## _wp16135

## Source details

**Canonical URL:** [_wp16135](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16135.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16135.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16135.pdf.json)

---

### I. Scale, overview, and key characteristics
- Nonperforming loans (NPLs) in Italy:
  - Tripled since the beginning of the crisis and, according to national definition, stood at €360 billion (18 percent of total outstanding loans) at end-2015.
  - Account for about a third of those in the entire euro area.
- Italian NPL categorization: impaired/bad debt or sofferenze (loans in a state of insolvency), substandard (incagli), overdue, and restructured.
- ECB/SSM 2014 asset quality review (largest 15 banks): aggregate nonperforming exposures ratio close to 22 percent (ECB’s broader exposure definition and post-AQR adjustments noted).
- Key drivers identified:
  - Over-indebted corporates following the crisis-related drop in output.
  - Banks generally low in capital buffers.
  - A highly complex legal system of corporate restructuring and insolvency and lengthy judicial processes.
  - A tax system that until recently discouraged NPL write-offs.
- Effects on the economy:
  - Gradual NPL write-offs; cost of risk and high administration costs for NPLs not fully factored into lending margins, weighing on profitability and impeding capital buildup.
  - Restricted bank financing and limited alternative corporate financing sources constrained investment and slowed the economic recovery.

### II. Distribution by size, sector, region, collateral, and provisioning
- Size distribution (bad debt, impaired loans, end-2014):
  - More than 75 percent of the total value of bad debt relates to loans above €250,000.
  - About 75 percent of the total number of borrowers with bad debt relate to loans of less than €75,000.
- Sectoral distribution:
  - About three quarters of bad debt are related to the corporate sector.
  - The service sector and less technology-intensive sectors are most affected.
- Regional distribution:
  - Pronounced north-south divide: by end-2014, most central and southern Italy regions saw bad debt (sofferenze) ratios increase above 20 percent (compared with most regions below 10 percent in early 2009).
- Collateral and provisioning:
  - Provisions stood at an average of 45.4 percent for total NPLs and 58.7 percent for bad loans as of December 2015.
  - Authorities estimate real estate collateral and personal guarantees against bad loans amounted to €87bn and €37bn respectively at end-2015.
  - Court times to access collateral are very long.

### III. Distribution across banks and data coverage
- System-wide characteristics:
  - Most banks have NPL ratios between 15 and 20 percent; some smaller banks face extremely high NPL ratios.
  - No clear correlation between NPL ratios (percent of total loans) and bank size.
  - As of December 2014, the five largest banks in Italy accounted for two-thirds of total bad debt and NPLs.
  - Large cooperative banks (banche popolari) faced the highest NPL ratios with the lowest coverage ratio at end-2014.
- Data samples:
  - Full SNL sample referenced: 402 Italian banks (summary data).
  - Full coverage time-series sample used for regressions: 62 banks with comprehensive data since 2005.
  - Some regressions used smaller bank counts (e.g., 326 and 57) due to limited time-series observations.

### IV. Econometric analysis — drivers of NPL buildup
- Methodology:
  - Dynamic panel regressions (fixed effects and system GMM) of NPLs (problem loans in percent of total gross loans) on macro and bank-specific variables; year dummies included; system GMM used to correct feedback biases.
  - Macro variables included lagged inflation, NEER percentage change, lagged real GDP growth, lagged percentage change in house and stock prices.
  - Bank variables included lagged equity/assets ratio, lagged ROE, lagged loan growth, lagged Tier 1 ratio.
- Key regression findings:
  - Both bank-level and macroeconomic factors significantly affected asset quality.
  - Lower past profitability (lagged ROE) associated with higher NPL levels.
  - Higher past lending (lagged loan growth) associated with higher NPLs.
  - Macroeconomic factors: lower growth, exchange rate appreciations, and falling house prices are significantly associated with higher NPLs (house prices significant in fixed effects, not in system GMM).
  - Exceptional duration and intensity of the recession had a profound impact on asset quality; incipient recovery may help reduce NPLs but moderate growth and inflation outlook imply additional measures will likely be needed.

### V. Default risk dynamics and Dynamic State Space Model (DSSM) insights
- Corporate and bank default probabilities peaked in mid-2012 and have come down substantially since mid-2012, while NPLs continued to rise until 2015Q3.
- DSSM properties and estimation:
  - DSSM allow time-varying coefficients and capture structural changes over time; estimation via Kalman Filter recursion.
  - State vector θt = (β1,t, …, βn,t)′; xt = (1, x1t, …, xnt); Ft as state transition matrix; Vt and Wt drive variances; nonzero Wt makes coefficients time-varying.
- DSSM empirical findings (Italy):
  - Time-varying “true elasticities” recovered for corporate default risk drivers; coefficients change over the sample.
  - Variables highlighted: Unemployment (share of potential workforce not employed), STIR (change in short-term interest rates), GDPreal (yoy change in real GDP growth), GovConsumption.Exp (change in government consumption expenditure).
  - Estimates suggest:
    - Further output losses would fuel credit risk considerably more than in the past.
    - Higher interest rates would now have a much more substantial impact on corporates than in the past.
  - Increases in interest rates would negatively impact the market value of banks’ holdings of Italian sovereign assets, given the substantial rise in sovereign bond volumes on banks’ books.
- Additional note: Italian banks have significant unrealized capital gains from sovereign bond holdings (BoI Financial Stability Report April 2015 referenced).

