Shelf-eligibility and data disclosure discussed
Representatives from Lewtan and Dentons recently discussed the considerations for implementing Regulation AB 2 in a live webinar hosted by SCI (view the webinar here). Topics included privacy concerns, data requirements and practical implications going forward.
Q: What is the background to Regulation AB 2?
A:
Ned Myers, svp, Lewtan: Reg AB 2 rules were initially proposed back in 2010 and sections were re-proposed two additional times, most recently in February 2014. On 27 August, the US SEC adopted the rules in the same open meeting where it adopted a set of rules for NRSRO reform.
The new regulation partially satisfies the Dodd-Frank rule makings around 939a and 942b, which are the removal of credit ratings from the regulatory framework required for public shelf registration and to require asset-level disclosure where relevant information for an investor to independently determine the credit quality of an ABS pool is necessary.
The regulations expanded and clarified areas that were adopted under the initial Regulation AB. Unlike the original proposal, which had elements such as a cashflow waterfall disclosure, the SEC also chose not to adopt a requirement that all 144a issuers would need to provide investors with the same type of information upon request as if the offering were a public deal.
With those high-level aspects that were adopted, the timeline for implementation of Reg AB 2 is 60 days plus one year from the point of filing it in the Federal Register. That means the rules will come into effect for US public deals in November 2015, with the asset-level disclosure requirement taking effect one year later.
Q: How have the rules addressed privacy concerns?
A: Stephen Kudenholdt, partner, Dentons: Before this proposal, there was already a fairly established practice of providing detailed loan-level residential mortgage data to investors. For other asset types such as auto loans and credit cards, there was not any such practice at all.
One of the purposes of this proposal was to create a level playing field with a comparable level of asset-level detail for all asset types that investors would be able to analyse and review. With the initial proposal for loan-level data there were certain elements of the proposal that, even though in some ways followed then industry practices, raised some privacy concerns.
For example with residential mortgages, the proposal included data fields such as loan amounts, date of origination and five-digit zip code, as well as other data fields that include sensitive borrower information like incomes and assets. It was noted early on in the process that it was possible to essentially re-identify an original borrower by combining the data in the posted fields with other publicly available information.
That concern was so significant that in February the SEC requested comment on a further proposal to provide the loan-level data via a restricted issuer website. However, in the final rule, the SEC determined that, first and foremost, all asset-level data that is required under the rule should be readily publicly available to both investors and non-investors. Therefore the issuer website proposal was scrapped.
As for re-identification, the SEC has made asset-level data required under the final rule significantly less granular all across the board - not least by reducing zip codes from five digits to two - and has removed individual data fields, such as income levels, in favour of ratios such as debt-to-income to provide disclosure. The SEC has also worked with the CFPB on issuing interpretative guidance.
Q: What are the new asset-level data requirements and how do they differ from the original proposals?
A: Myers: Reg AB 1 was largely a principal-based regime, while Reg AB 2 will allow for a hardship exemption if an issuer is unable to provide the required asset-level field requirements. The data must be supplied in an XML format so it is machine readable.
Additionally, the SEC simplified the delivery of the data in a couple of ways from the original proposal to the adopted regulation. The original proposal had the concept of general fields and then asset class-specific fields, while the adopted version only has asset class-specific fields that apply to each asset class in question.
That change eliminated a lot of confusion around fields that were mortgage-related and did not apply in other asset classes, for example. Secondly, the SEC also required that the initial asset-level data provided and the ongoing data elements are identical, so in consuming the data the market can write one set of programmes for each asset class and then expect each issuer to provide all of the required fields in an identical format, both at the inception of the deal and during the life of the transaction.
Q: What are the differences between the SEC's requirements and existing market practices?
A: Kudenholdt: The five-digit zip code is one obvious example, because investors are used to having that level of granularity to be able to assess risks, so going to two-digit zips represents a significant change to existing market practices. There are only around a hundred two-digit zip code combinations, so there is significantly less granularity.
Other fields were also changed to mitigate re-identification risk, such as removing the sales price of the property, property value at origination, origination date and first payment date. Information about the borrower's bankruptcy and foreclosure history, as well as their income and debt information was also amended.
The final rules are more granular in requiring an exact borrower credit score. Originally a range of credit scores would have been used.
Q: What will be some of the most important asset class-specific requirements for the other asset classes?
A: Myers: Unlike for RMBS, what the SEC adopted for CMBS is actually very close to existing practice, where the CMSA format has provided asset-level data for years. With respect to re-securitisations, I think that it will be several years until we see that requirement coming into effect because the original securitisation that is then re-securitised will not require asset-level data until 2016.
