the aggregate CET1 ratio by 4.9 percentage points relative to the starting point.

View Full Size

SIs Solvency Stress Test

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Source: IMF staff.Note:¹ The adverse scenario in this graph reflects the “without NPL disposal plan”, which assumes that banks will not be required to undertake further disposal of their NPLs under crisis conditions.

• Download Figure
• Download figure as PowerPoint slide

62. The adverse scenario has a significant impact on banks’ capital ratios (Figure 12).

• Without the inclusion of NPL disposal plans, the average fully loaded CET1 capital ratio declines from 11.9 percent in 2018 to 8.2 percent in 2021.²⁶ Three banks would see their capital ratios drop below the thresholds, with an average CET1 ratio decline of 9.3 percentage points.²⁷ The asset share of the three banks is 8.5 percent of the sample’s assets, and the resulting shortfall in capital is about 0.2 percent of GDP.²⁸ The results do not change materially with the inclusion of NPL disposal plans.

• The aggregate leverage ratio would decline by 1.4 percentage points from 5.8 percent to

4.4 percent. The three failed banks would see their leverage ratio decline below the 3 percent threshold.

• Credit risk was the largest contributor to the decline in capital ratios. The decline in the average CET1 ratio related to credit risk provisioning amounted to about 5.3 percentage points. The increase in the sovereign yields also has an important impact through valuation losses (1.2 percentage point decline in CET1) and NII reduction related to higher funding costs (0.9 percentage points). Heightened credit risk (partly driven by PiT shifts to risk parameters), valuation effects in FX exposures, and the projected path of loans that is still positive for some banks, contribute to an expansion of RWAs, thereby lowering CET1 by 1.2 percentage point across the sample banks. The pre-provision revenue, including mainly aggregate NII, non-interest income, and non-interest expenses, increases the aggregate CET1 ratio by 4.7 percentage points relative to the starting point. While the resulting capital shortfall against the threshold is small at about 0.2 percent of GDP, capital needs to bring the CET1 ratio of the 9 SIs in the stress scenario back to the end-2018 level of 12 percent is about 2.2 percent of the GDP.

• The FSAP conducted sensitivity analysis using alternative scenario assumptions. To assess the potential impact of higher loss rates under the stress scenario, the FSAP calculated banks’ CET1 ratios if loss rates were further increased by 20 percent.²⁹ Furthermore, CET1 ratios were evaluated against a higher threshold of 7 percent, which incorporates the capital conservation buffer (CCB). The number of bank failures do not change in the adverse scenario if a CET1 threshold of 7 percent was used or the LGD rates were increased by 20 percent. However, combining the two new assumptions (CET1 threshold of 7 percent and a 20 percent increase in LGD), an additional bank will fall slightly below the 7 percent CET1 threshold.

A. Debt Portfolios Sensitivity Test

66. This exercise applies to certain exposures in banks’ debt portfolios. Debt categories in FV portfolios are subject to mark-to-market changes due to changes in the yields. Valuation changes associated with the increase in yields are applied to all own-sovereign, financial and non-financial corporate, and securitized debt, according to the modified duration approach (duration for all these portfolios are available, except securitized debt, which is a small part of LSIs debt portfolio). In addition, we add provisions for those exposures held in AC portfolios according to the increase in spreads, using an LGD of 45 percent.

67. To calibrate the scenario, we compute on a daily basis, the 30-day change in yields since 2000. We choose k^th percentile of the monthly increase in the average of LT (10-year) and ST (3m) yield. Note that the average yield is not directly used in the scenario, it is just used to help us choose the percentiles. As such, the changes in the yield curve slope are historically accurate. Three calibrations are used (Table 5):

• 95th percentile of year 2011 (scenario 1)

• 98th percentile of 2000–2018 (scenario 2)

• 95th percentile of 2000–2018 (scenario 3)

^{Table 5.}

Italy: Debt Portfolio Sensitivity Analysis Scenarios

Source: IMF staff calculations.

¹This scenario also corresponds to the 99.5 percentile over the 2000–18 period.

^{Table 5.}

Italy: Debt Portfolio Sensitivity Analysis Scenarios

		ST	LT	Average	Date
Scenario 1	95th percentile of 20111	3.2	1.4	2.3	12/19/2011
Scenario 2	98th percentile, 2000–18	2.1	0.4	1.2	10/14/2008
Scenario 3	95th percentile, 2000–18	0.8	0.8	0.8	8/16/2012

Source: IMF staff calculations.

¹This scenario also corresponds to the 99.5 percentile over the 2000–18 period.

View Table

^{Table 5.}

Italy: Debt Portfolio Sensitivity Analysis Scenarios

		ST	LT	Average	Date
Scenario 1	95th percentile of 20111	3.2	1.4	2.3	12/19/2011
Scenario 2	98th percentile, 2000–18	2.1	0.4	1.2	10/14/2008
Scenario 3	95th percentile, 2000–18	0.8	0.8	0.8	8/16/2012

Source: IMF staff calculations.

¹This scenario also corresponds to the 99.5 percentile over the 2000–18 period.

As seen in the table, scenario 1 and 2 assumed a flattening yield curve, as the shock for ST yield is higher than LT yields.

Results

68. Results show that banks are vulnerable to mark-to-market losses arising from Italian government bond holdings (Table 6, and Figures 13 and 14). In the most stressful scenario, which corresponds to the rapid rise in yields observed in 2011, capital losses amount to 300 bps for the group of LSIs and almost a quarter of the group by assets see their CET1 ratio fall below 7 percent with a total capital shortfall of 0.02 percent of GDP. Estimated capital needs to bring the sample banks’ CET1 ratio back to the starting ratio of 15.2 percent is 2.1 percent of GDP. As regards the SIs, among the 9 SIs included in the test, no bank falls below the 7 percent threshold. These results highlight the higher concentration of Italian sovereign bonds in LSIs’ portfolios.³²

^{Table 6.}

Italy: Debt Portfolio Sensitivity Analysis Results

Source: IMF staff calculations.

^{Table 6.}

Italy: Debt Portfolio Sensitivity Analysis Results

		SIs			LSIs
		Scenario 1	Scenario 2	Scenario 3	Scenario 1	Scenario 2	Scenario 3
Aggregate CET1 impact (bps)		131	75	52	300	185	120
< 7% CET1
	#banks	0	0	0	10	3	2
	Share of assets	0%	0%	0%	22%	6%	6%
< 4. 5% CET1
	#banks	0	0	0	4	1	0
	Share of assets	0%	0%	0%	2%	0%	0%
Losses (share of assets)		0%	0%	0%	1%	0%	0%

Source: IMF staff calculations.

View Table

^{Table 6.}

Italy: Debt Portfolio Sensitivity Analysis Results

		SIs			LSIs
		Scenario 1	Scenario 2	Scenario 3	Scenario 1	Scenario 2	Scenario 3
Aggregate CET1 impact (bps)		131	75	52	300	185	120
< 7% CET1
	#banks	0	0	0	10	3	2
	Share of assets	0%	0%	0%	22%	6%	6%
< 4. 5% CET1
	#banks	0	0	0	4	1	0
	Share of assets	0%	0%	0%	2%	0%	0%
Losses (share of assets)		0%	0%	0%	1%	0%	0%

Source: IMF staff calculations.

View Full Size

LSIs: Debt Portfolio Characteristics

(End-June 2018)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Source: BdI and IMF staff calculation.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

LSIs: Debt Portfolio Sensitivity Test Results

(as of June 2018)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; European Central Bank; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

B. Loan Losses due to NPL Flows

69. The test considers new flows of NPLs for three portfolios: household loans (HH), non-financial corporates (NFC) and government loans (GG). The LSIs were subjected to a flow of new NPLs in line with the historically worst observed NPL flows. ³³ Default rates were adjusted to consider the fact that LSIs have higher PDs than the system and the starting point of each bank. The provisioning needs associated with the new NPL flows were set at 60 percent of the stock of new NPLs. In addition, we include the provisions required, if any, to bring the coverage ratio for the existing stock of NPLs up to 60 percent. The 60 percent is a conservative average, but one that is often used in stress tests; while 60 percent may be on the conservative side for a mortgage loan, it is on the lower side for consumer credit and a proportion of corporate credit.³⁴

70. Shocks were defined based on historical observations. For each loan portfolio, we choose the k^th percentile quarterly PD since 2000, available for the banking sector (system-wide). We then adjust this stress PDs to take into account the fact that LSIs have higher PDs than the system, such that:

Furthermore, we adjust the bank-specific stress PDs by the starting point PD for the same bank, such that the PD for the LSI sector as a whole reaches the desired level. In other words,

Taking the starting point PDs into account affects the distribution of losses and penalizes banks that already have higher PDs.