### VI. Accounting, tax treatment, and Deferred Tax Assets (DTAs)
- Historical tax and accounting features that affected provisioning:
  - Until 2013, a cap limited immediate tax deductibility to 0.3 percent of outstanding loans, with the remaining part treated as a deferred tax asset (DTA). This cap constituted a clear disincentive to provisioning.
  - Remaining provisions could be deducted in equal installments over a period of 18 years; DTAs were deductible from taxable revenues over a period of 18 years.
  - Since 2013, provisions and write-offs can be deducted in equal installments over five years, and with a higher tax rate; incentives for accelerated write-off have increased though still restrictive relative to other countries.
  - As of end-2014, DTAs not deducted from capital amounted to €43 bn.
  - As of August 2015, write-offs are now immediately tax deductible.
  - IAS 39 did not define when and how defaulted loans are to be written off; uncertainty under IFRS produced slow NPL write-offs and widespread application of a de-recognition rule beyond intended conditions.
  - Forthcoming IFRS 9 (scheduled for 2018) will include a clear definition of write-off and a definition different from loan cancellation; when applicable, banks need to write off earlier.
  - Accounting practices (use of original effective interest rate and treatment of indirect costs) reduce apparent NPL prices to potential acquirers and contributed to build-up of DTAs.

### VII. Banking sector capacity and composition of NPLs
- Larger banks:
  - Have disposed of NPLs, invested in internal NPL management, and set up decentralized AMCs; examples: UCG sold €2 bn in NPLs in fall 2014 and announced sale including a €2.4 bn NPL portfolio in February 2015; net of provisions, bad debts at UCG stood at about €36 bn in 2013.
- Medium and small banks:
  - Limited experience and capacity to deal with NPLs; lack risk management capacity, NPL management experience, and access to distressed debt markets.
- Sectoral composition:
  - Italian NPLs are mostly commercial, include a large number of micro and small enterprises, and are very heterogeneous (unlike other countries where NPLs were concentrated in real estate and easier to value).

### VIII. Legal and procedural obstacles to NPL resolution
- Two methods for dealing with NPLs: contractual (renegotiation or sale) and enforcement (individual enforcement such as mortgage foreclosure, or collective enforcement via insolvency).
- Italian insolvency regime complexities:
  - Multiple procedures and special regimes; lack of coordination and unified vision.
  - High procedural complexity contributes to litigiousness and debtor delaying strategies.
  - Informal/hybrid debt restructuring procedures often lack flexibility and aim for full payment; inability to bind dissenting creditors reduces efficiency.
  - Formal insolvency procedures can bind minority creditors and allow flexible solutions, though recent reforms imposed limits on haircuts unsecured creditors may accept in certain plans.
- Outcomes and delays:
  - In 2014, enterprise bankruptcies surpassed 15,000 cases (+10.7 percent vs. 2013).
  - Estimated more than 82,000 enterprises liquidated in insolvency processes since 2008, with associated loss of about a million jobs (Cerved 2015).
  - Enterprise reorganizations declined by 20 percent in 2014; for 2015, enterprise insolvencies stabilized at 14,700 cases and reorganizations continued to fall.
  - Liquidations on average lasted more than eight years, with preferential and unsecured creditors recovering 29 and 6 percent of claims, respectively.
- Institutional bottlenecks:
  - Judiciary overburdened and lacks specialization; enterprise courts do not have insolvency competencies; insolvency administrators lack specialization and are focused on liquidation rather than reorganization.

### IX. Recent policy measures and authorities’ actions
- Governance and sector reforms:
  - March 2015: reform of governance of cooperative banks — remove one vote per head and limitations on size of individual shareholdings; aim to promote consolidation, balance sheet clean-up, and improve profitability.
  - April 2015: protocol of intent by Ministry of Economy and Finance and ACRI to foster self-reform of banking foundations.
- Insolvency, fiscal, and procedural reforms:
  - August 2015 decree law: improve insolvency law, accelerate fiscal deductibility of provisions (deduct within a year instead of five), changes in civil procedures to strengthen debt enforcement, reforms to civil justice including streamlined enforcement and electronic processes, increased staff.
  - Decree-law of June 27, 2015 (converted August 6, 2015): interim finance, competing reorganization plans, competing bids in sale of businesses, strict deadlines for liquidations, electronic portal for sales.
  - Decree-law of May 5, 2016: out-of-court enforcement for secured credit in commercial relationships, reforms to secured transactions for enterprise assets (requires implementation).
  - Bank of Italy launched a new periodic survey to gather detailed information on bad debts, collateral, guarantees, and recovery procedures.
- Resolutions and backstops:
  - November 22, 2015: resolution of four lending institutions (Banca delle Marche, Banca dell’Etruria e del Lazio, Cassa di Risparmio di Chieti, Cassa di Risparmio di Ferrara).
    - Total market share: about 1 percent of system-wide deposits.
    - Total rescue cost amounted to €3.6 bn.
    - Bail-in of equity and subordinated debt: €798 mn losses imposed on junior bondholders (around half held by retail investors).
    - Bad loans written down to €1.5 bn from an original balance-sheet value of €8.5 bn (valuation of 17½ percent).
    - Resolution Fund outlays distributed as: €1.8 bn injected into “good banks” to reach target CET1 ratio of 9 percent; €1.7 bn used to write down most of the banks’ bad debt and set up the bad bank; €140 mn used to form the minimum capital for the bad bank.
  - Late January 2016: Italian authorities agreed with the EC on GACS to securitize and guarantee NPLs (appeared to supersede discussion of an AMC due to EU state aid concerns).
  - April 2016: Atlante fund created by Italy’s largest banks, nonbank financial institutions and banking foundations; by April 29, 2016 Atlante managed to collect €4.25 billion.
    - UniCredit SpA and Intesa Sanpaolo Spa each would take a €1 billion stake in the fund.
    - Atlante provided €1.5 billion for capital increase of Banca Popolare di Vicenza and €1 billion for Veneto Banca.
    - Atlante holds an 8 percent minority stake by Cassa Depositi e Prestiti (CDP).
    - Regulatory treatment of banks’ equity investments in Atlante (deducted from regulatory capital or not) is yet to be determined by the ECB and national authorities.
- GACS (Garanzia Cartolarizzazione Sofferenze) mechanics and expected scale:
  - Focuses on sofferenze (bad loans) and excludes past due loans and “unlikely to pay.”
  - Banks move bad loans at market values to SPVs for securitization; buy public guarantees for investment-grade rated senior tranches; sell in retail market.
  - Public guarantee priced at market terms based on CDS on Italian issuers; guarantee price staged: first three years based on three-year CDS; third until fifth year on five-year CDS; after five years based on seven-year CDS.
  - Market participants expect a positive albeit modest impact: transfer price for securitizing NPLs with government guarantees via GACS does not seem sufficient to close the 15–20 basis points pricing gap. GACS is widely expected to close this gap by around 2–3 percentage points.