For auto ABS, there have been several changes from historical market practice, not least that disclosure of asset-level data for auto loans and leases was not historically generally available. Because of the extra disclosure, issuers will need to make changes to their underwriting systems and then provide information on credit scores on an ongoing basis.
Auto leases bring a few extra pieces of information than auto loans, such as information around lease terminations and residuals. The bottom line for both loans and leases, however, will be integration deeper into origination and credit systems, providing fields that would not have historically been aggregated to investors.
Q: Will issuers be limiting themselves to the required fields or are investors going to be demanding more disclosure?
A: Kudenholdt: The fields required by Reg AB 2 provide a base line so that you can compare deal to deal, apples to apples. However, there is nothing in the rules to stop issuers from providing greater data.
Information provided in public offerings would be filed with the SEC and constitute publicly available data, but the tricky part will come if issuers look to provide additional data fields, which contain potentially sensitive information or information that might increase the risk of re-identification, such as five-digit zip. That is something that will take some time to play out.
Issuers may determine that it is simply necessary to provide more granular location information, such as a five-digit zip code, to investors that need that information in order to run their own models, make their own projections and effectively do their own diligence. I could envision a process under which issuers look to get some sort of relief from the SEC, which would enable issuers on an optional basis to provide more granular data to investors that meet certain criteria for being able to use it without that additional data having to be posted publicly on EDGAR.
Q: Are the regulations going to require fundamental changes in how information is collected and disseminated and what would the implications be for investors?
A: Myers: For asset classes like auto loans and auto leases, it is an entirely new paradigm for investors. The information in XML format is going to be publicly available through EDGAR, but you still have to do the work to understand that data.
Q: What are the new shelf eligibility requirements?
A: Kudenholdt: The requirements do not include some of the originally suggested requirements, such as risk retention and life of deal exchange. One of the most important additions is a depositor ceo certification regarding adequacy of cashflows.
There must also be a mechanism baked into the transaction requiring a third-party review of assets for breaches of representations if certain triggers are met, with mandatory mediation or arbitration of any disputes. And there must be a mechanism for investor communications to be facilitated.
Along with these requirements, another very important new requirement that applies to ABS issued under shelf registrations is the concept of what is called Rule 424(h) preliminary prospectus. That is a complete preliminary prospectus containing everything other than pricing-related information that must be provided to investors and filed on EDGAR three days before the first sale of a security in the offering is priced.
Q: What kind of uncertainties will face depositor ceos signing required certifications and how might they mitigate risks?
A: Myers: The SEC says the ceo must be the one certifying, despite many comments pointing out that ceos are not always best placed for that. Given a ceo might not be coming from the securitisation industry, if there are problems down the road, then there are two steps which can be put in place to offer protection.
While there is already authentication that cashflows are run under normal and stressed scenarios, I would submit that that should probably be preserved in an audited environment with the appropriate certifications. What is important is having proof that there was a rigour around validating the cashflow results at the inception of the deal.
The other thing that can be done is to prove that there is an objective approach to determining whether the new pool is like or unlike prior ones, supported by statistical evidence. If any fundamental changes make the current pool different to prior ones, then it is important for issuers to put together a statistical and objective regime to disclose that.
Q: How does the asset review provision work?
A: Kudenholdt: Currently, pre-offering asset-level reviews include things like comparison to underwriting criteria, data accuracy, property value and some other things along those lines. The new asset review provision is different in a number of respects.
Firstly, asset reviews must be geared more towards whether a representation or warranty was breached that could require a repurchase by the issuer. Importantly, all RMBS transactions today in the Rule 144a markets already embrace a post-issuance review procedure, so that when a loan triggers past a certain delinquency threshold it is mandated to be sent out for review by an independent reviewer to determine compliance with reps and warranties and most particularly with compliance with the underwriting criteria.
The market has already moved in the direction of adopting provisions that provide an asset-level review post-issuance, but what we are seeing actually emerge in transactions is quite different from the Reg AB 2 requirements. What we are seeing in transactions is a loan-by-loan review from day one for any loan that passes a certain delinquency trigger or other loan-level event, whereas under Reg AB 2 there is a pool-wide review only to take place if the entire pool performs at a level past a specified delinquency trigger.
Q: What is the practical impact of the new requirement for an annual compliance check?
A: Kudenholdt: One of the additional registration requirements is that you have a clean twelve-month history of having complied with the ceo certification requirements and the asset review procedures requirements in your transaction documents. They also added to the existing requirement for filing a shelf registration that you as an issuer must not have failed to file any required Exchange Act report during the past twelve months.
