We broadly agree with the impediments outlined in the DP. In addition we would like to highlight some additional impediments
1.1 Grouping of CLOs with CDOs of ABS
The data included in the DP includes an analysis of default rates for “AAA” and “BBB” rated classes of securitisations. CLOs and ABS CDOs share some common structural features as do many other types of securitisation, however the composition of CLOs are very different. While the underlying portfolios of CDOs of ABS consisted primarily of other types of ABS securities, CLO portfolios generally consist of:
(a) 90% or more senior secured loans and senior secured bonds to sub-investment-grade corporates; and
(b) the portfolios contain no ABS or synthetic exposures.
Whilst the category of assets securitised by managed CLOs is sub-investment grade corporate debt, overwhelmingly it is only the senior secured portion of such loans which compose a CLO portfolio. There is also a minimum rating requirement for the underlying assets going into the portfolio. The portfolios are actively managed in accordance with strict portfolio tests. Furthermore, mark to market haircuts are applied to portfolio assets whose ratings are “CCC” or below and to defaulted obligations which can result in the failure of the CLO to meet coverage tests thereby triggering cash flow sweeps in the interest priority of payments to repay principal to noteholders in order of seniority.
The default rates set-out in the DP do not accurately reflect the actual default rates of European CLOs. Default rates for European CLOs between 2007 and 2013 was 0.1%.
1.2 Basel II and new Basel Proposals for the securitisation framework.
We do not believe that the Basel Committee on Banking Supervision’s (“BCBS”) proposals have been very helpful. They fail to reflect the true position with regard to losses incurred in European securitisations.
We believe that capital neutrality should be a basic premise for the regulatory capital treatment of securitisation exposures. The capital charges applied to a pool of exposures prior to securitisation should be no worse than the total capital charge applied to the tranches of the securitisation, to reflect the unchanged economic risk across the pool.
In the case of a managed CLO, the portfolio is managed to ensure compliance with strict debt coverage tests and collateral quality tests which meet rating agency requirements. This added layer of protection is not available to direct investors in the loan portfolio, and the capital charge should reflect this. Any proposal to ascribe higher capital charges to banks investing in CLOs would be in our view detrimental, and could pose a real risk that bank investors would pull out of the CLO market, leaving the universe of CLO investors insufficient to keep CLO issuance economically viable. This would have a damaging effect on the ability of European sub-investment grade corporates to raise capital.
1.3 Solvency II and the EIOPA technical report
The proposed capital treatment of CLOs in the Solvency II regime gives us significant cause for concern. In September 2012, the European Commission requested EIOPA to review the calibration of capital requirements for investment in certain classes of long-term finance which provide management of long-term risk for insurers. In December 2013, EIOPA produced that report, entitled “Technical Report on Standard Formula Design and Calibration for Certain Long-Term Investments” (the “2013 Report”). In the 2013 Report, the classes of securitisation transaction which qualify as “Type A” and therefore attract lower capital charges expressly exclude CLOs other than SME CLOs. The criteria for qualification as “Type A” securitisation in the 2013 Report have now been included in Part I of EIOPA’s “Technical Specification for the Preparatory Phase” of Solvency II based on the working documents of the Level II delegated Acts to be published later this year, which was published on 30 April (the “Technical Specifications”). Our belief is that if CLOs were to remain in the “Type B” category proposed by EIOPA, there is a real risk that insurers required to use the Solvency II Standard formula may pull out of investing in CLOs as an asset-class.
We believe that the categorisation of CLOs as “Type B” by EIOPA is based i) on a misunderstanding of the nature of the risk-mitigating structural features of managed CLO transactions and ii) a focus on the widening spreads which resulted from rating downgrades of CLO tranches during the financial crisis - which downgrades were not borne out by default rates, and have since been largely reversed.
The 2013 Report states (at page 121) “The underlying of CLOs and CDOs is typically speculative-grade corporate debt”. As stated above, this is only part of the picture:
(a) the vast majority of CLO “2.0” portfolios consist of 90% or more senior secured loans and senior secured bonds to corporates; and
(b) the portfolios contain no ABS or synthetic exposures.
Whilst the category of assets securitised by managed CLOs is sub-investment grade corporate debt, overwhelmingly it is only the senior secured portion of such debt which goes into the CLO. There is typically a minimum rating requirement for the underlying assets going into the portfolio. The portfolios are actively managed in accordance with strict portfolio tests.