### X. Obstacles to resolution and market frictions
- Economic and market obstacles:
  - Uncertainties in recovery values lead to gaps between book and market values and disincentives to write-off.
  - Thin market for NPLs introduces a liquidity premium and heavily discounted transfer prices.
  - Weak profitability and low capital buffers constrain banks’ ability to absorb additional loan loss impairments or sell portfolios.
  - Relationship and related-party lending increased problem loans in medium-sized and small credit institutions, complicating resolution.
  - Prior tax treatment (cap at 0.3 percent until 2013) penalized aggressive tackling of problem loans.
- Accounting and valuation frictions reduce attractiveness of purchases (origination EIR, indirect costs treatment).
- Legal/institutional obstacles: lengthy judicial processes, lack of specialization, fragmented insolvency procedures, difficulties in enforcement and asset sales.

### XI. Strategic options and policy recommendations
- Overarching strategy: comprehensive and rapidly deployed “stick-and-carrot” approach combining supervisory, legal, fiscal, and market measures with strong disincentives/incentives for keeping/removing NPLs.
- Resolving asset quality and capital buffer uncertainty:
  - Apply a common asset classification system; subject banks not under SSM to capital assessment following an asset quality review; accompany with on-site inspections for smaller banks.
- Strengthen banks’ NPL management capacity:
  - Require comprehensive NPL strategies with ambitious targets and separation of NPLs from business-as-usual credit management.
  - Assign one executive committee member and one board member responsible for NPL strategy.
  - Develop comprehensive loan restructuring tools, more comprehensive NPL reporting frameworks, a code of conduct for debtor engagement, guidance on provisioning, licensing/regulation for nonbanks managing loans, and debt restructuring principles for multilateral workouts.
- Strengthened supervision:
  - Establish dedicated expert teams of supervisors to review NPL strategies and enforce compliance, including BoI guidance on write-offs.
  - Enforce common definitions to improve comparability and reduce forbearance.
- Time-bound fiscal incentives:
  - Consider time-bound tax incentives to encourage accelerated workouts and removal of bad debt and real estate assets from balance sheets, while minimizing ever-greening and misreporting.
- Development of distressed-debt markets:
  - Involve outside investors in corporate restructuring or NPL purchases to relieve banks of debt collection and foreclosure, conserve recovery values, facilitate debt/equity conversions, inject capital, and clean up the corporate sector.
  - Centralized, state-backed AMCs can help kick-start distressed debt markets and set uniform pricing standards, conduct regular auctions, and publish data to facilitate price discovery; however, AMC discussion paused due to EU state aid concerns.
- Intra-segmental cooperation:
  - Banks in the same segment (e.g., banche popolari) could cooperate in cleaning up balance sheets, possibly establishing a corporate restructuring vehicle (CRV) owned by banks but independently managed to build NPL management capacity.
- Use of informal workouts and restructuring practices:
  - Intensively combine legal reforms with informal debt restructuring and adoption of workout principles (e.g., “London Approach” or INSOL Principles) for the existing stock of NPLs.
- Legal reform priorities:
  - Rationalize and integrate informal, hybrid, and formal insolvency procedures; initiate insolvency earlier (consider “alert procedure”); expand restructuring possibilities (recapitalizations, debt/equity swaps) while preventing shareholder interference; revisit creditor priorities; streamline procedures to reduce appeals and delay; improve judicial specialization and professionalize insolvency administrators; implement flexible asset sale mechanisms (national platform Project “Common”); consider simplified/hybrid SME restructuring with triage approach; continue civil justice reforms and secured transactions reform implementation.

### XII. Specific reform and intervention timelines and outcomes (selected)
- February 2016: Government approved reform of small mutual banks, pushing consolidation under joint-stock (holding) companies with at least €1 billion in equity; deadline 18 months to implement reform.
- April 2016: Atlante raised €4.25 billion by April 29, 2016 and provided €1.5 billion for Banca Popolare di Vicenza and €1 billion for Veneto Banca capital support.
- November 22, 2015 resolution of four banks: total rescue cost €3.6 bn; bail-in €798 mn; bad loans written down to €1.5 bn from €8.5 bn (17½ percent); Resolution Fund disbursements €1.8 bn, €1.7 bn, and €140 mn.
- August 2015 decree-law accelerated fiscal deductibility and included measures to improve insolvency and enforcement.
- As of end-2014, DTAs not deducted from capital amounted to €43 bn; until 2013 immediate tax deductibility cap was 0.3 percent of outstanding loans; as of August 2015 write-offs are immediately tax deductible.

