The large stock of non-performing loans (NPLs) held by euro area banks should be more swiftly resolved, while avoiding fire sales. We make a case for a comprehensive European solution, combining various NPL resolution tools. Within the NPL resolution toolkit Asset Management Companies (AMCs) may offer significant benefits by bridging inter-temporal pricing gaps for asset classes such as commercial real estate loans. We outline elements of an EU-wide blueprint for country-specific AMCs, including state aid aspects, asset and participation perimeters, asset valuation, capital and funding structure, and governance. In addition to AMCs, internal NPL work-out will always play an important role in NPL resolution, complemented by private information and trading platforms, and securitisation schemes.

1. Introduction

The large stock of non-performing loans (NPLs) held on the balance sheets of euro area banks is a pressing financial stability issue for the euro area while it also represents sand in the wheels of the bank lending channel of monetary policy. The post global financial crisis surge in NPLs in the euro area peaked in 2013, when the aggregate NPL ratio reached 8%. While the average NPL ratio has declined gradually since then, by around one percentage point per year, differences across countries have been marked with six countries still having NPL ratios above 10%,^[1]These countries are Cyprus, Greece, Ireland, Italy, Portugal and Slovenia. significantly so in some cases.
There are many reasons why the resolution of NPLs in Europe needs to be accelerated. First, bank resources – capital, funding, management attention and human resources – are tied up by assets that are not producing income while the scope for new lending to productive ventures is also curtailed.^[2] See Aiyar et al. (2015) for a discussion of the possible impact of NPL resolution on bank capital and lending capacity. Related to this, the high stock of NPLs is associated with higher uncertainty about future bank profitability, leading to higher bank funding costs and commensurately higher costs of credit for all borrowers, even the soundest ones. Second, high stocks of NPLs usually indicate underlying solvency and debt overhang issues affecting the corporate sector. Such excessive indebtedness often means that corporate investment cannot keep pace with the expected recovery in the real economy.^[3] See, for example, Goretti and Souto (2013), Nkusu (2011), Balgova et al. (2016) for evidence that a high stock of NPL is associated with weaker economic growth. Moreover, keeping over-indebted and ultimately non-viable firms alive by not resolving NPLs in a timely manner generates artificial and unhealthy competition for viable firms in the market.
At the same time, caution is needed to avoid resolving NPLs too quickly as this may create fire sale conditions and put excessive pressure on bank capital levels. Moreover, premature liquidation of firms that might otherwise have remained viable after some restructuring and reorganisation may lead to a destruction of economic value. Overall, therefore, it is crucial to find the optimum speed of NPL resolution, which is likely to differ among countries and between asset classes.
As discussed by Constâncio (2017) and elaborated by Fell et al. (2016), asymmetric information and structural impediments are among the main causes of slow NPL resolution in the euro area. Fully efficient markets for distressed debt would swiftly clear NPLs from bank balance sheets. However, transparency around the quality and real value of NPLs is very limited, and the duration and outcome of legal processes to recover value from NPLs is highly uncertain. NPL transaction volumes in the euro area thus remain a small fraction of the entire NPL stock (Deloitte, 2016) and there is a wide gap between prices that banks wish to achieve (in line with their provisioning levels) and prices that investors are prepared to pay.
Against this backdrop, it is clear that a comprehensive approach to NPL resolution, involving some degree of coordination at the European and national level, is necessary. The NPL problem cannot be solved by any single policy measure be it supervisory, macroprudential, or structural in nature. Appropriately robust supervisory guidance as published by ECB Banking Supervision (ECB, 2017) is essential to improve banks’ management of the NPL problem. But it must be complemented by structural reforms to enhance the recoveries and the net present value of NPLs, and by complementary measures to facilitate the development of NPL markets. Only when banks can use the full set of potential NPL resolution tools can the current inaction bias be overcome, thereby minimising the undesirable side effects of liquidating NPLs. The remainder of this article discusses the elements of such a comprehensive strategy, with a particular focus on asset management companies.

