EUROPEAN COMMISSION

Brussels, 14.3.2018

SWD(2018) 75 final

COMMISSION STAFF WORKING DOCUMENT

IMPACT ASSESSMENT

The development of secondary markets for non-performing loans by removing undue impediments to loan servicing by third parties and the transfer of loans (Part 1/2)

And

Accelerated Extrajudicial Collateral Enforcement (Part 2/2)

Accompanying the document

Proposal for a DIRECTIVE OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL

on credit servicers, credit purchasers and the recovery of collateral

{COM(2018) 135 final}
{SWD(2018) 76 final}

Table of contents

1.Introduction: Political and legal context

2.Problem definition

3.Why should the EU act?

4.Objectives: What is to be achieved?

5.What are the available policy options?

6.What are the impacts of the policy options?

7.How do the options compare?

8.Preferred options

9.How will actual impacts be monitored and evaluated?

List of references

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

2.Organisation and timing

3.Consultation of the RSB

4.Evidence, sources and quality

Annex 2: Stakeholder consultation

Context

1. Coverage and representativeness of the consultation replies

2. The role of NPLs and NPL markets

3. Measures targeting NPL investors

4. The economic function of loan servicers

5. Policy measures targeting loan servicers

Annex 3: Who is affected and how?

1.Practical implications of the initiative

2.Summary of costs and benefits

Annex 4: Analytical methods

1. A stylised view on demand and supply of the NPL market

2. Cross country analysis

3. Quantifying the impact of the different policy options on NPL secondary markets- explanation and assumptions

4. Translating the scores for the assessment criteria into an overall ranking of the policy options

Annex 5: Market overview

1. Nature and size of the NPL market

2. Nature and size of the Loan servicing market

Annex 6: The regulatory framework of NPL transfers and loan servicers

1. A stocktake of Rules in the EU Member States: Results of the questionnaire to Member States

1.1Executive summary

1.2 Background

1.3. Assessment of answers to the questionnaire

1.3.1 Loan servicing activities

2. OBSTACLES FLAGGED IN THE PUBLIC CONSULTATION

Glossary

1.Introduction: Political and legal context

1.1. The need to address Non-performing Loans in the EU

Following the financial crisis, the regulatory framework for banks has changed substantially. The European Union has taken the lead in implementing reforms agreed globally at the level of the G20 and in the Basel Committee with the objective of reducing risk in the banking sector, reinforcing financial stability and avoiding that taxpayers have to contribute financially to the costs of failing banks. In addition to these measures, the institutional arrangements for the supervision and resolution of banks in the EU have been strengthened fundamentally with the establishment of the first two pillars of the Banking Union (BU): the Single Supervisory Mechanism (SSM) and the Single Resolution Mechanism (SRM). 1 As a result of these measures, the EU banking sector is in a much better shape today than in previous years.

Nevertheless, several challenges remain to be addressed, including how to decisively address the high stocks of non-performing loans (NPLs) and other non-performing exposures (NPEs) 2 . NPLs have piled up in parts of the EU banking sector in the aftermath of the financial and sovereign crises and ensuing recessions. High levels of NPLs in parts of the banking sector pose significant risks to financial stability and the overall economy in the EU, unlike in other major economies such as the United States or Japan which have previously taken a number of actions to reduce the level of NPLs and repair banks’ balance sheets. 3  

High NPL ratios 4 can weigh on a bank's short- and longer-term performance through two main channels. First, NPLs generate less income than performing loans – thus reducing bank profitability – and may cause losses that diminish the bank's capital. In the most severe cases, these effects can put in question the viability of a bank with potential implications for financial stability. Second, NPLs tie up significant amounts of a bank's resources, both human and financial. 5 Banks saddled with high levels of NPEs have therefore only a limited capacity to provide new credit to viable businesses. Small and medium-sized enterprises (SMEs) are particularly affected by the reduced credit supply, as they rely on bank lending to a much greater extent than larger companies, thereby affecting economic growth and job creation. 6  For all these reasons, the Commission has for a long time highlighted the urgency of taking the necessary measures to address the risks related to NPLs.

While tackling NPLs is primarily the responsibility of national authorities 7 , there is also a clear EU dimension of the NPLs issue. Given the high level of economic and financial integration in the EU, and especially within the euro area (EA), there are important potential spill-over effects from Member States with high levels of NPLs to the economies of other Member States and the EU at large, both in terms of economic growth and financial stability. 8 Weak growth in some Member States due to elevated NPL levels might affect economic growth elsewhere. Also, weak balance sheets of just a few banks can negatively affect investors' general perception of the value and soundness of other EU banks. This can unnecessarily raise the funding costs for the sector as a whole, which may adversely affect the cost of credit to borrowers.

Addressing high stocks of NPLs and their possible future accumulation is therefore essential for restoring the competitiveness of the banking sector, preserving financial stability and supporting lending to create jobs and growth. This analysis is shared by a number of reports from European institutions, international organisations, and think tanks. 9  

1.2 Recent evolution of NPLs

The general improvement in NPL ratios over recent years continued in 2017, as did the quality of banks’ loans portfolios. The latest figures confirm the downward trend of the NPL ratio, which declined to 4.6% (Q2 2017), down by roughly 1 percentage point (pp) year-on-year (see Figure 1). This reduction was mainly the result of one‐off events that impacted all bank‐size classes, in particular smaller banks. However, the ratio remains elevated when compared to historical norms and to other regions 10 and the total volume of NPLs across the EU is still at the level of EUR 950 billion. 11  

The situation differs significantly across Member States (see Figure 2). Several countries still have high NPL ratios (9 had ratios above 10% in the second quarter of 2017), while others have rather low ratios (10 Member States were below 3%).

There is evidence of some progress in reducing NPL ratios in the most affected countries, owing to a combination of policy actions and a stronger macroeconomic environment. However, significant risks to economic growth and financial stability remain and progress is still slow, especially where it is needed the most. Structural impediments continue to hamper a faster fall in NPL stocks. Provisioning is often still too slow and insufficient to allow for effectively resolving and preventing any critical accumulation of NPLs in the future. Among other elements, activity on secondary markets for NPLs is also not yet sufficient to substantially contribute to NPL reduction efforts, notwithstanding the increased interest from certain investor groups and the increasing volume of NPL-related transactions.

1.3 Towards a comprehensive package of measures to address NPLs

A comprehensive and credible strategy to address NPLs is an essential and urgent step towards restoring the viability of – and hence investor confidence in – the EU banking sector. Pursuing a comprehensive strategy and taking determined action to address NPLs is also essential for the smooth functioning of the Banking Union and the Capital Markets Union (CMU) and for a stable and integrated financial system. In this way, the resilience of the Economic and Monetary Union to adverse shocks will be enhanced by facilitating private risk-sharing across borders, while at the same time reducing the need for public risk-sharing.

Integrating national and EU-level efforts is needed to address the NPL problem, both on the existing NPL stocks and on future NPL flows. Reflecting the EU dimension and building on previous work by the Commission and other competent EU authorities, the Council adopted in July 2017 an Action Plan To Tackle Non-Performing Loans in Europe. 12 It recognises that work in this area must be based on a comprehensive approach combining a mix of complementary policy actions, since the complexity of the problem simply does not lend itself to a single ‘silver bullet’ solution.

The Council Action Plan combines various measures by national governments, bank supervisors and EU institutions that improve the tools and incentives for banks to pro-actively address NPLs either by internal work-out or through disposal. In practice, this means enhancing legal frameworks relevant for both the prevention and resolution of NPLs, including the functioning of secondary markets. However, other measures such as improving the availability and quality of data on NPLs or improving the market infrastructure (eg. set-up of trading or information platforms) are equally important. If the right pre-conditions are present, tools such as Asset Management Companies are also an efficient way to allow resolution of NPLs while removing NPLs from the banking system in the short term.

The Commission has committed to delivering on the parts of the NPL Action Plan within its remit. Accordingly, the Commission announced in its October 2017 Communication on completing Banking Union a comprehensive package for tackling high NPL ratios, to be put forward by Spring 2018. 13  

This "Spring package" consists of the following measures:

·A Blueprint for how national Asset Management Companies (AMCs) can be set up in compliance with existing EU banking and State aid rules by building on best practices learned from past experiences in Member States.

·A legislative initiative to further develop secondary markets for NPLs, especially with the aim of removing undue impediments to loan servicing by third parties and to the transfer of loans to third parties.

·A legislative initiative to enhance the protection of secured creditors by allowing them more efficient methods of value recovery from secured loans through Accelerated Extrajudicial Collateral Enforcement (AECE). This refers to an expedited and efficient out-of-court enforcement mechanism which enables secured creditors (banks) in all Member States to recover value from collateral granted by companies and entrepreneurs to secure loans. 14  

·A legislative initiative amending the Capital Requirement Regulation (CRR), with regard to the introduction of minimum coverage requirements for incurred and expected losses on future NPLs arising from newly originated loans, in order to backstop potential under-provisioning of future NPLs and prevent their build-up on banks’ balance sheets.

·A way forward to foster the transparency on NPLs in Europe by improving the data availability and comparability as regards NPLs, and potentially supporting the development by market participants of NPL information platforms or credit registers.  15

The Council Action plan initiatives under the responsibility of other EU institutions and competent authorities include, among others:

·General guidelines on NPL management applicable to all EU banks;

·Detailed guidelines on banks' loan origination, monitoring and internal governance, addressing in particular transparency and borrower affordability assessment;

·Macro-prudential approaches to prevent the emergence of system-wide NPL problems, taking into account potential pro-cyclicality and financial stability implications of NPL policy measures;

·Enhanced disclosure requirements on banks' asset quality and non-performing loans.

1.4 Commonalities and interdependencies of the various measures

The legislative and non-legislative initiatives of the Council Action plan are interlinked and mutually reinforcing. They should create the appropriate environment for dealing with NPLs on banks' balance sheets. Some of them have an impact on the reduction of the current stock of NPLs, and all are relevant for reducing risks of future NPL accumulation. Their impact is expected to be different across Member States and affected institutions. Some will have a stronger impact on banks' ex ante risk assessment at loan origination, some will foster swift recognition and better management of NPLs, and others will enhance the market value of such NPLs.

The Commission's three legislative initiatives, namely i) statutory prudential backstops for loan loss coverage; ii) the development of secondary markets for NPLs, and iii) accelerated extrajudicial collateral enforcement mechanisms, mutually reinforce each other and also interact with the other measures of the Council Action Plan. For example, the prudential backstops initiative ensures that credit losses on future NPLs are sufficiently covered, making their resolution and/or disposal easier. These effects would be complemented by better developed secondary markets for NPLs as these would make demand for NPLs more competitive and raise their market value. Furthermore, accelerated collateral enforcement as a swift mechanism for recovery of collateral value would reduce the costs for resolving NPLs. These interactions are described in greater detail in the below box.

Figure 3 Commission's policy initiatives within the NPL Action Plan

The effectiveness of the three aforementioned legislative measures would increase if banks are adequately capitalised in the future. Better capitalised banks will be more eager to sell NPLs in the secondary market or to realise the collateral of a non-performing loan in a timely fashion. Furthermore, statutory minimum coverage requirements would provide strong incentives for banks' management to prevent the accumulation of future NPLs through better NPL management and stronger loan origination practices. This will reinforce the expected effects of the EBA’s and ECB’s work on banks' loan origination, NPL management, monitoring and internal governance practices. Work on NPL information and market infrastructure would further enhance the functioning of NPLs secondary markets. Lastly, measures related to loan enforcement would complement the Commission's November 2016 proposal for a Directive on business insolvency, preventive restructuring and second chance, by increasing the chances that viable businesses survive while non-viable activities are swiftly resolved. 16  

1.5 The scope of the impact assessment

The initiative to develop secondary markets for NPLs discussed in this text focuses on a specific issue that is not taken up in any of the other policy measures in the Action Plan, namely to remove impediments to transfers of NPLs from banks to other entities and to simplify and harmonise requirements for loan servicers. Being one of the four key areas in the Action Plan, stakeholders signalled the importance of secondary markets for the resolution of NPLs in the public consultation that preceded this impact assessment (see Annex 2) as did banks contributing to the EBA Risk Assessment Questionnaire. In the latter, they considered the lack of secondary markets for NPLs one of the two most important impediments to the resolution of NPLs. 17  

The initiative analysed here is unique among all measures following the Council Action Plan as it is the only legislative measure that targets an increase in demand and to raise competitive pressure on the demand side of the NPL market. Most other measures in the NPL Action Plan will also have an impact on NPL secondary markets. The introduction of prudential backstops would increase banks' incentive to sell NPLs. The establishment of AMCs has been historically one of the driving forces kick-starting secondary markets of NPLs, bringing in economies of scale, advantages of specialisation and improved pool valuation. 18 Functioning AMCs can thereby help to both expand and to smooth the supply side of the NPL market. Data standardisation and transaction platforms improve the matching process of demand for and supply of NPLs. Measures to improve insolvency and enforcement would increase the recovery value of NPL, increasing the value for both demand and supply side. Annex 4.1 describes how the different initiatives set up in the Council Action Plan should impact on prices and traded volumes by depicting a stylised view on demand and supply conditions on the secondary market for NPLs.

Figure 5: Loan transactions and NPLs across selected EU Member States

Source : COM calculations with data from EBA and various consultancies (see Annex 4.2).

Since most other measures of the NPL Action Plan will also have an impact on the secondary market for NPL loans, the ultimate impact of this work stream depends also on the success of the other measures in the NPL Action Plan. At the same time, the effectiveness of other policies is questioned without a functioning secondary market for NPLs. Especially the benefits of AMC and supervisory action could become fruitless if demand for NPLs is missing.

2.Problem definition

The consequence of missing or underdeveloped secondary markets for NPLs is that banks with high NPLs have limited scope to sell them to non-banks or only at high transaction costs leading to low prices. This holds in particular for smaller banks, which may have high NPL ratios and a strong incentive to sell, but find that search and transaction costs are over-proportionally high for smaller portfolios. 19 The prospect of low prices on loan sales means banks may realise losses, which erodes their capital base and therefore represents a disincentive to sell. 20  

If NPLs cannot to be disposed, they stay on banks' balance sheets and require provisioning, which reduces banks' profitability and business opportunities. NPLs also generate uncertainty about asset quality, which decreases investors' demand and increases banks' capital costs. The overall result of both effects is reduced credit supply and higher lending rates, which tend to disproportionately affect lending to SMEs. 21 High NPLs bind bank operating resources and potentially prevent banks from carrying out more productive uses. 22 This effect is particularly material in smaller banks having less specialised staff. Moreover, the difficulty to assess the value of a bank which has a large stock of NPL on its balance sheet holds back merger and acquisition activity in the EU banking sector, often described as oversized. 23 Finally, a higher stock of NPLs on banks' balance sheets as consequence of a lack of secondary markets for NPL would mean that banks' are more exposed to financial turmoil, i.e. the risk of financial instability is higher.

2.1    What is the problem?

Despite some momentum in recent years, secondary markets for non-performing loans hardly exist in Europe. According to the SSM, euro-area Member States do not have a developed NPL market except Spain and Ireland, whose state of development is characterised as medium. 24 Also outside the euro area, markets are small and underdeveloped in most EU Member States except in the UK. While a genuine single market for NPLs in the EU would be difficult to realise in view of considerable cross-country differences in other relevant areas, in particular insolvency law, investors have been looking for opportunities beyond borders, i.e. some of those active first in the UK and Ireland and then in Spain, have entered the market in Italy or several CEECs.

Markets tend to be characterised by comparatively small trade volumes, a few large transactions involving a limited number of active investors, large bid-ask spreads when counterparts enter negotiations and a lack of transparency on market prices. 25 At the same time, Member States with higher loan sales recorded a stronger decline in their banks' NPL ratios (see Figure 10 and Figure 11 ), suggesting that secondary markets for NPL are contributing importantly to reduce NPL ratios.

Apart from data of NPLs on banks’ balance sheets recorded by banking supervisors and central banks, which represent the potential supply of NPL, there are no official statistics to track NPL markets. Banks are not obliged to reveal them to statistical offices and often have no incentive to disclose details. Some consultancies collect data of individual sales from various sources and publish their information in reports. This data is used in this Impact Assessment. Information about realised market prices is generally not available, but treated as confidential by the parties involved in the transaction. See Annex 4.2 for a discussion of data availability and quality and Box 1 in Annex 5.1 for a review of issues with the data on loan sale volumes.

Major consultancies point to less than EUR 120 billion of transactions in debt sales in 2016 in the EU, which corresponds to estimated EUR 100 billion NPLs sold. 26 The market overview in Annex 5.1 documents that the secondary NPL market is concentrated in the EU, with a strong clustering in four countries (ES, IE, IT and the UK) and dominance of large buyers (23% market share of the top five buyers, largely US or UK domiciled, over the last 2½ years). 27 According to market sources, prices depend strongly on the characteristics of the underlying loans, varying from 5-10% of face value for unsecured consumer loans to 50-60% for secured (mortgage) loans. 28 Annex 5.1 reviews the existing data on market structures.

This apparent malfunctioning of the market is driven by problems of incentives for engaging in transactions and sufficient information about possible transactions that banks as sellers of NPLs and non-banks as potential buyers face. Information problem occur on both sides because the value of an NPL is difficult to establish given its dependence on the likelihood and amount borrowers will pay back, the value of any underlying collateral and the time and effectiveness of legal or out-of-court enforcement. 29 They lead to high transaction costs. This is visible in a high gap between prices offered and bid for NPLs, entailing disincentives for banks to sell as well as limited participation of potential investors. NPLs are not an established asset class investment funds traditionally focus on, implying that precedent their involvement they need to set up a new strategic orientation and investment mandates. A particular factor that can discourage NPL investors to enter the market is the difficulty to access third-party loan servicers. Loan servicers have been virtually absent in most EU Member States until recently. Their activity is segmented by country due to local regulations, which prevents them from realising scale economies.

Incentive problems give rise to market failures leading to high transaction costs

On both market sides, there are underlying incentive problems that lead to a wide bid-ask spread. 30 Buyers assume, and therefore discount, the sellers' incentive to overrate the quality of the product. 31 Buyers have less information about the quality of the asset than the sellers. Exposure to such information disadvantage about the quality of the asset will be reflected in a risk premium that reduces the bid price of the prospective buyer. At the same time, the selling banks anticipate that the potential acquirer assumes that the bank is under pressure to divest the NPL portfolio. Otherwise it would keep it on its balance sheet and take the losses. In this strategic setting, banks have an interest to demand a higher price than justified by the true value of the NPL portfolio whereas investors have an incentive to understate their preparedness to pay. 32  

Negotiation about the purchase of an NPL entails significant transaction costs to agree on a price and other contractual terms, in addition to the information costs the buyer has to carry to evaluate the NPL portfolio. The difficulty in closing the gap between bid and offered prices and arriving on an agreed price between supplier and buyer leads to a long negotiation period and often prevents that deals are concluded. The available data reported in Figure 6 reveals that about a quarter of the loan sales transactions initiated in 2015 and 2016 were still not concluded in September 2017.

Since market participants do not disclose actual bid and ask prices, the gap between them can only be estimated by means of a theoretical model that combines the main determinants of price formation on both market sides. Figure 7 shows how initial bid and ask prices could differ across EU Member States using the model presented in Annex 5.3 The difference from the bold 45 degree line indicates the size of the bid-ask spread.

Additional factors that influence bid and ask prices are listed in Box 2 in Annex 5.1. The other measures in the NPL Action Plan, such as the establishment of transaction platforms or data standardisation, would reduce transaction costs and therewith the bid-ask gap and thereby foster demand for NPLs. They would not address the mentioned distorted incentives, which lead to reluctance of banks to sell. Establishment of AMCs could also reduce transaction costs and the bid ask spread, for example through re-packaging of NPL portfolios from different banks. They could also be instrumental in re-balancing bargaining power on NPL markets, thereby addressing the incentive issue. Market structure would then be determined by few large players with market power on both demand and supply side of the market. While this may lead to higher NPL prices and lower bid-ask gap, it may also discourage market entry of further, especially smaller, investors and therewith not engineer an increase in demand for NPLs.

Limited buyers' participation and weak competition leading to lower bid prices and concentration on large NPL portfolios

Low demand for NPLs has led to small transaction volumes and low bid prices. 33 Market entry of new non-bank investors could enlarge the investor base and thereby increase demand and competitive pressure. Higher competition among NPL buyers, in turn, would increase bid prices, entailing larger incentives for banks to sell. Entry conditions for potential NPL investors are therefore a critical parameter to stimulate demand.