71. We consider two scenarios that differ in the severity of the shocks.

• Scenario 1: Maximum PDs for HH, NFC, GG since 2000, system-wide.

• Scenario 2: 80^th percentile PDs for HH, NFC, GG since 2000, system-wide.

As of June 2018, the system-wide new flows are around the median of new flows observed since 2000s.

72. The provisioning needs associated with the new NPL flows are 60 percent of the stock of new NPLs. In addition, we calculate the provisions required, if any, to bring the coverage ratio for the existing stock of NPLs up to 60 percent.

73. Results show that credit losses can be meaningful for some banks and are larger in the corporate portfolio. In the most severe scenario, capital losses amounted to 365 bps for the group of LSIs, with 14 banks (35 percent of the group by assets) falling below the 7 percent CET1 ratio. Capital needs to bring back banks’ CET1 ratio to 7 percent are about 0.1 percent of GDP (Table 7 and Figures 15 and 16). Estimated capital needs to bring the sample banks’ CET1 ratio back to the starting ratio of 15.2 percent is 0.26 percent of GDP.

^{Table 7.}

Italy: NPL Losses Sensitivity Analysis (End-June 2018)

Source: IMF staff calculations.

^{Table 7.}

Italy: NPL Losses Sensitivity Analysis (End-June 2018)

		Scenario 1	Scenario 2
Aggregate CET1 impact (bps)		365	311
<7% CET1
	# banks	14	12
	share of assets	35%	29%
Losses (share of assets)		1.9%	1.8%
<4.5% CET1
	# banks	6	4
	share of assets	9%	8%

Source: IMF staff calculations.

View Table

^{Table 7.}

Italy: NPL Losses Sensitivity Analysis (End-June 2018)

		Scenario 1	Scenario 2
Aggregate CET1 impact (bps)		365	311
<7% CET1
	# banks	14	12
	share of assets	35%	29%
Losses (share of assets)		1.9%	1.8%
<4.5% CET1
	# banks	6	4
	share of assets	9%	8%

Source: IMF staff calculations.

View Full Size

SIs: Debt Securities Sensitivity Test

(as of December 2018)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Source: Banca d’Italia; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

LSIs: Asset Quality, Loan Portfolios, Loan Loss Impact

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; European Central Bank; and IMF staff calculations.Note: ^*) Group 1 represents banks with assets share more than 4 percent; Group 2 between 1 percent and 2 4 percent, and Group 3 less than 1 percent.

• Download Figure
• Download figure as PowerPoint slide

C. Interest Rate Risk in the Banking Book

74. Most LSIs have a positive interest-rate gap on their banking book, helping NII in case of an increase in interest rates. The IRRBB exercise was conducted by the BdI, but detailed results were shared with the FSAP team. The average loss of economic value of equity due to a +/- 200 bps interest rate change is 2.6 percent, nevertheless, a number of small banks face losses greater than 20 percent of economic value of capital in the case of a -200 bps interest rate shock, and several come close.

D. Concentration Risk and SME Supporting Factor in Significant Institutions

75. The FSAP also assessed the role played by the SME supporting factor and concentration risks (Figure 17). The result refers to the starting point position of the banks as of end-2018.

• The CRR has introduced a deduction in capital requirements for exposures to SMEs by applying the so-called SME supporting factor of 0.7619. The FSAP assessed the impact of this provision, considering the relatively large size of the SME portfolio in Italy. The analysis assigned a risk weight of 92 percent (the average risk weight of SME exposure that is not subject to supporting factor) and of 100 percent. The result showed that, on average, the SME supporting factor boosted the CET1 ratio by 18 bps (assuming 92 percent RWA) or 27 bps (assuming 100 percent RWA). The maximum impact of SME supporting factor to CET1 ratio is 60 bps for one sample bank (assuming 92 percent RWA).

• Concentration risk was tested by assessing the impact to bank’s capital from the simultaneous default of their largest exposures. The test assessed banks’ resilience under the assumption of a simultaneous hypothetical default of the five largest borrowers of each bank. The analysis used the net amount of the exposures. The results are as follows:

• Using zero recovery rate, the simultaneous default of the five largest borrowers would cause the aggregate CET1 ratio of sample banks to decline by 3.4 percentage points from 11.9 percent to 8.5 percent. No banks would see their CET1 ratio below the 4.5 percent threshold.

• Using a provisioning rate of 65 percent, the simultaneous default of the five largest borrowers would cause the aggregate CET1 ratio of sample banks to decline by 2.2 percentage points from 11.9 percent to 9.7 percent. No banks would see their CET1 ratio below the 4.5 percent threshold.

View Full Size

SIs Sensitivity Test

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: FINREP, COREP, IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

A. NSFR-based Liquidity

78. The stable funding structure of the banking system (excluding capital and undrawn credit and liquidity facilities) differs for SIs from LSIs (Figure 18). Mainly, data (as of end-October 2018 for SIs and end-June 2018 for LSIs) highlights the following differences:

• Retail funding is the largest source of available stable funding (ASF) for LSIs (59 percent) but represents a much smaller share of total funding for SIs (28 percent of SIs’ total funding structure). Additionally, corporate funding makes 20 percent of LSIs’ total funding, and debt instruments make 9 percent of their total funding.

• In contrast, SIs’ funding sources include: central bank funding (13 percent of total funding), unsecured funding from financial institutions (17 percent of total funding), and unstable retail funding (17 percent). These sources of funding are not significant for LSIs.

View Full Size

Liquidity: NSFR and Structure of Funding

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

79. NSFR ratios are already above 1 for the majority of SIs and LSIs. The FSAP sample included 11 SIs and 59 out of 61 LSIs. Looking into the overall NSFR for banks, the data indicate that SIs’ ASF has a lower factor compared to the European bank average, due to their reliance on less-stable sources of funding (and wholesale funding) (Figure 18). Meanwhile, LSIs show a lower required stable funding (RSF) factor compared to European banks’ average due to the large share of government bonds on their asset side, which reduces their RSF factor, and contributes to higher NSFR ratios.

B. LCR-Based Liquidity Stress Test

80. In addition to Basel III LCR prescribed scenario that go beyond the Basel (CRR) parameters for the calculation of LCR are considered. These scenarios are as follows:

• Retail scenario simulates a (retail) deposit run. The key assumptions are: (i) 10 percent runoff rates for stable retail deposits and 15–20 percent for less stable retail; (ii) 75 percent (40 percent) run-off rates for operational (non-operational) deposits not covered by Deposit Guarantee Scheme; and (iii) 10 percent haircut on government securities for the calculation of high quality liquid assets (HQLA).

• Wholesale scenario simulates a wholesale deposit and wholesale market funding withdrawal. The key assumptions are: (i) 100 percent run-off rates for wholesale funding from other financial institutions; (ii) 50 percent run-off rates for operational and non-operational deposits not covered by Deposit Guarantee Scheme; and (iii)10 percent haircut on government securities for the calculation of HQLA.

81. Banks’ liquidity profiles are broadly comfortable, yet results indicate that banks are particularly vulnerable to a retail event (Table 8). The aggregate LCR (at 160 percent for SIs and 230 percent for LSIs) is comfortable (Figure 19). Reliance on retail deposits, considered as a more stable source of funding, is high particularly for LSIs. Moreover, almost all HQLA is composed of level 1 assets. When subjected to two severe retail and wholesale scenarios where outflows are stressed beyond LCR rates, 9 SIs accounting for 91 percent of the group’s assets fall below the 100 percent LCR in a retail stress scenario, although the average stressed LCR for the system falls just below the 100 percent threshold and the liquidity shortfall represents only 1 percent of liabilities. LSIs accounting for 26 percent of LSI assets fall below the 100 percent threshold in a retail scenario, although the system remains substantially above minimum thresholds.

^{Table 8.}

Italy: LCR Stress Test Results

Source: IMF and BdI staff calculations. Notes: ¹ Liquidity shortfall is the amount of liquid asset needed to restore LCR to 100 percent.

²Retail scenario simulates a (retail) deposit run. The key assumptions are: (i) 10 percent run-off rates for stable retail deposits and 15–20 percent for less stable retail deposits; (ii) 75 percent (40 percent) run-off rates for operational (non-operational) deposits not covered by Deposit Guarantee Scheme; and (iii) 10 percent haircut on government securities for the calculation of high-quality liquid assets (HQLA).