We also believe that EIOPA’s focus on rating downgrades as the main indicator of quality of an asset class is inappropriate. The 2013 report places CLOs in “Type B” on the basis that 72.3% of CLOs were downgraded between mid-2007 and end of 2012. We would dispute this figure as representative of the performance of European CLO tranches.
Firstly, a significant amount of downgrades were a result of a change in the default models in rating criteria and methodology (some of which has been subsequently revised), and not actual default rates. S&P had updated their criteria in September 2009, and acknowledged that “Virtually all of the “AAA” downgrades resulted predominantly from the application of the updated criteria, rather than transaction performance.”
Moody’s updated their criteria on 4th February 2009, also changing their default probability model , and reviewed CLO tranches against the new criteria. Of the 395 Aaa- rated CLO tranches in Europe at the start of 2009, 47% retained their rating, while 53% were downgraded .
Secondly, CLO transactions continued to perform well in relation to other asset classes following the downgrades to end of 2012, and S&P and Moody’s both upgraded a significant proportion of tranches in CLO transactions as a result. As stated in the S&P research paper relied upon by EIOPA , S&P subsequently upgraded a number of tranches in CLOs due to the operation of “structural deleveraging” (explained below). In November 2011 Moody’s announced that 81% of the European CLO tranches originally rated Aaa were back to their original ratings, as a result of revised rating criteria in June 2011 together with improved par coverage and credit quality . The downgrade figure of 72.3% relied upon by EIOPA is therefore not representative either of default rates in CLOs, or in fact downgrades in existing CLO tranches in Europe.
1.4 Risk Retention
The European risk retention requirements in the Capital Requirements Regulation (“CRR”) and the Alternative Investment Fund Managers Directive (“AIFMD”) have proved a significant challenge for investors in the CLO market. Despite the recommendations of IOSCO that an exemption from retention be considered for managed CLOs, the “one-size fits all” nature of the retention requirement was imposed and has presented significant difficulty for CLO structures. As managers have struggled to fund and hold the retention and the related capital, so investors have struggled with the requisite assurance that the transaction is compliant if any mechanism is used other than simply the manager holding and funding the retention from its own balance sheet (see below).
Similar requirements are due to apply to insurer investors once Solvency II comes into force, and to UCITs funds once the implementing regulations are made under the new UCITS Directive.
Apart from the structuring of the retention itself at the outset of the transaction, there are ongoing difficulties with the retention rules which potentially affect returns to investors and investors’ ability to remove a failing manager.
1.5 Cost of retention for CLO managers
CLO sponsors meet the retention requirement using one of two of the five retention options - option 405(1) (a) (“vertical slice”) or option 405(1)(d) (“first loss”).
As most CLO managers do not deal as principal or underwrite financial instruments, they need to hold capital in an amount of i) the greater of their capital requirement for credit risk, or ii) an amount representing a quarter of their fixed overheads, under CRR. Prior to the retention requirements coming into force, many managers did not take any credit risk, and were therefore calculating capital based on their fixed overheads.
In holding the retention by way of first loss, assuming an unrated tranche of subordinated debt, managers applying the standardised approach to credit risk would have to apply a 1250% risk weight to their securitisation position. With a typical deal size in a CLO 2.0 being around €400m, a 5% holding as a first loss position will require capital of €20m.
To hold a vertical slice of the transaction, the manager needs to hold 5% of each of the tranches sold to investors (including 5% of the first loss tranche). Assuming a transaction with a total issuance of €400m, including €45m of subordinated debt, the capital charge on a 5% vertical slice is set out below:
Class of Notes Amount (€m) Rating Credit Quality Step Risk Weight % 5% (€m) Capital Charge (€m)
Class A 215 AAA 1 20 10.75 0.17
Class B 50 AA 1 20 2.50 0.04
Class C 30 A 2 50 1.50 0.06
Class D 20 BBB 3 100 1.00 0.08
Class E 30 BB 4 350 1.50 0.42
Class F 10 B 5 1250 0.50 0.5
Sub Notes 45 N/A N/A 1250 2.25 2.25
Total €20 €3.52
The total capital charge to the CLO manager for holding a vertical slice on this example transaction is €3.52m, against a €20m capital requirement for holding a first loss position. Most CLO managers do not hold any of the underlying portfolio prior to the CLO transaction and the retention rule will always involve a significant increase in capital and costs, unlike bank originators who (i) can benefit from lower capital charges by securitising, or at least have those capital charges capped at the capital charge which applied to the underlying portfolio and (ii) are able to apply the IRB approach in calculating their capital charge for the retention, which will in many cases result in a lower capital charge than that produced by the standardised approach typically used by CLO managers.