New entrants could come from various sources. There are third-country investment funds that target distressed debt or special situations, but not yet active in Europe. There are also smaller NPL investors in European Member States that target their home market, but refrain from acquiring loans in other Member States. Finally, there are also a few European firms that acquire NPL portfolios from various European banks, but specialise on specific asset classes (see Annex 5.1). Institutional investors such as pension funds and insurance companies are usually not active as direct market participants, but according to market sources some hold shares in investment funds that buy NPLs. 34 Taking a standard market diagram, an increase in the investor base would translate into higher bid prices and higher demand (see Annex 5.1). Policy measures that stimulate banks to supply NPLs or improve matching process are subject to other NPL work strands in the Action Plan (see Annex 4.1).

Foreign firms have been the largest investors in NPLs. Among the largest 10 investors in global distressed debt are 9 domiciled in the US and one in Canada (see Table A.5.1 in Annex 5.1). Broadening the potential investor base would be essential to increase demand for European NPLs. Since smaller European banks have a large exposure to NPLs, there seems to be also a mismatch in the size of NPL portfolios between what smaller banks could sell and non-bank investors currently active in the markets are interested to buy. While there are some smaller to medium-sized European investors in the market, they seem so far to be specialised on specific asset classes or Member States, and realise somewhat smaller average transaction volumes. 35

Limited availability and limited geographical reach of loan servicers

A particular factor that can discourage NPL investors to enter the market is the difficulty to access third-party loan servicers. Many submissions to the public consultation supports the notion that access to loan servicers is important for NPL markets to develop. 36 Member States with high NPL volumes and relatively vivid loan sales such as IE, IT and ES have on average more loan servicers whereas in some other Member States such as FR and AT, loan sales are under-proportional and loan servicers play little role (see Annex 5.2).

Loan servicers take care of the "after-sale services", they administer the interest payments of debtors, collect the principal, send notices and conduct other activities that affect the recovery value of NPLs. They have a particular role in the administration of NPL portfolios once these are sold, because it is important for the buyer of an NPL portfolio to exclude the originating bank from the debt collection to take full ownership and resolve any possible moral hazard. 37 See Box 1 in Annex 5.2 for a review of the value added of loan servicers.

Most buyers of NPLs are investment funds or asset managers without loan servicing capacity. Their expertise is in asset valuation and risk taking. They require access to third-party loan servicers for managing NPLs. Since loan servicers request a fee for doing so, 38 high costs for loan servicing are a potentially important deterrent for non-bank investors to acquire NPLs.

Facilitating the expansion of loan servicers across borders would allow them to tap scale economies, compete for business and provide their services to NPL investors at lower prices. Loan servicers have been virtually absent in most EU Member States until recently. 39 Despite dynamic adjustment in the sector in the last two years, activity has remained fragmented along national lines. 40 Loan servicers are segmented by country, due to local regulations, and by asset class. It is known from the US market that loan servicing benefits from scale effects, which implies that small loan servicers are less efficient. 41 Though based on a small number of observations and subject to a number of methodological caveats, Figure 8 suggests that third-country loan servicers active in the EU are on average larger and more profitable.

Figure 8: Size and profitability of firms offering loan servicing in the EU

Source: Commission calculations with individual firm data derived from Orbis or company accounts. See Annex 5.2.

Figure 9: Problem Tree

2.2    What are the problem drivers?

The public consultation and a questionnaire sent to EU Member States about rules pertaining to NPL investors and loan servicers in their jurisdiction helped identify factors that discourage participation and limit incentives to conduct cross-border activity (see Annexes 2 and 6).

High entry costs from authorisation requirements for loan purchases in some Member States

Entry conditions are a critical parameter to augment the investor base for NPLs, to increase competitive pressure and thereby kick-start market development. 42 In several Member States, non-banks are required to have authorisation from a public body if they purchase loans from banks. Especially where a full banking license (as opposed to other more specific licensing requirements) and a physical presence in the Member State concerned are required this represent costly entry barriers for potential NPL buyers in some Member States. 43 Motivation for authorisation requirements in some cases is based on debtor or data protection concerns, in others on a definition of bank activity that includes factoring services, or links holding of a loan portfolio to credit creation. Significant compliance costs seem also due if the NPL purchase requires the establishment of a securitisation vehicle or investment firm. 44 Other costly barriers relate to registration in each Member State they want to be active in, administrative delays and limitations on the loans they are allowed to acquire. 45 Non-EU institutions face the same requirements as EU-domiciled investors in most, but not all, Member States.  46 (see Annex 6).

Table 1: Entry conditions for NPL investors

Entry costs differ depending on firm characteristics and Member State (see Annex 3.2 and the Box in the Annex). Market sources describe them as not insurmountable, though scarcity of data and large variation in the few observations made available to Commission services do not allow an in-depth assessment. Costs to obtain authorisation are estimated to be below EUR 100,000 in most cases, unless a banking license or a securitisation vehicle is required.  For example, if a NPL investor can perform under the regulatory regime of an investment fund, the regulatory start-up costs would range between about EUR 10,000 to about EUR 15,000. 47  

In addition to the actual compliance costs in monetary terms, authorisation and licencing procedures entail additional economic costs because they require potential market entrants to acquire legal expertise to understand and fulfil obligations. 48 Taking the investment fund industry as a benchmark, a recent Commission study suggests that direct regulatory fees could amount to less than 20% of the regulatory start-up costs, about 40% of the regulatory start-up costs might be attributable to compliance costs in terms of labour costs and another approximate 40% to pay external servicers for local facilities in the host country.  49 Market sources interviewed by the Commission assessed the average of total costs to enter a new NPL market at about EUR 60,000 to 100,000. Hence, compliance costs are deemed not particularly high in relation with total entry costs incurred by investment firms.

Different legal provisions regarding loan disposal and NPL resolution across Member States

A further obstacle to market entry stems from the legal differences and the uncertainty this creates for the loan acquirer about their rights with respect to loan enforcement from the ultimate debtor. 50 See Table 1 for an overview of specific provisions in EU Member States. The European Commission’s survey revealed that while all Member States allow the transfer of a loan, the legal instrument is different as it either entails the transfer of the credit rights or the transfer of the loan contract. Hence, entry and conduct rules for investors willing to buy NPLs differ across EU Member States and in some Member States by type of loan, implying that investors' interest to buy and therewith banks' ability to sell NPLs is fragmented across Member States and also by asset class. 51

Differences in the legal framework entail additional costs for investors active on different national markets. They mean potential foreign NPL non-bank investors need to identify and respect the relevant licensing requirements and compliance costs in each Member State they want to be active in. Moreover, loan acquirers need to adapt their business model to each legal framework, which implies that for each Member State where they want to buy NPLs they have to develop an idiosyncratic approach for the valuation of loans, their relationship with the debtor, the procedures to enforce loans and other parts of their business conduct.

Table 2: Different requirements to business activity of NPL investors across EU Member States

The costs for NPL investors that the fragmentation of legal frameworks across Member States entails could be a reason why about 75% of the investors that bought NPL portfolios in 2015 and 2016 did so in only one Member State (see Table in Annex 5.1). The replies to the public consultations reveal that stakeholders consider the legal framework, insolvency rules and local habits as obstacles for cross-border activity. Some respondents also referred to data issues or incentive problems as factors (see Annex 2).

A cost estimate of the impact of these factors is not possible given the lack of data and strong differences across Member States. Information from market sources suggests that costs for supervisory reporting, internal audit, risk compliance, credit management procedures and anti-money laundering differ substantially across firms. A common pattern is that costs for compliance with anti-money laundering legislation are sizeable, which is in line with the prominence of know-your-customers concerns in the replies to the public consultation.

High entry costs from licensing requirements and other obligations for loan servicers in some Member States

Loan servicers are exposed to challenges similar to NPL investors with respect to authorisation and licensing. Regulatory entry barriers are more widespread across Member States than those for NPL investors and often motivated by debtor and data protection considerations. For this reason, some Member States even request physical presence in the Member State. The request for authorisation and domiciliation in particularly allows supervision by domestic authorities. Since restructuring of a loan can entail new lending, some Member States request from loan servicers to obtain a banking license. 52 Some Member States also insist that NPL investors make use of loan servicers that are licensed and supervised by their authorities (See Annex 6) 53 . Non-EU loan servicers are permitted in all Member States, except one. 54  

Table 3: Entry conditions for loan servicers

* depends on decision of supervisor, ** if activity covers lending or refinancing.

For more details, see Annex 6 and Appendix 6.A.3.

While fees for obtaining a license vary across Member States, they are overall small compared to overall entry costs loan servicers face, estimated to amount to EUR 5-15 million. 55 From the limited information the Commission services were able to obtain, one-off fees for the licensing range from a few hundred to more than EUR 50,000. Annual licencing fees vary significantly as well, ranging from a few hundred euro per annum to more than EUR 30,000. Compliance costs for data reporting could add to these set-up and licencing fees as well as the costs to comply with anti-money laundering rules that may prove significant.

High costs of cross-border expansion from different licensing requirements and other obligations for loan servicers across Member States

Cross border expansion can help loan servicers to grow and realise scale effects, thereby allowing them to offer lower prices for NPL investors. Since this requires multiple authorisation processes and adjustment of the business models to national conditions, cross border expansion is made more costly by differences in licensing and conduct rules across Member States. Loan servicing activities primarily fall under the freedom of contract, and there are no formal legal definitions of 'servicing', 'managing' and/or 'debt collection' of loans in most Member States. The resulting legal uncertainty about definitions applied in other Member States could be a reason why few loan servicers pursue cross-border expansion strategies. Reduction of legal uncertainty by the adoption of legislation that governs the establishment of loan servicing firms, as part of the country's economic adjustment programmes, has been instrumental in fostering the market in Ireland. Comparable legislation has also been adopted in Greece and has led to the establishment of loan servicing firms in that country very recently. 56

A specific point to loan servicers is how their relationship with non-bank NPL investors is governed, which entails possible restrictions on which services they are allowed to offer to the latter. There are no explicit prerequisites that a creditor has to satisfy before outsourcing certain servicing functions unless outsourcing is deemed to affect core functions or services in almost all Member States. However, loan monitoring and refinancing, which are typical by-products of loan servicing are considered a core services in some Member States. 57 In this case, outsourcing to loan servicers is not allowed or tied to strict requirements. Moreover, the creditor cannot outsource the undertaking formal enforcement actions in the large majority of Member States, i.e. investor-linked servicers are not permitted to undertake such actions on the creditor’s behalf.

The variation in licensing costs across Member States referred to above can serve as an approximation of the entry burden to loan servicing markets. In the absence of direct cost estimates, evidence that they have a material impact can be derived from the observation that as regards cross-border entry or expansion of loan servicers, the most frequent approach in the last years has been the acquisition of existing national loan servicers, implying that expansion to a new market is difficult without national incumbents already being present.

Different borrower rights and legal requirements for privacy and data

Respondents to the public consultation stress data and information problems to be a major obstacle to the acquisition of NPLs. Restrictions on non-bank investors are motivated by concerns about bank secrecy and personal data protection. These restrictions impair the transfer of information from the bank to the non-bank investor. This, in turn, complicates the evaluation of the value of the loan portfolio before the sale is signed, i.e. the potential buyer has only access to limited and anonymised information from which he needs to assess the recovery value of the NPL and, to the extent loans are collateralised, the value of the underlying collateral. The existing fragmentation among Member States of rules on data protection renders data management by non-bank investors more difficult in case they aim administer loans from different jurisdictions at a central place in order to realise scale effects. 58  

Bank secrecy provisions generally contain an exemption that allows the bank to disclose data which are necessary and proportionate for selling the loan. The FISMA survey unveiled legal constraints in a number of Member States. For example, the transfer of confidential data is only allowed under the debtor's consent or an authority's decision in a few Member States (see Annex 6). Despite possible constraints from privacy and personal data protection, banks as sellers and non-banks and buyers have found ways to cope with them, for example by providing anonymized or aggregated data in the pre-transaction phase. Hence, rather than making it impossible to negotiate a trade of NPLs, they increase transaction costs.

Cross-border transactions could entail conflict of laws. For example if an NPL purchasers from a different Member State than the originating bank and the borrower sells the loan on to a further NPL purchaser in another country. The latter may then question which national law applies, especially if the NPL purchaser is in a third country. For cross-border transactions in the EU, the so-called Brussels I and Rome I rules apply. In case of a third country and of a borrower that is a consumer, the consumer law and jurisdiction should, in principle, also apply but the situation would need to be assessed on a case by case basis taking also into account the third country rules on conflict of laws. Therefore, parties to a cross-border transfer of loans have to do their due diligence based on a set of potentially applicable laws. This inflates the costs of legal opinions required for due diligence 59 .

Borrower rights, in particular consumer rights, are different across Member States and also defined via different legal means, e.g. through insolvency regimes, borrower and consumer protection laws, or authorisation and supervision procedures. These rights do not change with the transfer of a loan from a bank to a third party. 60 However, the legal protection of borrowers' rights might be affected if the transfer of the loan modified the contractual relationship, if it led to a change of applicable law to a different country's law, or if it became subject to a different regime of general rules on debtors' protection. This could be problematic, in particular, for debts referred to consumers. Both NPL investors and loan servicers therefore need to adjust their business models to the legal regime in each Member State they want to be active in. Table 2 lists a number of rules that NPL investors and loan servicers have to adhere to in the different Member States. The existence of various country-specific legal requirements implies that a single market for NPLs in the EU will still remain segmented along national lines even if authorisation and conduct rules for NPL investors and loan servicers are standardised.

Differences in borrower rights across the Single Market motivate authorisation processes for NPL investors and loan servicers in some Member States. For example, IE and EL explicitly request compliance with borrower protection rules in their laws that authorise loan servicers, SE mandates the authorisation process to the data protection authority. Other Member States may not explicitly require a license out of the motivation to safeguard borrower rights or personal data protection. Registration and licensing, nevertheless, gives them an opportunity to monitor and/or supervise behaviour of the firm, therewith act if compliance with national law is jeopardised.

Are the problem drivers significant?

Policy makers have control over regulatory costs, such as direct costs of obtaining an authorisation to do business, and indirect costs that emerge from rules that govern the conduct of business and may limit profit opportunities. While the rules in place determine costs of compliance for market entrants, not all entry costs can be influenced through policy measures. An important fixed-cost component in the entry decision of the NPL investor is the due diligence that has to be performed on any NPL portfolio targeted. Moreover, potential entrants need to invest in studying local legal conditions that determine the recover value of the loans as well as the legal rights and obligations they would have as loan owners.

The replies to the public consultation suggest that a large share of the respondents considers an EU framework helpful and that licensing rules should be part of it (see Annex 2). While market sources indicated that regulatory costs are not their main concern when deciding to enter NPL and loan servicing markets, this suggests that regulatory costs are not an insignificant part of entry costs.

Even if these regulatory costs may not be high in absolute terms, their impact on entry decisions is amplified by:

·expanding the disadvantage vis-à-vis incumbent market players that already benefit from an information advantage from past experiences. Entry decisions are surrounded by uncertainty and complex interactions among a multitude of relevant factors. Most importantly, entry costs have the character of sunk costs, i.e. they are foregone and not reversible if the plan to enter the market is aborted or the activity turns out not to be sufficiently profitable. Since it is uncertain that a deal with a bank can be closed, the sunk cost character of entry investments has a significant impact on the decision to enter NPL markets.

·becoming recurring for each national market that the investor wants to enter or expand to. Regulatory costs and different national rules undermine the possibility of NPL investors to enter smaller markets and of loan servicers to realise scale effects. They especially increase search costs if either of the two considers expanding activity to another Member State. The different legal rules across EU Member States may also discourage particularly foreign investors to enter EU markets, for example US investment funds that are used to face uniform rules and try to realise scale economies from conducting large transactions. For example, scale economies are well documented for US loan servicers. 61

·translating into search costs required to find out what regulatory requirements and related costs actually are, which again increases with each national market the investor wants to enter. The latter two are particularly important for foreign investors. In the NPL market, entry costs entail search costs, compliance costs, costs for advise on legal and administrative matters. Investors active in the NPL market report that costs to understand local conditions and the relevant legislation matter importantly.

Other problem drivers

It should also be noted that the public consultation revealed a number of further factors, which stakeholders consider important to foster activity on NPL and loan servicing markets, and which are not addressed here. Stakeholders also often stressed the impact of harmonised insolvency frameworks and improved debt enforcement as relevant for the development of secondary markets. Access to data is also an important factor for potential buyers in order to assess the value of an NPL portfolio. Market sources often refer to a lack of supply of NPLs from banks. These are subject of other work strands of the Action Plan, addressing specifically, a review of national insolvency frameworks, templates for standardised data through which banks facilitate buyers’ evaluation of loan portfolios and means to establish a transaction platform. Taxation was also repeatedly listed. An obvious point is that banks are exempted from VAT while loan servicers are liable to VAT. Annex 6.2 summarises obstacles to the development of secondary markets for NPLs flagged in the public consultation, including those that are subject to other work strands of the NPL Action Plan.

2.3 How will the problem evolve?

Without supportive policy measures, one could expect that the investor base in NPL grows along its past trend and that loan sales may not increase by much as compared to the level of EUR 100 to 150 billion per annum as realised in the last years. Loan sales may even decline as past activity was concentrated in a few countries, and in some of them the stock of NPL has already declined (e.g. UK, DE) whereas loan sales in some Member States with high NPL ratios would remain at the modest level realised in the past. The consequence would be that revenues from NPLs sales remain modest and the NPL stock on banks' balance sheet declines gradually.

One could assume a decline in the NPL ratio along the GDP growth path. Credit growth has been sluggish in countries with high NPL ratios. This could continue and may feed back into relatively weaker economic growth in these countries. 62 If national markets become attractive (for example through supervisory pressure on banks to sell NPLs or reform of insolvency law that increase recovery values) market demand could develop endogenously and when faced with high opportunity costs of non-action, Member States may adopt legislation to support this process. The cases of Ireland and Greece suggest that even if the NPL problem is recognised by policy makers as requiring action, it takes time until for example laws that govern market entry into loan servicing markets are in place.

3.Why should the EU act?

3.1    Legal basis

Article 114 of the Treaty on the Functioning of the European Union (TFEU) confers to the European institutions the competence to lay down appropriate provisions that have as their objective the establishment and functioning of the single market. NPL purchases are a form of capital movement under the free movement of capital principle, which is applicable to investors from third-countries like the US as well. The problem that the initiative under consideration aims to address is related to different rules in the Member States as regards the rules for third parties acquiring NPLs from banks, as well as rules for offering loan servicing services (see Annexes 6 and 7) that restrict both the free flow of NPLs within the EU and investment opportunities for third-country investors. Consequently, the development of a single market for NPL investors and for loan servicers faces obstacles and without measures at EU level, national markets for NPLs and loan servicers will remain fragmented and in most Member States underdeveloped.

3.2    Subsidiarity: Necessity of EU action

The analytical work leading to the NPL Action Plan demonstrated that NPL ratios are currently high in a substantial number of Member States, reaching unsustainable levels in a few cases. 63 This legacy stock of bad debt creates risks of cross-border spill-overs throughout the EU economy and its financial system. Moreover, the high stock of NPLs alters market perceptions of the European banking sector as a whole and represents negative externalities for the whole EU. 64 These factors have become even more relevant in the context of the Banking Union. By contributing to enhanced growth and reducing financial fragmentation, measures to address the existing stock of NPLs would be beneficial for the EU as a whole. 65

With respect to building up or expanding the investor base for NPLs, Member States have an incentive to act and reduce regulatory barriers to attract foreign investors and to facilitate domestic investors or loan servicers to enter the market. A few Member States have actively acted to address the high stock of NPLs (IE, ES, EL, PT, SI), in most cases fostered through an EU/IMF economic adjustment programme. 66 Yet, progress in these Member States remains slow and did not counteract the spill-overs of risk perceptions of a weakened EU banking system as a whole. Given the inability to address the issue through action at the level of individual Member States, the subsidiarity principle warrants action at EU level.

While policy measures at the national level are possible, they are likely to cement market fragmentation. IE and EL set up legislation that governs the licensing process of loan servicers. Both, however, followed different approaches: the Irish law requests compliance with light fit-and-proper criteria for the authorisation without inference on business conduct, whereas the Greek law is more demanding, requesting for example loan servicers to act through a local entity and to take care of socially vulnerable groups (see Annex 66).

Stigma effects are a very likely reason why Member States have not yet taken a more active stance in reducing regulatory barriers to attract foreign investors to the NPL market. While there is no hard evidence on such factors, a number of channels may be relevant. Ultimate debtors represent a larger social group than creditors and they may see little advantage from a transfer of the ownership of their debt to foreign investors. They may even experience uncertainty if their debt is transferred from a creditor they know to an unknown third party entity and if they encounter difficulties to clarify whether or not the rules protecting them would remain valid. Since foreign hedge funds are the most visible NPL investors, their reputation as short-term oriented profit maximising entities may discourage policy makers from taking action to facilitate their operations. Legislators also strive to keep consumer protection at high levels and may fear that the transfer of NPLs to non-banks could challenge this protection. This initiative aims at addressing impediments to loan transfers, while ensuring that borrower rights in Member States are preserved.