*Italic: Source — _wp16135*

### References ________________________________________________________________28

### _wp16135 - References ________________________________________________________________28

### I. Introduction — scale and overview
- Nonperforming loans (NPLs) in Italy have tripled since the beginning of the crisis and, according to national definition, stood at €360 billion (18 percent of total outstanding loans) at end-2015.
- NPLs in Italy account for about a third of those in the entire euro area.
- Italian NPLs include four categories: impaired/bad debt or sofferenze (loans in a state of insolvency), substandard (incagli), overdue, and restructured.
- The ECB/SSM 2014 asset quality review of Italy’s largest 15 banks produced an aggregate nonperforming exposures ratio for the 15 analyzed banks of close to 22 percent (noting ECB’s broader exposure definition and post-AQR adjustments).
- Key drivers identified for high NPLs:
  - Over-indebted corporates following the crisis-related drop in output.
  - Banks generally low in capital buffers.
  - A highly complex legal system of corporate restructuring and insolvency and lengthy judicial processes.
  - A tax system that until recently discouraged NPL write-offs.
- Banks have pursued a gradual process of writing off NPLs, influenced by high risk premia and targeted ROEs by investors; cost of risk and high administration costs for NPLs are not yet fully factored into lending margins, weighing on profitability and impeding capital buildup.
- The combination of restricted bank financing and limited alternative corporate financing sources has constrained investment and slowed the economic recovery.

### II. The nature of the problem
- The paper examines the distribution of NPLs by size, region, sector, and across banks; provides econometric evidence on drivers of NPL buildup; assesses obstacles to resolution; and discusses options to repair private sector balance sheets and support recovery.
- Findings emphasize a need for broad-based cleanup of bank balance sheets, decisive corporate restructuring, and exit of unviable firms, supported by economic, supervisory, and legal measures and reforms to corporate governance of Italian banks.

### III. Distribution of NPLs by size, sectors, and region
- Size distribution (bad debt, impaired loans, end-2014):
  - More than 75 percent of the total value of bad debt relates to loans above €250,000.
  - About 75 percent of the total number of borrowers with bad debt relate to loans of less than €75,000.
- Sectoral distribution:
  - About three quarters of bad debt are related to the corporate sector.
  - The service sector and less technology-intensive sectors are most affected.
- Regional distribution:
  - Pronounced north-south divide: by end-2014, most central and southern Italy regions saw bad debt (sofferenze) ratios increase above 20 percent (compared with most regions below 10 percent in early 2009).
- Collateral and provisioning:
  - Provisions stood at an average of 45.4 percent for total NPLs and 58.7 percent for bad loans as of December 2015.
  - Authorities estimate real estate collateral and personal guarantees against bad loans amounted to €87bn and €37bn respectively at end-2015.
  - Court times to access collateral are very long.

### IV. Distribution of NPLs across banks
- System-wide characteristics:
  - Most banks have NPL ratios between 15 and 20 percent; some smaller banks face extremely high NPL ratios.
  - There is no clear correlation between NPL ratios (percent of total loans) and bank size.
  - As of December 2014, the five largest banks in Italy accounted for two-thirds of total bad debt and NPLs.
  - Large cooperative banks (banche popolari) faced the highest NPL ratios with the lowest coverage ratio at end-2014.
- Data samples and coverage:
  - Full SNL sample referenced: 402 Italian banks (summary data).
  - Full coverage time-series sample used for regressions: 62 banks with comprehensive data since 2005.
  - Some regressions used somewhat smaller bank counts (e.g., 326 and 57) due to limited time-series observations in SnL for some banks.

### V. Contributing factors to the build-up of NPLs — econometric findings
- Methodology:
  - Dynamic panel regressions following Klein (2013): fixed effects and Generalized Method of Moments (GMM) regressions of NPLs (problem loans in percent of total gross loans) on macroeconomic variables and bank-specific variables.
  - Included macro variables: lagged inflation, NEER percentage change, lagged real GDP growth, lagged percentage change in house and stock prices.
  - Included bank variables: lagged equity/assets ratio, lagged ROE, lagged loan growth, lagged Tier 1 ratio.
  - Year dummies included; system GMM used to correct for potential biases from feedback effects.
- Key regression results and interpretations:
  - Both bank-level and macroeconomic factors significantly affected asset quality.
  - Lower profitability in the past (lagged ROE) is associated with higher NPL levels.
  - Higher lending in the past (lagged loan growth) is associated with higher NPLs, indicating faster loan book expansion on average results in worse asset quality.
  - Macroeconomic factors: lower growth, exchange rate appreciations, and falling house prices are significantly associated with higher NPLs (house prices significant in fixed effects, not in system GMM).
  - The exceptional duration and intensity of the recession had a profound impact on asset quality, exacerbated by bank-specific factors.
  - The incipient recovery may help reduce NPLs, but moderate growth and inflation outlook suggest additional measures will likely be needed.

### VI. Default risk, dynamics, and macro-financial links
- Default probabilities and timing:
  - Corporate and bank default probabilities peaked in mid-2012.
  - Default probabilities have come down substantially since mid-2012, but NPLs continued to rise until 2015Q3.
- State space model insights (DSSM; Box 1 referenced):
  - Estimates suggest:
    - Further output losses would fuel credit risk considerably more than in the past.
    - Higher interest rates would now have a much more substantial impact on corporates than in the past.
  - Increases in interest rates would negatively impact the market value of banks’ holdings of Italian sovereign assets, given the substantial rise in sovereign bond volumes on banks’ books.
- Additional note:
  - Italian banks have significant unrealized capital gains from sovereign bond holdings (BoI Financial Stability Report April 2015 referenced).

### VII. Policy implications and suggested strategy elements
- A potential solution requires coordinated actions addressing both bank and corporate balance sheets:
  - Broad-based cleanup of bank balance sheets.
  - Decisive corporate restructuring and enabling exit of unviable firms.
  - Economic, supervisory, and legal reforms to facilitate NPL resolution.
  - Reforms to corporate governance of Italian banks to align incentives (including issues related to foundations and cooperatives).
- Authorities’ measures already introduced:
  - Improving the Italian insolvency system.
  - Fostering consolidation within the highly fragmented banking sector.
  - Facilitating securitization and sale of NPLs.
- Expected timing and effects:
  - Many reforms are expected to yield benefits gradually over time; given high NPL ratios and moderate growth, additional measures may be necessary to accelerate clean-up.

