Like many investors around the world that cant access the Federal Reserve Banks discount window, Paris-based Natixis has a problem: what to do with all the subprime-debt-laden instruments stinking up its balance sheet.
Whereas nearly two years ago virtually no market existed for toxic loan securities, hundreds of billions of which were conjured in the middle of the last decade, there is now at least a semblance of a thaw.
In evaluating possible ways to salvage its cache of residential-mortgage-backed securities, Natixis, which has $600 billion in total assets, turned to a new form of Standard & Poors credit research, known as recovery analytics: S&P sifts through individual mortgages, examines cash flows and, using models and assumptions based on forward-looking projections, attempts to predict which issues are most likely to return the most cents on the dollar.
Thats right. The same rating agency that in large part failed to foresee the downside risk embedded in the RMBS market is now helping investors evaluate possible upside opportunity in some of the very same securities that were once deemed to be investment grade but have turned out to be anything but. Its the equivalent of the comics Charlie Brown, having hurt his back trying to kick the football, turning to Lucy for a rehab regimen.
Not only does Natixis rely on S&Ps research, explains the firms New Yorkbased managing director Ralph Daloisio, who runs the U.S. practice, but Daloisio and his team view the McGraw-Hill Cos. subsidiary as extremely committed to a complete makeover from the inside out as it seeks to restore investor confidence.
Having seen the various business cycles, we know that often the best ratings come after the worst of times, Daloisio says. The magnitude of this crisis has forced the rating agencies to drop the gauntlet on the quality of their information gathering.
That the rating agencies are stepping up due diligence owes to another gauntlet, the size of the Chrysler Building, that has been dropped on them.
Whether inexplicably validating subprime-laced securities or failing to spot train wrecks such as Goldman Sachs Groups Abacus 2007-AC1 synthetic collateralized debt obligation which regulators allege was booby-trapped but which nonetheless garnered a coveted triple-A rating the screwups of the rating agencies have resulted in an exceptionally brutal public scourging that has lasted nearly two years and recently intensified with the release in late April of a stack of damning e-mails uncovered by a U.S. Senate investigation.
Had credit rating agencies been more careful in issuing ratings . . . we maybe would have averted the crisis, declared Senator Carl Levin, a Michigan Democrat and chairman of the permanent subcommittee on investigations, publicly identifying the Big Three rating agencies as chief culprits in the financial crash. But they did not. Without credit ratings, Wall Street would have had a much harder time selling securities, because they wouldnt have been considered safe. The rating houses played a significant role in igniting the crisis, agrees MFS Investment Management chairman emeritus Robert Pozen: There were investors who saw the triple-A as a sort of Good Housekeeping seal of approval connoting something that wasnt there.
Criticism, mainly of New Yorkbased S&P and Moodys Investors Service, and of Fitch Ratings, which is headquartered in New York and London, has been as scathing as it has been relentless. And it hasnt come only from the U.S. The European Central Bank in early May suspended the use of credit ratings with respect to Greek bonds, telling banks it would pay full cash on the barrelhead regardless of ratings possibly the most forceful public rebuke of the value of a triple-A that the industry has ever seen.
Rating agencies point out that they never intended investors to take their ratings as decrees from on high, while also pointing to an underappreciated track record of downgrades that were never heeded. In 2006, S&P handed out what at the time was a record number of RMBS downgrades, but few on Wall Street seemed to notice the trend or the subprime threat until the middle of 2007, after the implosion of two credit hedge funds managed by Bear Stearns Asset Management. The fact that of the $32 trillion worth of debt subjected to S&P ratings, only $2 trillion wound up being downgraded to subinvestment grade (and not all of that has defaulted) has failed to garner any sympathy. After all, $2 trillion warrants a lot of attention. Some 91 percent of the triple-A securities backed by subprime mortgages issued in 2007 alone have been downgraded to junk status.
In recent months evidence has surfaced suggesting that Goldman and other banks specifically recruited analysts from rating agencies to help fine-tune their deals. Meanwhile, New York State Attorney General and gubernatorial hopeful Andrew Cuomo seems to be angling for a showdown with the banks over what he alleges was an effort on their part to deliberately dupe rating agencies to goose their deal grades.