3.3    Subsidiarity: Added value of EU action

An EU-wide framework for NPL buyers and loan servicers would help reap scale effects and reduce entry costs for firms intending to operate in different national markets. It would help overcome the coordination issue in EU NPL markets where weaknesses in demand and supply amplify each other. In particular, banks would benefit from a larger investor base since competition among investors would exert upward pressure on prices. Investors, in turn, would benefit from a unified European legal framework as it would reduce their entry costs and create more favourable conditions and infrastructure for their cross-border operations.

Under an EU framework for loan servicers, firms' activity would be less constrained by the size of the domestic market. Consequently, expansion to other markets would allow loan servicing companies to grow in size and to realise scale effects, and potentially charge lower fees to NPL investors. Moreover, NPL purchasers would no longer be required to build a new relationship with a loan servicer in each market, but they would be able to work with a loan servicer they worked with in other Member States.

So far, debt servicers and investors concentrate on a handful of national markets, probably those with highest profit margins. In order to reap opportunities in other markets, they need to acquire expertise about prevailing local regulation and about the availability of loan servicing firms. While these search and compliance costs may not be very high, they still represent an effective obstacle to entry if the target market is small.

Developing an EU-wide investor base for NPLs would also be key for the effectiveness of other measures in the NPL Action Plan since the efficiency of asset management companies in coordinating NPL supply, of bank supervision in earmarking NPLs, information templates to standardise information on NPLs depends on the potential demand for NPLs.

Measures at EU level would also be beneficial to overcome the stigma effects that Member States face when pursuing legislative changes at national level in this area. 67 However, these measures should not decrease the level of consumer protection, as the negative consequences of such a decrease could entail social and financial costs that might outweigh the possible benefits.

4.Objectives: What is to be achieved?

A first general objective of the NPL package is to limit risks to financial stability by reducing the stock of NPLs in the European banking system and by avoiding the build-up of NPLs in the EU banking system in the future. A second general objective of NPL reduction is the support of stable financing to the economy and therewith economic growth. Banks saddled with NPLs tend to face higher funding costs and capital requirements and lower profitability, which limits their ability to extend new credit. Persistently weak loan portfolios are thus a potential drag on financing of firms, households and ultimately economic growth. Functioning secondary markets for NPL would allow banks to sell NPLs to non-banks, thereby reducing risks of financial stability and liberate resources to expand lending to the economy.

The first specific objective of this initiative is to stimulate demand for NPLs by generating a larger investor basis, and consequently also greater competition among investors. Greater demand for NPLs and greater competition among investors is expected to contribute to both a higher volume of NPL transactions and higher bid prices on secondary markets. This should contribute to reducing NPLs in banks’ balance sheets, thereby enhancing banks' resilience, and ultimately improving lending to the economy as well as reducing risks to financial stability. While the immediate objective is to decrease the presently high level of NPLs in the EU and some of its Member States in particular, a larger participation of buyers on NPL markets will also be beneficial in case of future accumulation of NPLs on banks' balance sheets.

Mirroring an increase in buyers' participation on NPL markets, a second specific objective consists in a complementary increase in the capacity of loan servicing firms to absorb rising demand for their servicers from more loan sales by NPL investors. Demand would be most efficiently matched if prices for loan servicing are competitive and geographical reach expands to all Member States.

More specifically, to strengthen the demand side and competition on these markets, the initiative aims at: (i) facilitating market entry of NPL investors and loan servicers in MS with high NPLs and material obstacles to market entry, (ii) fostering the entry of smaller firms so that smaller banks have a higher chance of finding counterparts in NPL transactions, (iii) equal treatment across markets in Member States, allowing loan servicers to realise scale economies from cross-border operation, (iv)  enhancing competition through the entry of foreign firms. The first two items could be considered as means to address failure in (national) NPL markets, whereas the latter three may be instrumental in fostering a single NPL market.

The investor base would be largest if a single market was created, so that investors would not discriminate across conditions on various national markets, and if conditions would be supportive to foreign firms' market entry. The objective of a genuine single market seems very ambitious given the importance of national and local determinants on NPL markets. Still it would be desirable that both domestic and foreign investors can expand activity across borders as easily as possible, so as to come close to a shared investor base among all EU Member States. Harmonised licensing and conduct rules could deliver this. The same could be accomplished by converging rules at national level. This would result in increased transactions of NPLs, eventually leading to a reduction of NPLs in banks.

More competition on loan servicing markets should result in lower costs charged by loan servicers to NPL holders and create profit opportunities through lower administrative costs and/or allowing synergy effects with other business areas. Creating wider profit margins through lower administrative costs and/or allowing synergy effects with other business areas should also lead to higher preparedness to pay for NPL investors, implying a pass through to prices that banks can realise when they offer their NPL portfolio.

Changes to entry conditions for loan purchasers and loan servicers could change the relationship between borrower and creditor and their negotiation position, especially since the transfer of the loan entails that the borrower faces a new counterpart that she/he had not chosen and possibly not even known. If this new counterpart is located or authorised in a different Member State, the borrower may consider these rights undermined, even if the contractual obligations of the new creditor remain unchanged. It is therefore also an objective of the initiative (v) to ensure efficient supervision and (vi) adequate safeguards for borrower rights. This would require, in particular, that the borrower protection clauses of the original contract are fully maintained even in case the contractual terms are modified as a result of the transfer the loan and that the general set of consumer protection rules in force in the country of the debtor is fully maintained and adequately enforced even in the case the debt servicer operates in regime of passport.

It would be problematic if purchasers of NPL were outside the scope of existing data protection rules. Acquisition of NPL by non-EU funds will imply either a transmission of data (if the acquirer processes the data in the EU) or a transfer of data (if he processes them in a third country) from the banks to the acquirer. In both cases the rules of the General Data Protection Regulation (GDPR) apply with particular reference to rules on international transfers in the latter case. Since GDPR provides maximum harmonisation, Member States cannot raise protection standards. Since the GDPR contains a number of "opening clauses" which allow Member States to further specify its rules and since there might be differences between Member States' approaches to its implementation, there is a potential for downgrading of the actual protection data receives if the new data controller is in a different country. Such a downgrade would be avoided if the acquirer of the credit had to respect the personal data protection prevailing in the country of the borrower.

5.What are the available policy options?

5.1 Framing the policy options' features

Entry and business conditions for both potential NPL buyers and loan servicers currently vary across Member States, as described in Section 2.2 and in Annex 6. This situation can be improved through various channels, such as simplifying and reducing cross-country variation of authorisation and licensing procedures and reducing administrative barriers and cross-country variation in information sharing rules, etc. This section defines a set of dimensions with available alternative policy choices. This framing of policy features will subsequently be used to define the policy options.

Authorisation procedures for loan purchasers

The entry conditions for loan investors often take the form of an authorisation procedure by National Competent Authorities (NCAs). One of the main divergences among Member States is whether or not a banking licence is required for loan purchasers. A lighter authorisation procedure would rely only on fit-and proper criteria (e.g., good repute of directors, to consider capital requirements or professional insurance, organisational requirements on IT, risk control, internal audit, compliance office). The authorisation requirement for purchasers can be partially or fully waived in cases where the loan purchase is registered with a NCA and the loan is serviced by an authorised servicer. Different setups are possible as regards cross-border activity of loan purchasers (national authorisation, passporting, single authorisation, etc.).

Business rules for loan purchasers

Rules applicable to the activity of loan investors may stipulate whether some types of transfers are restricted or require additional authorisations. As regards the resolution of NPLs, restrictions may apply for loan purchasers with respect to rescheduling of the original loans (in some Member States a banking licence is required). Member States have different legal instruments to transfer loans and a number of civil law provisions that may impose constraints on what NPL purchasers can undertake, for example if debtors are in insolvency procedures. Making these subject to harmonisation appears not proportional and is therefore not considered part of the option set.

Scope of eligible loans

The set of loans non-banks are allowed to buy differ among Member States. While a few do not have any limitations, some, e.g. BG, FR, LT, PT, RO, SK have different rules for non-performing and performing loans. The purchase of consumer credit loans is not possible in BG while also BE, LT, NL, SI and SK have special protection for household loans or consumer credit. While the main public interest is in enabling the purchase of non-performing loans, it is difficult to control in practice because most loans are sold in large portfolios. Some of these portfolios contain both performing and non-performing loans.

While their interest in maintaining a customer relationship with performing borrowers creates an incentive for banks to limit purchases to non-performing loans, actual sales often include sales that do not fall under the formal definition of non-performing. This includes: Some loans are expected to become non-performing, others may be partly performing, some may be performing, but from debtors that did not service other loans, etc. Part of the wider market for loan sales are also loans supplied by public asset management companies as consequence of the wind up of credit institutions.  68 The actual share of non-performing loans in loan sales is unknown; the estimations shown in Annex 4.2 suggest it could be between 70 and 80%. The counterpart of about a third of NPLs on banks' balance sheet are households and the share of loan sales owed by consumers was at least 11% in 2015-16. 69

Authorisation procedures for loan servicers

Similarly to loan investors, the authorisation procedure for loan servicers may or may not be linked to a banking licence requirement. Alternatively, the authorisation may merely rely on a fit-and-proper check (as in the case of investors: repute of directors, capital requirements or professional insurance, IT requirements, risk control, audit, and compliance). Cross-border activity of loan servicers can also be subject to different regulatory setups (national licensing, passporting, single authorisation, etc.).

Business rules for loan servicers

The scope of activity of loan servicers may be more or less broad. It usually includes such activities as direct contact with the debtor, but may in some cases go as far as out-of-court or judicial recovery. Again, servicers' role in rescheduling re-payment of the loan may require modifications in some Member States, in order to clarify whether a banking license is required in this context. Rules on outsourcing of activities can state whether the outsourcing institution maintains responsibility and accountability. Similarly to above, loan servicers' activity will be constrained by national rules, for example on debtor protection, which will continue to be set at the Member States' level.

Protection of debtor rights, privacy and data protection

An important challenge of this policy initiative is the potential conflict with debtor rights protection and personal data protection laws. 70  Some Member States set specific conditions or limit the scope of business activity for loan purchasers and loan servicers, others govern these issues in other laws, independent from the authorisation process of these entities, or impose the same rules on banks and loan servicers. In particular, the transfer of loans could cause issues with personal data protection if the processing of data involved in the change of creditor is carried out in non-compliance with data protection rules. If the debtor becomes subject to an insolvency procedure, NPL investors have very different rights and obligations in the different Member States. Since the issue of insolvency and debtor protection is covered in a different NPL work stream, this Impact Assessment discusses the effects in terms of gaps that could emerge if licensing and conduct rules for NPL investors and loan servicers are changed. Member States that currently use authorisation procedures as a means to ensure debtor rights protection may need to enact new legislation to maintain the desired level of protection through other means, e.g. through adopting more specific borrower protection rules. 71

Since changes to the authorisation regime of NPL purchasers and loan servicers could interfere with borrower rights, additional safeguards for the borrower could be warranted, in particular as regards consumers. As a matter of principle, borrower protection rules stemming from the contractual relationship as well as from legislation in the borrower's home country should be maintained. In addition, these risks could also be mitigated by introducing a requirement of notification of the debtor about the change of creditor, as well as about the applicability of national and EU rules on debtor rights protection and civil law rules about loan contracts. It could also be envisaged to strengthen supervision of the entities' actual conduct to further ensure safeguarding borrowers' rights. As regards cross-border operations of NPL purchasers and servicers, administrative cooperation mechanisms for NCAs and dedicated contact points for debtor appeals located in the debtor's Member State of the debtor should also be part of the policy initiative.

5.2 Identified regulatory best practices

The different elements described above need to be combined in a consistent way, yielding a regulatory regime that would be on average lighter and more comparable across Member States. Since it is neither proportionate nor feasible in the short-to-medium run to amend parameters such as civil law and insolvency regime, they represent the external context (which will be a given) to the conduct of NPL purchasers and loan servicers in a new regulatory regime. Additional safeguards for borrowers may be required depending on the other modifications brought by the new regulatory regime. The following list identifies regulatory best practices in a number of areas. These could constitute building blocks for a consistent new regulatory regime.

Licensing requirements for investors: In terms of making market access easier for NPL investors, one approach is to not foresee licensing requirements at national level (e.g. currently in CZ, ES, HR, LV, FI). Application of this principle to all EU Member States could potentially undermine the debtor protection specifically targeted by the rules in some Member States. A compromise solution is a lighter authorisation requirement for NPL investors, using a fit-and proper approach (e.g. IE and PT use such an approach for loan servicers, see below). Such authorisation requirement allows checking whether the applicant fulfils certain general conditions when it enters the market (see examples of fit-and-proper criteria in the previous section). Beyond the usefulness of the criteria themselves to ensure a certain level of quality and protection in the market, a light authorisation regime has the merit of establishing a first contact with the supervisor. This enables the supervisor to check, to review and, if necessary, to sanction the conduct of the entity. With respect to preserving the current level of debtor rights, lighter authorisation would therefore be consistent with a stronger role of the supervisor in controlling conduct.

Use of loan servicers by investors: Some Member States have close to no specific obligation on NPL investors, but require the use of an authorised loan servicer (IE, EL). Since most NPL investors delegate debt collection to loan servicing companies, which are the only ones to get in direct contact with the debtor, this requirement does not seem to lead to disproportionate costs. Since the outsourcing mandate to a loan servicer should not discharge the NPL investor from its responsibilities, rules need to clarify obligations and define how the investor monitors the loan servicer. Therefore, the licensing requirements, as discussed in the previous paragraph, could be less stringent where the NPL investor relies on an authorised servicer. If the NPL investor performs the loan servicing itself, the rules for both loan investors and servicers (if different) need to apply to it.

Types of loans eligible for disposal: Several Member States make a distinction between loan purchasers acquiring performing or non-performing loans (BG, FR, LT, PT, RO, SK) or have special rules for loans owed by households (BG and to some extent BE, LT, NL, SI and SK). Other Member States do not make this eligibility distinction. In Member States that have different rules for the sale of performing and non-performing loans this distinction tends to lead to higher transaction costs and lower interest of investors. This may be a cause of relatively low contribution of loan sales to the decline in NPLs during the period 2015-16 in BG and PT and the absence of notable loan sales in FR. 72 The reason why some Member States have different rules is that they consider ownership of a performing loan as similar to credit granting and therefore see a need to regulate them comparable to banks. This means that such investors are required to hold a banking license. The purchase and administration of loans, however, technically does not generate new credit. Moreover, non-bank investors do not refinance themselves through deposits and hence should not be subject to the same supervision or same restrictions on leverage or minimum capital as credit institutions. In economic terms, holding and administration of an existing loan is not similar to bank business and therefore should not necessarily require a banking license.

Apart from the cases described above, Member States do not restrict the sale of loans owed by households and instead provide protection for consumers and house owners through other means than authorisation conditions for non-bank loan purchasers or limitations on whether loans owed by consumers could be sold. The wider the coverage of eligible NPL, the fewer potential distortions between market segments need to be considered. From the perspective of the bank, NPLs from households or corporations weigh equally on their balance sheet and limiting the possibility to sell corporate loans would reduce the NPL’s potential market size by a third. From the perspective of a household, consumer credit and mortgage credit may be of same importance and it would be up to political preferences whether one or the other requires more safeguards in case of transfer to non-banks.

Authorisation requirements for servicers: Some Member States request authorisation of loan servicers along a fit-and-proper approach (IE, PT). They are in direct contact with the debtors and supervisors need to ensure that they comply with relevant rules relative to debtor protection, privacy and data protection. Although a few Member States do not have specific licensing requirements for loan servicing firms, such an approach might not be advisable at EU level due to the possible effects on the debtor rights. Due to risks for financial stability and importance of debtor protection, there is a strong interest to ensure that servicing firms have the organisational and technical capacity to operate in accordance with applicable laws and that they can continue business even against economic or legal headwinds. Hence, fulfilment of organisational requirements and possibly even request for indemnity insurance or loss absorbance capacity seems useful. Similar to the treatment of NPL purchasers, a lighter authorisation regime for servicers could be balanced by stronger supervisory rights.

Loss coverage of servicers: Most Member States do not request minimum capital for loan servicers. However, BG, EL and RO do and in some other Member States minimum capital requirements may emerge as consequence of the need to hold a banking license. In order to reflect the fact that loan servicers' activity is much narrower than that of credit institutions, it would be advisable to not subject them to capital requirements applicable to credit institutions. In order to secure that firms are able to compensate any damages related to their operations, one could consider a requirement of either indemnity insurance, or a capital buffer.

Borrower rights: Very often purchasers of NPLs are from a different Member State than the borrower or even from outside the EU. 73 In order to avoid that the cross-border transfer of a loan leads to uncertainty about which Member States' law applies, the standard approach is that the law that governs the contractual relationship between the borrower and the initial creditor, as well as the consumer protection rules of the borrower's home country continue to apply. This means that the borrower rights remain untouched from the transfer and the new owner cannot derive any additional rights if it is located in a country with a more creditor friendly regime. NPL purchasers and loan servicers would then need to adjust their business model and internal compliance standards to the law of the Member State of the initial loan contract, irrespective of their domicile, authorisation and passport. Currently, market participants rely primarily on consultancy firms and law firms to obtain such information. At least loan servicers set up domestic entities or cooperate with domestic firms to ensure compliance with national provisions.

Code of conduct for servicers: Ireland refers to borrower rights and to a code of conduct in its law that governs the authorisation of loan servicers. Sweden tasked its data protection authority with the authorisation of debt collection firms. Loan servicers need to observe existing legislation in Member States, especially insolvency law and borrower rights. And since they deal with personal information, they need to respect data protection laws. Other conduct rules that govern processes how they interact with debtors may also be warranted as regards their fundamental rights. If they take the form of legal obligations or enforceable codes, supervisors would be entitled to control and possibly sanction in case of misbehaviour. It is worth mentioning that some loan servicers have committed to self-set conduct rules that restrict their interactions with borrowers to certain limits, i.e. not using communication that is perceived as threatening or intrusion into privacy, or not spreading certain information. An industry association of loan servicers announced its incentive to set up conduct rules for the industry in its reply to the public consultation.

Rescheduling of loan repayment by servicers: Practice among most Member States, with notable exceptions in BG and DK, which place limits on the capacity of loan servicers in debt enforcement, is also that loan servicers can agree on rescheduling debtors' repayment of the loan outside insolvency proceedings. Often they can launch or participate in enforcement actions. It seems desirable that loan servicers do not face limitations in their efforts to reschedule the payment stream or establish a repayment plan, provided that is bilaterally agreed with the debtor. Even though this is in many regards not comparable to a new credit generation, some Member States such as AT treat it as such, which leads to a banking license requirement. Since outsourcing to external loan servicers is common practice in this business, other national practices consider that servicers' participation in debt resolution mechanisms (restructuring as well as enforcement) should not be hindered. In absence of requirements for such actions, there would, however, be a need rules ensuring that responsibility is not diluted, especially as regards the responsibility of the loan purchasers. Beyond these NPL-related cases, loan services would be bound by existing rules at EU and national level.

Supervision and cross-border activity: Given the lack of EU competence in supervision on this matter, national competent authorities play an important role as supervisors of loan servicers. If the regulatory regime allows cross-border activity, e.g. facilitated by an EU passport, there needs to be effective coordination between home and host supervisor. Since these activities are supervised at national level only, there are no experiences with the supervisory practice for cross-border transactions.

To facilitate cooperation host supervisors could be required to have a complaint office that would receive complaints from debtors about domestic and foreign loan servicers, with an automatic information exchange with foreign servicers' national competent authority. The home supervisor could be obliged to act once it receives a certain number of justified complaints. It could also be envisaged that the host supervisor may get the possibility to withdraw the passport. National competent authorities could make use of the Internal Market Information system (IMI) for their information exchanges. 74

Right to information: Best practices for additional safeguards for borrower rights are early information to the debtor about the loan transfer, information about possible legal defences and complaint methods. As in several MS the NPL purchasers and/or servicers have the same obligations vis-à-vis the debtor as the originating credit institution had it could also be envisaged to oblige purchasers/buyers to explicitly recognise that they assume the same obligation. It would also be important that in cases the transfer of loans from a credit institution to a non-bank reduces borrower rights, legal gaps are addressed.