³Wholesale scenario simulates a wholesale deposit and wholesale market funding withdrawal. The key assumptions are: (i) 100 percent run-off rates for wholesale funding from other financial institutions; (ii) 50 percent run-off rates for operational and non-operational deposits not covered by Deposit Guarantee Scheme; and; (iii) 10 percent haircut on government securities for the calculation of HQLA.

^{Table 8.}

Italy: LCR Stress Test Results

		SIs			LSIs
		LCR	Retail Scenario2	Wholesale Scenario3	LCR	Retail Scenario2	Wholesale Scenario3
System-wide LCR		156%	92%	110%	229%	157%	189%
Number of Banks w/
LCR<100%		0	9	1	1	10	7
Share of Group Assets		0%	91%	7%	0.5%	26%	16%
Liquidity shortfall (millions of euros)1		0	16,768	1,494	10	1,341	433
Liquidity shortfall (share of Liabilities of failed banks)		0.0%	0.9%	1.0%	1.0%	2.3%	1.2%

Source: IMF and BdI staff calculations. Notes: ¹ Liquidity shortfall is the amount of liquid asset needed to restore LCR to 100 percent.

²Retail scenario simulates a (retail) deposit run. The key assumptions are: (i) 10 percent run-off rates for stable retail deposits and 15–20 percent for less stable retail deposits; (ii) 75 percent (40 percent) run-off rates for operational (non-operational) deposits not covered by Deposit Guarantee Scheme; and (iii) 10 percent haircut on government securities for the calculation of high-quality liquid assets (HQLA).

³Wholesale scenario simulates a wholesale deposit and wholesale market funding withdrawal. The key assumptions are: (i) 100 percent run-off rates for wholesale funding from other financial institutions; (ii) 50 percent run-off rates for operational and non-operational deposits not covered by Deposit Guarantee Scheme; and; (iii) 10 percent haircut on government securities for the calculation of HQLA.

View Table

^{Table 8.}

Italy: LCR Stress Test Results

		SIs			LSIs
		LCR	Retail Scenario2	Wholesale Scenario3	LCR	Retail Scenario2	Wholesale Scenario3
System-wide LCR		156%	92%	110%	229%	157%	189%
Number of Banks w/
LCR<100%		0	9	1	1	10	7
Share of Group Assets		0%	91%	7%	0.5%	26%	16%
Liquidity shortfall (millions of euros)1		0	16,768	1,494	10	1,341	433
Liquidity shortfall (share of Liabilities of failed banks)		0.0%	0.9%	1.0%	1.0%	2.3%	1.2%

Source: IMF and BdI staff calculations. Notes: ¹ Liquidity shortfall is the amount of liquid asset needed to restore LCR to 100 percent.

²Retail scenario simulates a (retail) deposit run. The key assumptions are: (i) 10 percent run-off rates for stable retail deposits and 15–20 percent for less stable retail deposits; (ii) 75 percent (40 percent) run-off rates for operational (non-operational) deposits not covered by Deposit Guarantee Scheme; and (iii) 10 percent haircut on government securities for the calculation of high-quality liquid assets (HQLA).

³Wholesale scenario simulates a wholesale deposit and wholesale market funding withdrawal. The key assumptions are: (i) 100 percent run-off rates for wholesale funding from other financial institutions; (ii) 50 percent run-off rates for operational and non-operational deposits not covered by Deposit Guarantee Scheme; and; (iii) 10 percent haircut on government securities for the calculation of HQLA.

View Full Size

Liquidity Coverage Ratio

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Banca d’Italia; and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

C. Cashflow-Based Liquidity Stress Test

82. The cash flow-based liquidity stress test (CFLST) analyzes the liquidity risk exposure and risk bearing capacity of a sample of 11 SIs and 55 LSIs. The CFLST incorporates a set of embedded scenarios that allow the FSAP team to estimate the order of magnitude of potential liquidity needs of individual banks and the banking system under multiple stress scenarios. It also reveals the level of liquidity risk tolerance; i.e., under which circumstances banks would need additional liquidity support because of the mismatch of cash flows, and the absence of available counterbalancing capacity under stress. In addition, the CFLST highlights the liquidity risk exposures of banks in the banking system, such as potential reliance on unsecured short-term funding, wholesale funding from corporates, or holdings of less liquid assets in the counterbalancing capacity (CBC).

83. The CFLST focuses on two key indicators, the banks’ net-funding gap (NFG), and banks’ CBC. The NFG is defined as the difference between cash-inflows and cash-outflows in each time bucket, and the sum of these differences across buckets (i.e., the cumulated net-funding gap CNFG). The CBC is defined as the sum of cash inflows banks can generate under stress at reasonable prices in the respective bucket. The cumulated CBC is the sum of the counterbalancing capacities across time buckets. The analysis builds on data collected within the COREP templates (Maturity Ladder, C66.00).

84. The composition of SIs’ and LSIs’ CBC shows a high concentration in general government items, which could subject banks to market liquidity risks during stressed periods. For LSIs, 82 percent of the CBC is composed of general government items. Meanwhile, government items make 67 percent of SIs’ CBC (Figure 20). Such a large exposure to sovereign securities improves the quality of the composition of the CBC but exposes banks to higher sovereign risk.

View Full Size

Composition of CBC and Asset Encumbrance Ratios for SIs and LSIs

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Source: IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

85. High asset encumbrance (AE) ratios among some of the LSIs in the sample could hinder their ability to further tap wholesale funding markets (especially unsecured ones) and may be subject to heightened funding shocks. The AE ratios among SIs are 30 percent on average, higher than the LSIs’ (22 percent) (Figure 20). However, some of the LSIs in the sample have AE ratios above 40 percent. This, to a considerable extent, reflects funding constraints as well as a shift towards secured market borrowing (such as repos). Banks with high AE ratios may not only face higher outflows from short-term market and deposit funding during idiosyncratic and systemic liquidity events, but also could face difficulties in obtaining additional liquidity in the markets or central banks (as central banks do typically require collateral for funding operations).

86. Contractual liquidity risk exposure for SIs in the sample is very high, reflecting the reliance on demand deposits. Contractual outflows within the first month amounts to about 75 percent of total assets, while contractual inflows amount to about 38 percent of total assets. Thus, the cumulative net funding gap over the first 4 weeks reaches about 37 percent of total assets or 818 billion EUR. The heatmap reveals that 31 percent of liquidity exposure in the first month is concentrated in outflows from retail deposits, followed by 10 percent of contractual outflows from corporate deposits (Figure 21). Meanwhile, repos collateralized with zero percent risk-weighted securities would form around 12 percent of contractual outflows in the first month. On the other hand, the main drivers of contractual inflows are reverse repos against zero percent risk-weighted securities (15 percent of contractual inflows), and corporate and other inflows (10 percent of contractual inflows combined).

View Full Size

Heatmap of Contractual Cash Flows, SIs.

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: European Central Bank and IMF staff calculations. Data for 2018:Q3.

• Download Figure
• Download figure as PowerPoint slide

87. Liquidity risk exposures for LSIs is also high and concentrated in retail deposits (41 percent of contractual outflows). Concentration in retail deposit outflows is high in the first month of contractual outflows (37 percent in the first month, versus 41 percent in the total outflows horizon). Overnight risk exposure is also high for corporate deposits (12 percent of contractual outflows), and other deposits (8 percent of contractual outflows). Similarly, liquidity exposures to repos against 0 percent risk-weighted securities make 16 percent of contractual outflows in the first month of the horizon. In the case of contractual inflows, exposures to retail, corporate, and central bank inflows in the first month make a small portion of contractual inflows (around 9 percent of total inflows in the first month, versus 48 percent of contractual inflows for the entire horizon). Meanwhile, exposures to reverse repos against 0 percent RW securities are about 6.5 percent of contractual inflows in the first month.

88. Run-off rates are calibrated for inflows and outflows, depending on the length of the stress horizon, and the severity of the market stress. Specifically, run-off rates for outflows are less severe in a 5-day stress horizon, and more severe in a 3-month stress horizon. Additionally, run-off rates are higher for unsecured than for secured wholesale funding, as well as for non-insured deposits than for insured ones. Table 9 summarizes the calibration of the inflow and outflow parameters.³⁶ The inflows parameters are in principle 100 percent of the contractual inflows, except for inflows from loans to retail and for corporate customers (inflows 0 percent). This is in line with the objective of the CFLST to assume that banks will continue business as normal, i.e., analyze the ability of banks to cope with liquidity stress while maintaining their ability to lend to the real economy. In fact, when a bank cuts credit lines and/or stops granting loans, it may send a negative signal to the markets about its financial situation, which could lead to further outflows from that bank.

