- The Italian NPL mountain stands at EUR 360bn, provisions are EUR 146bn
- Total shortfall in provisions would be EUR 66bn, if NPLs were sold at 41 cents
- Issue is greater for banks outside the top 15, they have a EUR 42bn shortfall
- BMPS, UBI, Banco Popolare and Banca Carige would need to increase provisions greatly (EUR 12.8bn), to cover an NPL portfolio price of 41
- Consolidation and bankruptcy laws are vital for future of the banking sector
- Ongoing meagre growth outlook will negatively impact the NPL portfolios
DISCLAIMER: This report has not been prepared in accordance with the legal requirements designed to promote the independence of investment research, and that it is not subject to any prohibition on dealing ahead. This report is marketing communication and not investment research and is intended for professional and eligible clients only.Financials-Focus-Italian-Banks-Q-A.pdf (733 KB)
How big is the Italian NPL problem?
The number of non-performing loans (NPLs) within the Italian banking system is colossal, in total grossing EUR 360bn. This is 18% of the total outstanding consumer loans, and equates to 23% of Italian GDP.
But why do I hear Italian authorities quoting EUR 200bn?
The Italian authorities and institutions often quote a total NPL figure of EUR 200bn. However, this refers principally to ‘sofferenze’’ NPL positions. These are just a portion of the European Banking Authority (EBA) definition of NPLs. ‘Sofferenze’ is an Italian classification for bad debt that is the worst of the worst, principally with an exposure to any borrower in a position of insolvency or similar. As this classification excludes loans that are temporarily in arrears, this NPL number is naturally lower. As of December 2015, ‘sofferenze’ debt made up approximately 58% of the EUR 360bn Italian NPL mountain.
How are the NPLs at an individual banks level?
What are the value of provisions that have been set aside?
The Italian banks have set aside a sizeable amount of provisions. In total, provisions reached EUR 146bn at the end of Q4 2015, a value that is 41% of the gross NPL number. As provisions are currently 41% of the gross NPL value, this would mean that there would be limited financial impact to the Italian banks if the NPL portfolios would be sold at a price of 59.
The distribution of provisions for NPLs however, does significantly differ across institutions. Iccrea BancaImpresa for example, has EUR 788mn (71%) reserve for loan losses against a non-performing loan total of EUR 1,108mn. Whereas, Unione di Banche Italiane (UBI) for instance has EUR 4,237mn (32%) of provisions against EUR 13,233mn of non-performing loans.
Outside the top 15 banks in Italy, there is a huge NPL pile (EUR 91bn), of which provisions appear very low indeed (EUR 12bn). Subsequently the NPL dilemma is more acute for the smaller banks within Italy, many of which have had suffered difficulty recently in raising capital. The larger banks often draw headlines regarding NPLs, however the sector will have to address the huge amount of smaller, more regional based, banks to resolve the issue.
How much capital would be required by the Italian banking sector to address NPLs?
The average recovery rate for all NPL procedures was 41% during 2011-2014, according to a survey by the Bank of Italy based on the largest 25 Italian banks.
If the NPL portfolios of the Italian banks were priced at 41 (as a percentage of the NPL portfolio value), the amount of capital required for the banking sector would be EUR 148bn. On a net basis, this would be an increase of EUR 66bn given current provision levels. Alternatively, if we assumed a price of 25% on all NPL portfolios, the amount required by the sector would rise dramatically to EUR 124bn. This would equate to approximately 5.5% of the total Italian government debt outstanding.
How are capital needs distributed across banks?
If NPL portfolios were to a priced at 41%, EUR 66bn would be required by the banks in Italy to make up the provision shortfall. Of this EUR 66bn, EUR 45bn would be required from banks outside the top five banks in Italy (68%). We apply a price of 41, not just as this matches the recovery value from the Bank of Italy, but it would be a level whereby future NPL management by a bank would not impact significantly on the operations. At this price, it indicates that the larger banks are in a better position, when compared to the smaller domestic banks.
If we continue to apply an NPL portfolio price of 41 at an individual bank level. There would be an average provision shortfall equivalent to 21% of common equity tier 1 (CET1). This would be 2.5% on a risk weighted assets basis.