Possible best practices beyond the scope of this initiative: A number of provisions will not be touched as part of this initiative, although they have been identified as obstacles to the cross-border expansion of loan servicing and obstacles to NPL purchases. Standardisation of the legal instruments to transfer loans, other civil law provisions, debtor protection rights in national law or data protection law are outside the scope this initiative. They appear too heterogeneous to become standardised even if cross-border firms already active on different national markets would benefit from substantial cost savings if these were standardised, and the possible value added of standardisation seems uncertain. Though this limits the benefits this initiative can generate in terms of fostering cross-border market entry, existing firms active in several markets have been able to cope with these differences.

5.3    What is the baseline from which options are assessed?

In the baseline, current rules would continue. This means that specific entry barriers in some Member States would persist and conduct rules that discourage investor entry and the build-up of investor relationship with loan servicers would remain effective. Although there are no restrictions to invest in NPLs in most EU Member States, specific rules on notification to debtors, registration of collateral, localisation and licensing may effectively hold back investor entry into markets where they have not been active before. Especially foreign investors may remain reluctant to take exposure in smaller and lesser known markets.

The investor basis for the European NPL market would likely remain at its current size without additional incentives to boost the demand side. While the hedge fund industry recorded growth rates of almost 9% on average during 2015-17 and even double digit growth rates in 2013 and 2014, investment in private as well as in distressed debt has remained rather constant at both global and European level since 2015. Statistical data by Preqin, one of the main data collector on the alternative asset management industry, suggests that the European distressed debt market amounts currently to approximately EUR 20 - 25 billion, including part of the investors that target "special situations" (see Annex 5.1). Investment capacity of this magnitude is consistent with average NPL prices of 20 to 25% of face value and a continuation of loan sales of about EUR 100 billion per annum as observed in the last two years. Profitability in the industry will remain at the currently high rates significantly above 10% observed in the specialised investment fund industry (Annex 5.1) and in firms offering loan servicing ( Figure 8 and Table A.3 in Annex 5.2).

If other measures of the NPL action plan effectively expand NPL supply, the baseline scenario means that NPL supply would move along a constant demand curve with banks offering NPLs for sale and competing for a constant investor pool. Without additional demand, banks would not be able to realise higher prices, which increases their incentive to keep NPLs on their balance sheet and evergreen them to the extent possible.

The economic consequences of the above described scenario are manifold. First, the underdeveloped NPL market is expected to mean that NPLs remain on banks' balance sheets, which constitutes an obstacle to mergers and acquisitions among banks, impairing the market-driven restructuring of the EU banking sector. Second, if banks keep high NPLs on their balance sheet, but do not have the capacity to deal with them in a sustainable way, losses from NPLs can contribute to triggering a bank failure. Third, in a situation where NPL levels would increase in some Member States, they would face limited demand and therewith limited scope to sell NPLs to investors. Finally, in case of stress in the banking sector, banks with low NPLs and a similar asset structure as those with high NPLs risk being penalised by investors in bank debt, leading to higher funding costs for sound banks. Historic episodes have shown that in times of market turmoil investors are not discriminating sufficiently between banks' different asset quality.

Recent trends in NPL volume data show that loan sales in 2015 and 2016 contributed to a decline in the stock of NPLs in the EU by about EUR 200 billion. The exact contribution is impossible to be quantified because, inter alia, some loan sales also covered performing loans and there was no information about the breakdown, and because some loans were sold by AMCs so they did not reduce NPLs on the banks' balance sheet in the year they were sold, but earlier. 75 The EU's NPL ratio fell by 1.4 percentage points over these two years to 5.1% at end 2016. 76 Thus, upon continuation of this trend, which however cannot be taken for granted, it would take 3 to 4 years to reach a ratio of 2.2% and 1.8% in the EU. For comparison, the NPL ratio amounted to 1.7% in the US and in Japan in 2016 while the EU NPL ratio in 2008 was 2.2%.

The baseline assumption of a trend continuation means that loan sales will remain absent or small in a number of Member States and that NPL ratios are likely to remain at double-digit levels (see Figure 10 ) in Member States such as CY, EL, PT, BG. The baseline therefore implies risks to financial stability in those Member States where NPL ratios are high and for the EU banking sector as a whole even if the EU average NPL ratio would have fallen to an acceptable level after 3 to 4 years.

Note: Stand-alone banks and foreign controlled subsidiaries and branches. Extrapolation of missing observations by Commission services.

Source: ECB.

5.3    Description of the policy options

This section describes the three policy options that will be assessed in section 6. Table 4 below provides a broad overview of the coverage of these options across the main framing dimensions described in section 5.1. For each dimension and under each option, the table states whether the area would be covered by rules at national level, EU level, or both. The main features of each option are the listed under each area.

Table 4: Overview of the regulatory implications of policy options across five dimensions

5.3.1    Option 1 - Non-binding common high-level principles

A first option would be to establish non-binding high-level principles that would mostly target entry conditions and conduct rules of NPL investors and loan servicers. These principles cover criteria that investors would need to fulfil to serve as eligible counterparts for banks in NPL transactions and that loan servicers would need to fulfil to provide services to NPL investors. These principles would target areas that are most detrimental to market entrance and that differ strongly across Member States. It would be particularly useful to target a reduction in the regulatory burden in those Member States that have a high NPL ratio and few NPL sales: EL, PT, HR, CY, possibly also IT 77 , HU and SI.

In order to stimulate participation in NPL and loan servicing markets, the principles would favour a lighter entry authorisation approach for both investors and servicers. EU principles for loan servicers would include a licensing requirement, the freedom to establish a physical presence in any EU Member States and the respect for local consumer and data protection rules. The principles would not specifically address the Member States that require a banking license and physical presence of NPL investors, although the recommended approach would be to use lighter entry authorisations. It would further be recommended to lift limitations on the type of loans that non-banks can acquire, i.e. propose equal possibilities for banks to sell performing and non-performing loans or special types of loans, such as secured loans or loans owed by consumers. The principles could also suggest to Member States under which conditions loan servicers can reschedule loan repayments without generating a new loan. In order to reduce administrative delays, these rules would usefully also determine maximum requirements for public authorities to deal with them.

By nature, non-binding common principles would not introduce additional obstacles in those Member States in which market entry is already simple. This could be reached through non-legislative measures such as guidelines supplemented by country-specific recommendations in the EU semester and/or targeted support by the Commission's Structural Reform Support Service to those Member States most in need. While this option is non-binding on Member States, it would be recommended that Member States that deviate from the common principles adjust their national law accordingly. In cases where Member States establish an asset management company (AMC) that outsources the management of loans to third-part loan servicers and conducts auctions of NPLs, they could implement the principles by incorporating them into the eligibility criteria for loan servicers to provide services to the AMC and for NPL investors to participate in the auctions.

5.3.2    Option 2 - Binding common minimum standards with passporting

A second option would be binding common minimum standards for entry conditions and business rules for investors and loan servicers, including the possibility of operators established according to these standards in one Member State to provide services in other Member States ("passport"). These rules would be oriented along the best practices listed in section 5.2, but would not be defined as a general principle, but include minimum and/or maximum conditions at EU level, that Member States would need to respect when transposing the EU rule into national law. Member States would need to transpose these standards in national law and need to recognise authorisations of NPL investors and loan servicers from other Member States.

These rules would cover more subject matters as under Option 1, and would be more specific. They would establish criteria that investors would need to fulfil to serve as eligible counterparts for banks in NPL transactions and for loan servicers as counterparts to NPL investors. These rules would be based on fit-and-proper criteria (repute, capital requirements or professional insurance, complemented by organisational, IT, risk and compliance requirements) and would determine authorisation procedures and conditions for NPL investors and loan servicers. They would also govern outsourcing possibilities and limitations thereof. Another covered area would be the scope for loan servicers to re-schedule debt and their relationship to loan owners. The standards would require equal treatment of performing and non-performing loans or special types of loans, such as secured loans or loans owed by consumers. NPL investors would be incentivised to use authorised loan servicers. Both NPL investor and loan servicers would be obliged to comply with the civil law of the host country, when acting cross-border and to allow host supervisors to review their conduct. They could also commit Member States to reduce administrative burden of the licensing process and to refrain from enacting some limitations for NPL investors or loan servicers, such as the domiciliation request or the need for a banking license.

They would also commit applicants to not derive additional obligations from the borrowers than they had vis-a-vis the credit-originating bank. The standards would also set limits on what applicants can or cannot do, with enforceable conduct rules that supervisors can monitor and enforce. Since civil law provisions will not be altered, there are limits to additional safeguards that EU standards can introduce. It is, however, envisaged to set up cooperation among supervisors in home and host countries, so that complaints by borrowers about inadequate conduct can be effectively followed up by competent authorities.

Market entry would be stimulated because potential investors and loan servicers could apply for authorisation in one Member State, and would not have to request additional entry authorisation in other Member States.

5.3.3.    Option 3 - Binding single rulebook with passporting

A third option would be to harmonise entry and conduct rules in the EU for investors and loan servicers in the EU. This would result in uniform entry conditions in all EU Member States, which would spur market participation and allow the realisation of scale economies. The legal instrument would establish common specific fit-and-proper requirements that Member States require applicants to fulfil and would equally commit Member States to refrain from setting further national licensing and business requirements on NPL investors and loan servicers beyond the common EU rules and the applicable national laws as regards borrower rights.

The legislative instrument would introduce specific licensing and registration requirements for those investors that are not already authorised in the EU, for example as alternative investment fund. Financial firms and investment fund managers already authorised in the EU or other countries would not need special authorisation. Hence, only non-financial firms, private individuals and firms in specific jurisdictions 78 would need to apply. Eligibility criteria should be commensurate to the needs, for example covering fit-and-proper criteria and the obligation to respect all the national consumer, debtor and data protection rules. No distinction would be made between performing and non-performing loans or for special types of loans, such as secured loans or loans owed by consumers. In case the acquired NPL portfolio contains also underperforming or performing loans, there would be an obligation to outsource loan management to an authorised loan servicing firm. The rulebook would also include rules that govern the relationship between investors and loan servicers.

The legislative instrument would give a common definition of loan servicing, which could also be positive for market entrance because it would eliminate legal uncertainty. It would establish licensing requirements similar to those existing in some Member States, but with a larger number of loan servicers. These requirements should ensure that loan servicing firms are run by trustworthy managers and have sufficient IT and logistical capacity to offer loan servicing in a sustainable manner. Since NPL portfolios often contain also underperforming or performing loans, some criteria required in Member States for banking licenses should also be considered, for example those linked to bank secrecy. This would also contribute to a level-playing field between banks' in-house management of loans and that in loan servicing firms. Since the former are regulated, it would be inconsistent if the latter were not. The legislative instrument would request some fit-and-proper criteria for loan servicers, possibly supplemented by conduct rules on appropriate behaviour vis-à-vis debtor and data protection.

The binding standards would broadly cover the same issues as option 2. However, under this option the standards on these matters would be fully harmonised and Member States could not introduce more stringent standards to goldplate the common rules. The legislative initiative would also bring additional harmonisation of the supervisory framework. An EU Regulation would be the suitable legal tool to accomplish this.

5.4    Options discarded at an early stage

A less intrusive intervention than changing the regulatory environment would be to set up an information platform that stores rules governing market entry for NPL investors and loan servicers in all Member States. Such a register would be accessible for potential market entrants via a website. Maintaining this website could be done centralised by a European body, such as EBA or ECB/SSM, or decentralised by public authorities in the Member State coordinated via a common entry point or even by AMCs that exist or come into existence in Member States. Member States would need to ensure that the information is factually correct and updated if necessary. The institution maintaining and coordinating the website would have the task of ensuring standardisation of the presentation of national rules.

Such an information platform would be complementary to the data standardisation project in the Council Action Plan. Rather than facilitating potential buyers to assess the values of the loans they envisage to buy in a transaction with banks, it would reduce search and information costs for administrative barriers, thereby helping firms that consider entering NPL and/or loan servicing markets. It would follow the same approach as the single digital gateway (SDG), which aims to improve online availability, quality and findability of information and assistance services on EU rights and national rules concerning the operation and movement in the EU. 79

The option to establish an information platform is not further pursued because the benefits of reducing search costs for licensing conditions seem marginal compared to other search costs NPL investors and loan servicers have if they enter the market. While the costs of setting up such a structure would also be small, the value added is unclear. Member States that see value in providing such information can do so. For example Portugal is currently venturing this. Moreover, it is not clear whether the advantages of centralisation through an information platform are higher than information provision through competitive private firms. The information platform would crowd out the provision of the same service by consultancy firms.

Table 5: Maintained and discarded options - Standardise and simplify entry and conduct rules for potential NPL buyers and loan servicers

√ := Option maintained and discussed below.

6.What are the impacts of the policy options?

As indicated in the introduction, there is a general need to reduce NPL levels in European banks. The positive economic impact of reducing NPLs from banks' balance sheets will be an increased lending capacity of banks and improved financial stability resulting in increased market confidence, both likely to result in increased economic growth.

The different policy options address a narrow set of actors, consisting of the selling bank, the potential buyer, the loan servicer and the ultimate debtor. The main impact would therefore relate to these stakeholders.

The bank's sale of the loan to a non-bank may potentially affect the borrower and his/her rights. A positive example of a loan transfer beneficial to borrowers would be if the loan purchaser or servicer offered distressed debtors a more suitable payback profile of their loans. By way of negative example, it may also be that the loan purchaser or servicer would treat the borrower more strictly than the credit-originating banks (e.g., due to lesser reputational risks). However, even if the law applicable to the loan purchaser (e.g. the banking prudential framework would be replaced by another framework applicable to the buyer) changed with the sale, the same civil and commercial law, including the safeguards in the consumer protection rules, would continue to apply to the credit agreement based on which the loan is granted.

In view of the above, the transfer of loan from a bank to a non-bank investor does not change contractual obligations of the borrower. However, the change of the creditor means that borrowers are facing a new counterpart with whom they did not conclude contract, who may be less regulated than the originating bank and/or located and regulated in another Member State. Protection of borrowers is the most common reason for existing authorisation procedures in the Member States.

Similarly, information-sharing between investors and loan servicers could conflict with data privacy and business secrecy or, depending on the business model of the loan servicer or investor, lead to risks of excess profiling. This initiative therefore needs to ensure that purchasers and loan servicers comply with data protection rules in the country where the credit was originated. 80

Although the coverage of the policy options is limited to authorisation of loan purchasers and loan servicers and lightening a few selected behavioural constraints, such action at EU level may impact on Member States rules. For example, those Member States that require banking licenses (DE, FR, HU, MT, AT, SK) would no longer be able to do so; those that have different licensing regimes for performing and non-performing loans (BG, FR, PT, RO) would be expected to change the rules. A number of Member States (BE, BG, ES, HU, FI) would need to review the constraints they had put on some investment funds to buy NPLs and BG may need to generalise the permission to transfer consumer loans. Member States that have no specific authorisation regime for NPL investors and/or loan servicers (i.e. CZ, EE, ES HR, LT, SI, FI) would need to designate competent authorities in the transposition of the law.

Other rules in Member States would remain unchanged and would constitute a limit to the conduct of cross-border NPL investors and loan servicers. For example, BE, SI and SK could keep the condition that debtors consent is required, BG and DK their restrictions on debt enforcement 81 . Mandatory notification, notarial certification and registration is required in many Member States and untouched from the coverage of the policy options.

Direct environmental impacts are not expected, while indirect effects would occur only under specific circumstances. For example, a more dynamic NPL secondary market and better possibilities for banks to offload them should create more space for bank lending to environmentally-friendly, or more broadly sustainable projects. Moreover, since these projects are not immune to become non-performing, a better NPL market also allows banks to sell them, which might increase their willingness to fund them in the first place.

Figure 12: Assessment of effectiveness and efficiency

6.1    Option 1 - Non-binding common principles for NPL investors and loan servicers

The option could be implemented relatively quickly if pursued outside legislation at EU level. For example, the implementation could build on guidelines agreed by Member States. Whether such guidelines are able to deliver a quick reduction in the NPL ratio would depend on whether those Member States most concerned are willing and able to amend national legislation along the lines set out in the common principles. This represents the major risk and drawback of this option. At European level, Member States could be incentivised to take action through country-specific recommendations under the European semester, through an accessible list of best practices, or via technical support from the Commission's Structural Reform Support Service. Efforts towards raising political awareness and creating political acceptability for related legislative measures would also be helpful.

Address failure in (national) NPL markets

Common principles would aim to address the most material entry barriers, such as those linked to the requirement to apply for a banking license or to establish a local entity. Consequently, market entrants would benefit from cost savings from less demanding requirements and faster administrative processes. Monetary cost savings might be in the order of a hundred thousand euro if a banking license or structure of a securitisation vehicle is no longer required and there may also be further cost savings if no local entity has to be set up or capital requirements are lower. 82 Still, assuming a reasonable adoption of the initiative by Member States, the monetary cost savings are dwarfed by the large returns in NPL transactions. 83 Though actual cost savings are incremental, it could contribute to foster participation in so far underdeveloped markets. Hence, increased market entry is expected to be concentrated in countries with currently high requirements on the applicant, with long administrative delays and without meaningful loan sales.

Since entry costs are mostly fixed costs and have a sunk cost character, the magnitude of the benefits would be higher for market entrants that are smaller in size and have lower capital positions. Consequently, common principles are likely to constitute a particular incentive for small firms to enter previously ignored markets. One possible business strategy for such firms would be to specialise on bidding for smaller banks' portfolios.

Since implementation of the common principles would lead to convergence in standards across Member States, treatment of NPL investors and loan servicers would become more equal. In those EU Member States where existing rules already fulfil common principles, potential market entrants would neither face lower costs nor better incentives, but firms based in these Member States would face lower barriers to expand activity to markets with hitherto higher entry barriers. Hence, EU firms that consider expanding their NPL purchasing or loan servicing activity would be immediate beneficiaries of lower entry costs. Similar benefits would emerge for such firms based in third countries.

While helping to reduce entry costs and to let market conditions to converge somewhat, common principles would not create a single secondary NPL market because not all Member States may follow them fully. They could for example maintain or introduce different rules for sales of performing and non-performing loans or have special rules for different types of loans even if the common principles contained a clause that requested equal treatment of all types of bank loans. Moreover, non-binding common principles would not allow introducing a passport and would therefore not allow firms to operate in other Member States without still meeting national authorisation requirements in each Member State, and these authorisation requirements will still diverge to a certain extent. As a stand-alone measure, the reduction of burden in relation to the common principles is unlikely to generate a large incentive for foreign NPL investors and loan servicers to enter new markets. It may nonetheless contribute to more entry at the margin, especially if combined with other policy measures that lead to a greater supply of NPLs, lower information and transaction costs or higher recovery values of the underlying loans.

Specific for loan servicers: The reduction of entry costs for loan servicers through common principles should incentivise some loan servicing firms to expand their cross-border activity. The greater ease to establish a loan servicing firm, the availability of more loan servicers and lower costs of loan servicing through more competition among loan servicers would collectively further boost the NPL markets.

Impact on borrowers

If under this option, more NPL are sold from banks to third party entities, more borrowers are likely to face a third-party loan servicer. The latter would be operating under supervision of the debtor's national authorities and in accordance with national rules. Member States will therefore be able to maintain the desired level of borrower rights, even in cases where the authorisation regime becomes lighter as a result of the implementation of the principles (also discussed in the following section).

Safeguards for borrower rights

In those Member States where the implementation of common principles would lead to lighter authorisation regimes, it will be up to national competent authorities to ensure adequate supervision, for example by effective follow-up to borrower complaints.

This option would not have a large impact on supervision and the costs thereof. Member States and those in which a larger number of market participants will be active would receive more applications and need to supervise more market participants. They would need to build up additional administrative capacity.

The main advantage of this option would be that it would leave the highest degree of flexibility to Member States on how to best accommodate them to other applicable laws, in particular those affecting borrower rights. When implementing the common principles, Member States could review whether they entail gaps in borrower rights and adjust either these laws or those laws that implement the EU principles accordingly. This could cause some costs for the public sector.

All in all, the immediate economic effects coming from common principles seem limited. For instance, if the reduction of entry barriers helps kick-start market developments in some Member States, other Member States could follow suit. In addition, once tangible benefits from developing NPL markets are realized, Member States may consider reducing entry barriers further. These indirect implications cannot be quantified.

Table 6: Impact of non-binding common principles (assessment relative to baseline)

6.2    Option 2 - Binding common minimum standards with passporting

Address failure in (national) NPL markets

The advantage of common standards and passporting would be that investors and loan servicers could establish entities in other countries or provide services across the EU/ EEA without the need for further authorisation if they are already active in one country. The resulting saving of additional compliance costs, legal certainty and avoidance of administrative delays would have a positive impact on the incentive of incumbent market players to expand activity to other Member States. Since the magnitude of cost and time saving would depend on the Member State targeted, market entry would be over-proportionally stimulated in Member States with currently high entry barriers. 84 This effect would be larger than under option 1.