^{Table 9.}

Italy: Scenario Parameters: Run-off Rates for the Major Components of In- and Outflows

Source: IMF staff calculations.

^{Table 9.}

Italy: Scenario Parameters: Run-off Rates for the Major Components of In- and Outflows

Outflows/Inflow	Range of run-off factors (in percent) across mild, medium, severe and most severe scenarios
Unsecured LT/ST Issuances and financial deposits	30–100
Secured issuances	30–100
Stable retail deposits	0.5–10.5
Unstable retail deposits	1.25–29.3
Operational corporate deposits (NFCs)	1.25–38
Non-operational corporate deposits (NFCs)	20–55.7
Repo across all collateral classes*	100
Deposits outflows from financial institutions	50–100
FX-Swaps in-/outflows	15–100
Derivative in-/outflows	25–100
Retail / corporate inflows	0
Central bank inflows	100
Other entities inflows	0–30
Financial institutions inflows	30–100
Other inflows	0–100
Committed lines provided by the bank (FI)	50–100
Committed lines provided by the bank (non-FI)	50–100
Margin calls (derivatives)	2–10
Adverse market outflows (derivatives, HLBA)	0–100
Outflows due to rating downgrades	0–100

Source: IMF staff calculations.

View Table

^{Table 9.}

Italy: Scenario Parameters: Run-off Rates for the Major Components of In- and Outflows

Outflows/Inflow	Range of run-off factors (in percent) across mild, medium, severe and most severe scenarios
Unsecured LT/ST Issuances and financial deposits	30–100
Secured issuances	30–100
Stable retail deposits	0.5–10.5
Unstable retail deposits	1.25–29.3
Operational corporate deposits (NFCs)	1.25–38
Non-operational corporate deposits (NFCs)	20–55.7
Repo across all collateral classes*	100
Deposits outflows from financial institutions	50–100
FX-Swaps in-/outflows	15–100
Derivative in-/outflows	25–100
Retail / corporate inflows	0
Central bank inflows	100
Other entities inflows	0–30
Financial institutions inflows	30–100
Other inflows	0–100
Committed lines provided by the bank (FI)	50–100
Committed lines provided by the bank (non-FI)	50–100
Margin calls (derivatives)	2–10
Adverse market outflows (derivatives, HLBA)	0–100
Outflows due to rating downgrades	0–100

Source: IMF staff calculations.

89. The cashflow liquidity stress test runs a large set of embedded scenarios of increasing severity, for a 4-week, 3 months as well as 5-day time horizon. The same scenarios are applied for SIs as well as LSIs. The following scenarios are included for each time-horizon: a baseline scenario and four stress scenarios with increasing severity (mild market stress, medium market stress, severe market stress and a most severe market stress). Each of the stress scenarios is combined with three different approaches to the counterbalancing capacity (Figure 22):

• Full CBC: fully endogenous liquidity supply by the central bank as long as banks have unencumbered eligible collateral;

• Full CBC with market haircuts: a full CBC is assumed, but market-specific haircuts³⁷ and bank-specific market price effects are imposed on elements of the CBC.

• Partial CBC with market haircuts: non-marketable components of the counterbalancing capacity (i.e., credit claims and committed lines provided to the banks) are disregarded; market haircuts are imposed again.

View Full Size

Stress Test Scenarios

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

90. The results of the liquidity stress test for a 4-week horizon show that liquidity shocks can be substantial, but that most SI banks are able to cope even under the highest stress scenarios. SIs are resilient in the face of a four-week shock, as liquidity remains ample in the system (Figure 23), and potential liquidity needs become prevalent only under the most severe scenario. However, the liquidity risk exposure of the sample is relatively high for SIs, as the average scenario impact can range from -3 percent of total assets in a mild market, to -12 percent of total assets in a most severe market, and total CBC as a share of their total assets can decline from 15 percent to 5 percent. Potential liquidity needs could reach 12 billion euros in the most severe market scenario.

View Full Size

Cash Flow Liquidity Stress Test

(in billions)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Source: IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

91. Some LSIs on the other hand are less resilient when faced with the same adverse shocks for a 4-week horizon. While most LSIs are also resilient to significant liquidity shocks; potential liquidity needs become prevalent even in a mild market scenario for a few of them (Figure 23 and Tables 10 and 11). Specifically, LSIs face CBC shortfalls which could range from 20 million euros in a medium stress market scenario, to almost 1.1 billion euros in the most severe market scenario. Additionally, the scenario impact can range from -3 percent to -13 percent of LSIs’ total assets.

^{Table 10.}

Italy: Banking Sector Liquidity Cashflow Stress Test Results, One Month Horizon

Source: IMF staff calculations.

^{Table 10.}

Italy: Banking Sector Liquidity Cashflow Stress Test Results, One Month Horizon

	SIs	LSIs
	Mild Market	Medium Market	Severe Market	Most Severe Market	Mild Market	Medium Market	Severe Market	Most Severe Market
Scenario 1
Remaining CBC as a share of total Assets	15%	12%	10%	8%	21%	20%	20%	18%
Number of banks unable to maintain a positive CBC	0	0	0	1	0	2	2	4
Scenario 2
Remaining CBC as a share of total Assets	15%	10%	7%	5%	17%	15%	14%	12%
Number of banks unable to maintain a positive CBC	0	0	0	2	0	3	7	11
Scenario 3
Remaining CBC as a share of total Assets	13%	10%	7%	5%	16%	14%	13%	11%
Number of banks unable to maintain a positive CBC	0	0	0	3	1	4	7	15

Source: IMF staff calculations.

View Table

^{Table 10.}

Italy: Banking Sector Liquidity Cashflow Stress Test Results, One Month Horizon

	SIs	LSIs
	Mild Market	Medium Market	Severe Market	Most Severe Market	Mild Market	Medium Market	Severe Market	Most Severe Market
Scenario 1
Remaining CBC as a share of total Assets	15%	12%	10%	8%	21%	20%	20%	18%
Number of banks unable to maintain a positive CBC	0	0	0	1	0	2	2	4
Scenario 2
Remaining CBC as a share of total Assets	15%	10%	7%	5%	17%	15%	14%	12%
Number of banks unable to maintain a positive CBC	0	0	0	2	0	3	7	11
Scenario 3
Remaining CBC as a share of total Assets	13%	10%	7%	5%	16%	14%	13%	11%
Number of banks unable to maintain a positive CBC	0	0	0	3	1	4	7	15

Source: IMF staff calculations.

^{Table 11.}

Italy: Banking Sector Liquidity Cashflow Stress Test Results, Three Months Horizon

Source: IMF staff calculations.

^{Table 11.}

Italy: Banking Sector Liquidity Cashflow Stress Test Results, Three Months Horizon

	SIs	LSIs
	Mild Market	Medium Market	Severe Market	Most Severe Market	Mild Market	Medium Market	Severe Market	Most Sever eMarket
Scenario 1
Remaining CBC as a share of total Assets Number of banks unable to maintain a positive CBC	12% 0	10% 0	5% 2	2% 5	20% 3	19% 4	14% 14	10% 23
Scenario 2
Remaining CBC as a share of total Assets Number of banks unable to maintain a positive CBC	8% 0	5% 0	3% 3	1% 6	16% 3	14% 6	8% 25	4% 30
Scenario 3
Remaining CBC as a share of total Assets Number of banks unable to maintain a positive CBC	10% 0	7% 0	2% 6	-2% 8	15% 4	13% 8	7% 27	3% 32

Source: IMF staff calculations.

View Table

^{Table 11.}

Italy: Banking Sector Liquidity Cashflow Stress Test Results, Three Months Horizon

	SIs	LSIs
	Mild Market	Medium Market	Severe Market	Most Severe Market	Mild Market	Medium Market	Severe Market	Most Sever eMarket
Scenario 1
Remaining CBC as a share of total Assets Number of banks unable to maintain a positive CBC	12% 0	10% 0	5% 2	2% 5	20% 3	19% 4	14% 14	10% 23
Scenario 2
Remaining CBC as a share of total Assets Number of banks unable to maintain a positive CBC	8% 0	5% 0	3% 3	1% 6	16% 3	14% 6	8% 25	4% 30
Scenario 3
Remaining CBC as a share of total Assets Number of banks unable to maintain a positive CBC	10% 0	7% 0	2% 6	-2% 8	15% 4	13% 8	7% 27	3% 32

Source: IMF staff calculations.

