Do you remember the Ford Pinto? The 1970s car had a nasty tendency to explode into flames in rear end collisions. But the piece de resistance was when litigation exposed a Ford internal memo that showed the company was not only aware of the problem, but had run the math and concluded reinforcing the car would be more costly than compensating victims.
Similar logic is at work in the protests coming from the asset backed securities sector over plans to reform their ratings. Mind you, one of the ideas that is generating stiff resistance is an incredibly tame change, namely, that a structured finance instrument is a structured instrument by adding a special designation, such as “sf”.
Now why is this change warranted? Even though research has found that ABS have less desirable performance characteristics than corporate bonds, banks were required to hold only 1/5 the capital against mortgage backed securities rated AAA or AA than they would be required to hold against corporate bonds. For institutional investors, the problem appears to be more mundane: they evidently didn’t differentiate much (in terms of credit exposure) between ABS and other rated instruments.
In case you have any doubts that the differences aren’t trivial, please read the excellent paper by Joshua Rosner and Joseph Mason, “Where Did the Risk Go? How Misapplied Bond Ratings Cause Mortgage Backed Securities and Collateralized Debt Obligation Market Disruptions.” Among the issues they raise are:
1. Corporations are dynamic, and companies can and do take action to forestall credit downgrades (particularly below the investment grade level, which greatly restricts access to capital). Indeed, with a corporate instrument, investors can experience an upgrade. But with a structured vehicle, the only changes possible are adverse
2. When mortgage pools underperform, they will not be downgraded as quickly as under traditional methods
3. The statistical loss performance of asset pools is skewed compared to corporate investments
Their analysis is exhaustive, and includes a section that argues vehemently (for an academic paper) against the rating agencies’ First Amendment exemption from litigation for their opinions.
It’s not hard to conclude that ABS need to be scored differently, yet the industry reaction is vociferous. I was appalled by the Bloomberg story, “SEC `Scarlet Letter’ Drive Hurts Asset-Backed Market .” Normally, Bloomberg is pretty even-handed, but this piece was one-sided and alarmist. It was telling that the first quote came from the head of the American Securitization Forum, the industry lobbying group.
What was most disturbing was the that problem that the industry was worried about was that a new system would force investor to rethink how they use these instruments. My God, if it really is true that they haven’t done so already, despite the considerable evidence that some of these products need a black-box warning, and this move can force the lazy to get off their duffs and make badly needed changes in investment policies, it is exactly what is needed. Similarly, banking regulators want their charges to hold more equity, so they would not oppose this move.
The whole discussion has an Alice in Wonderland quality, and it says a great deal about the overrated quality of institutional investors. The reaction is something along the lines of “Oh, if the rating agencies tell us these instruments are different, then gee, we might have to treat them differently.” Huh? It takes the SEC hitting you over the head for you to recognize that these instruments aren’t what you thought they were?
These sources pretend to be investment professionals and they didn’t understand these ABS well enough to know they didn’t perform the same as corporate credits? Worse, even if they naively thought so pre-2007, they still carry the same beliefs despite repeated instances of ABS being downgraded multiple grades in a single rating review (and some CDOs going from AAA to junk)? That never happens with corporate debt issues except for accounting fraud or a major disaster, yet it is routine in ABS land.
Despite the evidence above that his prod is badly needed, the securitization proponents are reacting as if a patient that has just had a quadruple bypass shouldn’t be pushed to quit smoking, go on a diet and start exercising. Oh, it will restrict access to credit! Oh, it will hurt the banking industry! Now isn’t the time to do such a horrid thing! By definition, there is never a good time to toughen regulations. When the markets are weak, the incumbents protest that the change is another blow when they are already down (even when, as in this case, the changes are in response to a mess they helped create). And of course, when markets are favorable, the argument goes, “Things are just fine. Obviously, there is no need for change.”
The telltale sign that the article was unduly influenced by ABS backers: evidently no reporter bothered talking to Mason and Rosner or anyone familiar with their work. Deleveraging isn’t pleasant. Get with the program.
From Bloomberg:
Regulators’ plans to add a letter to credit ratings of asset-backed debt may constrict the $4.6 trillion market and choke off consumer credit at a time when Federal Reserve Chairman Ben S. Bernanke wants more lending to bolster the economy.
The U.S. Securities and Exchange Commission may recommend this week that Moody’s Investors Service, Standard & Poor’s and Fitch Ratings include a new designation to the scale created by John Moody in 1909, according to people familiar with the plans. The changes may force investors to reassess the way they gauge the risk of securities backed by mortgages, student and auto loans and credit cards, said one of the people, who declined to be named before the announcement. The action could force banks to add capital to guard against losses or curb lending.
