Response to discussion on simple standard and transparent securitisations
Go back
In fact, in a “tranched cover”, credit protection is bought directly from investors, in the form of a financial guarantee backed by cash collateral or personal (direct/unfunded) guarantees as credit enhancement, without issuing notes and tranching the underlying portfolio using the Supervisory Formula Approach (SFA). This structure can also cover the risk of a new origination portfolio, hence facilitating new securitizations which in turns would provide substantial new lending to the real economy. In this case, the underlying portfolio is built under certain assumptions (rating distribution, geographical area, sector concentration) pre-agreed with the “protection provider”, but under the bank’s usual credit processes.
This instrument ensures some advantages and a certain degree of flexibility (e.g. no need to create a Securitisation Special Purpose Entity (SSPE), no need to sell the underlying loans, no need to issue notes). Moreover, in the new regulatory environment, where capital costs have substantially increased to supervised entities in view of the additional regulatory requirements, any financial instrument is very “sensitive” to regulatory changes. A regulation which includes or excludes certain forms of synthetic securitizations from the High Quality Securitisation (HQS) definition has therefore the potential for hampering the economic viability and incentives for lenders to deploy new lending and for investors to share aggregated credit risk, with a positive impact on financial stability.
In a market and regulatory constrained environment, UniCredit synthetic securitizations, in the form of “tranched cover”, were able to attract and broaden a diverse investors’ base, ranging from national/supranational guarantee funds and mutualistic entities (so called “confidi”) to specialized institutional investors.
Based on these considerations and the positive experience with the “tranched cover” technique, UniCredit strongly believes that synthetic securitizations should not a priori be excluded from the HQS framework for simple, standard and transparent securitizations. Rather, UniCredit suggests to identify an HQS framework specific for synthetic securitization which:
- meets the EBA Guidelines on Significant Credit Risk Transfer relating to Articles 243 and Article 244 of Regulation 575/2013 and
- adequately addresses the objectives of the EBA and the EU Commission on simplicity, standardisation and transparency.
Therefore it is necessary to identify a specific HQS framework for synthetic securitization.
Here below are listed the main reservations (Note 1) which have been raised by the European Commission on synthetic securitization. For each reservation, UniCredit provides its viewpoint.
1. Complex
Synthetic securitizations are often structured without the need of an SPV, using simple contractual framework. For example, “tranched cover” transactions are characterized by peculiar features that ensure simplicity of transaction documentation which may consist of one single contract constituted by no more than 20 pages. Correspondingly, the relevant underlying assets may consist of loan assets such as SME loans which are critically important to the European economy. Synthetic securitizations are also characterized by the absence of an active portfolio management on a discretionary basis and/or a “cherry picking” practice.
Generally, transaction complexity is unrelated to the fact that a transaction is synthetic. Synthetic transactions represent a means of transferring risk and, in that way, are no different than other forms of securitisation.
2. Less transparent
UniCredit believes that notwithstanding their private and bespoke nature, synthetic securitizations can still be considered transparent for the following reasons:
o Synthetic securitization process typically includes an extensive due diligence during which the bank provides all engaged counterparties, who have previously signed Non-Disclosure Agreements (NDA), with necessary information not only regarding the underlying loan portfolio but also detailed information regarding its credit policies and disclosure regarding credit risk modelling.
o Ongoing transparency is also ensured in the form of detailed investor reports which containing all information about portfolio performance.
o If investors/noteholders decide to trade their synthetic securitization exposures on the secondary market, UniCredit will, subject to similar NDA commitments, provide the same data package to the new eligible investor as that available to the original investor.
3. Investor losses difficult to understand
This is untrue because the transaction documentation clearly specifies the cascading of investor losses and more recent transactions avoid such mechanisms that prevent charging the losses to the most junior tranches. In fact, the market and the current legislative framework discourage the creation of very complex transactions.
4. Recourse cannot be granted
This is untrue because it is possible to have the recourse on assets in case of default of the originator, if contractually agreed.
5. Additional counterparty risk
This is untrue because in some cases, from an originator point of view, the counterparty risk is minimal: guarantees provided by protection sellers can be called quickly upon the occurrence of a credit event, considering that they are typically fully cash-collateralized and therefore bear no counterparty risk for the bank/originator. In addition, as noted above, security interest and segregation may in fact be features of synthetic transactions so originator risk/exposure may be no different in a synthetic transaction than more conventional structures.
6. Hampering of investor’s rights
Investor’s rights are safeguarded and pre-agreed between the parties to a synthetic transaction; typically an external auditor is appointed for the benefit of the investors. Investor’s rights are specifically defined in transaction documents and structures, regardless whether they be synthetic or otherwise.
7. Additional risk modelling
This is untrue. Synthetic securitizations can be, and generally are, structured in a simple and transparent way. They are often structured as private transactions in the form of “tranched cover” e.g. whereby credit protection is bought directly to investors as a financial guarantee backed by cash collateral or personal guarantee as credit enhancement without issuing notes. This structure ensures some advantages and a certain degree of convenience in comparison to traditional securitizations (e.g. no need to create a SSPE, no need to sell the underlying loans, no need to issue notes).
Modelling requirements are generally defined by the structures and the underlying assets of transactions rather than their synthetic nature.
Based on the above considerations, according to UniCredit, in order to be HQS compliant, synthetic securitisations shall be subject to the following high-level conditions, to be subsequently elaborated into eligibility criteria:
A) Simplicity
- Simple legal framework: HQS Synthetic securitizations shall be structured without the need of an SPV, using simple structures and contractual frameworks (financial guarantee fully funded or unfunded if the protection provider is a Supranational Entity). For example, “tranched cover” transactions are characterized by features that ensure simplicity of transaction documentation which may consist of one single contract constituted by no more than 20 pages;
- Simple alignment of interest: HQS synthetic securitizations shall be subject to the retention rule, under which the originator retains a certain percentage of all tranches according to current regulation (Regulation (EU) No 575/2013 and Bank of Italy, Circular 285);
- Simple cascading of investor losses: HQS synthetic securitizations shall deploy a simple and sequential process of amortization of the tranches. The amortization should affect firstly the senior tranches and after that, when the senior is repaid, the mezzanine and after the junior;
- No dependence on rating agencies: HQS synthetic securitizations originated by banks applying the Advanced Internal Rating Approach (AIRB) may perform tranching using the Supervisory Formula Approach set forth by regulation, therefore, eliminating external dependence on credit ratings;
- No hampering of investors rights and contractual recourse to assets: HQS synthetic securitizations shall ensure that investor rights are safeguarded and pre-agreed between the parties; investors shall be granted i) the right to appoint a verification agent, ii) subrogation rights as well as iii) the possibility to have the recourse on assets in case of default of the originator, if these conditions have been a priori contractually agreed;
- Simple recourse to safe guarantees: HQS synthetic securitizations shall ensure that guarantees are provided by protection sellers which may be activated quickly upon the occurrence of a credit event, which has to be clearly and univocally defined;
- No counterparty risk: HQS synthetic securitizations shall ensure that guarantees provided by protection sellers are fully cash-collateralized and therefore bear no counterparty risk from the perspective of the protection buyer (i.e. the bank). Personal guarantees should be provided by Supranational entities.