92. On aggregate, SI banks are moderately resilient to shocks over the longer, three-month horizon, while LSIs face liquidity strains in the events of such shocks. Under the most severe market scenario, the average impact of a scenario can amount to -18 percent of SI’s total assets, and potential liquidity needs could reach up to 100 billion euros. LSIs face further challenges to maintain a positive CBC, as potential liquidity needs can arise early in a mild market scenario (-1.4 million euros) and reach almost 11 billion euros in the most severe market scenario. Additionally, 32 out of the 55 LSIs in the sample are not able to maintain a positive CBC in case of a most severe market stress.

93. SI and LSIs are resilient in the event of a five-day stress horizon. The liquidity stress test results for the five-day time horizon show that all SI and LSIs banks can maintain a positive CBC during a five-day stress horizon. As for liquidity, it remains ample enough to sustain the severity of the shocks: CBC as a share of total assets declines to 7 percent in the most severe scenario for SIs, while it remains as high as 17 percent of total assets in the most severe scenario for LSIs.

94. The analysis reveals that banks are resilient to liquidity shocks, but less so in a prolonged stress environment. Most of the SIs and LSIs remain liquid at the one-month horizon, even in the severe scenario, but less so at the three months horizons. The banks that feature a negative CBC after stress are very heterogeneous with respect to the causes of their liquidity problems: some have low initial CBCs, other feature relatively large shares of credit claims in their initial CBC; some have high outflows due to committed lines to customers, others due to outflows from deposits of financial institutions or other deposits.

A. Drivers of Profitability

96. This section looks into the underlying reasons for low profitability in the Italian banking system. The analysis relies on a dataset of 381 banks from S&P’s SNL database, and covers the changes in the above-mentioned sample of banks from 2010 to 2017. The coverage of the dataset is inclusive to banks with different business models: it includes 99 joint stock (SpA) companies, 26 mutual (“popolari”) banks, and 254 credit cooperative banks (BCCs). The banks are grouped by size: the largest banks are those with balance sheets larger than 30 bn euros; the mid-sized banks are the ones with balance sheets between 4 bn euros and 30 bn euros; and the smallest sized banks are those with balance sheets smaller than 4 bn euros.

97. A system-wide view on profitability shows that, despite strong fluctuations during the past years, profitability rebounded in 2017–18 to levels close to the EU median (Figure 24). However, the rebound in profitability has been largely driven by the group of large banks. Profitability for medium and small banks continue to be low, although it has significantly improved since the twin crises.

View Full Size

Banks’ Profitability

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: SNL and IMF staff calculations.

• Download Figure
• Download figure as PowerPoint slide

98. Low profitability for small- and mid-sized banks is hampered by a mix of structural and cyclical factors. Small- and mid-sized banks retain high operating costs compared to large banks, and to the EU median in general. The reduction in operating costs by large banks was not equally matched by small- and medium-sized banks. Figure 24 depicts the consolidation efforts made by banks in Italy over the period of 2010–17.

99. Cyclical factors, such as low and negative interest rates, also act as headwinds to banks’ profitability. Data shows that net interest income as a share of operating income declined significantly for all banks from 2010 to 2017, as the policy rates declined in the EA (Figure 25).

View Full Size

Drivers of Profitability, 2010–2018

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: SNL data and IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

100. The FSAP looked at the changes to the main contributing factors to profitability, proxied by return on assets (ROA), in the Italian banking sector between 2010 and 2018. The waterfall charts reported in Figure 25 present the changes in drivers of profitability between 2010 and 2017 to the Italian banking sector, as well as large-sized banks, medium-sized banks, and small-sized banks. The waterfall charts focus on the following drivers of profitability: non-interest income, operating expenses, provision, other non-operating items, taxes, net interest income, and an “other” category, which is a residual. All items are divided by total assets.

101. The improvement in systemwide profitability between 2010 and 2018 was mainly driven by the performance of large banks. Profitability rebounded in the banking system between 2010 and 2018, driven by increases in non-interest income and other non-operating items, an improvement in tax revenue, and the reduction of operating expenses and provision needs. However, the breakdown of the banking system by size shows that on net, only large banks’ profitability has improved over the period. In comparison, ROA of medium- and small-sized banks is lower in 2018 compared to 2010.

102. The decomposition of the change in ROA between 2010 and 2018 shows that net interest income has had the most negative impact on banks’ profitability. The waterfall charts show that for the banking sector, large-sized banks, and small-sized banks, net interest income’s contribution to the change in ROA has been significant.³⁸ However, medium-sized banks appear to have been able to moderate the impact of low interest rates on their NII, with a resulting positive contribution between 2010 and 2018. On the other hand, all banks’ profitability has benefitted from an increase in non-interest income, reinforcing what was above-mentioned regarding substitution of interest income with other sources of income.

103. Medium-sized banks’ profitability continues to be negatively impacted by the increased need for provisioning. As for large banks, the waterfall chart shows that provisions decreased in 2018 compared to 2010. This is because provisioning expenses significantly increased between 2012 and 2014 but have been on a declining trend since.

104. Large banks have reduced their operating expenses in an effort to boost profitability. On net, the banking system has reduced its operating expenses between 2010 and 2018, as the largest banks reduced their expenses the most. Meanwhile, operating expenses increased for medium and small banks, further pressuring profitability.

B. The Impact of TLTRO on Banks’ Profitability

105. The FSAP conducted analysis to estimate the support that the ECB’s TLTRO provides to banks’ profitability. Italy’s banks are main users of these operations, with about 200 billion euros of the ECB’s TLTROs allotted to Italy’s SIs. Banks currently benefit from a negative interest rate of -0.4 percent on TLTRO II borrowings (estimated at about 2 percent of SIs’ NII on average). Furthermore, replacing TLTRO funding with market-based sources will be costlier for banks. In March 2019, the ECB announced a new series of operations (TLTRO-III), starting in September 2019 and ending in March 2021, each with a maturity of two years. TLTRO-III will continue to support the banking system, reducing the need for banks to tap more expensive funding sources in the near term.

106. The market cost of an equivalent amount to TLTRO financing depends significantly on the funding mix and market conditions. If banks used short-term funding while maintaining comfortable LCRs and NSFRs (e.g., with the ECB’s main refinancing operations window at a zero percent rate or short-term retail deposits at a 0.38 percent rate),³⁹ they will incur relatively low funding costs. However, longer-term market funding would be costlier (Table 12). Banks’ CDS spreads have risen 25 percent on average in the period after May 2018: bonds issued by banks during the second half of 2018 were priced at much higher prices compared to early 2018 or 2017. More recently, banks’ CDS spreads have declined. Based on reported market prices for bond issuance in the second half of 2018, retail deposit rates, and the ECB’s main refinancing window, Table 13 and Figure 26 present an illustrative range of estimates of the market-based cost of an equivalent amount to TLTRO financing.

^{Table 12.}

Italy: Bond Issuances and Spreads

Source: Bloomberg, IMF staff estimates. Note: ¹ UniCredit did not provide an equivalent for the spread via the EUR mid-swap rate. IMF staff roughly estimate it to be around 190 bps via the 5yr EUR mid-swap rate.

^{Table 12.}

Italy: Bond Issuances and Spreads

(Bond Issuance, May 2018-January 2019)
Issuer Name	Date of Issuance	Amount Issued	Maturity	Payment Rank	Spread Vs. EUR 5yr Mid-Swap Rate
UniCredit	11/27/2018	USD 3 billion	12/4/2023	Sr Non-Preferred	420 bps
UniCredit	1/15/2019	USD 2.5 billion	1/14/2022	Sr Non-Preferred	365 bps
Intesa	8/30/2018	EUR 1.1 billion	8/30/2023	Sr Preferred	188 bps
MPS	1/29/2019	EUR 1.1 billion	1/29/2024	Secured	190 bps
Bond Issuance, prior to May 2018
UniCredit	1/18/2018	EUR 1.8 billion	1/18/2023	Sr Non-Preferred	70 bps
UniCredit	4/12/2017	USD 1.2 billion	4/12/2022	Sr Preferred	200 bps1
Intesa	3/20/2018	EUR 1.5 billion	3/20/2028	Sr Preferred	77 bps

Source: Bloomberg, IMF staff estimates. Note: ¹ UniCredit did not provide an equivalent for the spread via the EUR mid-swap rate. IMF staff roughly estimate it to be around 190 bps via the 5yr EUR mid-swap rate.