2. The benefits of Asset Management Companies (AMCs)

Asset management companies (AMCs) have often been used to manage distressed assets arising from systemic financial sector stress (Cerruti and Neyens, 2016) and have a proven track record in making significant contributions to the clean-up of banking sectors suffering from NPL problems. Examples include AMCs that were established in the aftermath of banking crises in Sweden (in the early 1990s),^[4]See Jonung (2009) for an account of the rationale for the AMC in Sweden and its role in the management of the banking crisis. in Korea (in the late 1990s)^[5]See He (2004). and, more recently, in the euro area countries Ireland (2010), Spain (2012) and Slovenia (2013). One of the common features of these systemic AMCs is that governments have been strongly involved in their creation, by providing capital, facilitating funding, and passing legislation governing the design and operations of the AMCs.^[6]AMCs may also be created in the process of restructuring or resolution of a single bank, often without government support. Such AMCs are often, somewhat loosely, described as ‘bad banks’. … Continue reading
The main function of systemic AMCs is to provide a “bridge” for the inter-temporal pricing gaps which emerge when market prices for NPLs and the underlying collateral are temporarily depressed. This may happen because of heightened risk aversion and a drying up of liquidity in the market, but, ultimately, market prices recover as economic conditions improve. Bridging this inter-temporal pricing gap is accomplished by removing a significant share of NPLs, usually belonging to a specific asset class such as commercial real estate, from bank balance sheets and working them out over a specified time horizon to maximise their recovery value. The transfer price paid to banks by the AMC is usually set at long-term (‘real economic’) value, thus avoiding the fire sales that would result from NPL disposals into illiquid markets where the risk premia required by outside investors are unusually high. Shielding banks from fire sale conditions can be especially beneficial if several banks are attempting to resolve their NPLs at the same time: systemic AMCs, in other words, can provide an important coordination role. Other benefits of AMCs are related to a swift reduction in uncertainty surrounding the profitability and solvency of banks once NPLs are transferred to the AMC. This, in turn, has a positive impact on bank’s funding and capital costs.
AMCs do not offer a panacea for systemic NPL problems and their success depends both on their design and the prevailing economic circumstances. Past experience suggests that several success factors should be present if an AMC is to accomplish its objectives. First, AMCs tend to be best suited for particular asset classes, notably fairly homogenous NPLs of a certain size, such as commercial real estate. Second, asset valuations and the resulting transfer prices should be realistic, thereby limiting the risk that AMCs run losses and deplete their capital while giving some room for manoeuvre with respect to asset resolution. A well-designed governance structure, with a strong mandate, is another essential ingredient for a successful AMC. There are numerous examples of AMCs failing because of political interference with their activities. The lifetime of the AMC should be finite and defined at its inception to ensure that the AMC does not become a self-perpetuating enterprise. Dedicated legislation is often necessary to lay down its governance structure and mandate. Finally, a basic premise for the success of AMCs is that asset values start to recover in the medium term. This, in turn, implies that authorities pursue sound macroeconomic and financial policies.

3. The merits of a blueprint for national AMC

In the EU the scope for establishing system-wide, government-sponsored AMCs is restricted by the EU legal framework governing state aid to the financial sector, as well as by other institutional and possibly fiscal constraints. More specifically, the Bank Recovery and Resolution Directive (BRRD) and the State Aid communications of the European Commission^[7]See Communication from the Commission on the treatment of impaired assets in the Community banking sector (‘Impaired Assets Communication’, 2009/C 72/01) and Communication from the Commission on … Continue reading regulate the participation of governments in AMCs. The complexity of these rules and their interplay is one of the reasons for developing a blueprint for system-wide, government-sponsored AMCs in the EU. Besides clarifying in detail how such AMCs would need to be designed in order to be compatible with the EU legal framework, such a blueprint should identify international best practices and explain how these best practices can be applied in those EU countries that may benefit from setting up an AMC.