A further advantage is a more positive impact on market structures than in option 1. Smaller investment firms or investment funds considering to enter the market or to expand activity cross border would benefit more than larger ones, which are already in the market, because the fixed-cost character of entry costs is relatively more important for smaller amounts invested. Hence, the measure could stimulate competition for smaller NPL portfolios. While the immediate objective of the measure would be to increase competition among investors for NPL portfolios, a uniform passporting/mutual recognition rule would also increase competition on the supply side as investors would face more offers from banks.

Foster a single NPL market

Legal differences across Member States could be further reduced, as compared to the previous option, if entry conditions for investors and loan servicers are defined in an EU Directive and the standards do not provide scope for different rules for performing and non-performing rule or specific forms of loans. It is possible that the use of passporting will only be possible if entry conditions converge in Member States. This would also imply they could become stricter in those Member States with currently rather lenient ones. In this case, the common standards with passporting/mutual recognition are expected to still deliver lower entry barriers on EU average than under the baseline. A possible discouragement seems less relevant for loan servicers, because domestic licensing or registration is standard in most Member States.

The availability of a passport could help attract third-country investors because one-stop licensing would allow them to access multiple EU NPL markets from a single subsidiary. In particular, this would incentivise NPL investors to enter markets in which licensing is currently overly cumbersome.

Specific for loan servicers: The possibility to expand activity across borders via passporting seems particularly beneficial for the loan servicing sector, where the scope to realise scale effects is significant. The likely ultimate outcome is lower loan servicing fees charged to NPL investors.

Impact on borrowers

Under this option, NPL sales to third parties would become more common. Borrowers would be likely to face a third-party loan servicer, some of whom would be authorised in another Member State and operating in the debtor's country with a passport. Borrowers would be protected from misconduct of a NPL purchaser or loan servicer through NCAs supervision in the debtor's country, in cooperation with the servicer's home country NCA. Some negative impacts on the borrowers' welfare could result from the fact that the home-host supervisor cooperation could be less effective in dealing with misconduct from loan servicers operating under a passport compared to a situation with national authorisation and supervision. These problems would be most likely in the first years of the cooperation framework, and would be expected to disappear over time.

Safeguards for borrower rights

Supervisors in host countries would need to set up effective procedures to deal with complaints from borrowers and reinforce cooperation with supervisors in home countries, including options to withdraw passports or licenses in case of lack of compliance with rules in host countries.

This option would therefore have an impact on supervision and the costs involved because competent authorities in Member States would need to supervise more cross-border firms. This entails higher complexity of actual supervision, higher responsibility vis-à-vis host countries and coordination needs with authorities in host countries. Competent authorities in Member States with low entry barriers could expect to receive more applications from firms domiciled outside the EU and follow up supervision of these entities, which are likely to be active in other EU Member States too. The regime with passport may therefore lead to higher costs of supervision than in option 1.

The binding common standards could lead to gaps in the legal protection of borrowers if they interact with other laws at national or EU level to the detriment of the borrower. This would be the case, for example, if the authorisation of domestic loan servicers had explicit provisions to conduct rules and the common standards had not or if the national laws impose specific conduct with respect to data protection and the EU rules do not. Implementation of the Directive would give some leeway for the Member State to transpose the rules in a suitable manner, within the limits the Directive allows. For cases beyond this, Member States would need to adjust other laws to maintain the desired status of borrower rights. The likelihood of such adjustment to be necessary is higher than in option 1.

The legislative process for common standards with passport/mutual recognition may be lengthy if Member States consider that maintaining their country specific regime is important. In addition to the time required to engineer agreement at EU level, Member States' implementation would be time-consuming.

Overall, option 2 is more effective than option 1, but at the expense of higher likely hosts to preserve borrower rights and more intrusion in Member States existing legislative framework and sovereignty.

Table 7: Impact of binding common principles with passport (assessment relative to baseline)

6.3    Option 3 - Single rulebook and common market supervision

Address failure in (national) NPL markets

In order to become quickly effective, the Regulation would not strive to harmonise the legal tools available to transfer loans. In view of the different legal traditions in Member States, this appears too challenging to be accomplished within a reasonable time span. Moreover, changing the fundamentals of the civil legal system does not look proportionate to the underlying problem and it does not seem achievable politically within a reasonable time span.

Instead, the single rule book would harmonise entry rules, thereby undo specific differences in entry conditions across Member States and particularly the costly obstacles to entry that exist in a few Member States. The single rule book would therefore particularly improve entry conditions in Member States facing high entry barriers by removing entry barriers that result for example from the need to obtain a banking license, to operate via a local entity or to use specific legal vehicles to hold NPLs such as a special purpose vehicle in securitisation arrangements or a specially created investment fund. The removal of some of these obstacles would significantly reduce entry costs and it could make investors more responsive to the NPL supply by banks. In this respect, the single rule book would be more effective than minimum standards.

Participants already active in one Member State would face no additional administrative costs or administrative delays when expanding activity to other Member States. Market entrants could realise monetary cost savings as high as in the case of a Directive and, compared to the baseline, particularly if they use the passport to expand activity to Member States with currently high entry costs and long administrative delays.

This would lead to a significant increase in the NPL investor base if investors expect supply to be sufficient and profit opportunities to be satisfying. As an isolated measure, the lowering of entry costs is expected to have a limited impact on entry decisions given that entry costs are small compared to other costs in the purchase process such as search and information costs to evaluate the value of the NPL portfolio for sale, costs for legal advice and compliance with different legal instruments in the Member States to transfer loans.

The market conditions for loan servicers will depend not only on the introduction of a passport but also on the demand for their services by NPL investors. The latter effect is particularly important in the short run. Over the longer term, the benefits of a more competitive market for loan servicing would become increasingly important. One example is the securitisation of loans, where lower costs of loan servicing contribute to more securitisation activity. Banks would also benefit from lower costs by outsourcing loan servicing to specialised firms. Finally, loan servicers are more IT intensive and smaller in size than banks so they may contribute to the pace of innovation and to technical progress.

The impact on the market structure is uncertain. Larger and more efficient secondary markets for NPLs could foster structural change in the banking sector of those countries with a high stock of NPLs. 85 As regards the impact on NPL investors and loan servicers, a single market framework could accelerate market adjustment, possibly encouraging entry of smaller NPL investors. The case is less clear cut for loan servicers. On the one hand, the standardisation of market entry may incentivise smaller firms to enter, accelerating the trend of the last years. On the other hand, the loan servicing market has dynamically evolved in the last years with a number of firms merging or being acquired. Hence, higher competitive pressure may lead to on average loan servicers.

Foster a single NPL market

The uniformity of rules would establish a level playing field of NPL investors and loan servicers, which is conducive to an intensification of competitive pressure among them. It would also avoid any market segmentation for different types of loans such as performing or non-performing-loans or loans owed from specific counterparts. This should contribute to lower bid prices for NPL portfolios.

Similar to the previous option, the single rule book might lead to tighter rules in Member States with currently lenient ones. For Member States with low NPL levels and consequently a low NPL supply, this measure would be close to neutral. These Member States would become beneficiaries in case NPL problems were to emerge in the future.

Nonetheless, compared to the status quo, harmonised rules would set an incentive for NPL investors to bid for NPLs from banks domiciled in different Member States because cross-border activity will be facilitated. Hence, it will also bring advantages for incumbent market participants. Moreover, the single rulebook helps attract market entry from smaller NPL investors and third-country investors. The former would benefit because the reduction in compliance costs has an over-proportionate effect for them, the latter because they face a larger market. The single rule book may also have a strong signalling effect on foreign investors. Similar effects may also emerge under option 2, but at lower intensity.

Harmonisation of business conditions is expected to have a considerable impact on loan servicers as they would be able to economise on licensing costs when entering several Member States. To the extent that legal uncertainty from different definitions of loan servicing across Member States deterred the expansion to new markets, a common definition would eliminate this obstacle. In addition, the resulting notion of a single market may induce third-country loan servicers to enter EU markets because licensing will be unified and therefore easier. This measure would reduce the costs for an established loan servicer to expand to other EU markets and it would also incentivise third-country loan servicers to enter the EU market.

Since loan servicers' business model is characterised by scale effects, a single market regime is likely to yield lower average costs than option 2. The cursory information about profitability in loan servicing firms active in EU Member States suggests that especially firms are more profitable the more assets they manage (see Figure 8 and Table A2 in Annex 5.2).

Cross-border expansion of EU firms and market entry of third-country firms would also lead to more competition among loan servicers. Both effects are expected to reduce the costs incurred by NPL investors when they delegate debt collection to a third-party loan servicer. The passport would also increase the pool of loan servicers an NPL investor can choose from when bidding for an NPL portfolio. The benefits from new market entries should be particularly visible in Member States having low numbers of incumbent loan servicers, especially if these are owned by competing NPL investors.

Impact on borrowers

Under this option, NPL sales to third parties would become more common. Some borrowers would be likely to face a third-party loan servicer, some of whom would be authorised in another Member State and supervised by a NCA in a different Member State according to a common rule book. Borrowers would be protected from misconduct of a NPL purchaser or loan servicer through NCAs supervision in the debtor's country, in cooperation with the servicer's home country NCA. There could be negative impacts on the borrowers' welfare due to ineffective home-host supervisor cooperation. As a result of more harmonised supervision rules across EU Member States, it should be less severe and disappear faster over time than under Option 2.

Safeguards for borrower rights

As compared to the previous two policy options, there could be a need for further convergence of business practices from a single rule book to be followed by national authorities in the authorisation and supervisory process or indeed a single institution in charge of the authorisation process and supervision.

When a network of competent national authorities is in charge of authorisation and supervision, it will be a challenge to establish harmonised supervisory practices. A single supervisory body may have lower coordination needs, but would rely on expertise on the ground and need to consider the impact of legal rules for the transfer of loans and debt protection provisions that remain national. It is not evident whether supervisory costs would be higher or lower than in the case of a Directive with passporting rights.

To ensure political acceptability of an easier market access for foreign loan servicers, it would be warranted to include in the Regulation obligations for loan servicers to observe local consumer and data protection rules. Alternatively, the Regulation could specify enforceable conduct rules for loan servicers.

In Member States that used to ensure debtor protection through the authorisation and supervision of NPL investors or loan servicers, there could be a need to implement new laws that uphold the desired level of debtor protection through other means. The required adjustment of national law and its implementation will entail one-off costs at national level. These are likely to be higher than in case of a Directive because Member States would have no possibility to consider them in the transposition of the Directive, but would need to channel all adjustment needs into amendments to other laws.

Overall, option 2 is more effective than option 2, but at the expense of higher likely hosts to preserve borrower rights and more intrusion in Member States' existing legislative framework and sovereignty.

Table 8: Impact Single rulebook and common market supervision (assessment relative to baseline)

7.How do the options compare?

Figure 12 listed criteria against which the effectiveness and efficiency of the different policy options are assessed under each option's assessment in the previous section. The Tab le 9 below summarises these individual assessments and makes it possible to do an overall comparison of the options.

Overall effectiveness was assessed by aggregating the benefits of the options in addressing failures in NPL markets and in fostering a single market. Overall efficiency was evaluated by comparing these benefits to the aggregate costs in terms of maintaining the same level of borrower rights. See Annex 4.4 for the detailed calculation. Coherence of the options was assessed with regard to broader Commission priorities 86 , but also in the specific context of the related measures of the Council NPL Action Plan (see the section 1 for a discussion of the interdependencies between the various initiatives of the Action Plan). Proportionality was assessed by looking at what measures are necessary in order to achieve the stated objectives, also taking into account the magnitude of the underlying problem.

Table 9: Summary of options and their effects

All the policy options are expected to improve the situation over the baseline, even if their degree of effectiveness can only be estimated. In particular, the lower effectiveness of option 1 against the stronger solutions (options 2 and 3) stems from the Member States coverage. The former would target material changes in a few Member States only while the latter would likely cause changes in most, if not all, Member States. On the flipside, targeting only the most material issues in a few Member States under option 1 would lead to lower need to adjust other legislation to maintain their preferred debtor protection rights, whereas option 3 would require Member States to adapt significantly national legislation. Box 1 below summarises the modelling work undertaken to quantify the magnitude of impacts of the different options in terms of increasing NPL sales. The details and quantification of other effects is provided in Annex 4 and 5, with evidence from the stakeholder consultation presented in Annex 2.

Though options 2 and 3 are more efficient in fostering a single market, i.e. increasing investor base and firms active as loan servicers, than the non-binding option 1, it could take more time before they are effective. Common principles could be implemented particularly quickly in Member States if they accompanied the set-up of an asset management company (AMC) that collected NPLs from banks, outsourced their interim management to third-party loan servicers and conducted auctions to ultimately sell them to non-banks. In this case, principles could be introduced in form of eligibility rules for loan servicers and participants in NPL auctions rather than through legislation.

A relevant trade-off emerges in the choice between options 2 and 3 due to their legislative instrument, in terms of speed of enactment and effectiveness. A Regulation can be considered superior in securing harmonised market conditions and uniform conduct of participants across Member States. Whereas a Directive would allow Member States to reflect country-specific conditions and preferences in their implementation and a Regulation does not, the Directives' additional degree of freedom may also lead to gold plating and reduce effectiveness. 

Trade-offs can also emerge between the opening up of competition to foreign entities and the effectiveness of national safeguards for borrower rights. Borrowers' rights do not change with the transfer of loans and, subject to applicable civil law, the borrower can use against the new owner the same defence available against the original creditor. However, the borrower is exposed to higher uncertainty because he/she is facing a new counterpart and starts from a weak bargaining position since he/she has not delivered on contractual obligations beforehand. Since the counterpart is located in a different Member State, the borrower may not know it and not understand how it is regulated. At the same time, opening access to foreign competitors appears essential to stimulate competition on NPL markets.

Options 2 and 3 score better on coherence than option 1, since there is a risk that the common principles will neither achieve a more resilient financial sector, nor will they lead to more homogeneity across the CMU. The scope of options 2 and 3 to accomplish these overarching objectives is comparable.

Option 1 may not be able to solve the issue of high NPL in some Member States and is therefore considered the least proportional, despite yielding some improvement over the baseline. Option 3 is more intrusive on Member States' sovereignty and existing legislative framework, and therefore less proportional than option 2. While Table 11 suggests a clear ranking of the different options in terms of proportionality, this is based on the assumption that reducing NPL is possible and risks for borrowers can be contained by adequate accompanying policies and Member States are prepared to take the costs of these accompanying policies. The ultimate choice will need to depend on the weight given by political preferences to the trade-off between engineering the most efficient or effective option on the one hand and the risk to borrowers and the costs to contain these risks on the other hand.

Overall, the comparison of the effects of the different policy options is inconclusive as regards whether option 2 or 3 is superior. The former has a somewhat lower effectiveness but with a better proportionality, while the reverse is true for the latter. The ultimate choice will depend also on policy preferences. The key question is whether the priority is to opt for a measure that is most effective, or for a measure that maintains more room for national discretion or minimises the scope of policy intervention to the most urgent challenges.

Table: Simulation results: NPL sales in billion EUR on the EU and in selected Member States two years after the measure is in place

8.Preferred options

The comparison of the effects of the different policy options shows that option 2 and 3 have different strengths in addressing different issues, and do not lead to clear conclusions as to a preferred option in terms of the selection criteria used in the assessment. It will therefore require political considerations to prioritise the choices, based on the impacts and trade-offs presented in the preceding sections.

A single rulebook Regulation that fully harmonises entry and conduct rules for investors and loan servicers would deliver an outcome closest to a single market. Such a single rule book is the most effective measure to increase the investor base for NPLs and reduce the currently high stock of NPLs in the EU and some of its Member States in particular, while ensuring a level playing field among firms from different Member States. It could imply however that certain national specificities cannot be taken into account, and that market entry could become more costly in those Member States in which market entry is already simple, and should be formulated such that it minimises the additional obstacles. Still, some Member States would be required to review whether enactment of a regulation that alters authorisation and conduct rules for NPL purchasers and loan servicers would lead to gaps in terms of borrower rights and data protection and would need to adjust other national laws accordingly.

Since a minimum standards Directive would allow Member States to maintain lighter regimes provided that they respect certain minimum conditions it might stimulate market entry more than a Regulation, depending on implementation in national laws. It could therefore be a good option in increasing the investor base for NPLs and competition on loan servicing markets while allowing Member States to maintain certain national specificities. At the same time it would be less effective in ensuring a level playing field and would give more discretion to Member States to goldplate the requirements, which would prevent these effects to materialise. Some Member States would need to adjust other national law if implementation of a Directive lead to gaps in terms of borrower rights and data protection.

The ultimate choice of the preferred option will depend on policy preferences and whether the priority is to opt for a measure that is most effective, or for a measure that maintains more room for national discretion or minimises the scope of policy intervention to the most urgent challenges.

Independent of the option chosen, the coverage of the initiative would be limited to introduce harmonised conditions for market entry and conduct of non-bank NPL purchasers and loan servicers.

·Loan servicers would need to fulfil fit and proper criteria with respect to their management, prove IT capacity and compliance with debtor and data protection obligations.

·A definition of loan servicing would clarify that loan servicers are not originating credit so that they do not require a banking license.

·Their relationship with the NPL investor/purchaser would need to be clarified and Member States should supervise them given the loan servicers' interaction with the ultimate debtor.

·Home and host supervisors would need to cooperate.

·Rules for NPL purchasers should be simple, possibly not going beyond registration and fulfilment of fit-and-proper criteria. Currently, non-bank investors do not face entry barriers in several Member States, while in others banking licences are required. If entry conditions for NPL purchasers would be left outside the scope of EU measures, the obstacles in other Member States would continue to exist.

·There would be no limitation on the type of loan non-banks are allowed to acquire: performing and non-performing and independent from type of borrower

·A possibility to cover NPL purchasers without putting additional administrative burden on them would be to offer an exemption from authorisation if they delegate the servicing of NPLs to an EU authorised loan servicing firm. If they decide to service loans themselves, they could be treated as loan servicers, possibly restricted to loans with consumers as borrowers.

Two important aspects would be outside the scope of the initiative:

·It would not strive to harmonise debtor protection rules across Member States. A reassurance of the borrowers' position may be is needed, in light of the many replies to the public consultation that flag debtor protection as a concern of more active secondary markets for NPLs. In any case, for all credits, additional safeguards should avoid that the harmonisation of authorisation conditions undermines borrower rights, including the need to impose the respect of the national rules, reinforced information to the borrowers about their rights and legal defences and the possibility to file complaints to national authorities.

·In view of the different legal traditions in Member States, standardisation of the legal tools available to transfer loans appears too challenging. Changing the fundamentals of the civil legal system does not seem proportionate to the underlying problem and it does not seem achievable politically within a reasonable

9.How will actual impacts be monitored and evaluated?

The proposal is expected to follow normal implementation procedures. Ex-post evaluation of all new legislative measures is a top priority for the Commission. The Commission shall establish a programme for monitoring the outputs, results and impacts of this initiative one year after the legal instrument becomes effective. The monitoring programme shall set out the means by which the data and other necessary evidence will be collected.

In terms of indicators and sources that could be used during the evaluation the following monitoring indicators:

·NPL volumes and ratios: The relevant data is available from the ECB and from EBA for all Member States so it is possible to conduct analysis at country level and check, inter alia, whether Member States having hitherto high NPL ratios benefitted over-proportionally;  

·Loan sales in all Member States: this data is not collected officially so data collection and reporting would rely on Commission services, information from supervisors and consultancy firms;

·Composition of the NPLs, in particular amounts of secured and non-secured consumer credits and home loans;

·New purchasers of NPLs, number of smaller banks and banks located in Member States with hitherto low loan sales: This data would also rely on Commission services’ data gathering, information from supervisors and consultancy firms;

·Loan servicers authorised in all Member States and their cross-border activity: This will be sourced from national competent authorities; 

·Debtors’ complaints about misbehaviour of loan servicers signalled to national competent supervisors and supervisors' follow up. Special attention will be paid to complaints about misbehaviour of cross-border loan servicers and their follow-up by home and host supervisors;

·Supervisors' sanctions of non-compliance to NPL purchasers and loan servicers with respect to borrower rights and data protection.

An evaluation is envisaged 5 years after the implementation of the measure and according to the Commission's better regulation Guidelines. The objective of the evaluation will be to assess, among other things, how effective and efficient it has been in terms of achieving the objectives presented in this impact assessment and to decide whether new measures or amendments are needed. Member States shall provide the Commission with the information necessary for the preparation of that Report.

List of references

AFME (2017), Consultation response -Development of Secondary Markets for Non-Performing Loans and Distressed Assets and Protection of Secured Creditors from Borrowers’ Default, October 2017.