B) Standardisation
Since the inception of the synthetic securitisation market more than a decade ago, the structural features have moved from the initial phase of a completely bespoke nature (bilateral negotiations between the parties within the limits set by the European regulatory framework) to a more mature phase, reflecting market reaction to the financial crisis. At present, certain structural aspects are widely accepted by market participants and can be deemed as market standard in accordance with industry best practice.
In order to consolidate this market progress towards standardisation, the transaction documentation of HQS synthetic securitizations shall:
- clearly specify the cascading of investor losses, avoiding mechanisms that prevent charging of losses to most junior tranches;
- clearly define the amortization payment priority, whereby the seniority of the tranches determines the sequential order of payments;
- be written or approved by an external counterparty (i.e. external auditors) in order to manage the potential conflict of interests among all the investors and the originator.
If these conditions are met, qualifying synthetic securitizations can be defined as meeting the “standard” principle of the HQS synthetic securitisation.
C) Transparency
Before the execution of the transaction:
In order to meet the transparency principle of the HQS definition, synthetic securitisations shall include an extensive due diligence during which the bank provides all engaged counterparties, which have previously signed Non-Disclosure Agreement (NDAs), with a pre-defined minimum set of information regarding:
• the underlying loan portfolio
• detailed information regarding its credit policies and
• detailed information regarding credit risk parameters.
Following the execution of a transaction:
HQS synthetic securitisations shall:
• provide detailed investor reports which shall contain all information about portfolio performance.
• provide the same data package to the new eligible investor, if current investors/noteholders decide to trade their synthetic securitization exposures on the secondary market.
In sum, UniCredit experience and above conditions show that it is possible and highly desirable for competent authorities to design a HQS framework for synthetic securitisation. European authorities are encouraged to benefit from the competent and independent advisory from third parties such as the PCS Association which would facilitate the task to define market standards, based on market consensus and best market practices.
In addition, for the sake of completeness, qualifying ABCP transactions should also be included in the HQS framework because there are well developed and standardised structures in the market. UniCredit stands ready to share with the EBA staff its view in relation to HQS eligibility criteria for ABCPs.
Note (1): The EU Commission expressed objections to the inclusion of synthetic securitisation in the High Quality Liquid Assets definition for the Liquidity Coverage Ratio (see FAQ for delegated acts http://europa.eu/rapid/press-release_MEMO-14-579_en.htm?locale=en ). In section 26.1.2. , it is stated that ”re-securitisations are explicitly excluded [from high quality liquid assets], as they are [1] typically complex and [2] less transparent structures, where the [3] cascading of investor losses is very difficult to understand due to re-tranching. The same goes for synthetic securitisations, where the underlying exposures are not transferred to the special purpose vehicle. Instead, the transfer of risk is achieved by the use of credit derivatives or guarantees, while the exposures being securitised remain with the originator. The transfer of the assets to be securitised ensures that securitisation investors have recourse to those assets should the Securitisation Special Purpose Entity (SSPE) not fulfil its payment obligations. [4] Such recourse cannot be granted in synthetic transactions, due to the fact that only the credit risk associated with the underlying assets, rather than the ownership of such assets, is transferred to the SSPE. Such a structure also [5] adds counterparty risk on derivatives or guarantees, and [6] hampers investors' rights to the proceeds of the underlying exposures. In addition, most synthetic structures add to the complexity of the securitisation in terms of [7] risk modelling”.
In general, UniCredit deems that defining “standard” securitisations through a jurisdictional framework is a difficult exercise, but the experience of the common European standards and best market practices developed by the PCS initiative is setting a direction worth considering by competent authorities.
The PCS initiative has taught market participants that the imposition of standard criteria in the securitisation market needs to take into consideration both the potential benefits as well as the potential unintended consequences, including possible limitations in terms of market development in those jurisdictions excluded from the PCS scope. For example, jurisdictions or asset classes originating in certain jurisdictions excluded from the definition of “standard” would necessarily face more limited investor acceptance even though market dynamics and developments may, in the future, no longer validate the initial definition of “standard” transactions.
Considering the remarkable impact that market dynamics and developments may have as well as the wide variety of factors which need to be duly taken into account when defining standardisation criteria, UniCredit cautions against the strict adoption of a permanent and non-flexible jurisdiction-approach used as a benchmark for defining standardisation. UniCredit suggests that a more flexible approach be considered. Such an approach, in UniCredit view, would need to imply periodic updates of what the “regulatory standards” are, including limits - if any - imposed on the type of jurisdictions.
Here below UniCredit provides some considerations outlining the wide variety of the factors governing the eligibility of jurisdictions:
Empirical evidence about the country of location. Securitizations in EU and the European Economic Area (EEA) are usually single-country transactions, although a number of deals are backed by assets from more than one jurisdiction. UniCredit has seen true sale publicly placed single-country transactions with underlying collateral from Austria, Belgium, Bulgaria, the Czech Republic, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Latvia, the Netherlands, Norway, Poland, Portugal, Spain, Sweden and the United Kingdom. To our knowledge, transactions backed by assets from the following countries have not been placed so far with investors: Croatia, Cyprus, Estonia, Hungary, Iceland, Liechtenstein, Lithuania, Luxembourg, Malta, Romania, Slovakia and Slovenia. Nevertheless, they should not be excluded from the EBA perimeter. Other European securitizations from non-EEA countries that were placed with investors included loans originated in Russia, Switzerland, Turkey and the Ukraine. Hence, we note that there has been regular issuance from a number of non-EEA countries (e.g. Russia, although not placed in Europe and Switzerland) while many EEA jurisdictions have yet to make their securitization debut. Collateral performance across non-EEA securitizations has also diverged notably.
Main factors governing the eligibility of jurisdictions:
- Rather than applying limits to the type of jurisdiction (such as EEA only), UniCredit suggests policymakers to take into account the characteristics of the local securitization markets. In UniCredit’s view, aspects including the availability of historical data, statistical predictability of asset performance, local underwriting standards, loan documentation standards and local securitization laws should govern the eligibility of a jurisdiction for the purpose of standardisation rather than following a fixed jurisdiction-based approach. UniCredit suggests following a similar approach in relation to the use of mixed pools in European securitizations. Historically, transactions backed by assets from more than one jurisdiction have been in most cases CMBS but they can include a variety of asset classes. In contrast, ABS with mixed geographical pools have been infrequent and largely limited to Auto and SME ABS, as well as a very small number of RMBS. Mainly mixed pools backed by assets from jurisdictions with well-established securitization laws have been seen in the market. Public data indicate that underlying assets in multi-jurisdiction transactions have performed broadly in line with those assets underlying transactions from a single jurisdiction.