The BRRD states that public capital support to banks is allowed, outside of resolution measures, only if a stress test identifies that a bank needs additional capital to ensure its solvency under a so-called adverse scenario (‘precautionary recapitalisation’) and if this capital cannot be fully obtained from private sources. In addition, state aid can only be granted to solvent institutions and it must be approved by the European Commission.
The State Aid communications of the European Commission concern NPL-related measures – AMCs as well as asset insurance schemes – as part of the crisis management toolkit which can be used under certain conditions, in particular^[8]See Medina Cas and Peresa (2016) for a more detailed discussion of the necessary conditions.:

• Transfer prices of NPLs should not exceed their ‘real economic value’;
• The ‘real economic value’ should be assessed through an independent valuation exercise following a methodology that is compliant with the requirements of the European Commission, and;
• Bank capital losses resulting from the transfer of NPLs to an AMC should be shared among equity-holders and subordinated creditors of the concerned banks.

The BRRD and State aid rules further specify that state aid may be provided to banks in two forms: indirectly, as state participation in an AMC, and directly, as a capital injection into a bank. The overall amount of aid is determined by the capital needs identified under the adverse scenario (see Figure 1).
A European blueprint for national AMCs would not involve international risk-sharing among EU or euro area Member States. Fiscal constraints may, however, come into play in some of the EU countries currently facing a high NPL stock. Should the AMC become part of the general government sector, its liabilities may increase, in some cases, already high public debt levels. This may, however, be avoided if the AMC is majority-owned by private parties and the risks related to the underlying assets are not borne by the government.^[9]See Part IV.5 of Eurostat (2016), Manual on Government Deficit and Debt. Implementation of ESA 2010.

4. Key features of an AMC blueprint

This section introduces the key considerations for setting up successful AMCs in Europe, allowing them to maximise recovery values, whilst limiting risks to the state.^[10]A poorly designed AMC may, however, increase the risks to the state. Losses incurred by an AMC may burden the state balance sheet and adversely affect the value of residual NPLs remaining in banks. … Continue reading The main issues to consider in this context are the asset perimeter, the participation perimeter, the asset valuation, the capital and funding structure and, last but certainly not least, the governance of AMCs. The description below is of a cross-country nature, taking the interconnectedness between the various issues, international best practice and the legal constraints described above into account. Obviously there is a need to adapt this ‘blueprint’ to country-specific circumstances as appropriate.

4.1. Asset perimeter

The first consideration relates to the assets to be transferred to the AMC.^[11]Historically, AMCs have often been set up and associated with particular asset classes, such as NAMA in Ireland and Sareb in Spain, arising from specific economy-wide macro-financial developments. Given the overarching objective to maximise asset recovery values, assets transferred to an AMC should be limited to those assets where AMCs have a demonstrated track record in recovering value, such as commercial real estate, large corporate exposures and syndicated exposures.
The scale of asset transfers should strike a balance between the benefits accruing from economies of scale and the risk that the AMC may become overburdened with having to work-out too many assets within a relatively short period of time, in particular if they are insufficiently homogenous.^[12]As part of a comprehensive NPL resolution strategy, an AMC can only be expected to address part of the NPL problem and need not be scaled to the overall stock of NPLs in a given countries banking … Continue reading Moreover, limiting the size of the AMC helps mitigating funding and capitalisation challenges.
Only assets above a pre-determined gross book value threshold should be transferred, to avoid burdening the AMC unduly with many small exposures, which give rise to substantial operational challenges. Finally, it is often very useful to take a debtor-level approach, to ensure that all exposures of the banking system to a (partially) non-performing debtor are transferred to an AMC.^[13]Experience has shown that such a debtor level approach is warranted. A debtor may have an NPL with one bank, but performing loans with another. By taking all of the outstanding debt of a specific … Continue reading