*Italic: Source — _wp16135 - References ________________________________________________________________28*

### Box 1. Modeling Elasticities Dynamically Using State Space Models

### Box 1. Modeling Elasticities Dynamically Using State Space Models

### Dynamic State Space Models (DSSM): key properties and motivations
- DSSM allow modeling nonlinearities in the functional relationship between variables, and describing explicitly structural changes over time.
- Important because the influence of different macrofinancial factors on credit risk typically changes during periods of financial distress (“correlation breakdown”).
- DSSM help push imprecise measurements toward true values when the measurement of credit risk variables is blurred, thereby increasing precision.

### Time-varying regression and State Space representation
- A regression with time-varying coefficients can be represented by equation (1) in the source.
- Alternatively, the State Space form can be chosen as equation (2), where:
  - θ denotes the regression coefficients in the state vector θt = (β1,t, …, βn,t)′.
  - xt = (1, x1t, …, xnt) are the n regressors.
  - Ft is the state transition matrix from time t to t+1.
  - The variance of the multidimensional noise vectors vt and wt are driven by the time-variant matrices Vt and Wt.
- If the variances in Wt are larger than zero, the coefficients become time-varying. Hence, for the dynamic multivariate state space regression model, both Vt and Wt are nonzero.
- The signal-to-noise ratio indicates how adaptive the coefficients are.

### Estimation: Kalman Filter recursion
- The dynamic coefficients can be estimated using the Kalman Filter, a set of recursion equations determining, conditional on information available at time t, optimal estimates of the state vector.
- The estimation is performed recursively using equations (3), (4), (5), and (6) as presented in the source.
- Notation and elements from the source:
  - y_t, x_t, β_t, ε_t are used in the time-varying regression representation.
  - State-space matrices include H_t, F_t, v_t, w_t, V_t, W_t, and the prediction and update steps use P and R matrices as shown in the source equations.
  - The source notes /0VW> (indicating a condition on V and W).

### Empirical illustration notes (Italy)
- The source includes a Figure showing:
  - Left chart: “true elasticities of corporate default risk’s key drivers resulting from a dynamic State Space model (DSSM).” Variables:
    - Unemployment is the share of potential workforce that is not employed.
    - STIR is the change in short-term interest rates.
    - GDPreal is the yoy change in real GDP growth.
    - GovConsumption.Exp is the change in government consumption expenditure.
  - Lags are given in quarters.
  - Static coefficients reference t0 = 2005Q4.
  - Right chart: build-up of sovereign exposure and Italian banks’ reliance on ECB liquidity; series include ECB support and ECB Liquidity Support and Bank Holdings of Italian Government Securities (Billions of euros), Change in holdings of government securities (Jan 2011=0).
- The left chart shows dynamics in true coefficients with series labeled Unemployed.Lag0, GovConsumptionExp.Lag0, GDPreal.Lag0, STIR.Lag2 and a vertical axis ranging from -3.0 to 3.0; quarters shown include Q1-2006 through Q1-2014 (as labeled in the figure).

### Findings and interpretations from the DSSM application (as presented)
- The DSSM recovers time-varying “true elasticities” of corporate default risk drivers, demonstrating that coefficients change over the sample (dynamic coefficients estimated via Kalman Filter).
- The sovereign exposure chart and ECB liquidity series illustrate a build-up in bank holdings of government securities and use of ECB liquidity facilities over the period shown.

### Authorities’ Actions (measures to address NPLs)
- March 2015: Both houses of parliament approved a decree to reform governance of cooperative banks that had scored weakly in the FSAP. Reforms remove one vote per head and limitations on size of individual shareholdings; aim to promote consolidation, balance sheet clean-up, and improve profitability.
- April 2015: A protocol of intent signed by the Ministry of Economy and Finance and ACRI to foster self-reform of banking foundations; commitments include adopting appropriate portfolio diversification and limiting risk concentration and participation in bank capital.
- August 2015: Decree law enacted including measures to help banks offload NPLs:
  - (1) Improving insolvency law to shorten durations and increase survival of distressed enterprises and creditors’ recovery.
  - (2) Accelerating fiscal deductibility of provisions by allowing banks to deduct loan losses from their tax bill within a year instead of five.
  - (3) Changes in civil procedures to strengthen debt enforcement.
  - (4) Reforms to civil justice: streamlined enforcement procedures, increased use of electronic processes, and increased staff (through hiring from other administration areas).
- November 2015: Authorities resolved four small banks (see Box 2).
- Late January 2016: Italian authorities agreed with the EC on GACS to securitize and guarantee NPLs (see Box 3); this appeared to supersede discussion of an AMC owing to EU state aid concerns.
- February 2016: Government approved reform of small mutual banks, pushing consolidation under joint-stock (holding) companies with at least €1 billion in equity; deadline 18 months to implement reform.
- April 2016: Italy’s largest banks together with nonbank financial institutions and banking foundations created fund Atlante; so far raised €4.25 billion to backstop capital increases and purchase non-investment grade tranches of NPL securitizations.
- May 2016: Decree-law introduced out-of-court enforcement mechanisms for commercial lending relationships secured by immovable collateral; new framework for non-possessory security interests over movable assets; a registry of enforcement and insolvency procedures; and other improvements.
- Bank of Italy launched a new periodic survey to gather detailed information on stock of bad debts, related collateral and guarantees, and recovery procedures.