The banking industry is “very significantly concerned,” said George Miller, executive director of the American Securitization Forum, a New York-based group representing 370 companies that package assets into bonds. “If the rating itself is substantively changed, or the symbology is changed, it’s not just investment guidelines that have to be examined.”…
The SEC staff is recommending giving ratings companies two choices, the people familiar with the agency’s plans said. One option is to publish a report on how they came up with each ranking and how it could go wrong. The other would add a designation distinguishing the assessment of asset-backed debt from a corporate bond.
Florida State Board of Administration Interim Executive Director Robert Milligan, who oversees pension funds and local government investment pools with more than $159 billion of assets, said changing the ratings symbols for structured finance would force a review of whether funds in his state could buy such securities.
The International Organization of Securities Commissions, a Madrid-based association of global regulators, issued a code of conduct on May 28 that also recommended a change in ratings codes for asset-backed securities. Moody’s said in February it may add “.sf” to assessments to signify structured finance. S&P said a day after the IOSCO announcement that it may add an “s” to rankings.
“If, all of a sudden, the security has a scarlet letter on it, maybe regulators won’t judge that so well,” said James Grady, a managing director in New York at Deutsche Asset Management, which oversees $240 billion in fixed-income securities. “What will it do to liquidity, valuation and capital requirements?”….
The number of collateralized debt obligations failing since October has reached 186, with $202 billion of assets, data compiled by S&P and Bloomberg show. That’s 40 times the total for the previous four years, according to the rating company.
To be considered “well-capitalized” under U.S. regulations, banks are required to hold five times as much capital against corporate debt than they are for commercial or residential mortgage-backed securities rated AAA and AA by S&P, Fitch Ratings and Moody’s…..
The treatment granted to asset-backed securities helped fuel a record $7 trillion of sales in the past three years, excluding debt packaged by government-chartered entities such as Fannie Mae, according to Asset-Backed Alert, an industry newsletter…
The changes are opposed by ratings companies, banks, securities firms and bond buyers. Seventy-one percent of investors surveyed by Moody’s, representing $9 trillion in assets, didn’t want new assessments, the company said in May. A special designation might diminish the value of investors’ holdings by making debt from different asset classes harder to compare, the survey said….
Investors may steer clear of structured-finance in the future, said Richard Metcalf, director of the corporate affairs department of the Laborers’ International Union of North America.
“ If there are going to be certain financial products which are flagged as having increased volatility or risk, then we may have to look at whether such products fit into our portfolios and, if so, where,” Metcalf said.
23 comments
Comments are closed.
The (main) problem with the proposal is that it will create massive headaches for investors, regulators and issuers, without actually providing any useful information. The big lesson of the credit crisis is that investors relied too much on the triple-A ratings instead of doing their own analysis. How is this idea going to help that?
It’s hard to believe that many if any investors simply didn’t know they were buying a structured product. The point is that many investors didn’t know or didn’t care what that entailed.
What the industry is pushing for instead is supplementary ratings of non-default risk, covering things like loss given default, rating volatility, collateral quality (regardless of structural features) and sensitivity to changes in assumptions. Moody’s and Fitch have already begun consulting on them (S&P is looking at the idea, but has so far only gone down the SEC favoured identifier route). Surely these would be far more useful to investors and would achieve the purpose of an identifier without ghettoising all of structured finance because the rating agencies cocked up in certain sectors.
It’s important to recognise that outside of those sectors (far too many, it’s true), structured ratings have actually been more stable than corporate ratings. There’s still sensitivity to assumptions, and it’s right to flag those up prominently, but it makes no sense to think that anything meaningful will come from labelling all structured products alike.
“How is this idea going to help that?”
Don’t know, but since it will flag products that are structured and thus by definition more opaque, it will force investors to provide a justification to regulators or to their bosses why they believe the product has a certain level of risk. Whether that’s the “ghetto”, I don’t know.
Wouldn’t that aim be better served by management and regulators simply demanding explanations of the risks in portfolios, which is something they surely should be doing anyway? It’s not like securitisations hide their structured nature, so you need a special marker to realise an ABS is an ABS.
This seems to be delegating responsibility to the agencies all over again. Simply using a signifier seems to compound the mistake of overreliance on headline ratings. “Oh it’s a AAA.sf. Better watch out. That’s a AAA. That’s OK.”
Also I don’t agree that structured products are “by definition” more opaque. If you buy the senior unsecured bond of a public bank, all you have to go is their quarterly reports, which as we’ve seen over the last year are far from transparent. If you buy that bank’s RMBS, on the other hand, you get a monthly breakdown of portfolio composition, delinquencies, defaults, recoveries, reserve levels, administration costs and so on. Now of course there are more risks to take into account when buying an RMBS, but that’s complexity, not opacity. And certainly there are some structured products that are opaque – again, my problem is not with identifying risks, but with slapping a label on all structured products and pretending it means something useful.