Which banks are in the most trouble?
The continual requirement to increase provisions, significantly dilutes the ability to retain earnings and chokes bank balance sheets. A number of banks in Italy still require provision increases if they wish to cover a larger portion of their NPL portfolio. If we compared the Italian banks on a risk weighted asset basis, Banca Monte dei Paschi di Siena, Banco Popolare, Unione di Banche Italiane (UBI) and Banca Carige all would have significant capital gaps based on an NPL portfolio sale price of 41 (see table, page five).
From a capital perspective, for the year 2016, UniCredit, Veneto Banca, Banca Popolare di Sondrio, Banca Popolare di Vicenza and Iccrea BancaImpresa all trail peers on their common equity tier one values. On average the fifteen Italian banks in the sample, trail the top 25 European bank median by 113 basis points. Capital issuance is both a obligation under regulatory capital requirements, and prerequisite for banks to increase their provisions for their NPL portfolios.
Why do Italian banks find their NPLs difficult to sell?
Italy’s inefficient bankruptcy system is one of the key obstacles for addressing the NPL problem. Italy severely lags its European peers in the time it takes to deal with bad debts. In a June 2015 survey by the Bank of Italy, bankruptcies in their sample took on average 8.2 years to conclude, although there was considerable differences across the country (standard deviation was 2.4 years). The share of liquidation procedures that lasted longer than two years fell from 95% before 2005 to under 60% after 2005, an improvement but still worryingly elevated. Foreclosures take on average 55 months in Italy to complete, according to the ECB. Which is in stark contrast to Germany, Spain and the Netherlands, which all have typical foreclosure durations of less than 12 months.
Why have I seen sales of NPL portfolios around 35%?
Banks in Italy are rumoured to desire to sell their NPL portfolios at a price of approximately 35%, according to reports in July. However, a general figure is difficult to determine, as it dependent on many dynamics. Firstly, the structure of the NPL is crucial. The amount of collateral and the current amount of time in arrears will determine the amount investors are willing to pay.
Furthermore, the length of proceedings in Italy, also erodes the value that investors are willing to purchase the portfolios for. The high bankruptcy timeline negatively impacts the NPL portfolio price. As not only does it reduce the present value of collateral and future settlements, but also reduces the motivation from private investors to spend years struggling through Italian courts. The potential purchasers of NPLs, are currently wanting to pay around 27% for NPL portfolios, according to reports.
How do we see the future of NPLs in Italy, will they continue to be a problem?
The IMF calculates that Italian GDP will only return to its pre-crisis levels around 2025. As a result, Italy’s economy has effectively lost a ‘double decade’ in GDP terms. Weakness in GDP growth can restrict the Italian bank’s ability to generate sufficient profits, causing them to struggle to write-off their bad debts, obtain new financing and strengthen their capital buffers.
We believe that GDP will be just in positive territory for the next two years, but it is likely to slow to a crawling pace. Based on their historical relationship it is likely that NPLs will start rising again given the weak outlook for GDP growth.
How could the NPL issue be resolved?
Selling NPLs to the private sector would be the silver bullet for the Italian banking sector. This would clear the NPLs off the balance sheets of the banks, freeing the banks to continue with normal operations and enhance the Italian domestic recovery. This would be directly assisted by an improvement in the bankruptcy timeframes in Italy. There is no reason why bankruptcy proceedings in Italy should not be below two years.
A reduction in the process time would generate a higher price for the NPLs, and lower the provision shortfall the banking sector currently experiences. This would reduce the impact to the entire sector, as a lower provision shortfall would significantly assist both the Italian banks and the Italian economy. There has never been a greater opportunity, as European focus intensifies, for the issue to be addressed by the Italian government. The Italian authorities have already taken some measures, that should assist in speeding up bankruptcy proceedings (Q2 2015 – Law Decree No. 83, and Q2 2016 – Law Decree No. 59). Both legislations are a step in the right direction. These however, are likely to take considerable time to assist in the reduction in NPLs that Italy currently needs.
What are banks doing in the meantime?