As well as their own capital cost, most CLO managers will also require funding for the purchase of the retention positions. This comes with additional costs to the CLO manager - even if the retained positions are used as collateral in a secured financing they are likely to require significant initial margin to be paid to the counterparty. These additional costs to CLO managers will reduce the issuance of new CLOs which will in turn reduce the availability of capital to European corporates.
We do not have a view as to whether synthetic securitisations should be excluded.
Our view is that the default definition should be consistent with the CRR definition. The market benefits from consistency across regulatory frameworks.
Underlying Portfolio: We are of the view that such limitations should not be included. Portfolio diversity should be encouraged. Limiting the jurisdiction of the underlying portfolio would inhibit portfolio diversity. In addition, geographical limitations do not reflect underlying credit quality particularly where such limitations are arbitrarily drawn based on geography or economic groupings. However, if such limitations are included EEA and non-EEA G10 countries should be included.
In CLOs, certain types of decisions such as the ability to accelerate and enforce are rights which vest solely with the most senior tranche of notes. Some decisions such as optional redemption of the structure are left to the vote of more junior classes of notes. In a CLO any exercise of an optional redemption by junior classes of notes requires that there are sufficient proceeds to repay all interest, principal and expenses which are senior to the most junior class of unrated debt. This ensures no optional redemption is exercisable by the most junior class of notes unless all rated debt can be repaid. If all decisions are left solely to the senior tranches, this could have negative consequences for the market. In order to obtain AAA ratings there must be a market for more junior classes in order to provide the subordination necessary to support the AAA rating. In addition, certain actions related to key “money terms” of a securitisation require all classes to vote. Disenfranchising junior classes in these instances, would allow the more senior classes of notes to unilaterally modify the economics of a securitisation with no consideration of the effect on more junior tranches. A prohibition on junior classes voting would diminish the appeal of the more junior classes of notes.
Offering documents which comply with the Prospectus Directive are required to describe all material terms. While we do not object with the premise of making the transaction documents available (CLO transactions documents are made available to the market following the closing of the transaction), the practicalities of distributing these prior to issuance would be problematic. As is the case with most transaction timelines, documents are being negotiated and agreed up to the issue date of the securities. Investors are required to make their investment decision based on the offering document. Issuer, arrangers and collateral managers are required to include full, accurate and complete disclosure in the offering document. This is the case in Europe and the United States. Securitisations have multiple documents comprising many pages. Providing investors with the documents prior to issuance could distract investors from a full and complete review of the offering document which not only describes the material terms of the transaction in great detail but also highlights material risks which investors should be aware of. It should also be noted that under the European Credit Rating Agency Regulation, it has been proposed that transaction documents be posted to a public website set-up by ESMA. Again any such disclosure should be required after the transaction has closed.
We do not believe that imposing a large exposure limit is helpful. It is important to note that granularity of a portfolio does not necessarily equate to better credit quality. A granular portfolio of poorly underwritten assets does not create a less risky portfolio. CLOs are already subject to large loan exposure limits in the various tests which the CLO must satisfy (typically such limit is set to approximately 3%).
Criterion 2: We do not share the view that a securitisation should be excluded on the basis that it is actively managed. A regulated CLO manager adds an expertise to the transaction and monitors each credit in the portfolio. A CLO manager performs in depth credit analysis on each asset. The CLO manager has knowledge and experience in corporate credit and represents the CLO on creditor committees and in work-out scenarios. These are regulated entities responsible for ensuring the CLO can repay its obligations to investors. We also disagree that active portfolio management adds a layer of complexity to a transaction. Investors investing in CLOs analyse both the Tests (as defined below) and the performance of the CLO managers in the same way and with the same rigour that they analyse a static portfolio of assets. There is a great deal of information available to investors on past performance of CLO managers as well as their approach to credit selection and work-outs. This information is widely available from a variety of public sources.