Ginger,
Interesting how YOU were the one who stated that requiring an “SF” tag with a rating would have the effect of “ghettoising” all structured financial instruments. Seems to me that “scarlet letter” is not be a good analogy for the using the SF designation. Rather, it’s more akin to a warning label,- “Warning- this product is multi-tiered and therefore may require a more complex level of due diligence
on the part of any buyer”.
“Wouldn’t that aim be better served by management and regulators simply demanding explanations of the risks in portfolios, which is something they surely should be doing anyway?”
Well then what’s the problem? You seem to be against doing something which will cause E by saying that people should be doing E anyway.
Sure a lot of problems in the world could be solved if people did their jobs perfectly. Only they don’t.
The point is that the existing rating system is deeply embedded in legal and regulatory frameworks and investment mandates. Changing the rating scale for all structured products without changing the regulations (which in the case of the EU are written into law and would take a long time to change) would force Basle II banks to deduct their entire ABS holdings from capital regardless of risk, which would make most of the European banking system insolvent. It would also force real money investors to sell holdings they are perfectly happy with, which would only exacerbate the crisis, unless they were able to change their mandates (again, not a simple process).
“”Warning- this product is multi-tiered and therefore may require a more complex level of due diligence
on the part of any buyer”.”
Which is precisely what the supplementary ratings do, while providing useful information to investors, management and regulators about how the specific risks of a structured security differ from corporate bonds and other structured products.
“Well then what’s the problem? You seem to be against doing something which will cause E by saying that people should be doing E anyway”
See my first para. I’ve got no problem with regulators or management tailoring their requirements to specific risks. That is in fact what is happening – the Basle committee is going to increase the capital requirement for CDOs squared and similar products, for instance, and is revisiting trading book treatment of ABS to reflect the vastly higher price volatility seen over the last year.
Indeed, with a corporate instrument, investors can experience an upgrade. But with a structured vehicle, the only changes possible are adverse.
Actually, structured tranches can be upgraded too. If a deal performs well and defaults come in below forecast, then there’s a natural upward drift to the tranche ratings.
You don’t even need to exceed forecasts in some cases. Depending on the payment and reserve structure , amortisation often builds up credit enhancement so that a subordinated bond performing exactly in line with expectations (or even slightly worse) is upgraded over time. This happened to two Dutch RMBS and a Portuguese SME CDO yesterday, for instance.
It is true that a AAA tranche can only be downgraded, but it will deleverage over time (again, depending on the payment structure, but in almost all cases other than a master trust).
I should add that my comments about European banks apply in much the same way to European insurers under Solvency II, and insurers are by far the largest real money buyers of ABS on this side of the Atlantic.
I’m lost; this discussion is that if you make these technical changes that the EU financial companies will go broke. How can they go broke? Unless they are already broke, but are kept afloat by accounting gimmicks.
What am I missing here?
GK
“Changing the rating scale for all structured products without changing the regulations (which in the case of the EU are written into law and would take a long time to change) would force Basle II banks to deduct their entire ABS holdings from capital regardless of risk.”
Well I guess we have to have a notion of what percentage of the structured deals are bad, right? If most of the deals are bad, then it’s better to exclude everything, rather than the present situation, which is to include everything. So sure the optimal is to include only the good and exclude the bad, or even better to weight everything precisely according to its risk, but that’s not going to happen.
So I still think you are sort of saying, either we do this perfectly, or we don’t do anything at all. And that doesn’t seem to me to be a wise choice, since the first isn’t going to happen, so we’ll be left with the second.
The reason that structured products need different ratings is that all tranched products (except the super senior tranche) are leveraged on the downside, but carry fixed rates on the upside. For this reason their ratings can never be comparable to corporate ratings.
If BaselII needs to be fixed at the same time that .sf is added to the ratings — then the industry should be calling for a coordinated change — not for no change at all.
“The reason that structured products need different ratings is that all tranched products (except the super senior tranche) are leveraged on the downside, but carry fixed rates on the upside. For this reason their ratings can never be comparable to corporate ratings.”
Actually, tranched products are much like corporates, set up with a debt side and an equity side. The equity in both a corporation and a CDO (for example) provide subordination to the debt holders and provide a levered return to the equity equal to the excess over cost of liabiities times the leverage ratio (simple version). If a company goes bankrupt, it’s equity much be wiped out before the corporate bondholders are touched much like the equity of a CDO must be wiped out to touch the rated liability tranches. So, tranche debt in structured products are really no levered to the downside compared to their overall ratio than corporate bonds are.