- Another important factor is the currency denomination of the underlying loans with respect to their credit performance. Transactions backed by local currency exposures (e.g. a securitized pool of Norwegian auto loans issued in NOK that underlies a EUR-denominated Norwegian auto ABS tranche) are typically hedged using cross currency swaps to avoid cash flow mismatches. Different is the case of local loans originated in a foreign currency (FX loans), which are subject to currency fluctuations. The rapid depreciation in the local currency versus the foreign exchange loan currency would expose local borrowers to additional risk. For example, to a large extent, mortgage and loan origination in countries such as Russia and the Ukraine historically took place in foreign currency, typically USD. As a result of the RUB depreciation, some Russian ABS transactions backed by USD were exposed to higher defaults as the borrowers’ monthly installments made up a larger share of their wages. This situation was replicated in other jurisdictions as well.
- True sale and segregation in legislation. The majority of securitization transactions incorporate the “true sale” concept that isolates the securitized assets from the insolvency estate of the originator. When it comes to the acquisition process of the SSPE, local laws governing securitization transactions should be in place to ensure a true sale of assets and their segregation from an originator’s estate. UniCredit is aware that the legal framework can differ to a high degree across jurisdictions with respect to the definition of bankruptcy remoteness.
- National laws. In addition, in well-established securitization jurisdictions, the laws governing securitization transactions protect the SSPE’s title to the purchased receivables in the event of fraudulent sale or conveyance. In such cases, investors would normally retain recourse against the seller of the property rather than the SSPE. Hence, limits with respect to jurisdictions that lack a comprehensive legal framework governing such circumstances could be considered but would need to be regularly revised, since legal frameworks change overtime and are tested via local courts.
Regarding the specific criteria of the Pillar II (from 7 to 14), in particular criteria no 8, UniCredit is of the view that greater definitional clarity is needed with regard to “appropriately mitigated” and “genuine hedging.” That said, a carve out should be provided for a structure that: a) has a credit enhancement that already covers the interest rate risk or b) has a natural hedge (i.e. implicit interest rate risk coverage) between the asset (sold portfolio) and the SPV/issuer liability (the note placed to the market). This is for example the case of a portfolio with loans bearing fixed rate payments against notes with fixed rate portfolio or of a floating rate portfolio with loans bearing cash-flows linked to the same reference index of the notes.
UniCredit fundamentally supports disclosure of all relevant information to investors on a timely basis. This is critical to the functioning of any successful securities marketplace. Investors must be provided with the information necessary to make an informed investment decision. As a result, disclosure documentation must be provided to investors prior to issuance and this is the practice across the global securities markets. Beyond this, UniCredit is not supportive of further regulatory requirements regarding the timing of disclosure prior to issuance.
As for securitisation, the adherence with market transparency standards should be certificated at issuance date and guaranteed by an impartial third party constituted with this specific mission. It is important in fact that a periodic quality check of the transaction features within the qualifying securitization framework both for regulatory purposes and secondary market disclosure is run to the benefit of investors.
The PCS initiative has promoted the PCS label certifying standardisation, data/information accessibility and comparability. On this respect, the PCS Secretariat could represent a relevant interlocutor with the authority also in the future, with the aim to contribute to help investors to have, along similar lines as the ECB eligibility process, a quality check of the transaction features within the qualifying securitization framework both for regulatory purposes and secondary market disclosure.
The goal should be to provide timely and meaningful information to investors who can then assess the underlying asset quality and structural features. Belated and incomplete documentation and deal information that is provided at short notice adds an additional analytical burden for investors, which could initially discourage them from investing in the transaction altogether. This relative lack of information will be reflected in a higher return required by investors for the additional risk. Therefore, making deal documentation available to investors in a timely manner is important for issuers wishing to attain the lowest possible borrowing costs. That said, however, a specific timetable for the required provision of disclosure information to investors is not necessarily required and may add unnecessary burden to issuers and investors, particularly given the clear economic incentives that already exist for issuers to be as complete and timely as possible in their disclosure. For example, within accepted issuance practice, sophisticated investors may either require less formal disclosure or may perform their own due diligence processes which may preclude the need for fixed disclosure schedules. In addition, for frequent issuers or issuers that are known in the market, the level of disclosure and timing may be relatively less important than less frequent issuers, new issuers or relatively complex structures involving non-traditional assets.
As a result, any regulatory requirement regarding disclosure timing should be flexible enough to cover the range of anticipated issuances and should consider the already existing economic incentives imposed on issuers by the marketplace.
Provision of on-going data
The emphasis should be on the relative appropriateness of the disclosure, content and the responsibilities and liabilities imposed on those involved in the securities placement/distribution process rather than on the ability to provide on-going data. Having said that, it should be nevertheless noted, however, that the provision of on-going data is also important for the successful maintenance of a liquid ABS marketplace.
To this end, UniCredit has been involved from the beginning in the successful establishment of the European Datawarehouse, supported by the ECB. Prior to its establishment, collateral-level data were often either entirely unavailable or inconsistent from deal to deal. Loan-level data were often not provided by European issuers due to privacy and competitive concerns. While loan transparency before issuance does not necessarily translate into better investor participation, a lot of information was lost when data were aggregated. Therefore, for loan/collateral level information, the relevance of data may depend on its immediacy but more importantly on the ability to access and compare relevant on-going data/information.
Additional considerations
In addition to our view in reply to Question 6, UniCredit has some specific considerations with reference to the specific criteria of the Pillar III (from 15 to 22):
- criteria no 18 : the proper pricing of each transaction should be evaluated by each investor on its own portfolio assumptions (default, arrears, prepayments forecast). Accordingly, no cash-flow model should be provided by the originator or the sponsor in order to avoid, as happened in the past, that they can unduly influence investor’s decision and securitization’s market. The originator or the sponsor can only provide market participants with all necessary information that allow to build in-house cash-flow models or resort to external service providers;
- criteria no 20 : in order to avoid excluding assets which have a short seasoning and that are crucial for the current economy development (for example renewable energy etc.), UniCredit suggests that the EBA investigates the opportunity to reduce the requirement for historical performance data from 5 to max 3 years.
All of the criteria in Pillar III should also consider the implications for short-term assets and securitisations utilizing such assets (such as ABCP). UniCredit stands ready to share its view in relation to transparency standards for ABCPs with the EBA.
The granularity concept has to be adjusted to reflect the specificity of the underlying asset classes. An adjustment of the granularity concept is considered crucial in order not to hamper the possibility to set-up securitizations on SMEs loans or on corporate portfolios where the EBA cannot apply a one-size fits all approach. In particular, the proposed single debtor concentration limit of 1% seems punitive when applied to pools of SME Leasing and Corporate exposures that have been historically structured with higher levels of concentrations. This is even more true when considering that the limit is referring to a group of connected clients.
