For the last two years or more, one word has been used in conjunction with securitisation more than any other transparency. Critics argued that a lack of transparency in securitisation contributed to, or even caused, the credit crisis, while nearly everyone agrees that improved transparency was necessary, but not sufficient, to restart stalled primary markets.
An array of industry initiatives have been undertaken to improve transparency, particularly in RMBS, with limited results so far. Frustrated by the lack of progress, regulators and central banks in the US and Europe have in recent months proposed strict regimes which would greatly increase disclosure in the market.
At the core of all the new disclosure regimes is something that many investors, particularly those lower down the capital structure, had clamoured for, to little effect, for many years before the credit crisis loan level performance data.
Before the crisis, a handful of issuers, mostly of subprime RMBS, provided loan level data to investors on request. The vast majority, however, shrugged off investors demands. With spread differentials between the highest quality and lowest quality deals being a handful of basis points, they had little incentive to put the work in.
When the credit crisis struck, however, the tables were turned. Now issuers needed all the investors they could get. Moreover, securitisation came under attack from all sides as opaque and riddled with collateral that was not what it seemed. Once politicians stopped demonising securitisation, realising it was needed to keep credit flowing, they focused on transparency as the key to its renaissance.
Indeed, one of the very first large scale regulatory responses to the credit crisis, the amendments to the Capital Requirements Directive known as CRD II, will require bank sponsors to give prospective investors "all materially relevant data on the credit quality and performance of the individual underlying exposures, cashflows and collateral supporting a securitisation exposure as well as such information that is necessary to conduct comprehensive and well informed stress tests on the cashflows and collateral values supporting the underlying exposures".
There is no mechanism in CRD II to enforce compliance by originators. Instead, the burden is on investors, who must prove to regulators their ability to analyse their securitisation holdings on a continual basis, and potentially face a steep capital penalty if the originators do not hold up their end of the bargain.
Despite this impetus, it was the in US where progress was quickest, thanks in part to the lack of inter-jurisdictional complications. The American Securitisation Forums Project Restart created a template for residential mortgage backed securities, followed by a unique loan identifier allowing loans to be tracked as it was securitised, called and resecuritised.
In Europe, progress has been slower, despite the European Securitisation Forums efforts to get originators to sign up to its standardised data template. Many of the largest UK and Dutch issuers have agreed, but almost all are only committed to using the template for future issuance and even then, many key data fields are optional. Moreover, for many jurisdictions no formal template yet exists, beyond the European Central Banks draft proposal for ABS in its repo facilities.
"I dont see any improvement in reporting quality," says Reto Bachmann, head of European securitisation research at Barclays Capital. "Its just as good or bad as it was before. Deals issued this year are exempted from the CRD, but you would think its in the issuers interests to prove they can deliver the requirements."
Although, or perhaps because, the industry in the US has been quicker off the mark, the US regulatory response has been slower. Some conflicting disclosure requirements have been included in financial reform legislation, but this has yet to be signed into law. The SEC did not lay out its plans for disclosure until April this year, but it did so in a sweeping manner if fully implemented, the SECs new rules would bring about by far the most radical change to securitisation disclosure in the world.
In addition to the basics of standardised loan level disclosure the SEC has prescribed templates for a wide range of asset classes the new regime implements a broad conception of transparency. Issuers would have to provide final prospectuses, minus pricing information, five business days before the first sale of a securitisation offering. Cashflow models would be made public. Private placements would have substantially the same disclosure standards as public deals. Repurchase requests from the trustee would be monitored by a third party who would confirm whether or not the sponsor complied with its repurchase obligations.
The Bank of England has also taken a broad view of transparency in its proposed eligibility criteria for ABS. For instance, it plans to require that in addition to the prospectus, "all underlying closing transaction documents" be made available, and updated whenever they are changed. Furthermore, the Bank proposes to require monthly reporting of the mark-to-market value of any swaps in a transaction, something that many investors would welcome.