Financial writers have produced volumes of detailed accounts placing the root causes of the great financial meltdown on the doorstep of the rating agencies. Pension funds and states, meanwhile, have filed suits against the agencies, alleging negligence. The Securities and Exchange Commission has gone so far as to purge various references to rating requirements from the rule books (see Raters Biggest Fans: Money Funds, page 40). And Congress, while holding off on tackling wholesale reform of rating agencies, has nevertheless proposed at least one legislative measure that has the Big Three standing at attention. If passed in its current form, the reform bill would subject the agencies to a tougher standard of legal liability, with plaintiffs merely having to show a failure to conduct a reasonable investigation, which is easier to prove in court than outright fraud.
The rating agencies themselves are racing to stay ahead of the pitchfork-wielding mob, acknowledging deficiencies; making substantive changes, such as more firmly cordoning off sales and rating surveillance teams; and taking steps to make their processes more transparent.
Weve brightened the line that divides the business side from the analytics, says Ian Linnell, Fitchs global head of structured finance.
S&P, to its credit, has published online hundreds of pages of new methodology explanations and stress-test scenarios, while inviting feedback from investors on ways it can produce more, and more useful, information. Its almost information overload.
So be it. Rating agencies are practically begging users to do their own research and stop treating ratings as implicit guarantees. Broker-dealers and money managers came to rely on ratings almost as a safe harbor.
Rating agencies believe that, beyond letter grades, their informational offerings will be wanted going forward. Some of the new S&P research, such as recovery analytics, is being given out for free, although in much the same way a heroin dealer gives out free samples to potential junkies.
Unlike the investment banking industry, the rating agencies are vocally supporting reforms to end certain questionable business practices; they are publicly owning up to, not privately lobbying against, important changes.
We know we have a credibility problem, says S&P president Deven Sharma. He asserts that he welcomes the elimination of certain SEC rules requiring ratings, even though it could undermine his companys livelihood. Its in our interests to improve our standards.
Sharma isnt the only rating executive to demonstrate candor about failures and the need for improvements.
Moodys CEO Raymond McDaniel would not agree to be interviewed for this article, but in a speech he delivered in Davos just months after the crisis began to unfold in 2008, he admitted to a panel: In hindsight it is pretty clear to us there was a failure in some key assumptions . . . the key assumptions failed in part because the information policy, completeness and veracity feeding the agencies . . . was deteriorating.
Relying on information provided by the beneficiaries of the financings without proper verification that was the rating agencies Achilles heel, says Natixis Daloisio, who is also chairman of the American Securitization Forum, an industry trade association based in New York.
If trusting without verifying was akin to a painful and debilitating lower-extremity injury, then the fact that the agencies allowed revenue-generation interests to outweigh proper dedication to due diligence could be compared to a spinal cord severing, and its one of the primary reasons so much unsuitable subprime RMBS paper infiltrated and destabilized the financial system.
Among the most unscrupulous elements of the financial crisis that is at least being acknowledged by the agencies: ratings shopping, an open secret on Wall Street in the mortgage securitization heyday of 200506. Issuers deliberately simultaneously engaged all three of the biggies in discussions to possibly rate a deal. Banks, basing their decision on who appeared willing to hand out the highest rating, would give out assignments worth hundreds of thousands of dollars in fees to those agencies. To stifle the shopping practice, Minnesota Democratic Senator Al Franken recently proposed an amendment to financial reform legislation that would set up an independent entity to select rating agencies. Meanwhile, the SEC is taking its own steps to stamp out rating shopping, requiring raters of structured-finance deals to disseminate all pertinent deal information with the unhired rating agencies. The measure took effect in June, and although there are some specifics to be worked out (such as how to maintain password-protected Web sites), there is a more pressing question: Who is going to take the time to use that shared information to produce a research report for which the agency might not be compensated?
Sketchy as it may seem, mandated information sharing is still a major development. Though the Big Three are the best-known agencies, there are 11 nationally recognized statistical rating organizations, or NRSROs. Under the new SEC rule, if two NRSROs are hired to rate a new RMBS issue, then all of the necessary data to do that rating becomes available to the other NRSROs, thus fostering, potentially, a new market for unsolicited ratings ratings that presumably would be more independent-minded, less corroded by conflict and more valuable to the buy-side marketplace.
This, in other words, might be the future of ratings.