View Table

^{Table 12.}

Italy: Bond Issuances and Spreads

(Bond Issuance, May 2018-January 2019)
Issuer Name	Date of Issuance	Amount Issued	Maturity	Payment Rank	Spread Vs. EUR 5yr Mid-Swap Rate
UniCredit	11/27/2018	USD 3 billion	12/4/2023	Sr Non-Preferred	420 bps
UniCredit	1/15/2019	USD 2.5 billion	1/14/2022	Sr Non-Preferred	365 bps
Intesa	8/30/2018	EUR 1.1 billion	8/30/2023	Sr Preferred	188 bps
MPS	1/29/2019	EUR 1.1 billion	1/29/2024	Secured	190 bps
Bond Issuance, prior to May 2018
UniCredit	1/18/2018	EUR 1.8 billion	1/18/2023	Sr Non-Preferred	70 bps
UniCredit	4/12/2017	USD 1.2 billion	4/12/2022	Sr Preferred	200 bps1
Intesa	3/20/2018	EUR 1.5 billion	3/20/2028	Sr Preferred	77 bps

Source: Bloomberg, IMF staff estimates. Note: ¹ UniCredit did not provide an equivalent for the spread via the EUR mid-swap rate. IMF staff roughly estimate it to be around 190 bps via the 5yr EUR mid-swap rate.

View Full Size

Replacement of TLTRO

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Bloomberg; European Central Bank; Banca d’Italia; and IMF staff estimates.

• Download Figure
• Download figure as PowerPoint slide

^{Table 13.}

Italy: Illustrative Estimates of Annual Market-based Cost of an Equivalent TLTRO Funding

(as of end-2018)

Sources: Bloomberg; European Central Bank; Banca d’Italia; and IMF staff estimates. Note: ¹This estimate assumes banks will replace TLTRO liquidity by using half of their excess reserves and financing the rest.

^{Table 13.}

Italy: Illustrative Estimates of Annual Market-based Cost of an Equivalent TLTRO Funding

(as of end-2018)

Types of Funding	Funding Cost (percent)	Gross TLTRO	TLTRO, Net of 50% of Banks’ Excess Reserves1
		Impact of Gross TLTRO Replacement (Percent of banks’ equity)	Impact of Net TLTRO Replacement (Percent of banks’ equity)
Retail deposit rate	0.4	0.5	0.4
ECB’s MRO	0.0	0.0	0.0
Long term bond – low volatility period (Intesa bond issuance, 2018)	1.8	2.2	1.8
Long term bond – high volatility period (UniCredit bond issuance, 2018)	4.2	5.2	4.2
Funding Mix 1: 30% Retail, 30% MRO,40% Bonds (Intesa price)		1.3	1.0
Funding Mix 2: 30% Retail, 30% MRO,40% Bonds (UniCredit price)		2.7	2.2
Funding Mix 3: 50% MRO, 50% Bonds (Intesa price)		1.1	0.9
Funding Mix 4: 50% MRO, 50% Bonds (Unicredit price)		2.6	2.1

Sources: Bloomberg; European Central Bank; Banca d’Italia; and IMF staff estimates. Note: ¹This estimate assumes banks will replace TLTRO liquidity by using half of their excess reserves and financing the rest.

View Table

^{Table 13.}

Italy: Illustrative Estimates of Annual Market-based Cost of an Equivalent TLTRO Funding

(as of end-2018)

Types of Funding	Funding Cost (percent)	Gross TLTRO	TLTRO, Net of 50% of Banks’ Excess Reserves1
		Impact of Gross TLTRO Replacement (Percent of banks’ equity)	Impact of Net TLTRO Replacement (Percent of banks’ equity)
Retail deposit rate	0.4	0.5	0.4
ECB’s MRO	0.0	0.0	0.0
Long term bond – low volatility period (Intesa bond issuance, 2018)	1.8	2.2	1.8
Long term bond – high volatility period (UniCredit bond issuance, 2018)	4.2	5.2	4.2
Funding Mix 1: 30% Retail, 30% MRO,40% Bonds (Intesa price)		1.3	1.0
Funding Mix 2: 30% Retail, 30% MRO,40% Bonds (UniCredit price)		2.7	2.2
Funding Mix 3: 50% MRO, 50% Bonds (Intesa price)		1.1	0.9
Funding Mix 4: 50% MRO, 50% Bonds (Unicredit price)		2.6	2.1

Sources: Bloomberg; European Central Bank; Banca d’Italia; and IMF staff estimates. Note: ¹This estimate assumes banks will replace TLTRO liquidity by using half of their excess reserves and financing the rest.

107. Costs of replacing TLTRO by market funding can have a meaningful impact on banks’ profitability. Based on reported market prices for bond issuance, retail deposit rates, and the ECB’s main refinancing window, we calculate the potential impact of replacing TLTROs on banks’ equity. Namely, we consider that if banks replace TLTRO by five-year bond issuance, then the price of issuance could range from “188 bps + mid-swap rate” to “420 bps + mid-swap rate”. Additionally, we consider the scenario when banks can use up to half of their excess reserves at the ECB for their financing needs (“net TLTRO” scenario). If banks were to use funding mixtures, replacement costs could vary from 1–3 percent of banks’ equity on aggregate, but the impact varies significantly by bank. The cost from TLTRO refinancing could also be passed-through to lending rate or mitigated by deleveraging, which are not considered in this analysis.

A. Aggregate Balance Sheet Developments in the Non-Financial Sector

111. The interconnectedness analysis conducted by the FSAP indicates that households are substantial net lenders. Households are by far the ‘richest’ institutional sector in Italy with overall household debt well below European peers and accumulated net financial assets of about 200 percent of GDP by end-2007, up from 156 percent of GDP at end-1995. Together with real assets the net wealth of Italian households can be measured at about 5½ times the GDP (Bank of Italy, 2015). Over the past three years, despite declining real estate prices, households have benefitted from favorable credit conditions (Bank of Italy, 2018c). Their real gross disposable income grew at an average rate of 1 percent per year. Until recently, employment has been increasing at a three-year average rate of about 1 percent—the fastest growth in a decade—and labor market conditions have tightened. Households’ bad debt to banks account only for about one-fifth of the economy-wide bad debt, or just above 5 percent of bank loans to households, posing relatively modest risks to the banking sector. According to Banca d’Italia (2019) the share of vulnerable households in recent years has been around 11–12 percent, down from a peak of about 19 percent during the 2012 recession. Households in Italy are therefore not likely to be a major source of macro-financial risks to banks.

View Full Size

Recent Evolution of Net Financial Assets by Sector

(Percent of GDP)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

• Download Figure
• Download figure as PowerPoint slide

112. The non-financial corporate sector gross debt to surplus ratio is close to the euro area average (text chart). After peaking in 2012, it has fallen to just above the 2017 levels. With net financial assets at -112 percent of GDP, the corporate sector is slightly more indebted than the euro area average (see text chart in preceding paragraph). Corporate indebtedness imposes a significant drag on total factor productivity growth (Anderson and Raissi, 2018) that has been persistently anemic in Italy, declining more in frontier manufacturing firms (OECD, 2017; Kangur, 2018), and lagging the EA.

View Full Size

Non-financial Corporates Debt-to-Surplus Ratio

(Debt to gross operating surplus)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: OECD.Note: OECD (2019) defines non-financial corporations debt as the sum of securities other than shares except financial derivates, loans, and other accounts payable.

• Download Figure
• Download figure as PowerPoint slide

113. The corporate sector has been the primary source of bad debts for banks and faces highly differentiated lending rates. Bad loan inflows from the corporate sector have declined notably (text chart). And while accommodative monetary policy eased financing conditions, bank credit growth to NFCs has been weak. Such dynamics reflect the selectiveness of banks (Banca d’Italia, 2019) as tight credit conditions have been especially applied to riskier firms who face higher cost of borrowing and tend to have lower profitability (e.g., analysis by Cerved, 2017 concludes that SMEs face highly differentiated lending rates depending on their size and riskiness). As a result, lending to corporations as a share of GDP has steadily declined.

View Full Size

Bad Debts by Institutional Sector

(percent of bank loans, in percent)

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: Haver Analytics, Bank of Italy, and staff calculations.

• Download Figure
• Download figure as PowerPoint slide

114. Dominated by small firms, the Italian corporate sector is generally less profitable than its euro area peers. Gross national income (GNI) of NFCs before interest and taxes—a national accounts measure of profitability—has declined throughout the past decade and remains about 7 percentage points below the euro area average as share of gross value added (Figure 27). As Italian corporates generally bear heavier effective tax burden compared to its European peers, the decline in after-tax capital return has been more pronounced. The decline in these national accounts measures of profitability are mostly reflected in lower dividends—distributed income of corporations—as well as net investment. Such internal adjustment pattern of lower profit margins and reductions in real quantities is the result of firms facing rigid nominal wages, a high unit labor cost (ULC) gap with their European peers by over 30 percent and low productivity (see Kangur, 2018).⁴²

View Full Size

Non-Financial Corporate Sector in National Accounts Statistics

Citation: IMF Staff Country Reports 2020, 238; 10.5089/9781513552217.002.A001

Sources: CERVED; and IMF staff calculations.Note: The calculations and source data for capital return are taken from Garcia-Macia (2019).