Balgova, Maria, Michel Nies and Alexander Plekhanov (2016), The economic impact of reducing non-performing loans, http://npl.vienna-initiative.com/wp-content/uploads/sites/2/2017/02/EBRD-The-Economic-Impact-of-Reducing-NPLs-WP_193.pdf, EBRD Working Paper No. 193, October 2016.

BancaIFIS (2017), Market watch NPL, The Italian Scenario, July 2017. https://bebeez.it/en/2017/07/19/secured-bad-loans-portfolios-prices-drop-to-33-in-italy-pushed-by-massive-offer/ .

Barba Navaretti, Giorgio et al (2017), Non-performing loans. Banks, Regulation, and the real sector. European Economy 1/2017  http://european-economy.eu/ .

Baudino Patrizia and Hyuncheol Yun (2017), Resolution of non-performing loans – policy options, BIS, Financial Stability Institute, FSI Insights on policy implementation No 3, October 2017

Bruegel (2017), Analysis of developments in EU capital flows in the global context, Draft Final report, prepared by Grégory Claeys, Maria Demertzis, Konstantinos Efstathiou, Inês Gonçalves Raposo, Pia Hüttl, Alexander Lehmann, FISMA/2016/032/B1/ST/OP September 2017.

Ciavoliello, L.G. et al, (2016), "What's the value of NPLs? Bank of Italy Notes on financial stability and supervision No. 3. https://www.bancaditalia.it/pubblicazioni/note-stabilita/2016-0003/index.html?com.dotmarketing.htmlpage.language=1 .

Deloitte (2017a), CEEC NPL markets in full swing, A record year in 2016, https://www2.deloitte.com/content/dam/Deloitte/global/Documents/About-Deloitte/central-europe/ce-non-performing-bank-loans-study-npl.pdf , January 2017.

Deloitte (2017b), Regulation driving changes in European loan portfolio markets, Delevreaging Europe H1 2017.  https://www2.deloitte.com/uk/en/pages/financial-advisory/articles/deleveraging-europe-market-update.html  .

Djankov, Simeon et al., (2008),"Debt Enforcement Around the World" Journal of Political Economy Vol 116, No. 6.

Drew Dahl , Andrew P. Meyer , Michelle Clark Neely ( 2016), Scale Matters: Community Banks and Compliance Costs,  https://www.stlouisfed.org/publications/regional-economist/july-2016/scale-matters-community-banks-and-compliance-costs .

Ecofin Council Conclusions on Action plan to tackle non-performing loans in Europe, http://www.consilium.europa.eu/en/press/press-releases/2017/07/11-conclusions-non-performing-loans/

Europe Economics (2009), Study on the Cost of Compliance with Selected FSAP Measures, Final Study for European Commission DG MARKT.

European Banking Authority (2016), EBA Report on the dynamics and drivers of nonperforming exposures in the EU banking sector, https://www.eba.europa.eu/documents/10180/1360107/EBA+Report+on+NPLs.pdf , July 2016.

European Banking Authority (2017), EBA Risk Assessment Questionnaire as of September 2017. EBA Staff, Risks and Vulnerabilities.

European Banking Federation (2017), EBF comments on the EC consultation on the development of secondary markets for NPLs and protection of secured creditors, October 2017.

European Central Bank (2016), Financial Stability Report December 2016.

European Central Bank (2017a), Financial Stability Report May 2017.

European Central Bank (2017b), Financial Integration Report April 2017.

European Central Bank and Single Supervisory Mechanism (2017), Stocktake of national supervisory practices and legal frameworks related to NPLs, June 2017.

European Commission (2017) Reflection Paper on the Deepening of Economic and Monetary Union, 31 May 2017, https://ec.europa.eu/commission/publications/reflection-paper-deepening-economic-and-monetary-union_en .

European Commission (2017a), A macroeconomic perspective on non-performing loans (NPLs), Quarterly report on the Euro Area, Volume 16 No 1, pp. 7-21.

European Commission (2017b), Impact Assessment Cross border investment funds, November 2017.

European Commission (2017c), Post-Programme Surveillance Report, Portugal, Summer 2017, DG ECFIN Institutional Paper No 61, October.

European Systemic Risk Board (2017), Resolving non performing loans in Europe, https://www.esrb.europa.eu/pub/pdf/reports/20170711_resolving_npl_report.en.pdf , June 2017.

European Systemic Risk Board, Advisory Scientific Committee (2014), Is Europe Overbanked? https://www.esrb.europa.eu/pub/pdf/asc/Reports_ASC_4_1406.pdf June 2014

Experian (2017), Debt Collection Agencies: The Keys to Improved Recoveries?, Decision Analytics Briefing Paper. http://www.experian.com/corporate/experian-decision-analytics.html .

EY (2017), Navigating the Italian credit opportunity, Non-performing loans and new credit tools,  http://www.ey.com/Publication/vwLUAssets/EY-Navigating-the-Italian-credit-opportunity/$FILE/EY-Navigating-the-Italian-credit-opportunity.pdf .

Federal Reserve Board et al., (2016), Report to the Congress on the Effect of Capital Rules on Mortgage Servicing Assets, prepared by Board of Governors of the Federal Reserve System, Federal Deposit Insurance Corporation, Office of the Comptroller of the Currency and National Credit Union Administration, Washinton DC, June 2016.

Fell, John (2017), Microeconomic perspectives for NPL. NPL meeting Venice 15 September 2017 http://www.nplmeeting.it/wp-content/uploads/2016/05/2017-08-24_NPL-Meeting_Venice-15-September_John-Fell_final16_9-format.pdf .

FSC (2017), Report of the FSC Subgroup on Non-Performing Loans, Council of the European Union, Brussels, 31 May 2017, 9854/17.

International Monetary Fund (2014) Supervisory Roles in Loan Loss Provisioning in Countries Implementing IFRS, September 2014.

International Monetary Fund (2015a) A Strategy for Resolving Europe’s Problem Loans and Technical background paper , Staff Discussion Note No 1519, September 2015.

International Monetary Fund (2015b), Euro Area Policies, Selected Issues, IMF Country Report No. 15/205.

Jassaud, Nadège and Kenneth Kang (2015), A Strategy for Developing a Market for Nonperforming Loans in Italy, IMF Working Paper 15/24, February 2015.

KPMG (2016), European Debt Sales, Portfolio Solutions Group,  https://assets.kpmg.com/content/dam/kpmg/xx/pdf/2016/09/european-debt-sales.pdf .

KPMG European Transactions Dashboard,  https://app.powerbi.com/view?r=eyJrIjoiZjkyOTBhOGItZDRiMS00ODcwLTk5YTMtYWQyNTY1ODNlMmI1IiwidCI6ImRlZmYyNGJiLTIwODktNDQwMC04YzhlLWY3MWU2ODAzNzhiMiIsImMiOjh9&ed2f26df2d9c416fbddddd2330a778c6=lgkzkosh-lfhhdwzd%20 .

Levitin, A.J. and T. Twoney (2011), Mortgage Servicing, Yale Journal on Regulation, Vol. 28,  https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1324023 .

McNulty, J.E., L. García-Feijoo, and A. Viale (2017), The regulation of Mortgage Servicing: Lessons from the Financial Crisis, Contemporary Economic Policy, Forthcoming.

Mohaddes, Kamiar, Mehdi Raissi and Anke Weber, (2017), Can Italy grow out of its NPL overhang? A panel threshold analysis, Economics Letters, Vol. 159, pp. 185–189.

Oliver Wyman, 2016, A model for efficient mortgage servicing.  http://www.oliverwyman.com/content/dam/oliver-wyman/v2/publications/2016/dec/Oliver%20Wyman%20-%20A%20Model%20for%20Efficient%20Mortgage%20Servicing.pdf .

PWC (2017), Portfolio Advisory Group Market update Q4 2016, https://www.pwc.co.uk/deals/assets/portfolio-advisory-group-market-update-2017.pdf , March 2017.

S&P Global intelligence (2017), EMEA Servicer evaluation industry report 2016, Global Credit Portal.  http://www.fbs.it/file/news/allegati/news_125.pdf .

Thao Le (2016 ), Information Loss in Mortgage Securitization: Effect on Loan Modification, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2798849  June 2016.

Vienna Initiative (2012), Working Group on NPLs in Central, Eastern and Southeastern Europe, https://www.imf.org/external/region/eur/pdf/2012/030112.pdf

Wilson, Nick et al. (2012), Estimating the cost of new public health legislation, Bull World Health Organ. 2012 Jul 1; Vol. 90(7), pp. 532–539, https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3397705/

Annex 1: Procedural information

1.Lead DG, Decide Planning/CWP references

Directorate-General for Financial Stability, Financial Services and Capital Markets Union.

The initiative is included in the Commission Work Programme 2018 as agenda planning item PLAN/2017/1121.

2.Organisation and timing

Work on the Impact Assessment started in July 2017 with the first meeting of the Steering group held on 2 October 2017, followed by three further meetings on 8 November and 4 December 2017 and 26 January 2018.

The Inter Service Steering Group was formed by representatives of the Directorates General Competition (COMP), European Political Strategy Centre (EPSC), Economic and Financial Affairs (ECFIN), Internal market, Industry, Entrepreneurship and SMEs (GROW), Justice (JUST), Communications Networks Content and Technology (CONNECT), Taxation and Customs Union (TAXUD), the Legal Service (LS) and the Secretariat General (SG). 

3.Consultation of the RSB

The draft report was sent to the Regulatory Scrutiny Board on 6 December 2017. The Regulatory Scrutiny Board delivered a negative opinion on 12 January 2018. A revised drat was sent to the Regulatory Scrutiny Board on 29 January 2018.

Changes introduced following the first opinion of the Regulatory Scrutiny Board

·A new introduction common to all three legislative initiatives on NPL was introduced. It explains the NPL issue in a wider context and elaborates on the linkages between the various initiatives in the NPL Action Plan in greater detail.

·Differences in borrower rights were introduced as problem drivers as well as discussions how they interact with changes in the authorisation regime for NPL investors and loan servicers. This was also taken up in the discussion on the general impact of the initiative.

·Specific objectives have been aligned with the assessment criteria.

·The concrete provisions were further specified, including a description of the range these provisions could take and a discussion of best practices and how they could be combined to a consistent regulatory regime.

·Assessment criteria for the policy options were made less abstract by connecting them to the desired impact on the main stakeholders.

·Possible adjustment needs in Member States are shown in the impact section.

·The evaluation of the impact of the different policy options was restructured and expanded.

·The presentation of the comparison of the impacts was simplified. The translation of detailed assessment criteria into rankings is described in a new annex 4.4.

·Coherence and proportionality were added as assessment criteria.

·The set of preferred options was narrowed and the link to concrete provisions this entails described in more detail. 

·The evaluation framework was made consistent with that of the other two NPL legislative initiatives and indicators for monitoring progress were added.

·A table with information provided by Member States about their authorisation regime for NPL purchasers and loan servicers was added as appendix to Annex 6.

Changes introduced following the second opinion of the Regulatory Scrutiny Board

·Clarification of the coverage of performing and non-performing loans in the three options.

·Addition of new elements for reviewing the success of this initiative. The additional indicators cover effectiveness of supervision and compliance with borrower rights and data protection.

4.Evidence, sources and quality

This impact assessment is based primarily on stakeholder consultations, the study of the FSC subgroup on NPLs and background documents prepared for the FSC subgroup, studies by EU and international organisations 90 and additional desk research of the Commission services. More specifically, sources include:

·replies by stakeholder to the following consultations:

oA public consultation on the inception impact assessment, 26 June 2017- 22 July 2017. 91

oPublic consultation on the development of secondary markets for non-performing loans and distressed assets and protection of secured creditors from borrowers’ default, 10 July to 20 October 2017 (closed on 27 October) 92

oA questionnaire to EU Member States 7 April to 1 June (last submission received 4 October 2017)

·Feedback from stakeholders and researchers through phone interviews and e-mail exchanges with stakeholders.

·Feedback from stakeholders through bilateral meetings between the Commission services and stakeholders.

·Simulations with the pricing model (see annex 4.3)

·Cross-country analysis (see annex 4.2)

·Analysis of annual accounts of individual firms active in the loan servicing market [and non-public data about compliance costs and their determinants from firms] (see annexes 5.2 and 3.2)

·Analysis of the performance of investment funds investing in distressed debt

·statistics and data from various sources, including ECB, EBA, World Bank, ORBIS, Preqin.

·Market reports and dedicated studies by consultancy firms (Price Waterhouse Coopers, Deloitte, KPMG, Earnest and Young etc.);

·Analysis carried out for other projects in the European Commission, 93

·academic (economic) literature (see List of References);

For a detailed description of the methodological approach, analytical methods, and limitations of the evidence underpinning this impact assessment, see annex 4.

Annex 2: Stakeholder consultation

Context

A public consultation was launched on 10 July 2017 with end-date 20 October 2017. It combined questions on the subject of NPL secondary markets with questions on the Accelerated Loan Security, which was later re-labelled into Accelerated Collateral Enforcement..

Since the public consultation asked stakeholders to identify obstacles to the development of secondary markets for NPLs and to give their view on their importance, several responses gave details on rules in place in the various Member States. Annex 6 focuses on this information, complementing the information received from a similar questionnaire sent to Member States.

1. Coverage and representativeness of the consultation replies

62 responses were submitted to the public consultation. However, some responses focused on the second part related to collateral enhancement and did not provide any input to the part on NPL secondary markets. Among the particularities were that several subsidiaries of one group sent submission, which were similar and consistent, but not identical. These were counted as one reply. Some associations sent almost identical replies. Since these represented standpoints of different institutions, each submission was taken individually. Overall, 51 submissions contained views about the development of the secondary market for NPLs. 10 of them declared that their submission should not be made public.

The Commission received replies to the consultation from respondents in 16 countries. Most submissions came from Germany (10) and Italy (8 of which 3 from citizens). There were also numerous contributions from the UK (7) and Belgium (6), accountable to the domiciliation of consultancy and law firms in the former and the seats of European organisations in the latter. 4 contributions came from each Austria and Poland. The only submission from a country outside the EU was from China.

Figure A2.1: Number of respondents by Member State

Since the issue at hand has a single market dimension and since third-county investors have an important role in NPL markets, it is consistent that many submissions came from international actors. If international associations, cross-border firms and consultancy firms are counted as representing supranational interests, 29 submissions (57%) fall into this category and 22 would represent interests in a particular Member State. Although no submission came from US investors, those of consultancy and law firms may come close to representing views from third-country investors.

The type of respondents is mixed. 9 replies came from firms active on the demand side of the NPL market or associations thereof. Also 10 submissions came from banks or their associations, i.e. representing the supply side view of the NPL market. 8 submissions were received from other financial associations and 7 from law or consultancy firms. 6 public authorities replied and 5 citizens. The remaining submissions are attributable to social partners, consumer organisations and one SME.

Figure A2.2: Replies to the consultation by type of stakeholder

Submissions differed in character and granularity. Some replied directly to the questions, some added reasoning. The numbers and indications of frequency below relate to those replies where either a direct response was given or Commission staff was able to derive it directly from the text. In those cases, where this was not possible, the reply was not counted for the determination of relative weight of responses. While some responses could not be used to determine support or not for a specific question, the reasoning and background provided entered nevertheless into the qualitative assessment done in other parts of this impact assessment.

Given the small number of responses and their non-representativeness, all numbers can be taken as a tendency only. If it was possible to trace back differences in responses to characteristics of the stakeholders, this is indicated below.

2. The role of NPLs and NPL markets

The dominant majority of the replies affirms that the current size, liquidity and structure of secondary markets for NPL in the EU are an obstacle to the management and resolution of NPLs in the EU. Some even describe this obstacle as significant. It is, however, notable that 13 submissions (25%) disagree and among them are some firms active in the market or their associations. Some argue that markets work efficiently at national level, others that the market will develop.

According to respondents, internal and external factors are relevant for banks to decide whether loan sales should be a significant part of their strategy to manage NPLs. None of the responses said that external factors alone are relevant. Among the internal factors, the impact of NPL sales on banks' capital and provisioning, including tax rules on provisioning, and the role of supervisors are often mentioned, the administrative costs of internal work out is also frequently listed. A few replies also refer to reputational risks. Examples for external factors were given less frequently and often of general nature, suggesting that the existence of an efficient secondary market and fair prices would be beneficial for banks.

Only few submissions make a point on whether the lack of investors is an obstacle to market development and among those that do, a slight majority rejects the notion. More specifically, most submissions from actors on the demand side of the market do not address the specific issue, while most of those from banks and bank association, representing the supply side affirm that the lack of investors is an obstacle. Most submissions consider specialisation advantages and management capacity as the economic benefits of non-bank investors, followed in frequency by non-bank's general contribution to help offloading their high NPL level. Occasionally, it was also said that the involvement of non-bank investors could improve the recovery value of NPLs or that benefits are due to non-bank investors longer time horizon.

3. Measures targeting NPL investors

As regards obstacles for investors to enter the market, data and information issues are by far the most frequent reply. Second ranked are non-financial factors, often specified as taxation or IT issues, closely followed in terms of frequency by legal conditions/insolvency law and banks' behaviour. Some replies also indicate other financial factors as obstacles and give as example the link of NPL to securitisation markets.

Several stakeholders also list risks and concerns from the involvement of non-bank investors. These concerns are diverse and cluster around the issues of reputational risks and consumer protection, a possible information disadvantage over banks including as regards local knowledge, a shift of losses from the regulated banking sector to unregulated entities, the impact of investors' short time horizons and high return requirements in the non-bank financial sphere.

The frequency with which main benefits and risks of NPL markets were indicated can be taken as approximation of their importance. The benefits most frequently listed relate to scale and liquidity on a deep and large market, and to specialisation gains. As regards risks, consumer/debtor protection and data protection and privacy are very frequently indicated; less often did respondents see risks from moral hazard. There are also several references to the equal treatment of investors and the efficiency of the legal framework, which do not fall into a risk/benefit categorisation.

For a clear majority of those respondents that give a view, differences in national rules pertaining to NPL sales are an obstacle to the development of NPL markets. This view is shared among firms active on the demand side of the market and other stakeholders. The opposite view is held by 37% of those respondents that reply to this question. Those that give reasoning argue that either there are no significant obstacles or that the different legislative frameworks or economic developments justify the differences.

As regards the nature of obstacles for cross-border activity, the dominant number of responses refers to the legal framework, insolvency rules and local habits. Much fewer respondents regard data issues or incentive problems as underlying drivers. When asked whether differences in these benefits and risks across Member States justify national differences in the framework for the secondary markets for loans, the majority agrees. Among those that consider national rules an obstacle, 40% finds them justified while 60% do not.

Some stakeholders hold necessary additional rules to safeguard consumer/debtor protection while others that think current rules are existent and should be maintained. The number of both views is broadly equal. A non-negligible number advocates specific rules for banks, non-bank investors and debt collection firms. Statements on the need to improve or maintain data protection level are not frequent, but generally affirmative.

While a majority supports an EU framework for NPLs, the minority share is sizeable and it is not possible to attribute a specific characteristic to this minority. For example, while the few individual firms on the demand side of the market are supportive to an EU framework, three associations related to demand side of the market are not. Banks and their associations, representing the supply side of the NPL market are also split. The dissenting minority consists of national actors and those with a supranational perspective. While most dissenting respondents are located in a Member State with a low NPL ratio, some respondents from these Member States supported an EU framework. It is also notable that among those that support an NPL framework at EU level, some make this conditional on a good design that is not overburdening the market players, takes local determinants appropriately into account and is targeted to obstacles and disincentives.

The consultation replies reveal a broad range of issues an EU framework for NPL sales should cover. In order of frequency in which issues were mentioned: The link of such a framework to insolvency law and debtor protection is very frequently flagged. Several responses advocate measures to standardise the legal process of loan transfers. Some submissions propose measures to facilitate data transfer and data management. The request for licensing of NPL sellers appears in the replies at comparable frequency. A few replies also make a link to taxation and banks' capital requirements.

4. The economic function of loan servicers

Many respondents recognise advantages from the use of third-party loan servicers and considers them as important for the functioning of NPL markets. The dissenting minority refers to a lack of evidence and argues that internal work out in banks can be as effective as the outsourcing to loan servicing firms.

Respondents to the consultation see the role of loan servicers largely in managing NPLs, some consider they manage both performing and non-performing loans, and very few also attribute a role to them in securitisation and specialisation in real-estate loans. Many describe as valuable their services linked to monitoring, evaluation and information. Very few make a similar point with respect to other objectives such as the accomplishment of lower costs of the management or a higher recovery value of the NPLs.