Agencies could start to differentiate themselves in a closely followed second opinion market, particularly if credit views are starkly divergent. In April, when Redwood Trust came to market with an RMBS transaction the first such deal posthousing crisis it carried a Moodys-issued triple-A rating. In a first for the agency, S&P shortly thereafter issued an unsolicited commentary declaring that in its view that some pieces of the issue would not have warranted a triple-A had they rated it. Though this was only a commentary piece published for anyone who cared to read it as opposed to a formal, albeit unsolicited, rating S&P has signaled that, when warranted, it can bring unique research to bear. Eventually, the thinking goes, investors will be willing to shell out for such information. But is a subscriber-pays model, long ago discarded, really sustainable?
MFSs Pozen, who before he was chairman of the Boston-based fund manager was a senior executive at Fidelity Investments, says bluntly that it is not.
As large investors, we do our own credit research, Pozen says of his firms portfolio managers and analysts. If asked to pay for ratings, I dont think we would.
S&Ps Sharma is more optimistic about the future even one bound to be drastically different. When we meet with investors, we get asked the same question: How do we know you wont slip back? And the answer is, because of the many organizational changes weve made, changes to the process and to how we monitor the process, he says. Theres more transparency and more accountability. Ultimately, the marketplace, not regulators, will decide if the changes we made were enough. We think there will be a demand for our research across many areas. We think the future looks bright.
Scrutinizing debt instruments as a business unto itself goes back more than a century and a half, to the listings of railroad bonds published in Henry Varnum Poors American Railroad Journal in 1859. One year later, Poor, on a campaign to better publicize details of corporate operations for investors, published The History of Railroads and Canals of the United States, effectively issuing historys first known credit ratings.
A half century later, John Moody founded his own railroad investment guide. It used letter grades to assess risk. Continuing the historical connection between the railroads and the rating agencies, following the collapse of Penn Central Transportation Co. in 1970, federal law required bond issuers to secure a third-party assessment of an offerings creditworthiness. All aboard: The modern business of ratings was pulling out of the station.
But a whole new set of financial industry conflicts had been unleashed. After all, bond issuers wound up paying for the ratings, so what would prevent the rating agencies from becoming overly beholden to the men who buttered their bread?
We dont have to go back too far to find a monumental example of rating agencies undermined by conflicts just one decade, to the Enron Corp. debacle.
In the 2002 aftermath of that companys epic collapse, many large institutional investors expressed more than a little consternation at having been given a false sense of security by the surveyors of Enrons creditworthiness, which had been seen as sturdy enough. In fact, before November 28, 2001, the Big Three all rated Enrons debt as investment-grade. A few days later Enron, downgraded to junk, went bankrupt, a victim of a vicious circle of its own making. A 2002 Senate Governmental Affairs Committee report on the Enron episode took the SEC to task for not properly regulating the rating agencies. The agencies, meanwhile, complained in public testimony that they had been misled by Enrons corporate officers.
After Enron, rating corporate debt would become only a small part of what the agencies did; securitization deals soon drove the rating industry. As Columbia Law School professor John Coffee Jr. told Congress in 2007: Structured finance accounts for a major share of total revenues . . . equally important, these amounts are paid by a small number of investment banks that know how to exploit their leverage.
In 2006, the same year President George W. Bush signed into law the Credit Rating Agency Reform Act a not exactly timely response to the accounting-scandal era an estimated $1.6 trillion in securitizations came to market. Most if not all of these deals required at least two rating agencies to be consummated. Congress had done its best to increase rating agency accountability and transparency. Of course, Wall Street banks and hedge funds were already learning everything there was to know about how ratings were assembled and bestowed.
Former rating agency analysts routinely ended up joining investment banks after a few years on the job. Shin Yukawa, a member of the Goldman Sachs team that created the Abacus deals, had been hired away from Fitch in 2005. Hedge funds also set out to crack the rating code.
As one version of a story goes, three rating agency employees walked into a hedge fund in 2005 as part of an ongoing investor education effort. Only one came back the other two were hired on the spot.