### Box 2 — 2015 Resolution of Four Lending Institutions: summary facts and outcomes
- Date announced: November 22, 2015. Institutions: Banca delle Marche, Banca dell’Etruria e del Lazio, Cassa di Risparmio di Chieti, Cassa di Risparmio di Ferrara.
- Total market share: about 1 percent of system-wide deposits.
- Total rescue cost amounted to €3.6 bn.
- Resolution consistent with EU Bank Restructuring and Resolution Directive (BRRD), transposed into Italian law on November, 16 2016. Bail-in tool mandatory only by January 1, 2016.
- Authorities’ rationale: banks play important regional role and protection of household and firms’ deposits was important to justify resolution over liquidation.
- Elements of the solution:
  - Bail-in of equity and subordinated debt: €798 mn losses imposed on junior bondholders, around half held by retail investors.
  - Each bank split into “good/bridge banks” (all assets except bad debts and all liabilities except those bailed in) and a “bad bank” containing bad debt.
  - A single bad bank created to take bad loan assets of all four banks. Bad loans written down to €1.5 bn from an original balance-sheet value of €8.5 bn, a valuation of 17½ percent.
  - Bad loans will be sold on the private market or managed directly; proceeds go to the Resolution Fund.
  - Resolution Fund outlays of €3.6 bn distributed as:
    - €1.8 bn injected into “good banks” to reach target CET1 ratio of 9 percent.
    - €1.7 bn used to write down most of the banks’ bad debt and set up the bad bank.
    - €140 mn used to form the minimum capital for the bad bank.
  - Rescue cost borne by the Resolution Fund; liquidity advanced by Intesa, Unicredit, and UBI.
  - Good banks retained original brand with suffix “Nuova”; bridge banks temporarily administered by BoI and to be sold quickly and transparently to highest bidder; sale proceeds retained by Resolution Fund.
- Public reaction: large uproar over bail-in of junior bonds held by retail investors; banks voluntarily set up a fund attached to national deposit insurance scheme to compensate many retail investors; because contributions are voluntary by banks, it does not constitute state aid.

### Box 3 — Public Guarantees (GACS) and Backstop Fund (Atlante): mechanics and scale
- GACS (Garanzia Cartolarizzazione Sofferenze):
  - Agreed with the EC in January 2016: government guarantees securitization of bad loans via multiple SPVs, one for each participating bank.
  - Scheme focuses on sofferenze (bad loans) and excludes past due loans and “unlikely to pay.”
  - Banks can move bad loans at market values to SPVs for securitization, buy public guarantees for investment-grade rated senior tranches, and sell them in retail market.
  - Equity and junior notes expected to be attractive to private investors and hedge funds.
  - Public guarantee priced at market terms based on credit default swaps (CDS) on Italian issuers with similar risk profiles. Price of guarantee increases over time to incentivize accelerated NPL recovery:
    - First three years based on three-year CDS.
    - Third until fifth year on five-year CDS.
    - After five years based on seven-year CDS.
  - Because guarantees are priced at market terms based on expected losses, they do not qualify as public support subject to EC approval under EU State aid regulations.
  - Market participants expect a positive albeit modest impact: transfer price for securitizing NPLs with government guarantees via GACS does not seem sufficient to close the 15–20 basis points pricing gap. GACS is widely expected to close this gap by around 2–3 percentage points.
- Atlante:
  - Fund funded by largest Italian banks, nonbank financial institutions and banking foundations, with an 8 percent minority stake held by Cassa Depositi e Prestiti (CDP).
  - Role: backstop facility for ongoing banks’ capital increases; buyer of last resort for non-investment grade tranches of NPL securitizations; can invest in real estate assets.
  - By April 29, 2016 Atlante managed to collect €4.25 billion.
    - UniCredit SpA and Intesa Sanpaolo Spa each would take a €1 billion stake in the fund.
    - Atlante provided €1.5 billion for capital increase of Banca Popolare di Vicenza and €1 billion for Veneto Banca.
  - Regulatory treatment of banks’ equity investments in Atlante (whether contributions will be deducted from regulatory capital) is yet to be determined by the ECB and national authorities. If contributions are deducted from regulatory capital, funding banks would have a smaller capital buffer in excess of prudential capital requirement.

### Obstacles to NPL resolution — Economic factors (as presented)
- Uncertainties in recovery values result in substantial gaps between book and market values, creating disincentives for banks to write-off NPLs.
- Reducing such uncertainties requires more conservative approaches to provisioning.
- A thin market for NPLs introduces a liquidity premium, leading to heavily discounted transfer prices upon sale.
- Weak profitability and, for a number of banks, low capital buffers provide limited space for balance sheet clean up.
- NPLs weigh heavily on bank profitability (Figure 7) and thin profit margins make it difficult for banks to digest additional loan loss impairments or sell portfolios in an illiquid market.
- Relationship and related-party lending have contributed to the build-up of problem loans in medium-sized and small credit institutions, making NPL resolution more difficult.
- Prior tax treatment (until the 2013 reform) penalized banks tackling problem loans aggressively: write-offs were only tax deductible in the state of insolvency; tax deductibility of loan-loss provisions from taxable income was limited (text cuts off in source).

*Source: _wp16135 - Box 1. Modeling Elasticities Dynamically Using State Space Models*