On an annual basis, the issuer must meet those criteria on a look-back basis. What is most different about the new rules is that it is possible for an issuer to lose shelf registration eligibility because they failed to file an Exchange Act report during the course of an up and running and effective shelf, which was not possible under the old rules.
Q: What details do we have about the Reg AB 2 pilot programme?
A: Myers: The SEC formerly announced on 6 October that there would be a pilot programme, with issuers and sponsors invited to apply to participate. The details of the pilot programme are not fully hashed and vetted out yet, but the expectation that the SEC disclosed will be that the issuers participating in this programme will have a new draft registration statement ready for review around February 2015.
Q: What are the systems and operational changes that issuers should be working on immediately?
A: Myers: From a timeline perspective, the very first thing I would suggest an issuer does is perform a gap analysis between the required field elements for their asset classes and those they have historically accessed, even if they have not been providing loan-level data for an aggregated format. That is also true for first-time issuers; look at all the fields that are required and see which ones you cannot provide.
Second, I would assume that originator systems and receivable accounting systems that were not designed with securitisation in mind will not be the likely way to fill those gaps. You will likely need a separate database and securitisation environment to comply with the required field elements of the asset-level disclosures.
Beyond the asset-level disclosures, the next thing I would recommend is developing a plan on how you are going to manage the reps and warranties tracking and other nuances of maintaining your shelf eligibility that could trip you up. Third, develop a system for the data integrity checks because all the fields are required.
I think the market will eventually penalise issuers who do not provide all the data in substantially a correct format. So, having a process to make sure the data is of high quality is important.
Fourth would be working on making sure that you can provide the required forms. Then lastly you can worry about getting that data and the ability to publish it in an XML format; once you've done the first steps, that becomes a much easier task.
Q: What is the outlook for US public and 144a markets?
A: Kudenholdt: I think it will be difficult to do public RMBS offerings because investors will want more granular data. Most of the securities are sold to very sophisticated institutional investors, who are accustomed to having their own proprietary models and running their own information and assessing the risk, so I do not think they are going to be happy with the two-digit zip code.
Ultimately, I hope the SEC would make an accommodation, because there are benefits to investors of being in the public market. There are additional protections and benefits under the securities laws for registered deals that investors do not get in Rule 144a offerings. And I would also hope that the SEC makes an accommodation to permit investors to have those protections but also get the information that the most sophisticated among them are going to need.
Myers: That is one of the unintended effects that Steve mentioned, which is by adopting a regime that is actually perhaps less robust than the existing market practice and at the same time leaving the 144a option open to avoid the low-level disclosure that was originally going to impact both public and 144a deals, I think that there is a risk that the private label mortgage market takes a step back. That would damage liquidity and the reach of the investor base that would be available in the public market.
It is not so dire in the other asset classes, where we already mentioned for example that CMBS is largely consistent with existing market practice. The public markets are attractive enough that the big issuers are likely to continue to access the broader investor base and take advantage of the public markets.
It is possible that some of the specialised finance companies or those who do not regularly issue in public may wait on the sideline to see how this takes hold. But I think there is a likelihood that the SEC, in addition to looking at other asset classes that might require asset-level disclosure, may revisit the 144a loophole at some point in the future.
Q: What are the issues that the SEC has left unresolved?
A: Myers: The initial proposal had nine asset classes where asset-level data was going to be required and it was very clear when Rolaine Bancroft made her comments that the Commission did not make a definitive determination that it was done with rulemaking around asset-level disclosure. I think that the political pressure to get Reg AB 2 out the door and complete some measure of Dodd-Frank rulemaking necessitated the Reg AB 2 that we got.
There is an open question with respect to student loans and equipment loans and leases. I think there will be a Reg AB 3, which will address that.
Cashflow waterfall disclosure is another open issue, which the SEC will likely conclude on. Additionally, the greater disclosure the market may require - particularly in RMBS, as we have mentioned - may call for some additional rulemaking.
Q: What are the best and worst aspects of these new rules?
A: Kudenholdt: The best aspect has to be that they were not extended to Rule 144a. It was very much the right decision by the SEC to be cautious with this rulemaking to avoid harming liquidity and harming the markets.
Another really good aspect of the rule is that investors are going to have more time to look at information. It is important to note, however, that although the rules have not been explicitly extended to Rule 144a offerings, practitioners in the space certainly do look to the public rules for guidance, so we will probably see some effect in the Rule 144a market in terms of getting disclosure documents to investors earlier in the process.
As for the worst aspects, the one-size-fits-all approach is not necessarily the right one. It is a noble objective to have all this data always freely available to the public, but it does not sit too well with these particular markets because there are investors who simply require further granularity and disclosure.