In view of these considerations and taking into account the difficulties of SMEs to fund themselves in the market UniCredit suggests to introduce a limit of 5% in term of concentration / global exposures towards a group of connected clients when considering SME, Leasing and Corporate counterparties. The concentration limit of 1% shall be applicable to pools of retails exposures.
Concerning the credit risk criteria, UniCredit deems there is a fundamental misconception of the framework. Criterion C, point ii defines the eligibility criteria for underlying exposures in terms of maximum individual risk weight (based on the Standardised Approach that has, in UniCredit view, a low-risk sensitivity). This is a major departure from best market practices.
While it is understandable that prudential regulation prevents credit risk excesses stemming from the underlying assets during the underwriting phase, regulatory limits should not be primarily imposed on collateral exposures where the risk is evaluated through a tranching of the liabilities.
It is also recalled that the Italian legislation does not foresee the use of the individual risk weight levels as an eligibility criteria for asset selection also to avoid cherry picking, a practice to be discouraged as indicated also in this document.
UniCredit suggests to the EBA to consider alternatives in terms of portfolio measures (e.g. max weighted average).
As expressed above, to avoid any potential adverse market consequence against specific type of securitisations, all securitisation techniques, though with the appropriate differentiations, have to be considered for the purpose of being recognized as simple standard and transparent, not only true sale securitisations. This is particularly the case for synthetic securitisations which continue to represent an important risk-sharing and credit risk mitigation instrument, crucial for the support to the economic recovery. As recalled in reply 2, in the current regulatory environment, where capital costs have substantially increased due to a multiplicity of requests to supervised entities, any financial instrument is very sensitive to regulatory changes. A regulation, which includes or excludes certain forms of synthetic securitisations or asset backed commercial papers (ABCP) from the HQS definition, has therefore the power to foster or vice versa hamper its economic viability, the related incentives for lenders to deploy new lending and for investors to share aggregated credit risk. That’s why the overall impact on the economy and on financial stability need to be carefully assessed by the regulator.
In addition, the qualifying framework will also contribute to the creation of the incentives for the market to increase the standards towards the best market practices.
The proposal to have differentiated capital requirements for qualifying and non-qualifying securitizations will allow for a more symmetric treatment of securitizations. As recalled by the EBA and other international institutions such as the IMF, some type of securitizations are treated asymmetrically vis-à-vis other asset classes (e.g. covered bonds), particularly in the EU.
Therefore, in UniCredit view, the qualifying securitization label must entail a distinctive more favorable RWA treatment compared to the non-qualifying securitization transactions. This kind of differentiation in capital requirement should be incorporated in the BCBS proposal of revisions to the securitization framework (currently ongoing) and should be referred to in all the available approaches:
- Internal Rating Based Approach;
- External Rating Based Approach;
- Standardized Approach.
More specifically, UniCredit would like to share some suggestions on some aspects related to the new approach proposed by the Basel Committee on Banking Supervision (“BCBS”) in its second consultative document issued on December 2013 and finalized in December 2014. This may a have a bearing on the EBA proposal on: i) the risk weight floor, ii) the maturity definition and iii) the use of IRB.
i) The EBA calibration of capital requirements for HQS should consider a different and more favorable risk weight floor (see also recommendation 6). The target floor could therefore be defined as follows:
a. For qualifying securitizations (including ABCPs) the risk weight floor should be no more than 10% in order to make it aligned with covered bonds requirements;
b. For non-qualifying securitization the risk weight floor should be 15% as already provided in the BCBS document on Revisions to Securitization framework.
ii) The definition of maturity proposed by BCBS is very conservative (it normally corresponds to the legal maturity). Given the high impact of maturity in the capital requirements calculation, the proposed approach leads to a significant overestimation of risk weights for many transactions; it is therefore proposed to use for all securitizations a different definition of maturity which takes into account the maturity distribution and the weighted average life of the underlying assets.
iii) reduce the reliance on external credit ratings by not discouraging the use of the IRB method under certain conditions. In UniCredit opinion, the BCBS requirements to use the IRB method especially for investment positions are too strict. In fact, especially for European banks working as sponsors or investors, the issue of data availability is relevant. UniCredit retains that it is important for all securitizations to ease the data requirements for the IRB method. The definition of more simple and homogenous criteria will help the IRB method to become a real option for European banks.
Hence, UniCredit welcomes the EBA intention to introduce “qualifying” securitization with a specific regulatory treatment that would lead to a more reasonable calibration across tranches. Also, with reference to recent regulatory developments, UniCredit appreciates the BCBS consultation launched last December on the “Criteria for identifying simple, transparent and comparable securitisations”. UniCredit encourages the EBA and the European competent authorities to work at international level in order to agree on a specific regulatory framework for HQS globally.
Question 2: Should synthetic securitisations be excluded from the framework for simple standard and transparent securitisations? If not, under which conditions/criteria could they be considered simple standard and transparent?
In UniCredit experience, the technique of synthetic securitisation has taken various forms but has served as a valuable means of managing both risk and capital. In Italy, this is evidenced through the positive experience with the so called “tranched cover” which represents an important risk-sharing and credit risk mitigation instrument.In fact, in a “tranched cover”, credit protection is bought directly from investors, in the form of a financial guarantee backed by cash collateral or personal (direct/unfunded) guarantees as credit enhancement, without issuing notes and tranching the underlying portfolio using the Supervisory Formula Approach (SFA). This structure can also cover the risk of a new origination portfolio, hence facilitating new securitizations which in turns would provide substantial new lending to the real economy. In this case, the underlying portfolio is built under certain assumptions (rating distribution, geographical area, sector concentration) pre-agreed with the “protection provider”, but under the bank’s usual credit processes.
This instrument ensures some advantages and a certain degree of flexibility (e.g. no need to create a Securitisation Special Purpose Entity (SSPE), no need to sell the underlying loans, no need to issue notes). Moreover, in the new regulatory environment, where capital costs have substantially increased to supervised entities in view of the additional regulatory requirements, any financial instrument is very “sensitive” to regulatory changes. A regulation which includes or excludes certain forms of synthetic securitizations from the High Quality Securitisation (HQS) definition has therefore the potential for hampering the economic viability and incentives for lenders to deploy new lending and for investors to share aggregated credit risk, with a positive impact on financial stability.
In a market and regulatory constrained environment, UniCredit synthetic securitizations, in the form of “tranched cover”, were able to attract and broaden a diverse investors’ base, ranging from national/supranational guarantee funds and mutualistic entities (so called “confidi”) to specialized institutional investors.
Based on these considerations and the positive experience with the “tranched cover” technique, UniCredit strongly believes that synthetic securitizations should not a priori be excluded from the HQS framework for simple, standard and transparent securitizations. Rather, UniCredit suggests to identify an HQS framework specific for synthetic securitization which:
- meets the EBA Guidelines on Significant Credit Risk Transfer relating to Articles 243 and Article 244 of Regulation 575/2013 and
- adequately addresses the objectives of the EBA and the EU Commission on simplicity, standardisation and transparency.