### 0.3 percent of outstanding loans, with the remaining part treated as a deferred tax asset (DTA).

### _wp16135 - 0.3 percent of outstanding loans, with the remaining part treated as a deferred tax asset (DTA).

### DTAs, tax treatment, and accounting issues
- Remaining provisions could be deducted in equal installments over a period of 18 years; DTAs were deductable from taxable revenues over a period of 18 years.
- Until 2013, a cap limited immediate tax deductibility to 0.3 percent of outstanding loans, with the remaining part treated as a deferred tax asset (DTA). This cap constituted a clear disincentive to provisioning.
- Since 2013, provisions and write-offs can be deducted in equal installments over five years, and with a higher tax rate; restrictive than in other countries, incentives for accelerated write-off have increased.
- As of end-2014, DTAs not deducted from capital amounted to €43 bn.
- As of August 2015, write-offs are now immediately tax deductible.
- IAS 39 does not define when and how defaulted loans are to be written off; the accounting regime under IFRS lacked clear write-off rules, producing slow NPL write-offs and widespread application of a de-recognition rule (loan cancellation) beyond intended conditions.
- Forthcoming IFRS 9 (scheduled for 2018) will include a clear definition of write-off and a definition of write-off different from loan cancellation; when applicable, banks need to write off earlier, opening the way for corporate restructuring or liquidation.
- Banks adopting IAS/IFRS use the original effective interest rate on the assets (usually much lower), reducing NPL prices; banks include indirect costs of managing NPLs in the financial statement of the year incurred, whereas potential acquirers deduct them immediately, thus reducing purchase prices.
- Slow NPL write-offs and accounting uncertainty contributed to the build-up of Deferred Tax Assets in Italian banks, as remaining provisions could be deducted in equal installments over a period of 18 years (and at a lower net present value).

### Banking sector capacity and NPL composition
- Larger banks: have disposed of NPLs, invested in internal NPL management, and set up decentralized AMCs; example: in fall 2014, UCG sold €2 bn in NPLs; in February 2015, UCG announced sale of its participation in UniCredit Credit Management Spa, including a €2.4 bn NPL portfolio; net of provisions, bad debts at UCG stood at about €36 bn in 2013.
- Medium and small-sized banks: limited experience and capacity to deal with NPLs; lack risk management capacity, NPL management experience, and access to distressed debt markets.
- Sectoral composition: Italian NPLs are mostly commercial, including a large number of micro and small enterprises, and are very heterogeneous (unlike other countries where NPLs were concentrated in real estate and easier to value).

### Legal factors affecting NPL resolution
- Two main methods to deal with NPLs: contractual (renegotiation or sale) and enforcement (individual enforcement such as mortgage foreclosure, or collective enforcement via insolvency).
- Effectiveness of enforcement influences feasibility and effectiveness of contractual methods; effective enforcement increases value and attractiveness of NPLs and incentives for market development.
- Italian insolvency regime: high complexity with multiple procedures and debt restructuring tools lacking coordination and unified vision; system divided between general insolvency regime and special regimes for large enterprises, small enterprises and individuals, and enterprises subject to special supervisory regimes.
- Procedural complexity likely contributes to high litigiousness and facilitates debtors’ delaying strategies.
- Informal or hybrid debt restructuring procedures (piani di risanamento and accordi di ristrutturazione) often lack flexibility and aim for full payment to creditors; inability to bind dissenting creditors reduces efficiency.
- Formal insolvency procedures (concordato preventivo and fallimento) can bind minority creditors and allow flexible solutions, though a recent reform imposed limits on haircut unsecured creditors may accept in certain plans.
- Obstacles to advanced restructuring solutions (e.g., debt/equity swaps) include shareholders’ interference and shareholders’ preemption rights; the 2015 reform gave courts power to appoint a special administrator but did not remove shareholders’ meeting powers over capital structure.
- Insolvency caseload and outcomes:
  - In 2014, enterprise bankruptcies surpassed 15,000 cases (+10.7 percent vs. 2013).
  - Estimated more than 82,000 enterprises liquidated in insolvency processes since 2008, with associated loss of about a million jobs (Cerved 2015).
  - Enterprise reorganizations declined by 20 percent in 2014.
  - For 2015, enterprise insolvencies stabilized at 14,700 cases and reorganizations continued to fall.
- Insolvency procedures have long delays: liquidations on average lasted more than eight years, with preferential and unsecured creditors recovering 29 and 6 percent of claims, respectively.
- Specific bottlenecks: litigation within insolvency (especially appeals), difficulties in collection and sale of assets, and insolvency cases initiated too late limiting recovery possibilities.
- Institutional challenges: judiciary overburdened and lacks specialization in debt enforcement, restructuring and insolvency; enterprise courts (tribunali delle imprese) do not have competencies on insolvency matters; no specific qualifications for insolvency administrators; professionalization limited with expertise focused on liquidation over reorganization.

### Recent reforms to insolvency and enforcement
- Decree-law of June 27, 2015, converted into Law of August 6, 2015, introduced numerous measures mainly related to insolvency law and civil procedure.
  - Changes include regulation of interim finance, introduction of competing reorganization plans and competing bids in sale of businesses, strict deadlines for liquidations, and a new modality of debt restructuring agreement for financial claims.
  - Reforms sought to reduce length of enforcement procedures, improve criteria for accepting bids, facilitate payment in auctions and assignment of property to creditors.
  - Use of information technology for search of debtor assets and introduction of an electronic portal for sales were included (see Art. 532; Arts. 432-bis and 490; Art. 161-ter of the code of civil procedure).
- Decree-law of May 5, 2016 introduced out-of-court enforcement for secured credit in commercial credit relationships, affecting immovable collateral and using conditional transmission of ownership (“patto marziano”); also reformed law on secured transactions for enterprise assets but requires implementation actions to be operational.
- Reforms have improved legal framework quality but challenges remain in the institutional framework (judicial specialization, professionalization of insolvency administrators).