Structured-finance deals were booming by the middle of the decade, and Moodys quickly became the leader. The firms annual revenues grew from $1 billion in 2002 to more than $2 billion in 2006. Between 2003 and 2006, Moodys profits more than doubled, from $363 million to $754 million. In October 2009 a McClatchy Newspapers investigation quoting former employees exposed a Moodys culture in which executives who tried to steer the company away from dubious behavior were marginalized. The Senate subcommittee that recently finished a probe of the rating agencies unearthed an e-mail from 2007 in which a Moodys employee told a JPMorgan Chase & Co. banker that a colleague at the rating agency was looking into some adjustments to his methodology that should be a benefit to you folks.
These days, Moodys at least appears to be trying to address cultural deficiencies. In a panel discussion on the future of ratings held by the American Securitization Forum earlier this year, Moodys chief regulatory and compliance officer Michael Kanef insisted that his company had taken steps to strengthen analytical integrity and further separate the side of the house that secures new business from the research and monitoring side.
In a lawsuit against Moodys filed March 10, Connecticut Attorney General (and U.S. Senate candidate) Richard Blumenthal alleged that the firm punished employees who expressed dissenting views. Moodys employees who internally raised concerns about the manner in which Moodys was rating structured finance securities were marginalized within the company, received less compensation and were demoted, the state attorney generals complaint says, citing the example of an unnamed managing director who complained vociferously about credit risk methodologies applied to CDO ratings, even calling for a moratorium on rating certain RMBS-tied CDOs, only to find himself admonished by superiors.
Moodys culture allowed it to dominate the RMBS CDO market. But S&P was in hot pursuit. Harold Terry McGraw III, McGraw-Hills CEO, entrusted the running of the rating business to two longtime veterans of the firm: Vickie Tillman, who joined S&P as a municipal bond analyst in 1977 and ultimately became head of all fixed-income ratings, and Joanne Rose, head of structured product ratings since the late 1990s.
According to Blumenthals lawsuit against S&P, also filed on March 10, top executives at the rating agency were troubled by losing business to Moodys and succumbed to bottom-line pressure. The lawsuit alleges that S&Ps quest for profits unduly influenced its rating methodology, and points to an internal communication from an unnamed S&P senior executive who wrote in 2004: We just lost a huge . . . RMBS deal to Moodys due to a huge difference in the required credit support level . . . which was at least 10 percent higher than Moodys . . . I had a discussion with the team leaders here and we think that the only way to compete is to have a paradigm shift in thinking.
The Senates cache of incriminating e-mails paints a similarly troubling picture.
In April 2006 there was an e-mail exchange among S&P employees and Goldman employees relating to one of the deals in the Abacus CDO series. One exchange included e-mails copied to Fabrice (Fabulous Fab) Tourre, the only Goldman employee so far charged with fraud by the SEC. In one of them an S&P analyst gripes that the counterparty criteria is totally messed up.
Implicitly nudging his colleague to back off, another S&P employee wrote: [Goldman has] done something like 15 of these trades, all without a hitch. You can understand why theyd be upset . . . to have me come along and say they will need to make fundamental adjustments in the program.
Some of S&Ps most granular analytical activities during the credit bubble have come into question, cutting to the heart of what may have caused the crisis.
Each deal that S&P rated per its process involved a team of a dozen or so analysts (there are more than 1,000 credit analysts working at the company) crunching a series of data points some 70 in all drilling down to the individual mortgages that went into each pool being securitized. But the research aimed at evaluating borrower-level creditworthiness relied on spreadsheet models fed by assumptions. A handful of these assumptions, such as first-year payment delinquency rates, proved to be way off the mark, S&P acknowledges. Assumptions about levels of fraud embedded in the pools also turned out to be way off base.
Were talking about reasonable historical assumptions, such as the percentage of late payments within the first six months, that went from being negligible, less than 1 percent, to levels never fathomed by anyone in the entire real estate industry, says Mark Adelson, S&Ps chief credit officer, who joined the company in May 2008. People walking away from homes that was something just never seen before, not in our lifetime. The entire mortgage market was fundamentally changing faster than our researchers could take these shifts into consideration.
In his book The Big Short, author Michael Lewis relates the experiences of protagonist portfolio manager Steven Eisman to underscore how S&Ps models were flawed. Lewis writes that in 2007, digging into how ratings were being constructed, Eisman called a contact at S&P and asked a question that few others seemed to be asking: What would happen to default rates if real estate prices fell? The man at S&P couldnt say; the companys model could not accept a negative number. The anecdote has all the dramatic makings of a profound tipping point, if not a smoking-gun moment.