"Its extremely difficult to get the documents at the moment," says Mark Hale, chief investment officer at Prytania. "If you have a stressed deal from three years ago, 50% of the cashflow can be going to the swaps. Its critical for valuation."
The most intriguing element of some of the new disclosure regimes is the requirement to make deal models public. Both the SECs revisions to Reg AB and the Bank of Englands proposed discount window criteria provide for issuer or arranger endorsed waterfalls, though their approaches are very different. So far, neither the European Central Bank nor the Capital Requirements Directive has made this disclosure compulsory indeed, the ECBs disclosure requirements in general take a much more constrained view of transparency, focusing on collateral data.
In some ways, the waterfall models are even more important than loan level disclosure, particularly for more subordinated tranches, as they allow investors to more accurately model the performance of deals when triggers are breached. As deals such as Granite have shown, investors can be exposed to considerable risk even when the collateral performs fairly well
"While it can be argued that investors always had access to the waterfall information since the priority of payments was described in the offering documents, converting the narrative of such calculations into their parallel algorithms can be a daunting task," says Ann Kenyon, a partner at Deloitte in a note on the SECs proposals. "One potential positive of this requirement may be to enhance the clarity of the language in the document, and highlight potential areas of differing interpretation among the deal parties."
The SEC will require issuers to file on Edgar, the code for a programme, written in Python, which instantiates the cashflow waterfall and allows investors to plug in the loan level asset performance data. The Bank of England, by contrast will be satisfied that the deal models are "publicly available" if they are published on an approved data services initially Intex and ABSXchange or on an issuer maintained website. Currently these services, along with some other third parties, construct their own cashflow models based on an analysis of the available deal documents, opening up the possibility of error. Naturally, they charge for their services.
"There are enough vendors out there that produce cashflows before the deal goes to market that I dont think thats the big issue," says Usman Ismail, global head of sales at Lewtan in New York. "Do not take that capability away from the industry let the industry create the models. The industry has enough people to create models to create transparency. What you want to make sure is that all the documentation thats required to make the models is available. At least in Europe, one of the biggest things is to do with swaps. You find that not everyone makes the swap documents readily available. As a result, when you model a deal you have to make some assumptions as to what the swap documents are saying."
Another potential problem with the BoEs proposals is that the waterfall model could still be a black box, making it hard for sophisticated investors to tweak it to their satisfaction, and so far the Bank has given little detail on how it should be constructed. Still, any disclosure would be better than the current situation.
"Everyone will make use of it somehow," says Bachmann. "Some people would run the deal models themselves. Some people can see the Intex code and will modify the model to suit their own stress scenario. Some would just ask their IT department to set up a model and run it. At another level down, the investors wouldnt run them, but would use them to understand the capital structure, for example the effect of swap costs."
Despite the many improvements contained in the new disclosure regimes, investors are still pessimistic about the impact they will have.
"Whats holding back securitisation and may even kill it is regulatory risk," says Bachmann. "Im not sure loan tapes and deal models can overcome this."
Moreover, the fragmented nature of the regimes is worrisome for both originators and investors. Originators, if they want to access the widest possible investor base, have to ensure their systems can meet the demands of all the regulators, capturing the maximal amount of data and outputting it in various formats on multiple schedules. Meanwhile investors face the prospect of conflicting definitions and multiple data sources."Potentially it could provide a great deal of stability if its done well and is comprehensive," says Hale. "The increased costs compared to the wider gains for everyone are not particularly material. Conceptually, its a big net gain, despite some interim costs. The difficulty is that the likelihood is things will not be ideal, we wont get co-ordination and commonality well get disparate requirements and different standards. That will end up massively increasing peoples costs and adding to confusion, therefore significantly undermining the net gains that should come out of the whole process. There is a prize to persuade a lot more investors to become much more involved than they are now. That prize is still within grasp, but unfortunately at this stage I dont see that we are going to get there."