• Download Figure
• Download figure as PowerPoint slide

115. While the corporate sector debt-servicing capacity has improved in recent years, the sector remains sensitive to shocks. Corporate sector vulnerabilities have receded in recent years against the background of prolonged cyclical monetary accommodation and fiscal support. Italian corporates have benefited more than other countries in the euro area from the monetary stimulus that has contributed to a decline in interest expenses as a share of value added from around 8–10 percent pre-crisis to below 2 percent in 2017. This in turn has more than doubled the ratio of GNI before interest and taxes to interest expense that in 2017 reached its historical peak of 13.5 per cent (compared to an average of about 5 percent between 2002 and 2007). At the same time, lending rates are highly differentiated and remain high for smaller and more vulnerable companies, unit labor costs remain elevated and corporate profitability and net investment low. Important segments of the corporate sector can thus remain vulnerable to shocks, contributing to higher debt-at-risk and posing risks for banks’ asset quality.

B. Financial Vulnerabilities at a Micro-Level

Overview of the Corporate Sector through the Lens of Micro-data

116. The corporate sector stress test relies on firm-level balance sheet data. The firm-level database, compiled by CERVED, provides detailed balance sheet and income statement data on nearly the whole population of Italian non-financial corporates. The sample covers the period 2002–17 and comprises of more than 10 million observations for 1.5 million firms.⁴³ The coverage reflects the structure of the Italian corporate sector, with about 80 percent of firms in 2017 comprising of micro-firms with less than 10 employees and 1 percent of firms comprising of large companies with more than 249 employees.⁴⁴ According to the Structural Business Survey conducted by Eurostat micro-enterprises account for about 28 percent of the value added of the total business economy.

Sample Coverage by Firm Size 2002–2017

Note: the sum of micro, small, medium, and large firms can exceed the reported total due to some firms switching the size categories over time.

Sample Coverage by Firm Size 2002–2017

Firm size		Observations	Firms
Total		10,214,845	1,531,853
	Micro	8,682,340	1,454,908
	Small	1,127,716	234,468
	Medium	327,110	55,853
	Large	77,679	11,335

Note: the sum of micro, small, medium, and large firms can exceed the reported total due to some firms switching the size categories over time.

View Table

Sample Coverage by Firm Size 2002–2017

Firm size		Observations	Firms
Total		10,214,845	1,531,853
	Micro	8,682,340	1,454,908
	Small	1,127,716	234,468
	Medium	327,110	55,853
	Large	77,679	11,335

Note: the sum of micro, small, medium, and large firms can exceed the reported total due to some firms switching the size categories over time.

117. CERVED database provides consistency across time, firm size, and other characteristics. While the coverage of firms increases gradually across time, the structure of the database remains broadly consistent. Two structural features warrant highlighting. First, the database standardizes accounting presentations across industries that in the Italian law can follow different reporting standards. Second, firms with simplified accounting rules are not obliged to report their full liability structure. As these are mostly micro firms and tend to be more vulnerable compared to sample average, relying on sub-components of debt can bias the analysis of debt-at-risk. This, however, will not affect the analysis of debt-at-risk based on aggregate debt. The database is cleaned by dropping all firms for which at any point in time observations on total assets, tangible fixed assets, or sales are zero or missing. Following Anderson and Raissi (2018) we drop a small number of observations where total assets in any given year either increase or decrease by more than 50 times compared to the previous year to alleviate problems with extreme outliers. Finally, with a decline in interest rates in recent years an increasing number of mostly micro-firms—reaching to 40 percent of total firms and holding up to 9 percent of total debt in 2017—report zero interest expenses. This is exacerbated by data rounding up to thousands of euros that would lead to an optimistic bias if the observations were treated as zeros. To alleviate this issue, we drop observations with zero interest payments if financial debt is positive for two consecutive years. In a very small number of cases we impute interest expenses based on an average effective interest rate (in case of non-zero debt).

118. Cyclical factors have supported balance sheet recovery while profitability remains structurally low.⁴⁵ A period of economic recovery, driven largely by cyclical factors such as a supportive monetary environment⁴⁶ and corporate tax incentives, have allowed firms to improve their liquidity positions and reverse the crisis-led trend of declining net investment. Over the period of 2015–17, following the twin crisis, liquidity positions of Italian corporates improved markedly (Table 14). Corporates have used this time to reduce their leverage both by reducing gross debt stock that now stands at close to 2007 levels and by building up equity. Still, profitability has been slow to recover. In 2017 both average and median EBIT over assets remain about 6–7 percent below average pre-crisis levels and about 17–18 percent below the 2007 peaks. While median profitability remains below the pre-crisis average levels for all firm size classes in a range of 14–2 percent, average EBIT per assets remains particularly low for micro and large companies. It is also noteworthy that across all indicators median values are worse than average and show smaller improvement, indicating a smaller set of high-performing firms and a longer tail of weaker firms.

^{Table 14.}

Italy: Firm Characteristics in Micro-Data

Source: CERVED; and IMF staff calculations.

^{Table 14.}

Italy: Firm Characteristics in Micro-Data

Share of Firms by Firm Size
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	100	100	100	100	100	100	100	100	100	100	100	100	100	100	100	100
	Micro	85	85	85	85	85	85	85	86	86	86	86	86	86	86	81	78
	Small	11	10	10	11	11	11	11	10	10	10	10	10	10	10	14	17
	Medium	4	3	3	3	3	3	3	3	3	3	3	3	3	3	3	4
	Large	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
Liquidity Ratio by Firm Size (Median, in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	4.1	4.0	4.0	4.2	4.3	4.1	3.5	3.6	3.6	3.5	3.2	3.5	4.0	4.6	5.3	5.5
	Micro	4.4	4.3	4.3	4.5	4.7	4.5	3.8	3.8	3.8	3.7	3.4	3.6	4.0	4.6	5.2	5.3
	Small	2.8	2.7	2.8	3.1	3.1	2.9	2.3	2.6	2.7	2.6	2.6	3.1	3.8	4.6	5.8	6.4
	Medium	2.8	2.8	2.9	3.1	3.1	2.8	2.2	2.5	2.6	2.3	2.5	3.1	3.6	4.3	5.3	5.9
	Large	2.3	2.3	2.6	2.5	2.6	2.5	2.1	2.3	2.4	2.1	2.4	2.8	3.1	3.8	4.2	4.7
Liquidity Ratio by Firm Size (Average , in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	5.6	5.5	6.1	6.1	5.9	5.8	5.2	5.6	5.6	5.5	5.9	6.3	7.0	7.4	8.0	8.4
	Micro	7.2	7.0	6.9	7.2	7.2	7.1	6.1	6.0	5.9	5.9	5.7	5.9	6.5	7.0	7.6	8.0
	Small	7.2	6.6	7.2	7.2	7.0	7.0	6.1	6.5	6.6	6.7	6.9	7.6	7.9	8.7	9.6	10.2
	Medium	5.7	6.3	6.3	6.8	6.8	6.2	5.4	6.1	6.6	6.1	6.4	7.2	7.7	8.5	9.4	10.1
	Large	4.5	4.3	5.4	5.0	4.7	4.9	4.4	5.0	4.8	4.7	5.6	5.9	6.7	6.9	7.2	7.5
Debt-to-Equity Ratio by Firm Size (Median)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	3.6	3.5	3.6	3.6	3.7	3.7	3.1	3.0	3.0	3.0	2.8	2.7	2.6	2.6	2.6	2.7
	Micro	3.3	3.3	3.3	3.4	3.4	3.4	2.9	2.8	2.9	2.8	2.6	2.6	2.5	2.5	2.5	2.6
	Small	5.1	5.1	5.2	5.3	5.5	5.5	4.2	4.0	4.1	4.1	3.7	3.6	3.4	3.4	3.5	3.4
	Medium	3.9	3.9	3.9	3.9	4.1	4.2	3.1	3.0	3.0	3.0	2.8	2.6	2.6	2.5	2.5	2.5
	Large	3.6	3.5	3.5	3.5	3.6	3.7	3.1	2.9	3.0	2.9	2.8	2.6	2.5	2.4	2.3	2.3
Debt-to-Equity Ratio by Firm Size (Average)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	2.7	2.6	2.5	2.6	2.6	2.7	2.4	2.3	2.3	2.4	2.4	2.3	2.2	2.1	1.9	1.9
	Micro	3.8	3.8	3.7	3.6	3.7	3.7	2.7	2.8	2.9	2.9	3.1	2.9	2.9	2.8	2.6	2.3
	Small	3.4	3.1	3.4	3.3	3.1	3.4	2.7	2.6	2.7	2.7	2.7	2.5	2.3	2.2	2.2	2.1
	Medium	2.7	2.7	2.5	2.5	2.7	2.7	2.3	2.2	2.2	2.3	2.3	2.1	2.0	1.9	1.8	1.8
	Large	2.3	2.1	2.1	2.1	2.1	2.3	2.2	2.1	2.0	2.1	2.1	2.0	1.9	1.8	1.7	1.8
EBIT per Total Assets by Firm Size (Median, in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	3.6	3.3	3.3	3.3	3.8	4.0	3.4	2.5	2.6	2.7	2.3	2.4	2.7	2.9	3.1	3.3
	Micro	3.2	2.9	2.9	3.0	3.5	3.7	3.1	2.3	2.4	2.5	2.1	2.2	2.4	2.7	2.8	2.9
	Small	5.1	4.6	4.4	4.3	4.7	5.0	4.4	3.4	3.3	3.4	3.1	3.4	3.7	3.9	4.0	4.0
	Medium	4.9	4.4	4.4	4.2	4.6	4.9	4.2	3.1	3.3	3.4	3.0	3.4	3.8	4.0	4.1	4.2
	Large	4.7	4.6	4.5	4.5	4.9	5.0	4.4	3.5	3.8	3.8	3.5	3.8	4.1	4.4	4.5	4.6
EBIT per Total Assets by Firm Size (Average, in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	4.4	4.1	4.3	4.5	4.9	5.2	4.3	3.2	3.7	3.4	2.9	3.0	3.2	3.7	4.0	4.3
	Micro	2.9	2.4	2.2	2.2	2.8	3.5	2.3	1.2	1.4	1.5	0.8	0.8	1.1	1.3	1.7	2.2
	Small	4.9	4.1	4.1	4.2	4.3	4.9	3.9	2.6	2.8	2.8	2.1	2.6	3.1	3.7	4.1	4.7
	Medium	4.2	4.4	4.2	4.2	4.8	4.8	3.7	2.6	3.2	3.1	2.5	3.1	3.8	4.2	4.7	5.0
	Large	4.9	4.7	5.2	5.5	6.0	6.1	5.5	4.4	5.1	4.6	4.3	4.0	4.0	4.6	4.7	4.7