4.2. Participation perimeter

Participation in the AMC should not normally be fully left at the discretion of the concerned banks, as the case for the AMC rests on its achieving a critical mass of assets. Purely voluntary participation may result in inaction, on account of first-mover disadvantages, or cherry-picking of NPLs by participating banks. The authorities should thus introduce incentives to transfer the assets, be it through moral suasion, supervisory (macro- or microprudential) or accounting measures, or by sharing in the AMC’s upside.
Only banks holding significant exposures to the asset class(es) captured in the asset perimeter should participate, whilst level playing field concerns must be satisfied. Objective and transparent criteria, linked to the overall objective of the AMC, should be laid down to identify these banks. Less significant exposures or exposures held by very small banks may be best worked-out by other means.
Non-participating banks may still be willing to contribute equity to the AMC, given that they are likely to reap indirect benefits from its establishment, e.g. a positive impact on asset price developments

4.3. Asset valuation

State aid rules require that a valuation exercise needs to be conducted at the time of the asset transfers, to establish the market value and real economic value of the assets. The valuation process should be run by an independent expert, following a methodology established in agreement with the European Commission and subject to oversight by the authorities.
The valuation process should start once the possible asset and participation perimeter has been determined. Initially, that perimeter is likely to be broader than the final scope of the AMC as some assets may be unsuitable for resolution within the AMC.
The assumptions of the valuation methodology should be realistic and account for all expected cash inflows and outflows associated with the assets. In particular, the legal, tax, maintenance, and servicing costs should be included in the estimates of the real economic value. In line with state aid rules, the government should be appropriately remunerated for taking on the risk that ultimate recoveries may fall short of estimated real economic value.^[14]In practice, this is captured by a risk premium included in the discount rates. For example, NAMA used the Irish sovereign yield curve with a mark-up of 170 basis points to discount future cash flows … Continue reading
The valuation should include a viability test on the underlying assets and debtors. Such a test would identify assets that need to be liquidated rather than transferred to the AMC for recovery, and would inform the future course of action for individual assets.

4.4. Capital structure

The capital structure of the AMC should ensure that the AMC remains unconsolidated with the general government sector. This is particularly important for Member States with limited fiscal space. A public-private partnership model, with the majority private equity stake provided by private sector participants has achieved this goal in the case of NAMA and SAREB. Whilst government still puts up equity, its stake would remain below 50%, thus not giving the government effective control over the operations of the AMC.
The total capital level should be calibrated to ensure that the equity layer is sufficient to absorb unexpected losses on the AMC’s assets, so that the majority of risks and rewards from the resolution of the assets would not rest with the government. In any case, the equity requirements when setting up an AMC should remain limited, provided that it’s overall size remains constrained by the appropriate asset and participation perimeters mentioned above. Moreover, given that asset transfers have to be done at real economic value, AMCs should not make major unexpected losses during their lifetime.

4.5. Funding structure

The funding structure of the AMC should minimise costs and liquidity risk. This can be achieved by issuing government-guaranteed senior bonds which can be used as payment-in-kind to purchase NPLs from banks. Senior bonds may be short-dated (one-year), with restrictions on transferability and an implicit roll-over guarantee, to mitigate roll-over risks. With the government guarantee, senior bonds may be structured to meet the eligibility criteria for use in Eurosystem credit operations although the ECB obviously will decide on this on a case-by-case basis.^[15]See in particular requirements for marketable assets, laid down in Articles 62 to 71 of ECB (2015). This may further expand the range of funding options for the banks.
Appropriate controls should be put in place to ensure that the AMC redeems senior debt according to schedule, rather than building cash reserves or diverting resources to other interests.