Therefore it is necessary to identify a specific HQS framework for synthetic securitization.
Here below are listed the main reservations (Note 1) which have been raised by the European Commission on synthetic securitization. For each reservation, UniCredit provides its viewpoint.
1. Complex
Synthetic securitizations are often structured without the need of an SPV, using simple contractual framework. For example, “tranched cover” transactions are characterized by peculiar features that ensure simplicity of transaction documentation which may consist of one single contract constituted by no more than 20 pages. Correspondingly, the relevant underlying assets may consist of loan assets such as SME loans which are critically important to the European economy. Synthetic securitizations are also characterized by the absence of an active portfolio management on a discretionary basis and/or a “cherry picking” practice.
Generally, transaction complexity is unrelated to the fact that a transaction is synthetic. Synthetic transactions represent a means of transferring risk and, in that way, are no different than other forms of securitisation.
2. Less transparent
UniCredit believes that notwithstanding their private and bespoke nature, synthetic securitizations can still be considered transparent for the following reasons:
o Synthetic securitization process typically includes an extensive due diligence during which the bank provides all engaged counterparties, who have previously signed Non-Disclosure Agreements (NDA), with necessary information not only regarding the underlying loan portfolio but also detailed information regarding its credit policies and disclosure regarding credit risk modelling.
o Ongoing transparency is also ensured in the form of detailed investor reports which containing all information about portfolio performance.
o If investors/noteholders decide to trade their synthetic securitization exposures on the secondary market, UniCredit will, subject to similar NDA commitments, provide the same data package to the new eligible investor as that available to the original investor.
3. Investor losses difficult to understand
This is untrue because the transaction documentation clearly specifies the cascading of investor losses and more recent transactions avoid such mechanisms that prevent charging the losses to the most junior tranches. In fact, the market and the current legislative framework discourage the creation of very complex transactions.
4. Recourse cannot be granted
This is untrue because it is possible to have the recourse on assets in case of default of the originator, if contractually agreed.
5. Additional counterparty risk
This is untrue because in some cases, from an originator point of view, the counterparty risk is minimal: guarantees provided by protection sellers can be called quickly upon the occurrence of a credit event, considering that they are typically fully cash-collateralized and therefore bear no counterparty risk for the bank/originator. In addition, as noted above, security interest and segregation may in fact be features of synthetic transactions so originator risk/exposure may be no different in a synthetic transaction than more conventional structures.
6. Hampering of investor’s rights
Investor’s rights are safeguarded and pre-agreed between the parties to a synthetic transaction; typically an external auditor is appointed for the benefit of the investors. Investor’s rights are specifically defined in transaction documents and structures, regardless whether they be synthetic or otherwise.
7. Additional risk modelling
This is untrue. Synthetic securitizations can be, and generally are, structured in a simple and transparent way. They are often structured as private transactions in the form of “tranched cover” e.g. whereby credit protection is bought directly to investors as a financial guarantee backed by cash collateral or personal guarantee as credit enhancement without issuing notes. This structure ensures some advantages and a certain degree of convenience in comparison to traditional securitizations (e.g. no need to create a SSPE, no need to sell the underlying loans, no need to issue notes).
Modelling requirements are generally defined by the structures and the underlying assets of transactions rather than their synthetic nature.
Based on the above considerations, according to UniCredit, in order to be HQS compliant, synthetic securitisations shall be subject to the following high-level conditions, to be subsequently elaborated into eligibility criteria:
A) Simplicity
- Simple legal framework: HQS Synthetic securitizations shall be structured without the need of an SPV, using simple structures and contractual frameworks (financial guarantee fully funded or unfunded if the protection provider is a Supranational Entity). For example, “tranched cover” transactions are characterized by features that ensure simplicity of transaction documentation which may consist of one single contract constituted by no more than 20 pages;
- Simple alignment of interest: HQS synthetic securitizations shall be subject to the retention rule, under which the originator retains a certain percentage of all tranches according to current regulation (Regulation (EU) No 575/2013 and Bank of Italy, Circular 285);
- Simple cascading of investor losses: HQS synthetic securitizations shall deploy a simple and sequential process of amortization of the tranches. The amortization should affect firstly the senior tranches and after that, when the senior is repaid, the mezzanine and after the junior;
- No dependence on rating agencies: HQS synthetic securitizations originated by banks applying the Advanced Internal Rating Approach (AIRB) may perform tranching using the Supervisory Formula Approach set forth by regulation, therefore, eliminating external dependence on credit ratings;
- No hampering of investors rights and contractual recourse to assets: HQS synthetic securitizations shall ensure that investor rights are safeguarded and pre-agreed between the parties; investors shall be granted i) the right to appoint a verification agent, ii) subrogation rights as well as iii) the possibility to have the recourse on assets in case of default of the originator, if these conditions have been a priori contractually agreed;
- Simple recourse to safe guarantees: HQS synthetic securitizations shall ensure that guarantees are provided by protection sellers which may be activated quickly upon the occurrence of a credit event, which has to be clearly and univocally defined;
- No counterparty risk: HQS synthetic securitizations shall ensure that guarantees provided by protection sellers are fully cash-collateralized and therefore bear no counterparty risk from the perspective of the protection buyer (i.e. the bank). Personal guarantees should be provided by Supranational entities.
B) Standardisation
Since the inception of the synthetic securitisation market more than a decade ago, the structural features have moved from the initial phase of a completely bespoke nature (bilateral negotiations between the parties within the limits set by the European regulatory framework) to a more mature phase, reflecting market reaction to the financial crisis. At present, certain structural aspects are widely accepted by market participants and can be deemed as market standard in accordance with industry best practice.
In order to consolidate this market progress towards standardisation, the transaction documentation of HQS synthetic securitizations shall:
- clearly specify the cascading of investor losses, avoiding mechanisms that prevent charging of losses to most junior tranches;
- clearly define the amortization payment priority, whereby the seniority of the tranches determines the sequential order of payments;
- be written or approved by an external counterparty (i.e. external auditors) in order to manage the potential conflict of interests among all the investors and the originator.
If these conditions are met, qualifying synthetic securitizations can be defined as meeting the “standard” principle of the HQS synthetic securitisation.
C) Transparency
Before the execution of the transaction:
In order to meet the transparency principle of the HQS definition, synthetic securitisations shall include an extensive due diligence during which the bank provides all engaged counterparties, which have previously signed Non-Disclosure Agreement (NDAs), with a pre-defined minimum set of information regarding:
• the underlying loan portfolio
• detailed information regarding its credit policies and
• detailed information regarding credit risk parameters.
Following the execution of a transaction:
HQS synthetic securitisations shall:
• provide detailed investor reports which shall contain all information about portfolio performance.
• provide the same data package to the new eligible investor, if current investors/noteholders decide to trade their synthetic securitization exposures on the secondary market.
In sum, UniCredit experience and above conditions show that it is possible and highly desirable for competent authorities to design a HQS framework for synthetic securitisation. European authorities are encouraged to benefit from the competent and independent advisory from third parties such as the PCS Association which would facilitate the task to define market standards, based on market consensus and best market practices.