So far this year, Italian banks have been trying to increase their tier one capital as well as increasing their provisions against NPLs. However, as shown with the Q2 2016 cash calls of Banca Popolare di Vicenza and Veneto Banca, investor demand has severely waned for bank equity of the smaller Italian banks in 2016. Consequently, relying on capital calls to address the NPL issue is not a long-term strategy.
What impact does the NPL situation have on the cost of funding for Italian banks?
When looking at the Italian bank debt market, investors are currently pricing senior unsecured debt as though an investor friendly solution can be found to solve the NPL issue. This is allowing Italian banks to have relatively low funding costs when compared to their last 52 weeks performance. Although the outcome of the NPL solution is highly uncertain, it seems that the risk of a spread widening is tilted to the upside. Furthermore, the correlation between provisions and the cost of funding is interesting to note. Developments in provisions will be a driving factor for spread movements going forward.
How do we see other recent Italian NPL initiatives?
A number of initiatives have been raised in 2016, such as a securitisation scheme (GACS from Q1 2016) and the Atlante fund (from Q2 2016). The GACS (Garanzia Cartolarizzazione Sofferenze) scheme could help support moving NPLs off the balance sheets of Italian banks, especially if bankruptcy laws are adjusted or private investor sentiment improves. This would reduce the pricing discrepancies between the price the banks wish to sell at, and what investors are willing to pay. A reduction in this discrepancy (which at present is large) will significantly lessen the impact of NPLs on the Italian banks.
The Atlante fund could though potentially be dangerous for the sector, in our view. The Atlante fund works well in the short term, in supporting urgent cash calls of the smaller Italian banks, helping the sector to increase CET1 ratios, all without using state money. However, larger banks, such as UniCredit which has gained EUR 1.1bn selling minority stakes in parts of the business in July 2015, cannot be expected to continue to throw money into the fund. Especially as to date, the fund has simply been used to purchase the equity of smaller Italian banks, which private investors did not want.
The GACS Scheme:
An approach, agreed by Italy and the EC, involving the securitisation of bank NPLs. The securitisation assists banks’ to remove non-performing loans off their balance sheets.
The Atlante Fund:
A private contributed fund which acts as a buyer of last resort for banks that struggle to:
- raise equity capital in the private markets
- and / or are unable to sell junior tranches of their securitised NPLs
Further details: ABN AMRO
How has the Italian government assessed the NPL situation?
The Italian government has indeed been proactive in discussing the NPL issue and how to recapitalise the sector. It raised the issue with the European Commission and senior European leaders throughout this year. Albeit in part forced by the application of the Bank Recovery and Resolution Directive (BRRD) on 1 January 2016. This proactive policy stance is an important step, but continuous government focus is required. The government is willing to step-in and recapitalise the banks, but the BRRD regulation is a bottleneck to the situation (see next question). The fact they have discussing the issues with European policy makers though is a positive step.
Could bail-in be enacted on the Italian banks?
Indeed, a bail-in could be legally enacted on the sector. Whereby an Italian bank bail-in would allow for additional aid from the state and resolution funds to be accessed, albeit with significant impact on outstanding bank investors.
At present if bail-in was to occur in Italy, all Tier 1 and subordinated debt would realistically be wiped out for the banks in question. This is because under BRRD, an 8% loss of total liabilities is required for bail-in before any public funds can be used. We calculate that 8% would exceed many of the Italian banks Tier 1 and subordinated debt; and would even touch the senior unsecured holders in some cases.
Furthermore, in a resolution case, one would assume an even lower value for CET1 than present values. This is because in resolution, the value of assets often decline as parts of a business are attempted to be sold off quickly. The current CET1 values apply a going-concern assumption, but this would deteriorate in a resolution scenario. This deterioration would cause further impact to Italian bank investors in bail-in, especially senior unsecured holders.
The European authorities stick to the view that state aid cannot be given without bail-in being enacted (see box left). This most likely will lead to a showdown shortly with the Italian government against the European authorities as the NPL concern propagates.
EU Court of Justice 19th July 2016 Is bail-in before state aid legal?