Almost all stakeholders that see advantages flag benefits from specialisation. Some submissions argue that benefits derive also from scale effects, local expertise and expertise in collateral management or help with restructuring debt. A few argue that loan servicers can help NPL investors in their bargaining process with banks, while a few others argue that the realisation of advantage is depending on the nature of the outsourcing firm. Points made in this respect by some respondents are that outsourcing to third-party loan servicers creates new risks, potential conflicts of interests and impacts the reputation of the outsourcing firm. The impact of outsourcing on debtor and data protection is also regularly listed in this context.

Stakeholders' views are almost equally divided on whether there are obstacles for banks or non-bank investors to access third-party loan servicers or not. Several responses affirm that country-differences matter. One respondent remarks that absence of loan servicers would have no impact on NPL transactions taking place. Another one sees advantages from ownership of loan servicers, while a third considers it a disadvantage if loan servicers were owned by competitors on the NPL market.

As regards the impact on the ultimate debtor of an involvement of third-party loan servicers, a clear majority considers that it represents a challenge to existing debtor protection rights. While those that see no risk for debtors or consider existing rules as sufficient are a minority, some of them make the case that the obligations of the debtor do not change if a loan servicer becomes involved. Others explicitly refer to the reputation of loan servicers as a challenging factor.

5. Policy measures targeting loan servicers

A clear majority considers differences in business practices in loan servicing as significant. Views are almost equally split on whether the differences are justified or caused by financial regulation. Among those that do not consider them significant are also firms active in the market and one respondent argues that market entry of international firms has led to a convergence of industry practices regardless of the local market. As regards activity in several jurisdictions, relevant differences are seen as caused by legal tradition and consumer protection rules. Several respondents flag the difficulty and costs to learn and adjust to local conditions and some stress also the relevance of differences in licensing rules in this respect.

A substantial majority of responses indicates that it would be warranted to remove these obstacles and that this would have a positive impact on NPL markets. The few dissenting comments argue that consumer legislation requires differences across Member States to be maintained or that it could have a harmful impact on lending markets.

A substantial majority supports an EU framework for loan servicers. Only a small minority either objects or abstains and among them also respondents active in the market or representing interest of market participants. Almost all respondents that support an EU framework advocate it should cover a licensing regime and about half of them propose it regulates the supervision of loan servicers. Several also advocate measures to access data and improve data transparency. Many mention taxation, debtor protection and insolvency law as framework conditions which, if harmonised, would also improve conditions for international loan servicing firms. A few responses say that market standards and simplification should be covered. Single stakeholders also add as warranted coverage of loan servicers' remuneration structures, qualification requirements for their managers and staff, respect for local rules, debt collection guidelines and suspension rules.

Annex 3: Who is affected and how?

1.Practical implications of the initiative

Investment funds and investment firms that intend to purchase NPLs from banks should face reduced costs in getting authorisation if needed and lower compliance costs when buying NPL from banks in different EU jurisdictions. This is particularly the case for the smaller funds and investment funds where compliance costs are disproportionately larger. Investors would also benefit from availability of more loan servicers and lower costs of outsourcing the management to NPLs to loan servicers caused by higher competition on the loan servicing market. Higher competition among NPL investors should lead to declining profit margins in this industry.

Loan servicers, debt collection firms and financial firms considering to enter this business line should face reduced costs in getting authorisation and lower compliance costs when managing NPLs outsourced from NPL investors to them. Firms acting in different jurisdictions would benefit particularly from the passport, which eliminates to request authorisation in each jurisdiction. Tapping markets in different jurisdictions allows them to realise scale economies. More competition among loan servicers and scale economies should lead to declining fees for loan servicing. Average size and market concentration is expected to rise while profit margins should decline in the medium term as the result of more contestable loan servicing markets.

Banks would face a larger investment base and the more intense competition among investors would lead to higher bid prices for NPLs. This increases their profits respectively reduces the loss they would derive from selling NPL portfolios to non-banks. Banks located in Member States with hitherto high entry barriers for NPL investors and a small investor base would benefit over-proportionally. Smaller banks would have proportionally larger benefits because the larger the investor base, the smaller the size of the average investor and the smaller the investor the more likely it is that they bid for smaller NPL portfolios held in smaller banks.

Institutional investors such as insurance companies or pension funds are unlikely to be enticed to enter NPL markets. They are offered a greater range of attractive investment opportunities in investment funds that buy NPLs as a result of the initiative.

Third country firms would face lower entry costs from licensing if they buy NPLs from EU banks or provide loan servicing to NPL investors. The passport offers them to conduct business in all EU Member States.

Consultancy firms and law firms may see part of their business and profit opportunities erode since potential market entrance will require fewer services and legal advice from them in an environment of less burdensome and more harmonised entry and business conditions.

Debtors should in the first place not be affected because their obligation to pay back their debt and interest is independent from whether the NPL is held by a bank or transferred to a non-bank. However, they may face welfare losses from uncertainty because facing a counterpart they had not chosen and do not know, especially if the counterpart is authorised in a different Member State. While they know the conduct of banks from past relationships, they have less certainty about how the new creditor or loan servicer will behave. Debtors may see a loss in value of their customer relationship with their bank if the bank decides to sell its loan to a non-bank. This could in turn increase the incentive of the debtor to avoid the loan becoming non-performing. There is also a possibility that if NPL markets are established, debtors attribute a smaller value to their customer relationship with their bank. This may have an impact on their selection of banks and conceivably also on their willingness to take a bank credit.

The potential impact on highly indebted households is hard to foresee as it will depend on the behaviour of loan servicers. If the latter help indebted social groups more than banks to arrive at a more suitable payback profile of their loans, debtors may benefit. The opposite is possible if loan servicers apply existing debtor protection rights in a stricter way than banks.

The enhanced environment for banks to offload NPLs from their balance sheets through loan sales should be positive for SMEs since it will create room for banks to expand lending to viable companies. Similar the impact on highly indebted households, the impact on highly indebted SMES will depend on the behaviour of loan servicers. If the latter help them more than banks to arrive at a more suitable payback profile of their loans, they may benefit. The opposite is possible if loan servicers apply existing debtor protection rights in a stricter way than banks.

The public sector benefits from lower NPL on banks' balance sheets. This reduces the fiscal costs of a banking crisis. It also reduced the costs of banking supervision because one critical element of supervision becomes less sizeable. The targeted reduction of compliance costs could reduce administrative burden for the public sector. Some Member States may face rising demand for authorisation from third-country firms that intend to make use of the passport, but may need to be authorised by the supervisor in one Member State for doing so. Unless the fees charged for authorisation and supervision contain an implicit subsidy for the applicant, the impact should be budget neutral even in those Member States facing an increase in requests. If Member States see their preferred state of debtor protection eroded through the EU framework for NPL investors and loan servicers, they would warrant to complement the policy options at EU level through policy measures at national level in a way that keeps their preferences in place whilst help develop the NPL secondary market.

2.Summary of costs and benefits

The heterogeneity of the conditions for market entry among Member States, as well as national supervisory requirements related to the size and legal form of any market participant considering to buy NPLs, complicate to a great extent the quantification of the benefits of changes to regulatory standards. Since most investors are hedge funds or private equity investors, regulatory fees in the asset management industry seem useful to serve as comparison term. If a NPL investor can operate under the regulatory regime of an investment fund, the regulatory start-up costs would range between about EUR 10,000 to about EUR 15,000. 94 A Commission study suggests that direct regulatory fees could amount to less than 20% and about 40% of the regulatory start up costs might be attributable to compliance costs in terms of labour costs and to pay external servicers for local facilities in the host country.  95 Annual ongoing costs, for supervision, were estimated at about the same amount. Since market sources interviewed by the Commission assessed the average of total costs to enter a new NPL market at about EUR 60,000 to 100,000, compliance costs are deemed not particularly high in relation with total entry costs incurred by investment firms.

Table A.3.1: Compliance costs of cross-border asset management firms

Note: Scenario A describes an asset management company relying on in-house legal advice and in-house fund administration, whereas Scenario B shows an asset management company outsourcing legal advice and fund administration to third parties.

Source: European Commission 2017.

Loan servicers are subject to costs for licensing in most Member States, with requirements and compliance costs differing across Member States. The NPL report of the ESRB (2017) refers to market entry costs ranking up to about EUR 5 to 15 million, but does not specify the share of licensing costs or country differences. Market sources having replied to European Commission's inquiries indicate that licensing fees vary strongly across Member States and differences across firms in the same country suggest that both firm- and country specific factors matter.

From a very limited sample of replies, it became evident that actual one-off fees for the licensing vary from a few hundred euro (in several Member States such as Finland, Ireland, Sweden) to more than EUR 50,000 (for example in Czech Republic and Italy). Annual licencing fees range significantly as well, from a few hundred euro per annum to about 34,000 annual fees (for supervision of loan servicers charged in IE). Compliance costs for data reporting could add to these set-up and licencing fees the costs to comply with anti-money laundering rules, that may prove significant.

While fees for a banking license may not be particularly high, especially in those Member States that do not require a full banking license, a banking license carries additional compliance costs in terms of direct labour costs, necessary to ensure compliance with all rules applicable to credit institutions, including capital costs to fulfil minimum capital requirements. 96 In the absence of available examples for EU banking sectors, Dahl et al. (2016) in a US study on compliance costs found that small banks paid USD 100,000 to USD 170,000 for personnel expenses and USD 64,000 - 90,000 for other costs linked to compliance. 97  

According to market sources, some Member States' supervisory framework (for example Hungary and Romania) require a set-up social capital for loan servicers amounting to EUR 500,000 for NPLs acquisition and debt collection firms, respectively. Greece requires loan servicers to maintain capital of EUR 100,000. With a standard estimate of 10% costs for equity, this would translate into up to EUR 10-50,000 additional capital costs per annum if a banking license is required. Absolute amounts will be different for each individual case, also depending on the share of capital without a banking license. Similar considerations apply if NPL investors are required to set up a securitisation vehicle or investment fund structure and this needs to be supported with capital.

Cost structures relating to compliance depend strongly on the applicable legislation and type of firm. A 2009 study by Europe Economics 98 based its analysis of compliance costs emerging from various EU Directives on extensive interviews with the financial industry. The results for the asset management industry are shown in the tables below. The one-off costs are not fully comparable to licensing costs since they relate to the investment costs of existing firms to comply with new regulation and not of new firms to comply with existing legislation. They may nevertheless be indicative of the types of costs involved.

Table A.3.2 The drivers of one-off compliance costs in the asset management industry by Directive

Note: FCD := Financial Conglomerate Directive, CRD := Capital Requirements Directive, MiFID := Markets in Financial Instruments Directive, 3AMLD := Third Anti-Money Laundering Directive

Source: Europe Economics 2009

Table A.3.3 The drivers of ongoing compliance costs in the asset management industry by Directive

Note: FCD := Financial Conglomerate Directive, CRD := Capital Requirements Directive, MiFID := Markets in Financial Instruments Directive, 3AMLD := Third Anti-Money Laundering Directive

Source: Europe Economics 2009

Ongoing supervisory fees for banks depend, in the euro area, on the size of the bank and its risk exposure, by means of a fixed and a variable component. By means of example, a non-systemic bank with total assets of EUR 1.6 billion and risk exposure of EUR 700 million would have to pay a fee to the SSM of about 10,000 in 2017. In the chosen example, about a tenth of it is due to the fixed component. 99

A recent study by Dahl et al (2016) decomposes the compliance costs of smaller US banks by cost type and size of bank. It demonstrates the importance of personnel expenses as well as strong scale economies underlying costs for personnel and data processing. Also the share of costs for accounting and consultancy decline with firm size.

Table A.3.4: Compliance costs in small US banks

Cases of actual licensing costs are only available in form of anecdotal evidence. A market source indicated costs of EUR 60-100,000 to enter a market of which less then EUR 10,000 are caused to obtain a license as NPL investor. Market sources indicated a banking licenses in a Nordic country requires a guarantee depending on the turnover. It would amount to around EUR 500,000. A German bank founder reported to a newspaper costs of EUR 700,000 to EUR 800,000 to obtain a banking license in Germany. 100 Other online sources suggest that starting an offshore bank demands between $150,000 to $250,000 and requires $1 million in capital, depending on the jurisdiction. For founding a bank in the USA, the amounts would be four times as high. 101  

Cost savings would be very different across Member States depending on their licensing regime, which sometimes entails only a partial banking license. The table below categorises the examples given in the text below. Given the anecdotal character of some, their country-specific nature and different sources, they are not comparable.

Table A.3.5: Overview of administrative costs by type of financial institution

Note: all numbers in EUR unless otherwise indicated.

The Table below translates the scarce information about costs that NPL investors and loan servicers entail if they expand activity into estimates of potential cost savings if any of the preferred option is implemented. Given the poor data quality, the numbers should only be seen as indicative. They will be very different in dependence of the individual firm. The numbers only cover regulatory charges. Labour costs and legal fees would multiply the amount.

The required re-writing of law and its implementation will entail one-off costs at EU and national level, especially for Member States that used to ensure debtor protection through the authorisation and supervision of NPL investors and would need to implement new laws that uphold the desired level of debtor protection through other means. Costs of writing new legislation are substantial. Using data from New Zealand, Wilson et al. (2012) estimate costs of a new law to amount to USD 2.6 million USD and that of a regulation at about USD 400. They refer to a similar study that point at costs in the US amounting to less than USD 1 million, but find that this study is likely to underestimate costs. As negative outcome, one can assume that each EU Member State finds it necessary to adjust existing legislation and encounter costs as high as those found in the study from New Zealand. This would be broadly EUR 60 million 104 and may represent an upper bound because not all Member States would need to adjust national law and for those that do, it will concern most of the time adjustment of existing law rather than completely new law.

Some Member States may face rising demand or authorisation and licensing from NPL investors and/or loan servicers, which would imply higher administrative costs. Standard practice is that the public sector charges a fee for the licensing process and supervision that fully covers these costs. Recent work by the Commission services on the licensing of investment funds and crowdfunding (see corresponding impact assessments) identified licensing costs in the range EUR 5,000 to 10,000. One Member State calculates costs of supervision of a loan servicer at EUR 34,000. Given that supervision of loan servicers might be more expensive if the entity acts in different Member States, it would be reasonable to assume that costs increase up to EUR 50,000 p.a.. NPL investors may face lower supervisory costs if they use an EU supervised loan servicers. These costs could apply to each new entrant, but would need to be seen in conjunction with cost savings from a single point of entry rather than requiring to bear the costs for each Member State the entity wants to be active in.

All policy options are expected to lead to compliance costs as regards the implementation of the new law, the relevant formalities and training for NPL investors and for loan servicers. Significant compliance costs are expected in particular when it comes to audit and management, monitoring, supervisory and licensing fees.

While the establishment of principles or rules at EU level would entail compliance costs, NPL investors and loan servicers would no longer be exposed to the costs of compliance to national rules. Overall, the EU compliance costs should be lower than the average compliance costs across EU Member States. Especially NPL investors and loan servicers operating in several EU Member States should benefit from lower compliance costs.

Given the fact that national rules for loan servicers are generally tighter as compared to national rules for NPL investors, compliance with common rules is likely to trigger higher costs for loan servicers as compared to NPL investors. Citizens/consumers are not expected to bear direct costs. Banks wishing to sell NPLs may be exposed to indirect costs resulting from the regulatory change if these affect the price bids of NPL investors.

In both groups of market players (i.e. NPL investors and loans servicers), the cost of complying with the regulatory change is expected to be greater for small players as compared to bigger players as those costs constitute a greater share in revenues. In the case of small NPL investors, this initial cost may be compensated over long term with greater revenues linked to expanded activity across country borders. The described compensation is less likely for loan servicers, whose operations are characterised by economies of scale. Consequently, the loan servicing sector is expected to consolidate following the regulatory change, with potentially negative effects for small businesses but with efficiency gains at an aggregate level.

Over longer term, i.e. once market participants have adjusted to the regulatory change, cost savings are expected for the industry on activities such as application for a licence, calculation of regulatory fees, regulatory reporting, marketing. A further beneficial impact on costs is expected thanks to a lesser need for legal advice due to harmonisation and transparency of rules..

For market participants based outside the EU in particular, NPL investors and for loan servicers are expected to benefit not only from lower regulatory fees but also the potential search and legal counsel costs, facilitating access to EU markets.

From consumer perspective, the obligation to respect national rules for privacy, data and debtor protection for loan servicers is key to ensure an adequate and predictable level of consumer protection in all EU countries.

Annex 4: Analytical methods

1. A stylised view on demand and supply of the NPL market

The first part of this annex presents a conceptual framework to analyse the potential impact of policy measures on the demand for and supply of non-performing loans on secondary markets. To put the initiative analysed in this text into perspective, the appendix gives a schematic overview of the different NPL initiatives and of the specific failures they address, such as shortage of supply, lack of demand, information costs, valuation.

The presented framework is theoretical as data availability does not allow a derivation of full quantitative properties of demand, supply and of the market equilibrium. A further complication arises from the fact that non-performing loans are not a homogenous good, which is evidenced by different prices for secured (largely by real estate) and non-secured (largely consumer loans). 105 Finally, the structure of the NPL market is that of an oligopsony, with few large buyers conducting a small number of transactions. 106 Notwithstanding the mentioned difficulties, it seems possible to derive stylised characteristics of the NPL demand and supply function based on the incentives that investors (the demand side) and banks (the supply side) face.

The proposed framework considers a simplified model of the NPL market with a portfolio consisting of a large number of homogenous loans. Looking at the supply side, the higher the market price relative to the nominal value (the gross book value), the higher the volume of NPLs proposed for sale. One can also assume a price floor below which banks will not offer any NPLs. Determinants of this floor could be: i) a (non-negative) expected recovery value on the loan portfolio, and ii) the desire by a bank not to realise capital losses, which may arise if the market price is too much below the expected recovery value. Even if the debtor does not pay back his loan, the bank may anticipate that he pays back part of the loan and may therefore wish to keep the loan on its books. Banks may also refuse to sell NPLs if they esteem the customer relationship and hope for ongoing business with the debtor, which could become profitable again. 107 The mentioned mechanism may result in a price ceiling. Moreover, the offer price may reflect potential reputational costs for a bank linked to NPL sales and their negative impact on long-term relationships with its clients. Finally, strategic considerations may also determine the offer price. For example, the bank may anticipate that the bidder will try to exploit its weak bargaining position if it has a large pile of NPLs on its balance sheet. Therefore, the bank may strategically decide to enter negotiations with a somewhat inflated offer price.

In the chart below we assume that at a market price equalling the nominal value banks will offer 70% of NPLs for sale (i.e. keep 30% of the NPLs on their loan book). We also assume that banks will not supply any NPLs to the market at a market price equal or lower less than 20% of the nominal value.

Figure A.4.1: A stylised perspective on demand for and supply of NPLs

Looking at the demand side, potential investors are expected not to offer a price much below the highest recovery value of the NPLs offered because it would be refused by banks. Although investors' expectations on recovery values may differ from related expectations by banks, the discrepancy may not be too significant. At the same time, investors are likely to demand a price that covers their costs of administering the NPLs, which may justify the discount as compared to the recovery value. NPL investors are also likely to request an add-on that reflects funding costs and/or their internal requested rate of return. Both seem to be higher for non-banks than for banks.

For a zero market price, demand for NPLs is expected to amount to the entire loan book as investors are willing to take the total loan book if it is for free. As to the shape of the demand curve, the relationship between price and demanded volumes is expected to be non-linear for strategic reasons. For smaller shares of NPLs offered, the investor may anticipate a lemon issue: the counterpart could offer NPLs with the weakest recovery value and keep the higher quality NPLs on the balance sheet. The described mechanism could in the extreme case lead to a demand curve which is backward bending in parts, i.e. demand declines when prices decline. 108 While it is not possible to identify the range in which the demand curve is backward bending, one can at least assume that the demand curve is steeper (i.e. more elastic to changes in prices) for lower volumes of NPLs. The larger the share of NPLs sold, the less relevant the lemon issue. For larger shares of NPLs offered, non-linearity may occur due to efficiency gains in loan administration, for example by realising scale effects in loan servicing and debt collection.

NPL transactions are done with consultants, which charge a price for their services. The added value of the consultancy services is to match demand and supply, which is not trivial given the opacity of the market, the underlying lemon issue and the bilateral bargaining position of both the buyer and the seller, requiring a tailor-made contract that encompasses all information and incentive asymmetries. These transaction costs are reflected in a bid-ask spread, which can be charged either on the selling banks or on the buying investors. The chart above assumes that the transaction costs increase the costs for investors proportional to the price and move the demand curve northwards.