In addition, for the sake of completeness, qualifying ABCP transactions should also be included in the HQS framework because there are well developed and standardised structures in the market. UniCredit stands ready to share with the EBA staff its view in relation to HQS eligibility criteria for ABCPs.
Note (1): The EU Commission expressed objections to the inclusion of synthetic securitisation in the High Quality Liquid Assets definition for the Liquidity Coverage Ratio (see FAQ for delegated acts http://europa.eu/rapid/press-release_MEMO-14-579_en.htm?locale=en ). In section 26.1.2. , it is stated that ”re-securitisations are explicitly excluded [from high quality liquid assets], as they are [1] typically complex and [2] less transparent structures, where the [3] cascading of investor losses is very difficult to understand due to re-tranching. The same goes for synthetic securitisations, where the underlying exposures are not transferred to the special purpose vehicle. Instead, the transfer of risk is achieved by the use of credit derivatives or guarantees, while the exposures being securitised remain with the originator. The transfer of the assets to be securitised ensures that securitisation investors have recourse to those assets should the Securitisation Special Purpose Entity (SSPE) not fulfil its payment obligations. [4] Such recourse cannot be granted in synthetic transactions, due to the fact that only the credit risk associated with the underlying assets, rather than the ownership of such assets, is transferred to the SSPE. Such a structure also [5] adds counterparty risk on derivatives or guarantees, and [6] hampers investors' rights to the proceeds of the underlying exposures. In addition, most synthetic structures add to the complexity of the securitisation in terms of [7] risk modelling”.
Question 3: Do you believe the default definition proposed under Criterion 5 (ii) above is appropriate? Would the default definition as per Article 178 of the CRR be more appropriate?
In UniCredit opinion, the default definition proposed under criterion 5 is appropriate. Having said that, it is however not straightforward why a new definition is needed in this context, hence a clarification by the EBA would be appreciated.Question 4: Do you believe that, for the purposes of standardisation, there should be limits imposed on the type of jurisdiction (such as EEA only, EEA and non-EEA G10 countries, etc): i) the underlying assets are originated and/or ii) governing the acquisition process of the SSPE of the underlying assets is regulated and/or iii) where the originator or intermediary (if applicable) is established and/or iv) where the issuer/sponsor is established?
General considerations:In general, UniCredit deems that defining “standard” securitisations through a jurisdictional framework is a difficult exercise, but the experience of the common European standards and best market practices developed by the PCS initiative is setting a direction worth considering by competent authorities.
The PCS initiative has taught market participants that the imposition of standard criteria in the securitisation market needs to take into consideration both the potential benefits as well as the potential unintended consequences, including possible limitations in terms of market development in those jurisdictions excluded from the PCS scope. For example, jurisdictions or asset classes originating in certain jurisdictions excluded from the definition of “standard” would necessarily face more limited investor acceptance even though market dynamics and developments may, in the future, no longer validate the initial definition of “standard” transactions.
Considering the remarkable impact that market dynamics and developments may have as well as the wide variety of factors which need to be duly taken into account when defining standardisation criteria, UniCredit cautions against the strict adoption of a permanent and non-flexible jurisdiction-approach used as a benchmark for defining standardisation. UniCredit suggests that a more flexible approach be considered. Such an approach, in UniCredit view, would need to imply periodic updates of what the “regulatory standards” are, including limits - if any - imposed on the type of jurisdictions.
Here below UniCredit provides some considerations outlining the wide variety of the factors governing the eligibility of jurisdictions:
Empirical evidence about the country of location. Securitizations in EU and the European Economic Area (EEA) are usually single-country transactions, although a number of deals are backed by assets from more than one jurisdiction. UniCredit has seen true sale publicly placed single-country transactions with underlying collateral from Austria, Belgium, Bulgaria, the Czech Republic, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Latvia, the Netherlands, Norway, Poland, Portugal, Spain, Sweden and the United Kingdom. To our knowledge, transactions backed by assets from the following countries have not been placed so far with investors: Croatia, Cyprus, Estonia, Hungary, Iceland, Liechtenstein, Lithuania, Luxembourg, Malta, Romania, Slovakia and Slovenia. Nevertheless, they should not be excluded from the EBA perimeter. Other European securitizations from non-EEA countries that were placed with investors included loans originated in Russia, Switzerland, Turkey and the Ukraine. Hence, we note that there has been regular issuance from a number of non-EEA countries (e.g. Russia, although not placed in Europe and Switzerland) while many EEA jurisdictions have yet to make their securitization debut. Collateral performance across non-EEA securitizations has also diverged notably.
Main factors governing the eligibility of jurisdictions:
- Rather than applying limits to the type of jurisdiction (such as EEA only), UniCredit suggests policymakers to take into account the characteristics of the local securitization markets. In UniCredit’s view, aspects including the availability of historical data, statistical predictability of asset performance, local underwriting standards, loan documentation standards and local securitization laws should govern the eligibility of a jurisdiction for the purpose of standardisation rather than following a fixed jurisdiction-based approach. UniCredit suggests following a similar approach in relation to the use of mixed pools in European securitizations. Historically, transactions backed by assets from more than one jurisdiction have been in most cases CMBS but they can include a variety of asset classes. In contrast, ABS with mixed geographical pools have been infrequent and largely limited to Auto and SME ABS, as well as a very small number of RMBS. Mainly mixed pools backed by assets from jurisdictions with well-established securitization laws have been seen in the market. Public data indicate that underlying assets in multi-jurisdiction transactions have performed broadly in line with those assets underlying transactions from a single jurisdiction.
- Another important factor is the currency denomination of the underlying loans with respect to their credit performance. Transactions backed by local currency exposures (e.g. a securitized pool of Norwegian auto loans issued in NOK that underlies a EUR-denominated Norwegian auto ABS tranche) are typically hedged using cross currency swaps to avoid cash flow mismatches. Different is the case of local loans originated in a foreign currency (FX loans), which are subject to currency fluctuations. The rapid depreciation in the local currency versus the foreign exchange loan currency would expose local borrowers to additional risk. For example, to a large extent, mortgage and loan origination in countries such as Russia and the Ukraine historically took place in foreign currency, typically USD. As a result of the RUB depreciation, some Russian ABS transactions backed by USD were exposed to higher defaults as the borrowers’ monthly installments made up a larger share of their wages. This situation was replicated in other jurisdictions as well.
- True sale and segregation in legislation. The majority of securitization transactions incorporate the “true sale” concept that isolates the securitized assets from the insolvency estate of the originator. When it comes to the acquisition process of the SSPE, local laws governing securitization transactions should be in place to ensure a true sale of assets and their segregation from an originator’s estate. UniCredit is aware that the legal framework can differ to a high degree across jurisdictions with respect to the definition of bankruptcy remoteness.