“Burden-sharing by shareholders and subordinated creditors as a prerequisite for the authorization…of state aid to a bank with a shortfall, is not contrary to EU law.”
Therefore, why is bail-in not being enacted?
One of the key reasons for not enacting bail-in is the impact to the investor base, and especially retail investors. Italian retail investors hold one-third of the outstanding senior bank debt and 46% of total subordinated bank debt, according to the IMF. Therefore the impact of bail-in to the retail investor base would be enormous and could have severe political implications. For the Italian banking sector as a whole, the impact to retail investors with subordinated debt, would be approximately EUR 31bn. Meanwhile, the impact to retail investors with senior unsecured debt would be EUR 20bn per 10% loss.
Regarding institutional investors, the political impact may not be as large as with retail investors, however the long term impact could be pronounced. A bail-in action by the Italian government, could cause even further institutional investor flight from the country, exacerbating the problems Italian banks face in raising capital. Investors may demand higher compensation for their capital, and the funding costs for banks may rise too high to be sustainable. In this scenario, equity cash calls would continue to be extremely difficult.
Could retail investors be saved from bail-in?
Any government desire to protect retail investors and cause institutional investors to foot the bill could be met with severe institutional investor and legal resistance (as shown after the Novo Banco actions in December 2015).
One solution could be that a separate fund is created to compensate retail investors for their losses related to a bail-in. This fund could be funded privately by the Italian banks over a long time frame, and would not constitute state aid. This method is not untested as, in December 2015 when four Italian banks failed, Italian banks set up a fund to compensate retail investors who had lost money during bail-in proceedings. This method would help protect retail investors and help cushion the political fallout.
Additionally, the October 2016 referendum by Matteo Renzi for constitutional reform to restrict the role of the upper house (Senato) makes the Italian government particulary sensitive to upsetting retail investors. The Prime Minister has stated that he will step down should he lose the vote. Any decision to bail-in retail investors before this date, could mean that public opinion could turn on Mario Renzi and he would have to leave office.
How would the Italian resolution framework apply for any bail-in?
An updated Italian resolution framework does not actually come into affect until January 2019 (one of the latest in Europe). This planned Italian resolution framework will subordinate senior unsecured debt to depositors.
Thus, until this date, senior unsecured holders and uninsured deposits rank pari-passu for any resolution that may occur this year. Consequently, any bail-in in the next two and a half years could significantly impact uninsured deposit holders also. Once again amplifying the effect of any bail-in in the country.
Finally, could Italy follow aspects from the Spanish route?
Following parts of the Spanish route might be a solution, as the Italian situation looks rather similar to the situation faced in Spain in 2012 when the financial sector faced severe difficulties. These were also related to non-performing loans, as well as the large structural size of the banking sector. The Spanish banks needed to be recapitalised and consolidated, and in the end, a European Stability Mechanism (ESM) loan was given to recapitalise the banks, while the sector was restructured.
Bail-out occurred in Spain, however, under BRRD rules since January 2016, Italy would not be able to do this so easily. Nevertheless, portions of the consolidation actions in Spain could be applied to the Italian situation. The Italian banking sector is in drastic need of consolidation. In Q2 2016, there were approximately 640 banks operating within Italy, the IMF calculated. The Italian legislation from Q2 2015 (a move to convert cooperative banks into joint stock companies), and February 2016 (joint-stock corporations where minimum paid-in capital should be over EUR1bn), should assist in consolidation, but this will all take time.
The consolidation format could follow the Bankia example of 2010 in Spain, whereby the operations of seven regional savings banks were consolidated. As it would be a travesty not to use 2016 as a year to address Italian bankruptcy laws and the much needed consolidation in the sector. Throwing taxpayer money at the situation in summer 2016, without structural change, may continue to elongate the “zombie” bank syndrome which some smaller banks in Italy are facing.
Tomas Kinmonth, CFA
Fixed Income Strategist
ABN AMRO Bank N.V.
Group Economics | Macro & Financial Markets Research
Gustav Mahlerlaan 10 (HQ7000), 1082 PP Amsterdam, the Netherlands
Phone +31 20 628 1405 | Mobile +31 610 722 073