Both the demand and the supply curve may be affected by policy options listed in the NPL action plan. More efficient insolvency frameworks would increase the recovery value of NPLs, thereby shifting both demand and supply upwards. Market prices would increase, but the effect on volumes is uncertain, depending on whether banks or non-bank debt collectors could benefit more from the improved insolvency framework. Another element potentially leading to a changed market outcome is supervisory pressure on banks to disclose NPLs and/or to off-load them from their balance sheet. If banks face stronger incentives to provision NPLs, the book value declines relative to the nominal value, which reduces the gap to the market price. The mentioned supervisory pressure has the potential to shift the supply curve downwards, thereby decreasing prices and increasing volumes. An establishment of AMCs could have a similar effect as it would incentivise banks to supply more NPLs if transactions are arranged by a third-party with possibly larger bargaining power and smaller stigma effects. The direct consequence could also be lower transaction costs if the AMC economises on some of the activity that consultants are undertaking.

On the investor side, the existence of an AMC on the market would likely lower search costs but it could strengthen the bargaining position of their counterpart (i.e. the bank selling NPLs). AMCs and information platforms could also reduce transaction costs by providing impartial information to potential buyers, i.e. facilitating due diligence and reducing information asymmetries between the initial creditor and potential buyer of the debt contract. If the transaction costs are ultimately paid by the investor, the involvement of AMCs and information platforms will move the demand curve upwards, i.e. closer to the curve without transaction costs. More efficient securitisation markets would be largely to the benefit of lower funding costs for NPL transactions, which could move the demand curve upwards as investors could afford to pay a higher price for NPLs.

Lowering market entry conditions for investors into the secondary market for NPLs would shift the demand curve to the right, but the impact on the reservation price is unclear as it is uncertain whether new entrants would be able to realise higher recovery values. Both recurrent and one-off costs may constitute market entry barriers as they affect the result of a cost-benefit analysis undertaken by a potential entrant. As to one-off costs, they cannot be recovered if the firm is not able to do successful business (so-called sunk costs) and they include: obtaining authorisation and licenses, investments to become eligible for national conditions, search for loan servicer. Examples of recurrent costs include: debt collection, collateral use, compliance to conduct rules.

Figure A. 4.2: An economic perspective on different policies to tackle NPLs in EU banks

2. Cross country analysis

This part derives insights from the cross-country variation in selected NPL data. The comparison covers the EU Member States. For some exercises, data was not avaialable for all EU Member States and in some comparisons it turns out that the UK observations were outliers. In some of the latter cases, the observations for the UK were not considered. Overall, the quality of the data basis is weak. Despite the weak data quality, most of the found correlation look plausible and evidence of a systematic bias in the data that would distort the results could not be detected. This said, the resulting numbers should be best understood as illustrative only as they do not stand up to the requirements of rigorous robustness checks.

The only official statistics available for NPLs are volumes of NPLs on banks' balance sheets and their ratio to loans and advances on banks' balance sheets collected by EBA and ECB. Since the coverage of banks is larger in the ECB than in the EBA data file, ECB data was used for the analysis.

The only source for data on NPL sales are international consultancy firms and they collect the data from public sources, own business and business contacts. Data collected and made public by the different consultancy firms is broadly similar, but differs somewhat, which indicates that the underlying ground work is difficult and there are limited means to verify data. Issues emerging from data quality are discussed below. For the analysis in this section, the data of NPL sale by Member State 2015 and 2016 published in PWC (2017) and Deloitte (2017). For ratios, the loan sales data of the year was combined with the stock of NPLs in the same Member State's banks' balance sheet at the end of the previous year, i.e. 2016 transactions relative to the stock in 2015Q4.

Even if consultancy firms strive to have a high standard on data collection, they do not have the means and authority to verify data to the same extent public statistical offices can. Hence, there may be a bias in the data emerging from the possibility that some loan sales take place without any notification to the public. Another issue is that the available data is patchy, i.e. not all data fields are complete. For example, the amount traded is not disclosed or available in more than 15% of the transactions, in one micro data set the Commission services were able to check. The consequence is that observations for Member States with few transactions and a large share of transactions with unknown amount cannot be used. This led to the decision not to consider the observations for Belgium in the analysis below.

Another complication for any empirical analysis with this data stems from the observations that underlying transactions are very different in size and a few very large transactions will determine observations for some Member States. This may cause outliers to have a strong impact on descriptive and analytical statistics. Since the largest transactions are clustered in Member States with a larger number of transactions (UK, IT, ES, DE, IE), the risk of empirical results being determined by individual loan sales appears limited.

In almost 25% of the transactions, buyer respectively seller are not known. Hence it cannot be said whether the NPLs were sold or bought by non-banks or other banks. Very often loan sales combine the sale of non-performing with performing loans. Given the uncertain sourcing of the data, it may even be possible that some transactions are carried out with performing loans only. An example that combines several of these two issues is the sale of a portfolio with a face value of 11.8 billion GBP by UKAR to Prudential and Blackstone in May 2017. 109 The amount is equal to about 10% of the total annual turnover in 2016. The seller is not a bank, but a public AMC. The underlying 104,000 loans are performing and they had been originated by the bank Bradford and Bingley before it was put in public ownership in 2008.

While the share of NPLs in the reported loan sales is unknown, three different methods suggest it could be on average in the range of 70 to 80%.

·A first estimate stems from AFME and is reported in its reply to the public consultation. According to data from KPMG "74% of the total loans sales completed between 2015 and 1H 2017 in Europe represented either non-performing or a mixture of non-performing loans with other risk exposures (i.e. with performing, subprime, or re-performing loans)."

·A second estimate was conducted by the Commission Services with data from another consultancy firm shows a share of non-performing loans in loan transactions of 47% on average 2014-2016. It also reveals that for 34% of the loan amounts there is no information whether they are performing or non-performing and in 12% the loan amount was a mixed portfolio, consisting of a unknown share of performing and no- performing loans. If it is assumed that the ratio is the same in the unknown and mixed transactions as in the trades with a known breakdown, the ratio of non-performing loans in all loan trades would be 78%.

·A third method consists in a regression analysis that relates the loan sales in 15 EU Member States to the change in the volume of non-performing loans on banks' balance sheets (see chart below). The regression line suggests that for 100 billion EUR loan sales, the amount of non-performing loans declines on averge by 73 billion, i.e. an implied proportion of 73% of non-performing loans in total loan sales. If the outlying observation of the UK is not considered, the ratio would increase to 80%.

Figure A 4.3 Loan sales and change in NPL volumes across EU Member States

The chart below applies the same methodology, but does not take loan sale and NPL volumes in EUR, but as a ratio to the stock of NPLs and total loans, respectively. This is an implicit control for the size of the market and avoids that Member States with large NPLs have a dominant impact on the correlation. The correlation is insignificant and the R2 small unless the UK as outlier is excluded from the panel. If the UK observation is not considered, the regression line suggests that a 1 %-pt increase in the ratio of loan sales to NPL volumes decreases the NPL ratio by 0.3%-pts.

Figure A 4.4 Loan sales and change in NPL ratios across EU Member States

.

The next chart applies the same methodology, but uses a different data set, namely loan sales submitted and calculated by AFME on the basis of data collected from KPMG from public data sources. This data set includes an observation for SE, but misses some EU Member States (FR, LT). Numbers are broadly comparable with the exception of those for UK and HR. Still, the slope coefficient is similar and the share explained by the regression line somewhat higher, even if the UK is not excluded from the panel.

Figure A 4.5 Loan sales and change in NPL ratios across EU Member States (alternative data source)

A further cross-country comparison shows some correlation between loan sales and the bid-ask spread derived from the theoretical model presented in Annex 4.3. While the R2 is not particularly high, the correlation is significant at 5% level independent of whether the UK is included into the sample or not. Though the regression analysis gives no information about causality, it suggests a 10%-pts decline in the bid-ask spread would be consistent with an 3.3 to 4.5 %pt increase in loan sales relative to the outstanding NPL stock. If the initial NPL ratio or the market size are added as additional control variable, they does not come out significant, and do not change the significance of the bid-ask spread.

Figure A 4.6 Loan sales and bid ask spread across EU Member States

Since for some Member States there is not data about loan sales, it is also interesting to directly compare the bid-ask spread derived from the model with the change in NPL ratios. For both data is available for all Member States, bar CZ for which the NPL ratio in the ECB data set starts only with the observation of 2016Q1. The relationship between bid-ask spread and the change in the NPL ratio from 2014Q4 to 2016Q4 is not significant. This, however, changes, if the observations for those Member States, in which the NPL ratio increased over these two years are excluded from the panel (BG, EL and PT). Since the reasons for an increase in the NPL ratio are unrelated to the bid-ask spread that impacts the loan sales, such elimination of single observations from the data set seems justified. The correlation become significant and suggests Member States with a lower bid-ask spread were able to realise a relatively larger decline in their banks' NPL ratio. The slope coefficient suggests a 1 %pt lower bid-ask spread reduces the NPL ratio by 1.6%.

Figure A.4.7 Change in NPL ratio and bid ask spread across EU Member States

The table below shows the numerical results of the different specifications.

Table A4.1 Regression results, cross-country OLS, dependent variables Loan sales (upper panel) and change in NPL ratio (bottom panel)

There is also a broad correlation between the number of loan servicers active in a Member State and the volume of NPLs. The number of loan servicers is taken from Table A2 in Annex 5.2. The right-hand panel zooms in on smaller Member States, that are not clearly visible in the left-hand panel. The comparison suggests that the number of loan servicers relative to the amount of non-performing loans is small in Italy, France, and despite numerous authorisations recently in Greece. 110 Spain and Portugal are borderline cases. The UK is also an outlier since it has many loan servicers. This might be explainable by the role loan servicers have in supporting securitisation activity or the outsourcing of the management of real estate loans in the UK.

Figure A.4.8 Number of loan servicers and NPL volume across EU Member States

Although the cross-country comparisons produce plausible coefficients, the small number of observations and the data caveats listed above suggest that the result are best treated as illustration and not at statistical evidence. Results may not be robust and change once the analysis is re-run with observations for more countries or additional years.

3. Quantifying the impact of the different policy options on NPL secondary markets- explanation and assumptions

Inefficiencies in the pricing of NPLs show up as relatively wide spreads between the ask price from the sellers of NPLs and the bid price from buyers. Then, one of the goals of the difference policy options is to improve NPLs secondary market efficiency helping to reduce such spread. We observed that a reduction in the bid-ask spread is correlated with a reduction in the NPL ratio.

Pricing model for the bid-ask spread

We have implemented a theoretical model to calculate the bid ask spread on NPLs for the EU MS:

We apply the methodology proposed by Ciavoliello, et al (2016) and proceeded as follows:

1. The future value of a loan that performs and that matures at time n is Fn. This loan has cash flows (ft) from now until maturity in time n. We use the loan effective rate (i) to calculate the future value.

2. To calculate the present value of a performing loan (Gross Book Value or GBVu) we discount the future value of the loan (Fn) using the loan effective interest rate (i). Thus:

3. If the loan defaults or does not perform (NPL), the owner (bank) of the loan can only recover a percentage on the GBVu (recovery rate = rr). Then the Gross Book Value of the defaulted loan (GBVd ) is:

4. If the loan becomes non-performing it incurs in some costs to either default management or loss mitigation that we name indirect costs (ic). These costs are the fee that the loan servicer will charge for their services and it is a percentage of the Gross Value of the default loan. Then, the net value of the default loan (NBVd ) is:

5. Then, the bank with NPL has to provision for the losses in the loan. The provisions should be the difference between the GBVu and the GBVd :

6. To avoid further losses on the loan, the bank will be willing to sell the NPL at GBVd . Any price above this value will generate profits and any price below will further damage the bank profitability and its capital position. Then, our ask price estimation for the loan, the minimum price at which the bank would be willing to sell the loan, is GVBd :

7. The ask price will be higher if the NPL is under provision to avoid inputting further losses in the income statement. The more in need of capital and under provisioned the higher the ask price by the seller bank.

8. Potential NPL buyers need to take into consideration the GBVd , the indirect costs to recover the loan (ic) and its expected profit. This expected profit should be weighted by risk. However, for simplicity reasons, our assumption is a plain profit of 15% on top of the loan effective interest rate. Then, the bid price, the maximum price that the buyer is willing to pay will be:

9. The bid ask spread in the secondary market for NPLs will be:

Then the drivers of the differences in bid ask spread among EU MS will be:

a.The loan effective rate

b.The time to recover the loan

c.The recovery rate

d.The indirect costs

e.The provisions

f.The buyer expected profit

Calibration of the model

To provide an estimation of the differences in bid ask spread on NPLs among MS, we have gathered information from World Bank, Doing Business 2016.

The loan effective rate is calculated for every MS using the interest rate of new lending to non-financial corporations with a maturity of 1-5 years. This data is compiled by the ECB.

We use the time to recover the loan, the recovery rate and the indirect costs provided by The World Bank in its publication Doing Business 2016. The values of these variables for each MS are calculated based on the time, cost and outcome of insolvency proceedings in a given economy.

Time remaining to collect the cash flow from the NPL loan is provided either by NBER and Doing Business. NBER data is from 2006 whereas data from World Bank is from 2016. The values are not the same but the differences are small for all the Member States but the Czech Republic and Romania.

To make the data on recovery rate comparable across countries, several assumptions about the business and the case are used. The recovery rate is recorded as cents on the dollar recovered by secured creditors through reorganization, liquidation or debt enforcement (foreclosure or receivership) proceedings. The calculation takes into account the outcome: whether the business emerges from the proceedings as a going concern or the assets are sold piecemeal. Then the costs of the proceedings are deducted (1 cent for each percentage point of the value of the debtor’s estate). Finally, the value lost as a result of the time the money remains tied up in insolvency proceedings is taken into account, including the loss of value due to depreciation of the hotel furniture. Consistent with international accounting practice, the annual depreciation rate for furniture is taken to be 20%. The furniture is assumed to account for a quarter of the total value of assets. The recovery rate is the present value of the remaining proceeds, based on end-2015 lending rates from the International Monetary Fund’s International Financial Statistics, supplemented with data from central banks and the Economist Intelligence Unit. It is important to note that the drivers of the recovery rate, i.e. the cost, the time and the binary outcome of the process (the company continues to operate or is sold piecemeal) are derived from questionnaire responses by local insolvency practitioners and verified by the World Bank through a study of laws and regulations as well as public information on insolvency systems. In other words, the recovery rates calculated by the World Bank are not directly based on an average of observed recovery rates.

The estimated cost of the insolvency proceeding or indirect costs are reported as a percentage of the value of the insolvency estate, borne by all parties. Costs include court/bankruptcy authority costs, attorney fees, bankruptcy administrator fees, accountant fees, notification and publication fees, assessor or inspector fees, asset storage and preservation costs, auctioneer fees, government levies and other associated insolvency costs. These costs will be mainly the fee that the third-party loan servicers will charge. Once again there are small differences or not differences at all in these values between NBER and Doing Business for all Member States but Austria, Denmark and Poland. In these three countries the indirect costs reported by NBER are substantially higher than the ones we use in our calculations.

We have made the assumption that the bank provisions the difference between the Gross Book Value of the performing loan less the Gross Book Value of the non-performing loan. If the bank has a higher provision mean it is over provisioned if the bank has a lower provision it means the bank is under provision for that NPL. Banks that are in trouble because low profitability and higher capital needs tend to be under provision, which means they will ask for a higher price that if they were better provisioned.

Finally, our assumption for the buyer (investment fund) is that it will enter the secondary market if it can make a profit. When the buyers of NPLs enter the market they will use the services of third-party loan servicers. Then, the buyers of NPLs in the secondary market will take into account in their bid price the indirect costs, or costs associated with loan servicing, and the expected profit. For the expected profit, the IRR of the external investors (hedge funds, mutual funds, other non-bank investors, etc), we have made the assumption that the external investors identify buying NPLs as a risky business so they applied an excess return of 15% 111 on top of the rate of return adjusted for country risk in each country 112 . Then the IRR in each country is the sum of the excess return because of NPL specificities + the rate of return adjusted for country risk.

Scenario analysis

Once we have calculated the bid-ask spread in the NPL secondary for the current situation we estimated the effect that the different policy options could have on such spread. We distinguish between policy options that could increase the investor base and the policy options to improve the availability of third-party loan servicers. On the other hand, we have observed, using country data, that there is a correlation between the two year variation in the NPL ratio ( NPL loans / total loans) and the bid ask spread. Applying regression analysis we estimated that for 1% decrease in the bid ask spread there is a 0.88% decrease in the NPL ratio, which means that, “ceteris paribus”, there would be a reduction of 0.88% in the volume of NPLs every two years, or 0.44% each year (see Annex 4.2). Our assumption is that such reduction on NPLs will increase the volume of transactions on NPLs or NPL sales from banks to investors. To estimate the incremental volume of NPL sales for the next two years in each MS we multiply the reduction in bid ask spread times 0.88% (this is the correlation found) and times the volume of NPL in each country at the end of first quarter of 2017, the last available information from ECB.

The options are: common principles, passporting and rule book for both NPL investors and loan servicers. The next step has been to quantify such policy options in the NPL market.

Policy scenarios for NPL investors

To quantify the impact of NPL investors’ policy options on NPL market our assumption is that such policy will contribute to reduce risk perception by investors in EU Member States. Such reduction will occur through a convergence in the rate of return required by investors in a market that becomes European, because of the policy measures to improve market efficiency, therefore more efficient that those MS individual markets consider isolated. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Policy option A for NPL investors is to have minimum common standards for investors across EU Member States. To quantify the impact of such policy on NPL market our assumption is that such policy will contribute to reduce risk perception by investors in some EU MS, those with more entry barriers: BG, EL, IT, CY, HU, AT, PL, and RO. For these countries our assumption under policy option A is that they will adjust their return adjusted by country risk to the benchmark of 1.9%. The benchmark is the average lending rate of the countries with the lowest country risk. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Figure A.4.9: Lending rates as country-specific risk premium in investors' required return

Source: Commission calculations with ECB data of MFI interest rates of new lending to non-financial corporations maturity 1-5 years in October 2015.

Policy option B for NPL investors is to implement a common passport for NPLs investors. To quantify the impact of such policy on NPL market our assumption is that such policy will contribute to reduce risk perception by investors, specifically in those EU MS with country risk above the benchmark (1.9%). We assume that all the MS will benefit of the European market framework with a reduction of 50 basis points in the rate of return demanded by investors, besides those MS with country risk above the benchmark will converge by 50% to the benchmark. The benchmark is the average lending rate of the countries with the lowest country risk. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Policy option C for NPL investors is to implement a common rule book for NPLs investors. To quantify the impact of such policy on NPL market our assumption is that such policy will contribute to reduce risk perception by investors, such that the worst performers will adjust their rate of return to the benchmark but at the same time the best performers will also adjust to the benchmark which will penalize then. We assume that all the MS will benefit of the European market framework with a reduction of 50 basis points in the rate of return demanded by investors, besides those MS with country risk above the benchmark will converge by 50% to the benchmark whereas those MS below the benchmark will move to the benchmark reducing the distance by 25%. The benchmark is the average lending rate of the countries with the lowest country risk. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Policy scenarios for loan servicers

To quantify the impact of NPL investors’ policy options on loan servicers market our assumption is that such policy will contribute to increase the number of third-party loan servicers in MS which we associate with a reduction in the barriers of entry and in the costs of providing loan servicing. Such reduction will occur through a convergence in the cost of servicing that will improve market efficiency, therefore more efficient that those MS individual markets consider isolated. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Policy option A for NPL investors is to have minimum common standards for loan servicers across EU MS. To quantify the impact of such policy on NPL market our assumption is that such policy will contribute to reduce servicer costs in some EU MS, those with more barriers: EL,IT,CY and AT. For these MS our assumption under policy option A is that they will be able to adjust their cost by 10%. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Policy option B for NPL investors is to implement a common passport for loan servicers. To quantify the impact of such policy on NPL market our assumption is that such policy will contribute to reduce loan-servicing costs by 10% in all countries due to the higher size of the market and those MS with cost above the benchmark will close the gap by 50%. The best performers MS in terms of cost will be able to keep such advantage. The benchmark is the average cost among those countries with the best cost records. Such reduction will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.

Figure A.4.10: Indirect costs of loan recovery as indicator of costs of loan servicing

Source: Commission calculations with World Bank Doing Business 2016 data of insolvency – cost of recovery in %.

Policy option C for NPL investors is to implement a common rule book for loan servicers. To quantify the impact of such policy on NPL market our assumption is that such policy will contribute to reduce loan-servicing costs by 10% in all countries due to the higher size of the market and the cost gap between each country and the benchmark will close the gap 50%. The best performers MS in terms of cost will suffer an increase in their cost due to the convergence to the benchmark. The benchmark is the average cost among those countries with the best cost records. The reduction in costs will contribute to reduce the bid ask spread in the NPL market which will increase the volume of NPL transactions.