Management of a CLO portfolio is subject to collateral quality tests, overcollateralization tests and concentration limitation tests (the “Tests”) – these are rigorous rating agency tests measuring over-collateralisation and various portfolio characteristics which the manager is required to comply with in order to continue to be able to reinvest in new assets. These portfolio-level tests are already industry-standard and are particular to managed CLOs as opposed to securitisations of static portfolios. The CLO manager has to meet the Tests on an ongoing basis. The Tests and strict trading rules imposed on CLOs means that active management of the portfolio has a very limited effect (positive or negative) on the most senior tranches of a CLO.
The Tests also allow the CLO manager to provide detailed and transparent disclosure to investors on a monthly basis in respect of the portfolio. They cover data such as diversity of obligor by industry and geography, weighted average spread on the assets, weighted average fixed rate coupon, weighted average rating and weighted average life of the underlying assets.
Unlike traditional asset-backed securities, the underlying portfolios of CLOs are typically not purchased from one originator or seller but are typically sourced in the primary or secondary market by regulated investment managers who are independent of any originator or seller of the loans. The CLO manager is able to independently assess the quality of the portfolio and is free of the negative incentives which can arise in an originate-to-distribute securitisation model. This adds an additional level of credit analysis which is not a feature of other types of securitisation.
Risk-reducing characteristics of managed CLOs
We cannot stress too heavily the fact that the assets in a CLO portfolio are only a part of the performance of the CLO itself. As we state above, CLOs give investors the ability to invest in the loan market with the benefit of structural enhancements and active management which significantly reduce risk to the senior noteholders when compared with a direct investment, or with a static loan portfolio. The exclusion of managed portfolios fails to recognise the weight that should be given, uniquely in this asset class within the securitisation space, to structural deleveraging and active management.
In summary, structural deleveraging in a CLO interrupts the normal priority of payments in the event that the quality of the portfolio falls below a certain level. The debt coverage tests measure the amount of over-collateralisation in the CLO. In a managed CLO transaction, there is typically a maximum of 7.5% CCC (or below) rated assets in the portfolio. If these low-rated assets rise above that percentage, those assets will be treated as valued at market value rather than at the usual par value in meeting the debt coverage tests. If the debt coverage tests fail as a result, no interest can be paid out on the junior notes, the majority of the CLO management fees and all receipts from the assets will go to pay principal on the senior notes sequentially until the pool complies again with the coverage tests. The same applies to failure of the coverage tests as a result of defaulted assets. Thus the senior notes benefit both from the credit enhancement provide by the junior tranches and the protection of senior income and principal prior to any default. In a static-pool securitisation, whilst mechanisms may be built in to ensure excess income supports deficiencies on the senior notes, this is not done on a managed basis. This structural deleveraging was partly responsible for the subsequent upgrades on many of the AAA CLO tranches which were downgraded by S&P in 2009.
In addition, post the expiry of a CLO reinvestment period, to the extent that portfolio collateral is repaying and the manager is unable to reinvest these proceeds (the circumstances are usually related to note ratings or certain test compliance), these proceeds are ultimately repaid to note holders in order of seniority. This is another mechanism by which structural deleveraging can occur and is an inherent characteristic across European CLO 1.0 and 2.0 structures.