### Options and recommendations for bringing down NPLs
- Strategy overview: comprehensive and rapidly deployed strategy using a “stick-and-carrot” approach with strong disincentives/incentives for keeping/removing NPLs; combine supervisory, legal, fiscal, and market measures.
- Key recommendations and elements:
  - Resolving asset quality and capital buffer uncertainty:
    - Apply a common asset classification system; ECB’s 2014 Comprehensive Assessment revealed sizeable provisioning gaps among 15 participating banks (filled during 2015).
    - Subject banks not under SSM supervision to capital assessment following an asset quality review to clarify uncertainty, accompanied by on-site inspections for smaller banks.
  - Banks’ ability to deal with NPLs:
    - Require banks to maintain comprehensive NPL strategies with ambitious targets and separation of NPLs from business-as-usual credit management.
    - Assign one member of the executive committee and board of directors responsible for NPL strategy.
    - Develop comprehensive loan restructuring tools and consider more comprehensive NPL reporting frameworks, a code of conduct for debtor engagement, guidance on provisioning, licensing/regulation regime for nonbanks managing loans, and debt restructuring principles for multilateral workouts.
  - Strengthened supervision:
    - Establish dedicated expert teams of supervisors to review NPL strategies and enforce compliance with rules and guidance, including BoI guidance on write-offs.
    - Enforce common definitions to improve comparability and reduce forbearance.
  - Time-bound tax incentives:
    - Consider time-bound tax incentives to encourage accelerated workouts and removal of bad debt and real estate assets from balance sheets, while minimizing ever-greening and misreporting.
  - Development of distressed debt markets:
    - Involve outside investors in corporate restructuring or NPL purchases to relieve banks of debt collection and foreclosure, conserve recovery values, facilitate debt/equity conversions, inject capital, and clean up the corporate sector.
    - Centralized, state-backed AMCs can help kick-start a distressed debt market and set uniform market pricing standards, conduct regular auctions, and publish data to facilitate price discovery; however, authorities have put AMC discussion to rest for now because of EU state aid concerns and uncertainty impacting banks’ NPL resolution plans.
  - Intra-segmental cooperation:
    - Banks in the same segment (e.g., banche popolari) could cooperate in cleaning up balance sheets, potentially establishing a corporate restructuring vehicle (CRV) owned by banks but managed independently to build NPL management capacity.
  - Use of informal debt restructuring and workout practices:
    - For the existing NPL stock, combine legal reforms with intensive use of informal debt restructuring and adoption of debt restructuring principles by the financial sector (e.g., “London Approach” or INSOL Principles).

### Recommendations for legal reform (detailed)
- Rationalize and better integrate informal, hybrid, and formal insolvency procedures to create a seamless continuum for debt restructuring and insolvency.
- Initiate insolvency procedures earlier by establishing adequate incentives and disincentives for debtors; consider an “alert procedure” as recommended by the Rordorf Commission.
- Expand restructuring possibilities in insolvency, including recapitalizations and debt/equity swaps, while preventing shareholder interference that blocks these solutions.
- Revisit creditor priorities to increase rate of return for unsecured creditors; rationalization needed given proliferation of creditor priorities and expanded priority of public creditors.
- Streamline insolvency processes to reduce appeals and opportunities for delay while protecting participants; improve specialization of judiciary, reinforcement of court officials, and quality of insolvency administrators.
- Implement increased flexibility in sale of assets: Ministry of Justice projected a national platform for insolvency and enforcement sales with certificates to creditors corresponding to projected claim values (Project “Common”) to enable trading or use in auctions.
- Consider a simplified/hybrid debt restructuring mechanism for SMEs based on a triage approach and clear debt restructuring principles for over-indebted SMEs.
- Continue reforms to increase efficiency of individual enforcement procedures, including electronic processes and extension of best court management practices nationwide.
- Reform secured transactions law to improve security interests over movable assets and corresponding registration systems; recent reform requires further technical implementation.
- Strengthen institutional framework: continue civil justice reforms, address court backlogs, increase specialization of judges in commercial and insolvency matters, strengthen enterprise courts, and professionalize insolvency administrators.
- Note: Legal reforms are more effective at reducing the flow of new NPLs than in resolving the existing stock; for current NPLs, legal reforms should be paired with intensive informal restructurings and market-based solutions.

*Italicized source attribution: IMF working paper content from the provided PDF excerpt.*

### Box 4. The Importance of Efficient Secured Transaction Laws

### Box 4. The Importance of Efficient Secured Transaction Laws

### Role and benefits of security rights
- Security rights provide certainty for creditors and increase access to finance.
- The importance of the secured transactions system has been highlighted by legal and economic experts and has given rise to detailed best practice recommendations (UNCITRAL, 2007).
- The possibility of using different types of collateral creates opportunities for more efficient financing arrangements, particularly through security interests over movable collateral.
- Efficient secured transactions affect the availability of credit for enterprises, particularly SMEs.

### Design features for movable-collateral regimes
- Flexibility in creation of security interests:
  - Security interests should cover not only specific assets (for example, machinery) but also categories of assets (for example, inventory, equipment and accounts receivable) without the need to specify each and every asset.
  - This flexibility allows the debtor to remain in possession of the assets and to continue using them, including selling inventory in the ordinary course of business.
- Notice-based public registries:
  - Security interests are recorded in public registries that are “notice based”: they provide notice to the public of the creation of the security interest.
  - Registries are computerized and allow for interaction with other registries and databases.
- Priority rules:
  - Priority of the security interest is a crucial element; registration establishes the point in time when the lender acquires priority rights versus other creditors and third parties.
  - Protecting the secured creditor’s priority position generates positive effects for borrowers.
- Swift enforcement:
  - Rapid enforcement of secured claims is critical for movable collateral because depreciation of assets is fast.
  - Enforcement mechanisms may include summary judicial procedures or out-of-court enforcement.
  - Fiduciary contracts can improve enforcement of security interests over movable collateral without resorting to the court system.

### Empirical and economic evidence
- According to recent research, loan-to-values of loans collateralized with movable assets are on average 21 percentage points higher in countries with strong collateral laws relative to immovable assets.
- Stronger collateral laws shift collateral composition away from immovable to movable assets (Calomiris and others, 2015).
- In terms of their effect on NPLs, secured transactions reforms do not reduce the stock of NPLs, but can reduce significantly the flow of new NPLs in the banking system.

*Source: Box 4, "The Importance of Efficient Secured Transaction Laws" (excerpt from the provided IMF working paper content).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp16135.pdf_