4.6. Governance and operations

Strong and sound governance is a critical success factor for an AMC. It should strike the right balance between the business flexibility needed to maximise recoveries, and constraints preventing diversion from the core mandate of the AMC.
The AMC should be established on the basis of legislation that lays down its objectives and decision-making bodies as well as its transparency and accountability rules. Historical experience suggests that AMCs should be free from political interference and budgetary pressures. In particular, they should not be established as a government agency or part of the civil service. Yet, public authorities should exercise oversight over some aspects of AMC operations, in particular with respect to compliance with its mandate and applicable regulations, whilst not interfering with daily business decisions.
The AMC should have a clear primary mandate to maximise the recovery values of NPLs on a commercial basis. It should be permitted to use any relevant legal tool or work-out strategy to achieve its goals, regardless of political or vested interests. Its lifetime should be finite and defined at the outset, alongside a credible business plan and measurable performance goals.
Risks that AMCs are diverted from their core mandate must be carefully controlled. For instance, political interests may attempt to use the AMC as a source of financing for state projects or as part of the social safety net. These risks can be partly mitigated through careful asset selection (for example, avoiding the transfer of loans to state-owned enterprises or residential mortgages), and through restrictions on operations of the AMC provided for in the legislation (for example, the AMC should not hold a banking license).
The operational overheads of the AMC should remain light. Wherever available, the AMC should be allowed to outsource services such as property management, legal services or collections to independent providers at market prices. Where servicing capacity is not available in the market, governments should implement necessary reforms to facilitate the build-up of the servicing industry.

5. Other elements of a comprehensive approach to NPL resolution

Besides robust supervisory oversight of AMCs, three additional approaches should be considered when designing comprehensive, multifaceted approaches to deal with large systemic NPL stocks; NPL transaction platforms, co-investment schemes and liquidation vehicles (see Figure 2).^[16]Internal work out of NPLs by the originating bank will always form part of NPL resolution. It requires banks to maintain or build necessary expertise. At the same time they may recover more value for … Continue reading
An NPL transaction platform has the potential to deliver some of the benefits of an AMC whilst avoiding most of the costs. The platform may act as a central hub for NPL sales by being a central repository for NPL data from participating banks. Data must be standardised and of sufficient quality for investor due diligence purposes. The platform should be enabled with uniform, standardised legal, documentation and transactional services. Ideally, the platform should be enabled to sell assets, subject to guidance, from participating banks.

The advantages of such a platform are significant and the platform is likely to have a discernible impact on market prices for NPLs by reducing information asymmetries. As the value of NPLs would become clearer, the rate of return expected by NPL investors would be expected to decline. Furthermore, investor costs, including, for example, shoe-leather costs can be reduced, through standardisation of data and processes, and the consolidation of NPL sales in one agency. Participation in such a platform, which may be encouraged by supervisors, may induce banks to resolve data problems. This could help resolve, in particular, the least transparent and most difficult-to-value assets, such as corporate and SME loans. A further impetus to prices may arise from transparency around completed NPL transactions. The establishment of the platform should also be an impetus for necessary services to be established / increased, for example, in relation to data quality improvements, transaction services, loan servicing, etc. The platform may even have a role in centralising and coordinating these activities. Operationally, a number of challenges around, for example, data confidentiality, would have to be overcome. At the same time, a precedent for such a platform already exists in the EU, in fact with a rather similar rationale.^[17]The ECB led an initiative to improve transparency in ABS markets by requiring loan-by-loan information to be made available and accessible to market participants and to facilitate the risk assessment … Continue reading
Fell et al. (2017) make the case for appropriately structured co-investment instruments, where the state co-invests, at market conditions, with NPL investors. Having the capacity to address information asymmetries and incentivise states to implement necessary structural reforms, this may, in turn, partially address wide bid-ask spreads. Through risk-sharing and by reducing the cost of carry, such instruments may enable NPL transactions to take place which might otherwise not have closed, in turn having the potential to increase the price that investors are willing to pay for NPLs. Co-investment structures are particularly effective in the context of securitisation, considering the significant advantages that securitisation has over direct sale, as a NPL resolution tool.
Finally, given the scale of the NPL problem and the elapsed time since some NPLs became impaired, it seems plausible that some loans, extended to SMEs as well as households, have little recovery value beyond the collateral. Given the time and costs of recovery, and the potential for some collateral to be of limited re-sale value, orderly liquidations may be required. Banks – as well as AMCs – are not typically well placed to take on this role. There may hence be a case for a public entity specialised in liquidating loans that have no or very little recovery upside.