In addition to those protections, we would suggest that a CLO which complies with the following criteria should be included in any “high-quality” class of ABS which receives preferential regulatory treatment:
(a) the securitised exposures must be managed on a continuing, discretionary basis by:
(i) an EEA investment firm which is required to be regulated in its home member State and which is subject to the Markets in Financial Instruments Directive (“MiFID”) or an affiliate thereof; or
(ii) a firm authorised under the Alternative Investment Fund Managers’ Directive (“AIFMD”) or an affiliate thereof; or
(iii) a firm or an affiliate thereof which would fall within (i) or (ii) above if its head office was situated in the EEA and which is subject to equivalent regulation in relation to the conduct of its business and its management of conflicts as a firm established in the EEA (for instance investment advisors registered under the US Investment Advisers Act of 1940, as amended);
(b) the CLO manager of the securitised exposures must undertake to the investors in the securitisation to comply with the regulatory requirements applying to it in relation to the management of conflicts of interest, in connection with its management of the securitised exposures (i.e. compliance with MiFID and/or AIFMD or equivalent regulations outside the EU);
(c) the securitisation must contain provisions whereby the interests of the CLO asset manager are appropriately aligned with the interests of the investors during the whole life of the securitisation. It is recognised that this may be achieved by a material part of the manager’s compensation for carrying out its duties being structured as an incentive fee, which will only become payable upon appropriate performance thresholds of the securitised exposures having been met;
(d) Investor reports should be provided monthly;
(e) In addition, the following portfolio characteristics could be provided for in a definition of CLOs to ensure that only certain types of structures would actually constitute a CLO:
(i) it contains a high percentage of senior secured loans and bond loans to corporates;
(ii) it does not contain any asset-backed securities or synthetic securities; and
(iii) it is managed by an independent investment firm or an affiliate thereof which satisfies paragraph (a) above and who independently reviews, and individually selects, each asset to purchase in the primary or secondary market (with no obligation to purchase from any individual bank or originator) ;
The assets in a CLO portfolio are purchased according to the eligibility criteria, which specify the conditions for individual loans, such as jurisdiction, rating, non-convertibility, tax and regulatory conditions etc., and as at the effective date the portfolio profile and collateral quality tests must be satisfied. CLO managers may also only invest in assets during the reinvestment period if following investment, the portfolio profile and collateral quality tests remain satisfied or if not satisfied, they must be improved following such reinvestment. These tests ensure diversification of assets by industry, limit maximum concentration in a single borrower or borrower type, and ensure quality of loan covenants etc. The active management of the portfolio ensures the continued compliance with these tests. Thus CLOs have built-in protection for the quality of the assets in the portfolio.
Criterion 3: (True Sale): This criterion is unclear and we do not believe it reflects the intention of its inclusion. The key concept here is that the assets should be isolated from the insolvency of the seller. Given the differing legal regimes of various jurisdictions and the numerous methods of transfer available, we would suggest that the requirement be re-phrased such that assets are transferred by way of a “true sale” (or similar isolation) such that the assets are isolated from the creditors of the seller including upon an insolvency of the seller.
Criterion 4: Homogenous in terms of asset type, currency and legal system: We do understand the need of a homogenous pool of assets however this requirement should be met by asset-type groupings, in the case of CLOs, corporate debt. We do not agree that there needs to be homogeneity in currency and legal system. Within the EEA itself there are different currencies and legal systems.
We have concerns over the possible “cliff effect” of classifying certain types of securitisations as simple, standard and transparent (“Qualifying Securitisations”).
In particular, there is a real danger that securitisations which are not Qualifying Securitisations are seen as sub-standard by the market and thus discourage investment in securitisations which fall outside of the Qualifying Securitisation designation.
The appropriateness of a framework which certifies some transactions as Qualifying Securitisations depends on the consequences of such a certification. An implicit regulatory stamp of approval will inevitably mean that investors rely to an extent on that stamp as a representation that the product is lower-risk, particularly if the associated capital risk-weights are lower than other similar products without such a stamp. This could actually result, not in investors taking necessarily better investment decisions, but instead investing in a smaller universe of securitisation instruments. The resulting positive correlation in an investor’s portfolio could mean potentially greater losses should default levels in an asset-class increase.
There is already concern in the market that the regulatory capital treatment of securitisations is unduly harsh and in some instances higher than a regulated entity holding the individual assets comprising the securitisation portfolio. Given the very low default rates for European CLOs the regulatory capital treatment of a “AAA” rated tranche of a CLO should not be treated as materially worse than a “AAA” tranche of a Qualifying Securitisation.
As stated above in question 9, we have concerns that the designation of some securitisations as Qualifying Securitisations has the potential to have a material adverse impact on securitisations which fall outside the scope of this designation. Any differentiation in terms of the capital requirements applicable to Qualifying Securitisations and those outside of scope increases the likelihood of such impact. The risk weights applicable in the current Basel proposals are in our view still too high across all securitisation asset classes. It should under no circumstance be more advantageous to hold an underlying pool of assets than investing in the same pool in a securitisation. Holding a position in a securitisation should result in a better capital charge when compared with holding a pool of underlying assets. If CLOs are not included in the “Qualifying Securitisation” category, we would encourage regulators to look at historical default rates on CLOs (and not CDOs) and reflect a capital requirement which is more in-line with historic default rates for the asset class.
Any capital requirements should be based on the credit risk of the underlying portfolio taking into account the subordination and structural protections included in a securitisation. As stated in question 10 above, the risk weighting of securitisation positions should not be higher than a direct investment in the underlying portfolio.