- National laws. In addition, in well-established securitization jurisdictions, the laws governing securitization transactions protect the SSPE’s title to the purchased receivables in the event of fraudulent sale or conveyance. In such cases, investors would normally retain recourse against the seller of the property rather than the SSPE. Hence, limits with respect to jurisdictions that lack a comprehensive legal framework governing such circumstances could be considered but would need to be regularly revised, since legal frameworks change overtime and are tested via local courts.
Question 5: Does the distribution of voting rights to the most senior tranches in the securitisation conflict with any national provision? Would this distribution deter investors in non-senior tranches and obstacle the structuring of transactions?
In UniCredit view, the distribution of voting rights have to be differentiated at various levels in order to guarantee the rights not only of most senior notes but also of the mezzanine and junior notes, otherwise both the primary and secondary markets cannot be developed for these notes. The redirection of voting rights to the most senior tranches would, unquestionably, adversely affect investor interest in non-senior tranches of securitisations. In fact, mezzanine and first-loss investor retain necessarily the greatest relative exposure and risk to the underlying assets. Limited voting rights is directly contradictory to this relative risk and exposure profile. In fact, a voting right reduction of mezzanine tranches represents a major obstacle for the restart of securitization market.Regarding the specific criteria of the Pillar II (from 7 to 14), in particular criteria no 8, UniCredit is of the view that greater definitional clarity is needed with regard to “appropriately mitigated” and “genuine hedging.” That said, a carve out should be provided for a structure that: a) has a credit enhancement that already covers the interest rate risk or b) has a natural hedge (i.e. implicit interest rate risk coverage) between the asset (sold portfolio) and the SPV/issuer liability (the note placed to the market). This is for example the case of a portfolio with loans bearing fixed rate payments against notes with fixed rate portfolio or of a floating rate portfolio with loans bearing cash-flows linked to the same reference index of the notes.
Question 6: Do you believe that, for the purposes of transparency, a specific timing of the disclosure of underlying transaction documentation should be required? Should this documentation be disclosed prior to issuance?
Disclosure prior to issuanceUniCredit fundamentally supports disclosure of all relevant information to investors on a timely basis. This is critical to the functioning of any successful securities marketplace. Investors must be provided with the information necessary to make an informed investment decision. As a result, disclosure documentation must be provided to investors prior to issuance and this is the practice across the global securities markets. Beyond this, UniCredit is not supportive of further regulatory requirements regarding the timing of disclosure prior to issuance.
As for securitisation, the adherence with market transparency standards should be certificated at issuance date and guaranteed by an impartial third party constituted with this specific mission. It is important in fact that a periodic quality check of the transaction features within the qualifying securitization framework both for regulatory purposes and secondary market disclosure is run to the benefit of investors.
The PCS initiative has promoted the PCS label certifying standardisation, data/information accessibility and comparability. On this respect, the PCS Secretariat could represent a relevant interlocutor with the authority also in the future, with the aim to contribute to help investors to have, along similar lines as the ECB eligibility process, a quality check of the transaction features within the qualifying securitization framework both for regulatory purposes and secondary market disclosure.
The goal should be to provide timely and meaningful information to investors who can then assess the underlying asset quality and structural features. Belated and incomplete documentation and deal information that is provided at short notice adds an additional analytical burden for investors, which could initially discourage them from investing in the transaction altogether. This relative lack of information will be reflected in a higher return required by investors for the additional risk. Therefore, making deal documentation available to investors in a timely manner is important for issuers wishing to attain the lowest possible borrowing costs. That said, however, a specific timetable for the required provision of disclosure information to investors is not necessarily required and may add unnecessary burden to issuers and investors, particularly given the clear economic incentives that already exist for issuers to be as complete and timely as possible in their disclosure. For example, within accepted issuance practice, sophisticated investors may either require less formal disclosure or may perform their own due diligence processes which may preclude the need for fixed disclosure schedules. In addition, for frequent issuers or issuers that are known in the market, the level of disclosure and timing may be relatively less important than less frequent issuers, new issuers or relatively complex structures involving non-traditional assets.
As a result, any regulatory requirement regarding disclosure timing should be flexible enough to cover the range of anticipated issuances and should consider the already existing economic incentives imposed on issuers by the marketplace.
Provision of on-going data
The emphasis should be on the relative appropriateness of the disclosure, content and the responsibilities and liabilities imposed on those involved in the securities placement/distribution process rather than on the ability to provide on-going data. Having said that, it should be nevertheless noted, however, that the provision of on-going data is also important for the successful maintenance of a liquid ABS marketplace.
To this end, UniCredit has been involved from the beginning in the successful establishment of the European Datawarehouse, supported by the ECB. Prior to its establishment, collateral-level data were often either entirely unavailable or inconsistent from deal to deal. Loan-level data were often not provided by European issuers due to privacy and competitive concerns. While loan transparency before issuance does not necessarily translate into better investor participation, a lot of information was lost when data were aggregated. Therefore, for loan/collateral level information, the relevance of data may depend on its immediacy but more importantly on the ability to access and compare relevant on-going data/information.
Additional considerations
In addition to our view in reply to Question 6, UniCredit has some specific considerations with reference to the specific criteria of the Pillar III (from 15 to 22):
- criteria no 18 : the proper pricing of each transaction should be evaluated by each investor on its own portfolio assumptions (default, arrears, prepayments forecast). Accordingly, no cash-flow model should be provided by the originator or the sponsor in order to avoid, as happened in the past, that they can unduly influence investor’s decision and securitization’s market. The originator or the sponsor can only provide market participants with all necessary information that allow to build in-house cash-flow models or resort to external service providers;
- criteria no 20 : in order to avoid excluding assets which have a short seasoning and that are crucial for the current economy development (for example renewable energy etc.), UniCredit suggests that the EBA investigates the opportunity to reduce the requirement for historical performance data from 5 to max 3 years.
All of the criteria in Pillar III should also consider the implications for short-term assets and securitisations utilizing such assets (such as ABCP). UniCredit stands ready to share its view in relation to transparency standards for ABCPs with the EBA.
Question 7: Do you agree that granularity is a relevant factor determining the credit risk of the underlying? Does the threshold value proposed under Criterion B pose an obstacle to the structuring of securitisation transactions in any specific asset class? Would another threshold value be more appropriate?
The “Origination to Distribute” model showed that financial institutions were not adequately incentivized to follow a prudential approach during the underwriting phase. Moreover, the correlation of the portfolio of securitized assets is one of the main indicators that can explain the potential impact of idiosyncratic and systemic risks in the securitization. Against this background, UniCredit agrees that the granularity of the pool is a focal point to be observed as indicator for the minimum level of securitisation credit quality but, correspondingly, granularity should not be considered the unique focal point.The granularity concept has to be adjusted to reflect the specificity of the underlying asset classes. An adjustment of the granularity concept is considered crucial in order not to hamper the possibility to set-up securitizations on SMEs loans or on corporate portfolios where the EBA cannot apply a one-size fits all approach. In particular, the proposed single debtor concentration limit of 1% seems punitive when applied to pools of SME Leasing and Corporate exposures that have been historically structured with higher levels of concentrations. This is even more true when considering that the limit is referring to a group of connected clients.