But is it even true?
S&P has strongly taken issue with Lewiss version of Eismans story. In fact, S&Ps base case scenario model has always factored in the possibility of a decline in housing prices. (Eisman did not return a call; Lewis, for his part, does include in The Big Short a footnote saying that S&P disputed the account.)
The problem with S&Ps negative assumption was that it was not nearly negative enough. It was for a high-single-digit decline and nowhere near the 30 percent-plus drop in housing prices that ensued.
Today, Adelson explains, structured-finance ratings are subjected to stress tests that factor in a 35 percent decline in housing prices as well as other outside-the-model what if scenarios. S&P takes an issue and subjects its performance to Great Depressionlike unemployment, in turn creating new thresholds for minimum credit enhancements, cash cushions that accompany such securitization deals.
Recalibrating the models may seem minor, like adjusting the blades on a sausage-making contraption, but it is all part of a series of steps that S&P and the other agencies are making postcrisis to try to regain credibility.
In a new organizational structure introduced in mid-2008, S&P centralized its rating criteria group, which reports to Adelson. S&P also created an independent quality-assurance unit, which reports to chief quality officer Neri Bukspan, and established an independent policy governance group to develop and approve all rating policies and procedures. Moodys, meanwhile, named a new head of surveillance.
However, because structured-finance deals are way down or nonexistent right now, many of the efforts to buttress standards and practices in this financial hot zone are more moot point than meaningful sea change for now, anyway.
Then again, the industry is always just another boom cycle away from repeating the sins of the past.
The Wall Street Journal reported in March that the commercial-mortgage-backed market was set to warmly embrace its first deal of the year, a $500 million security underpinned by existing loans refinanced by Royal Bank of Scotland to meet stricter credit guidelines. We expect more deals like this in the coming months as dealers begin to line up, Paul Norris, senior portfolio manager at Dwight Asset Management in Burlington, Vermont, told the Journal. If deals are underwritten to a robust standard and rated triple-A, the deals will see good sponsorship.
This spring the business of rating securitized products was turned upside down, and David Jacob, the head of S&Ps structured-finance rating business, can hardly contain his excitement.
This is huge, Jacob says of the SECs Rule 17g-5, which mandates information sharing, in effect facilitating the creation of a new unsolicited ratings market. The rule took effect on June 2.
Well be able to come out with ratings in cases where we really have a material difference of opinion on a credit quality, Jacob explains.
He joined S&P a few weeks before Lehman Brothers Holdings September 2008 bankruptcy, following a short-lived retirement. Before that, Jacob had helped structure securitization products for Nomura Securities International in New York, working alongside Adelson.
That two former investment bankers joined a major rating agency in 2008 speaks as much to the severe displacement on Wall Street as it does to S&Ps desire to install some intellectual firepower in its senior ranks, but the additions of Adelson and Jacob appear to be taking the company in a new direction where the integrity and transparency of rating research are paramount. How to leverage all of the research conducted in the course of producing ratings is the million-dollar question.
We think investors are going to demand that we come in and do these unsolicited ratings, Jacob says.
Others in the industry not just MFSs Pozen beg to differ. Speaking at the ASF panel earlier this year, Daniel Curry, president of boutique rating agency DBRS, expressed doubts that, in an environment with more legal liability, agencies will volunteer ratings they werent paid to produce. I wouldnt hold my breath for unsolicited ratings, Curry said in a podcast of the session made available by the ASF.
Jon Van Gorp, an attorney with Mayer Brown who specializes in asset-backed securitizations, says he is encouraged by the SECs new information-sharing rule. This will be an effective device toward mitigating the conflicts of interest in the issuer-pay model, he says. I think any effort to try to stimulate an alternative business model should be welcomed.
Speaking at the ASF session, Michael Macchiaroli, an SEC associate director, flat out admitted that although the regulatory agency could have simply eliminated the issuer-pays model by decree, it decided not to.
This is a difficult problem, Macchiaroli said. We didnt want to destroy the ratings business to the contrary, we wanted more competition. We realized that the agencies depend on the issuer-pays model to maintain their operations. Our goal was to increase competition in an area, structured finance, where it made sense. We dont know who will do unsolicited ratings, but the least we could do was try.