6. Conclusions

The high stock of NPLs in the European Union calls for urgent policy action. Although significant and necessary progress has been made by microprudential supervisors in improving NPL measurement and management by banks, this is unlikely to be sufficient on its’ own. This article has outlined the broad range of NPL resolution options available to banks and policymakers, as well as some desirable extensions of the existing toolkit.
In particular, system-wide national AMCs may contribute to a speedy reduction of large, systemic NPL stocks in Europe. We see value in developing a European blueprint for national AMCs that clarifies how such AMCs can be established in full respect of the EU legal framework and drawing on international best practices. Appropriately designed, AMCs may offer substantial benefits and provide an important complement to more standard NPL resolution options such as internal work-out and direct NPL sales. Other tools which should be developed to allow a more comprehensive yet country-specific, bespoke approach to dealing with systemic NPL problems include an NPL transaction platform, co-investment schemes and liquidation vehicles.
It is very important to keep in mind, though, that all of these tools can only be successful if they are supported by appropriate legal and administrative framework conditions that facilitate debt enforcement and access to collateral, and by sound macrofinancial policies which help to promote economic recovery.

---

References

Aiyar, S., Bergthaler, W., Garrido, J., Ilyina, A., Jobst, A., Kang, K., Kovtun, D., Liu, Y., Monaghan, D., and Moretti, M. (2015). A strategy for resolving Europe’s problem loans. Staff Discussion Note SDN/15/19, International Monetary Fund.

Balgova, M., Nies, M., and Plekhanov, A. (2016). The economic impact of reducing nonperforming loans. Working Paper 193, European Bank for Reconstruction and Development.

Cerruti, R., and Neyens, C. (2016). Public asset management companies: a toolkit. World Bank Studies; Washington, DC: World Bank.

Constâncio, V. (2017). Resolving Europe’s NPL burden: challenges and benefits, keynote speech at “Tackling Europe’s non-performing loans crisis: restructuring debt, reviving growth”, Brussels, 3 February.

ECB (2015). Guideline (EU) 2015/510 of the European Central Bank of 19 December 2014 on the implementation of the Eurosystem monetary policy framework (ECB/2014/60) (recast).

ECB (2017). Guidance to banks on non-performing loans. ECB Banking Supervision, March 2017

Fell, J., Grodzicki, M., Martin, R., and O’Brien, E. (2016). Addressing market failures in the resolution of non-performing loans in the euro area. Financial Stability Review, ECB, November, 134-146.

Fell, J., Moldovan, C., and O’Brien, E. (2017). Resolving large stocks of NPLs: a role for securitisation and other financial structures? Special Feature in Financial Stability Review, ECB, May 2017.

Goretti, M., and Souto, M. (2013). Macro-financial implications of corporate (de)leveraging in the euro area periphery. Working Paper WP/13/154, International Monetary Fund.

He, D. (2004). The role of KAMCO in resolving nonperforming loans in the Republic of Korea. Working Paper WP/04/172, International Monetary Fund.

Jonung, L. (2009). The Swedish model for resolving the banking crisis of 1991-93. Seven reasons why it was successful. Economic Papers 360, European Commission.

Medina Cas, S., and Peresa, I. (2016). What makes a good ‘bad bank’? The Irish, Spanish and German experience. European Economy Discussion Paper 036, European Commission.

Nkusu, M. (2011). Non-performing loans and macrofinancial vulnerabilities in advanced economies. Working Paper WP/11/161, International Monetary Fund.