Subject: File No. 4-661
From: Bill Harrington

May 3, 2013

Mr. Losice:

Thank you for the link for the May 14 agenda. I will submit materials to aid the roundtable next week.

Also, I volunteer to serve as panelist for the third panel. Several topics fall into areas that I have assessed for the benefit of the SEC and other bodies. My goal is to help avoid another bank bail-out attributable to inaccurately rated securities backed by assets and derivatives.

For instance, Structured Credit Investor published the following interview on April 10 in which I discuss broad improvements to the market for securities backed by assets and derivatives.

Both the link to Structured Credit Investors and the interview text are included below.

Best regards,

Bill Harrington

P.S. Please advise if I am incorrect in assuming that the SEC  values the free speech of highly-informed and unencumbered private citizens at least as much as the self-serving free speech of rating agencies, bond counsel for securities backed by assets and derivatives, regulatory consultants, etc.

"Ratings differentiation

Call for improved assessment of derivatives risk

'Too big to fail' is emerging as a mainstream concern and with it come calls for the assessment of derivatives risk to be improved. For securitisation investors, better differentiation of risk in credit ratings is being put forward as one solution.
In a hypothetical scenario, where the securitisation industry could start over with regards to ratings frameworks, ex-Moody's svp William Harrington believes that ratings should be capped at single-A for deals that contain derivative hedges. At present, senior ratings don't distinguish between transactions that have a derivative at the top of the waterfall and ones that don't, even though downgrade risk is significantly higher for the former.
"Ratings caps would facilitate a more rational investment landscape that enables asset managers to look at their overall portfolio and identify which other factors - not just credit risk in the underlying - they might be sensitive to," he argues. "They could then hedge out their currency risk, for instance, on an exchange or leave the transaction unhedged and be compensated for associated risk. It would engender a better understanding of performance."
Harrington says that better differentiation of risk in ratings would create a clearer alignment of investment objectives across the spectrum from conservative institutional investors to risk-savvy sophisticated investors. He suggests that three distinct investment profiles could be targeted in this way: investors who would like to eliminate derivatives risk entirely; investors who can accept index/exchange risk but don't want derivatives risk; and investors seeking exposure to both index/exchange risk and derivative risks.
At issue is the limited number of ratings categories for structured finance, which means that hundreds of different outcomes converge on only 19 different ratings. "Different types of risks can be borne in securitisations and at present there is no way of distinguishing them. If variegated risks are reflected appropriately in a rating, it's then up to the investor which ones they can bear. A better way to gauge these risks would be to, say, designate them on a scale of one to a hundred," suggests Harrington.
This would also lessen the 'cliff effect' observable in current ratings approaches, whereby the minimal difference in expected loss between triple-A, double-A and single-A means that mild losses can move sharply down the capital structure.
Another issue that needs to be remedied, according to Harrington, is that information - such as differing underwriting standards, as well as nuances between asset classes and derivative type and counterparty - isn't typically disclosed to the end-users of ratings. Doing so would help investors and regulators gain a more granular sense of the risk involved.
"Publishing the vote tally in ratings committees would also help investors form opinions about contentious decisions, as well as follow rating patterns over time," he adds.
Harrington expects the drive towards central clearing of OTC positions will mean that derivatives return to their original function as hedging instruments. It may also force asset managers to scrutinise their derivatives documentation in more detail, thereby shedding light on how confirms are changing over time, for example.
"The broader issue is that assessment of derivatives risk needs to be improved," he observes. "It is analogous to the Y2K systems overhaul in that asset managers should be undertaking as much due diligence on derivatives risk as they do on credit risk. They now have two hedging options - via futures exchanges and OTC clearing."
Finally, event risk should be explicitly modelled by rating agencies, according to Harrington. This would include monitoring how many deals a counterparty is exposed to or whether any counterparty has an oversized exposure to a certain sector.
"There should be an upfront linkage between the ratings of a counterparty and the potential for flip clauses to be triggered," he concludes. "Ultimately, securitisations should be modelled according to whether they are fully hedged, partially hedged or unhedged. An overlay pertaining to where flip clauses are enforceable can be added where necessary."