One company that sees a bright future in the rating business is Morningstar, best known for evaluating mutual funds. Earlier this year the Chicago-based company, seizing on upheaval in the industry, announced it was creating a new unit to issue corporate credit ratings, carving off cash-flow research already done by its equity team. Later, in March, Morningstar announced a deal to acquire, for $52 million, Realpoint, a privately held boutique rating agency headquartered in Horsham, Pennsylvania, that specializes in CMBS deals.
After the meltdown we saw a real opportunity to not only enter the structured credit market but to redefine it, says Catherine Odelbo, president of Morningstars equity and credit research division. Culturally, Morningstar and Realpoint are a true fit in terms of common values. We both value our own independent research, and investors come first.
Though for now Realpoint will still focus on CMBSs, it may expand into new areas of structured finance that had been dominated by Moodys, S&P and Fitch.
Remember that CMBS deal mentioned earlier, the first of the year, the one restructured by RBS to meet more-strident requirements? It is an important deal in that it was the first pickle out of the jar in 2010, but the deal marks a first for Realpoint: RBS, the issuer, paid the firm to rate the deal. Before that, Realpoint had only been compensated by investors. (It gave the deal a triple-A.)
Perhaps the only way to end conflicts in the rating of securities is to create an entirely new breed of rating agencies, says Joseph Grundfest, a professor at Stanford Law School and co-director of the Arthur and Toni Rembe Rock Center for Corporate Governance. Grundfest has proposed an alternative regulatory strategy in which the SEC would create a new type of agency. Instead of NRSROs, there would be IOCRAs investor-owned and controlled rating agencies. These would have an inherently investor-oriented perspective, and if they were to fail in anticipating credit market issues, the investor community would at the end of the day have only itself to blame, Grundfest says.
Jules Kroll, founder of the private investigative powerhouse that bears his name, recently established K2 Global Partners, a risk consulting firm that currently is in the process of creating a new rating practice that would be investor-owned. As of late June, K2 still planned on issuing its first rating by the end of the summer. Last fall, Kroll told 400 attendees at a Council of Institutional Investors conference in Los Angeles that he was looking for up to 20 pension funds and endowments as backers for his new company. The financial institutions who control pension funds have got to insist that a new level of due diligence take place, Kroll said on CNBC in April. It doesnt require any regulation.
Kroll plans to take his research to new heights think forensic accounting, not modeling exercises.
But even a team of seasoned investigators examining a structured deal is still going to have to rely on bank issuers for certain pieces of information. As early as November 2007, when the credit crisis was starting to spread but almost a year before the real meltdown, thenS&P structured product head Rose wrote a research piece publicly lamenting the inability of her rating analysts to glean all the necessary information. Not surprisingly, issuers are sometimes reluctant to give us access to all of their data, and, more importantly, they are reluctant to make all of that data available to the public for use in comparisons and trending, she wrote. If market participants believe these things are important, they need to speak out and take action to make them a reality.
One company that is starting a new type of low-overhead, purely technology-driven ratings is Rapid Ratings International, established in New York in 2007 by James Gellert and Douglas Cameron, the firms CEO and COO, respectively. Their vision is for investors to be able to do their own analysis of creditworthiness in real time, using a robust trove of available data and heavy-duty technology synthesizing reams of data as never before.
Nonetheless, even if new players such as Kroll or Rapid Ratings build better mousetraps, or if Congress decides to step up regulation of rating agencies beyond raising the legal liability standard, there will be conflicts as long as issuers pay for ratings. Even under the ambitious Grundfest plan, issuers would be required to hire an NRSRO and an investor-owned rater for deals, but revenues would still be flowing from the issuer to the raters. Ten years out, when some new bubble forms, an investor-owned rating agency could get caught up in a revenue push that at the time seems reasonable.
Of course, an outright ban on issuers paying for ratings would be the death knell for the industry; a subscriber-pays model would not support a 1,300-person credit rating operation such as S&Ps. Even if investors were willing to pay, however, that model has its own potential conflicts of interest. If William Gross ran a rating agency in addition to the largest fixed-income fund on earth, what would prevent the appearance of some form of bias by him toward the securities he intended to load up on?
Managing conflicts, Sharma explains, is the essence of rehabilitating the rating agencies and financial services in general. Concludes Sharma, Its in our own best interests.