In view of these considerations and taking into account the difficulties of SMEs to fund themselves in the market UniCredit suggests to introduce a limit of 5% in term of concentration / global exposures towards a group of connected clients when considering SME, Leasing and Corporate counterparties. The concentration limit of 1% shall be applicable to pools of retails exposures.
Question 8: Do you agree with the proposed criteria defining simple standard and transparent securitisations? Do you agree with the proposed credit risk criteria? Should any other criteria be considered?
UniCredit broadly agrees with the proposed criteria defining simple standard and transparent securitisations, taking into account also the above considerations in answers 2 and 7.Concerning the credit risk criteria, UniCredit deems there is a fundamental misconception of the framework. Criterion C, point ii defines the eligibility criteria for underlying exposures in terms of maximum individual risk weight (based on the Standardised Approach that has, in UniCredit view, a low-risk sensitivity). This is a major departure from best market practices.
While it is understandable that prudential regulation prevents credit risk excesses stemming from the underlying assets during the underwriting phase, regulatory limits should not be primarily imposed on collateral exposures where the risk is evaluated through a tranching of the liabilities.
It is also recalled that the Italian legislation does not foresee the use of the individual risk weight levels as an eligibility criteria for asset selection also to avoid cherry picking, a practice to be discouraged as indicated also in this document.
UniCredit suggests to the EBA to consider alternatives in terms of portfolio measures (e.g. max weighted average).
Question 9: Do you envisage any potential adverse market consequences of introducing a qualifying securitisation framework for regulatory purposes?
UniCredit has been leading and supporting the PCS initiative since inception and would share its experience in this regard. The PCS label de facto has introduced, with increasing market acceptance, a qualifying securitisation framework towards best market practices. UniCredit is aware that there are pros and cons when introducing a qualifying securitisation framework. The cons are primarily for non-qualifying securitisations but overall, based on the PCS experience, UniCredit remains strongly convinced that there are more pros than cons in introducing a qualifying securitisation framework. This is particularly evident when the objective is the revitalisation of the European securitisation market, which has historically performed much better than the US comparable market.As expressed above, to avoid any potential adverse market consequence against specific type of securitisations, all securitisation techniques, though with the appropriate differentiations, have to be considered for the purpose of being recognized as simple standard and transparent, not only true sale securitisations. This is particularly the case for synthetic securitisations which continue to represent an important risk-sharing and credit risk mitigation instrument, crucial for the support to the economic recovery. As recalled in reply 2, in the current regulatory environment, where capital costs have substantially increased due to a multiplicity of requests to supervised entities, any financial instrument is very sensitive to regulatory changes. A regulation, which includes or excludes certain forms of synthetic securitisations or asset backed commercial papers (ABCP) from the HQS definition, has therefore the power to foster or vice versa hamper its economic viability, the related incentives for lenders to deploy new lending and for investors to share aggregated credit risk. That’s why the overall impact on the economy and on financial stability need to be carefully assessed by the regulator.
In addition, the qualifying framework will also contribute to the creation of the incentives for the market to increase the standards towards the best market practices.
Question 10: How should capital requirements reflect the partition between qualifying and non-qualifying?
UniCredit deems appropriate to use a separate calibration for “qualifying securitizations” which should entail lower risk weights and lower floors.The proposal to have differentiated capital requirements for qualifying and non-qualifying securitizations will allow for a more symmetric treatment of securitizations. As recalled by the EBA and other international institutions such as the IMF, some type of securitizations are treated asymmetrically vis-à-vis other asset classes (e.g. covered bonds), particularly in the EU.
Therefore, in UniCredit view, the qualifying securitization label must entail a distinctive more favorable RWA treatment compared to the non-qualifying securitization transactions. This kind of differentiation in capital requirement should be incorporated in the BCBS proposal of revisions to the securitization framework (currently ongoing) and should be referred to in all the available approaches:
- Internal Rating Based Approach;
- External Rating Based Approach;
- Standardized Approach.
More specifically, UniCredit would like to share some suggestions on some aspects related to the new approach proposed by the Basel Committee on Banking Supervision (“BCBS”) in its second consultative document issued on December 2013 and finalized in December 2014. This may a have a bearing on the EBA proposal on: i) the risk weight floor, ii) the maturity definition and iii) the use of IRB.
i) The EBA calibration of capital requirements for HQS should consider a different and more favorable risk weight floor (see also recommendation 6). The target floor could therefore be defined as follows:
a. For qualifying securitizations (including ABCPs) the risk weight floor should be no more than 10% in order to make it aligned with covered bonds requirements;
b. For non-qualifying securitization the risk weight floor should be 15% as already provided in the BCBS document on Revisions to Securitization framework.
ii) The definition of maturity proposed by BCBS is very conservative (it normally corresponds to the legal maturity). Given the high impact of maturity in the capital requirements calculation, the proposed approach leads to a significant overestimation of risk weights for many transactions; it is therefore proposed to use for all securitizations a different definition of maturity which takes into account the maturity distribution and the weighted average life of the underlying assets.
iii) reduce the reliance on external credit ratings by not discouraging the use of the IRB method under certain conditions. In UniCredit opinion, the BCBS requirements to use the IRB method especially for investment positions are too strict. In fact, especially for European banks working as sponsors or investors, the issue of data availability is relevant. UniCredit retains that it is important for all securitizations to ease the data requirements for the IRB method. The definition of more simple and homogenous criteria will help the IRB method to become a real option for European banks.
Question 11: What is a reasonable calibration across tranches and credit quality steps for qualifying securitisations? Would re-allocating across tranches the overall capital applicable to a given transaction by reducing the requirement for the more junior tranche and increasing it for the more senior tranches other than the most senior tranche be a feasible solution?
The new securitisation framework proposed by the BCBS is contributing to: i) a simplification of the hierarchy of approaches available for capital requirements calculation; ii) reduced reliance on external credit ratings; iii) increased risk sensitivity and reduced cliff effects. UniCredit fully supports these goals. Nevertheless UniCredit deems that the proposed calibration by the BCBS still carries serious risks: to disproportionately increase the costs related to securitizations (both synthetic and true sale) and to create an unbalance across tranches. The proposed BCBS framework is therefore likely to hamper the economic viability and effectiveness of these instruments, with potential detrimental effects on lending to the real economy.Hence, UniCredit welcomes the EBA intention to introduce “qualifying” securitization with a specific regulatory treatment that would lead to a more reasonable calibration across tranches. Also, with reference to recent regulatory developments, UniCredit appreciates the BCBS consultation launched last December on the “Criteria for identifying simple, transparent and comparable securitisations”. UniCredit encourages the EBA and the European competent authorities to work at international level in order to agree on a specific regulatory framework for HQS globally.