Source: CERVED; and IMF staff calculations.

View Table

^{Table 14.}

Italy: Firm Characteristics in Micro-Data

Share of Firms by Firm Size
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	100	100	100	100	100	100	100	100	100	100	100	100	100	100	100	100
	Micro	85	85	85	85	85	85	85	86	86	86	86	86	86	86	81	78
	Small	11	10	10	11	11	11	11	10	10	10	10	10	10	10	14	17
	Medium	4	3	3	3	3	3	3	3	3	3	3	3	3	3	3	4
	Large	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
Liquidity Ratio by Firm Size (Median, in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	4.1	4.0	4.0	4.2	4.3	4.1	3.5	3.6	3.6	3.5	3.2	3.5	4.0	4.6	5.3	5.5
	Micro	4.4	4.3	4.3	4.5	4.7	4.5	3.8	3.8	3.8	3.7	3.4	3.6	4.0	4.6	5.2	5.3
	Small	2.8	2.7	2.8	3.1	3.1	2.9	2.3	2.6	2.7	2.6	2.6	3.1	3.8	4.6	5.8	6.4
	Medium	2.8	2.8	2.9	3.1	3.1	2.8	2.2	2.5	2.6	2.3	2.5	3.1	3.6	4.3	5.3	5.9
	Large	2.3	2.3	2.6	2.5	2.6	2.5	2.1	2.3	2.4	2.1	2.4	2.8	3.1	3.8	4.2	4.7
Liquidity Ratio by Firm Size (Average , in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	5.6	5.5	6.1	6.1	5.9	5.8	5.2	5.6	5.6	5.5	5.9	6.3	7.0	7.4	8.0	8.4
	Micro	7.2	7.0	6.9	7.2	7.2	7.1	6.1	6.0	5.9	5.9	5.7	5.9	6.5	7.0	7.6	8.0
	Small	7.2	6.6	7.2	7.2	7.0	7.0	6.1	6.5	6.6	6.7	6.9	7.6	7.9	8.7	9.6	10.2
	Medium	5.7	6.3	6.3	6.8	6.8	6.2	5.4	6.1	6.6	6.1	6.4	7.2	7.7	8.5	9.4	10.1
	Large	4.5	4.3	5.4	5.0	4.7	4.9	4.4	5.0	4.8	4.7	5.6	5.9	6.7	6.9	7.2	7.5
Debt-to-Equity Ratio by Firm Size (Median)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	3.6	3.5	3.6	3.6	3.7	3.7	3.1	3.0	3.0	3.0	2.8	2.7	2.6	2.6	2.6	2.7
	Micro	3.3	3.3	3.3	3.4	3.4	3.4	2.9	2.8	2.9	2.8	2.6	2.6	2.5	2.5	2.5	2.6
	Small	5.1	5.1	5.2	5.3	5.5	5.5	4.2	4.0	4.1	4.1	3.7	3.6	3.4	3.4	3.5	3.4
	Medium	3.9	3.9	3.9	3.9	4.1	4.2	3.1	3.0	3.0	3.0	2.8	2.6	2.6	2.5	2.5	2.5
	Large	3.6	3.5	3.5	3.5	3.6	3.7	3.1	2.9	3.0	2.9	2.8	2.6	2.5	2.4	2.3	2.3
Debt-to-Equity Ratio by Firm Size (Average)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	2.7	2.6	2.5	2.6	2.6	2.7	2.4	2.3	2.3	2.4	2.4	2.3	2.2	2.1	1.9	1.9
	Micro	3.8	3.8	3.7	3.6	3.7	3.7	2.7	2.8	2.9	2.9	3.1	2.9	2.9	2.8	2.6	2.3
	Small	3.4	3.1	3.4	3.3	3.1	3.4	2.7	2.6	2.7	2.7	2.7	2.5	2.3	2.2	2.2	2.1
	Medium	2.7	2.7	2.5	2.5	2.7	2.7	2.3	2.2	2.2	2.3	2.3	2.1	2.0	1.9	1.8	1.8
	Large	2.3	2.1	2.1	2.1	2.1	2.3	2.2	2.1	2.0	2.1	2.1	2.0	1.9	1.8	1.7	1.8
EBIT per Total Assets by Firm Size (Median, in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	3.6	3.3	3.3	3.3	3.8	4.0	3.4	2.5	2.6	2.7	2.3	2.4	2.7	2.9	3.1	3.3
	Micro	3.2	2.9	2.9	3.0	3.5	3.7	3.1	2.3	2.4	2.5	2.1	2.2	2.4	2.7	2.8	2.9
	Small	5.1	4.6	4.4	4.3	4.7	5.0	4.4	3.4	3.3	3.4	3.1	3.4	3.7	3.9	4.0	4.0
	Medium	4.9	4.4	4.4	4.2	4.6	4.9	4.2	3.1	3.3	3.4	3.0	3.4	3.8	4.0	4.1	4.2
	Large	4.7	4.6	4.5	4.5	4.9	5.0	4.4	3.5	3.8	3.8	3.5	3.8	4.1	4.4	4.5	4.6
EBIT per Total Assets by Firm Size (Average, in percent)
Firm size		2002	2003	2004	2005	2006	2007	2008	2009	2010	2011	2012	2013	2014	2015	2016	2017
	Total	4.4	4.1	4.3	4.5	4.9	5.2	4.3	3.2	3.7	3.4	2.9	3.0	3.2	3.7	4.0	4.3
	Micro	2.9	2.4	2.2	2.2	2.8	3.5	2.3	1.2	1.4	1.5	0.8	0.8	1.1	1.3	1.7	2.2
	Small	4.9	4.1	4.1	4.2	4.3	4.9	3.9	2.6	2.8	2.8	2.1	2.6	3.1	3.7	4.1	4.7
	Medium	4.2	4.4	4.2	4.2	4.8	4.8	3.7	2.6	3.2	3.1	2.5	3.1	3.8	4.2	4.7	5.0
	Large	4.9	4.7	5.2	5.5	6.0	6.1	5.5	4.4	5.1	4.6	4.3	4.0	4.0	4.6	4.7	4.7

Source: CERVED; and IMF staff calculations.