Anon of 4:51 PM,
While what you say is technically accurate, the inference is incorrect, and Huh?’s less precise wording is far more accurate.
You will never get better than your initial rating with a structured product. The distribution of credit outcomes is skewed to the downside. With a corporate bond, you can get an upgrade, and perhaps most important, management can and will take action to prevent downgrades into the non-investment grade area if at all possible. The distinction between an entity that is actively managed and motivated to survive is different than that of a passive pool.
I suggest you all go read the relevant section of the Mason/Rosner paper. It makes the difference in behaviors quite apparent.
I am reminded of the 1990s UK sitcom “Keeping up Appearances”.
JdlO
“You will never get better than your initial rating with a structured product. The distribution of credit outcomes is skewed to the downside. With a corporate bond, you can get an upgrade, and perhaps most important, management can and will take action to prevent downgrades into the non-investment grade area if at all possible. The distinction between an entity that is actively managed and motivated to survive is different than that of a passive pool.”
As described above, this simply isn’t true, except for the most senior tranches. Even in passive (static in the terminology) pools you can and do get upgrades for subordinated tranches, and in managed pools (ie most CDOs in the years prior to the credit crisis) there was the theoretical possibility of upgrades through active management of the portfolio. Obviously, interests are/were not always aligned, but this is also true in corporate finance.
Once again, please don’t think that I’m saying the risk in a typical structured product is the same as in a typical corporate bond. What I’m saying is that the risks differ in a way that a simple identifier does nothing to elucidate. I’m absolutely in favour of more information being provided about the idiosyncratic risks of a given structured security, but a blanket identifier doesn’t help anyone but a complete idiot. I can’t stress this enough – the problem wasn’t that people didn’t know they were buying structured bonds, the problem was that they didn’t care why they yielded more than conventional bonds. An identifier tells you nothing about that, whereas complementary ratings do.
Anon 4:51:
Not like corporates:
The only truly debt-like tranche is the super senior. All other tranches are an equity-debt hybrid.
While some of this is over my head I suspect there are many investors like myself that trust(ed) their brokers advice. I was in an ABS as well as an ARPS that won’t auction. I think that a -SF identifier is an excellent idea to warn that maximum due diligence is required. When the investment community as well as their lobby mounts such an adversarial position – something stinks. Any investment business attempting to block clarity has suspicious motives.
Bruce
Coming from a bulge-bracket investment bank…I cannot tell you how many credit officers when approving derivatives facing structured finance vehicles cite their “AAA” status as a reason to approve swaps facing the senior tranches. Many of those swaps are now deeply underwater and the i-banks who wrote them are facing incredible, mounting exposures. If even credit officers at i-banks do not distinguish btwn AAA corporates and AAA-rated ABS, MBS and CDOs, then how can anyone think that adding this simple distinction will not help the situation?
Please correct me if I am misinformed but weren’t we talking, about 3 months ago, when this mess got ugly, of the need for investors to do their own due diligencene and not rely on ratings? Now we want to add a letter to an alphabet soup of ratings and think that solves the problem!
Let’s look back to a couple of good points that were raised about the entire system.
First, it seems to be a non-regulated, regulated industry. By that I mean that various governmental agencies approve the qualified raters (reluctantly by historical standards) and mandate that a significant number of purchasers of financial instruments base their investment decisions on the ratings issued by those chosen to perform the task.
Second, the agencies rating the securities are paid by the issuers that need their impramatur which is essentially the definition of a conflict of interest.
So now we tweek the system and proceed with business as usual?
How about doing something unconventional. Why not abolish the whole concept and tell the investor to figure it out. It might give rise to a whole new industry. For instance, the pension funds might band together and form their own internal rating system. If the issuers want an opinion from someone they can pay for it and disclose they paid for it. Or if the buyer wants a really unbiased opinion he can go on line and pay for that as well. Open up the system and let the market provide the information based on demand.
This is probably heresy in a world gone mad for reregulation. Nevertheless, you can only patch an old tire so many times. It seems as if the rating agency concept is a very leaky tire that can’t stand a whole lot more stress. Maybe it’s time to rethink the whole process.
Tom,
The problem isn’t merely regulatory. The requirement to hold investment grade securities (or stick with certain rating standards) is enshrined in legislation and in a tremendous number of investment agreements. Even the Fed uses ratings to determine what it will take at the discount window and what haircuts apply. At this juncture, getting rid of ratings would be like trying to unscramble eggs.
Yves,
I recognize the fact that there are a number of impediments to the idea but is that a valid reason to stick with a flawed system?
Eggs can be and have been unscrambled before. You might have to stage reform but that doesn’t mean you don’t attempt reform.