How Moody’s faltered

In the summer of 2002, John Diaz, a managing director at Moody’s Investors Service, was called before a US Senate subcommittee investigating the collapse of Enron. The senators wanted to understand why Moody’s had said that the energy trader’s debt was investment-grade in late October of that year – only to see the company default on its bonds four weeks later as it declared bankruptcy.

Diaz described his company as an information provider. Moody’s researched and analysed companies’ financial health, then offered investors its views, to help them decide where to put their money. “Ratings,” he told the subcommittee, “are a simple symbol system to express relative creditworthiness ... Moody’s ratings are designed to provide a relative measure of risk, with the likelihood of default increasing with lower ratings.”

Diaz expressed regret that he and his colleagues had not “discovered the information that would have allowed us to serve the market more efficiently in this instance”. If Enron had not misled the analysts, he said, the ratings would have been lower. In his concluding remarks, when suggesting areas in which the ratings process could improve, he called for more financial disclosure from companies, closer attention to liquidity risks by Moody’s, and a focus on companies’ “corporate governance and how aggressive or conservative are accounting practices”.

As it turned out, the accountants, not the rating agencies, came to shoulder much of the blame for Enron’s collapse and the economic downturn it wrought. The bookkeepers had shed their dull reputation, their ledgers and calculators, for life in the fast lane, advising companies and helping them design Byzantine financial structures ripe for breakdown. And they paid the price: Enron’s accountants, Andersen, closed down in 2002, shedding 85,000 jobs, and most of the other big firms split their accounting and consultancy practices.

Moody’s fared better. Despite being the subject of an Enron-related competition probe by the US Justice Department (which was dropped but led to a separate felony indictment for the shredding of subpoenaed documents), the company, which only listed in 2000, continued to grow and reported the highest profit margins of any company in the S&P 500 index, the creme de la creme of big US companies. It held that position for five years running; its shares rose 500 per cent in their first four years of trading – at a time when the rest of the market was down; and Moody’s earnings rose by 900 per cent in a decade.

Then, on August 16 last year, after an internal revision of its ratings practices, Moody’s made an announcement that heralded the beginning of the credit crunch.


Moody’s was the second investor-services project of John Moody. The son of a New England ice salesman, Moody moved to New York at the turn of the 20th century. The city was gripped by a banking boom and Wall Street was awash with railroad bonds, issued to finance the taming of the Wild West. Moody was a muckraking journalist, a sometime Wall Street analyst and a Catholic proselyte. In 1900, he had what he called his “eureka moment” and left his clerking job to begin publishing a regular manual evaluating the quality of stocks and bonds. It became enormously popular, but the venture went under in the 1907 stock market crash. In 1913, Moody began publishing a new manual, which analysed the creditworthiness of the US railroad companies themselves. Soon, he had broadened the mandate, writing about a wider range of industries, and a year later, the company was incorporated.

Moody modelled his ratings on credit-reporting systems established in the last decades of the 19th century. If a company’s bond was given a triple-A rating, it meant Moody believed that you could lend to the group with a very high chance of getting your money back, plus interest. The worse the chance, the lower the rating – from Triple A to Double A to Single A, then down to Baa. Ratings below that were considered junk – or high-risk.

It soon became essential that any financial product carry a rating, and by the 1920s, Moody’s rated almost the entire American bond market. In those years, the service was paid for by investors, through subscriptions. It was a system that made sense: in property sales, for example, surveyors are paid by the buyers, not the sellers. This wasn’t so different. But by the 1970s, the growing complexity of financial products and the sheer size of the debt market demanded staffing levels at Moody’s that a subscription model couldn’t sustain.

There was a further rationale for changing the model and charging the bodies that wanted to issue bonds to obtain a rating: the ratings had become essential to any deal. As debt markets grew, investors became more discriminating. Most bond investors required two ratings, in fact. Moody’s provided one and Standard & Poor’s, a more recent arrival on the scene, the other. It was a happy duopoly – and the lack of competitive pressure meant independence for the agencies. On the other side of the deal table, however, it led to frustration. Companies complained that agencies were aloof, opaque, secretive – the svengalis of debt. Analysts didn’t answer their phones. And when they did, the bond issuers complained, they weren’t helpful. By the 1990s, Moody’s was known for its bookish, academic atmosphere, which grated on many bankers, according to an industry survey.

Several factors began to change this culture. First, was the rise of Fitch, a third big ratings agency. Fitch had been recognised, along with Moody’s and S&P, as a “nationally recognised statistical rating organisation” (NRSRO) by the Securities and Exchange Commission (SEC) – Wall Street’s regulator – as early as 1975, but it wasn’t until the 1990s that it became a serious contender for ratings jobs. A third agency meant that banks could now “ratings shop” – and avoid Moody’s, which was known for giving lower ratings than its competitors.

A second factor for change was Brian Clarkson. Clarkson was born in Detroit and went to law school at the University of North Carolina, Chapel Hill, before joining Moody’s in 1991 at age 35. He started as a mid-ranking analyst in what was at the time an experimental backwater at the company: structured finance.

Rating a corporate bond, the bread and butter of the Moody’s business, is relatively straightforward. It essentially involves rating a company, and companies have assets, business models, balance sheets, managements and, most importantly, history. A rating is based on a qualitative and quantitative assessment. Rating a structured-finance instrument is different. Rather than having a whole, living corporation standing behind them, structured bonds are backed by pools of debts which are collected and sold on by banks. Most structured bonds are collections of mortgage loans.

The first mortgage-backed bonds were created in the late 1980s, well before Clarkson’s time, by a trader called “Lewie” Ranieri. Ranieri, the head of the mortgage trading desk at the former investment bank Salomon Brothers, was famous for the huge sums of money he netted for his employer and for the quantity of cheeseburgers he ate. What he struck upon in structured finance was a process of pure alchemy: a way of turning myriad messy mortgage loans into standardised, regimented and easy-to-assess bonds.

Ranieri knew that the magic of structuring was in the packaging. Packaged in the right way, mortgages could come to create a huge, new tradable bond market. And this is where the rating agencies came in. Structured bonds, like any other bond, needed ratings in order to be sold. But with a structured bond, the pools of debt could be built or modified in order to attain a particular rating. This wasn’t a matter of disguising the risk, rather a way of reapportioning it and allowing investors with different risk appetites to buy the right product for them. “The rating is what gives birth to the structure in the first place,” explains Sylvain Raynes, a financial modelling expert who was with Moody’s in the 1990s, when Clarkson joined. In some cases, the ratings are known before the bonds have even been inked. “You start with a rating and build a deal around a rating,” Clarkson told an investment magazine last year.

From the start, Clarkson was involved in the part of the structured finance world Ranieri created: mortgage-backed bonds. By 1997, Clarkson was in charge of the whole mortgage bond division. What he understood was that the success of the structured finance business didn’t only depend on the complicated models and analysis it used. As with any other deal-making process on Wall Street, it was about people. If Moody’s wanted to reassert itself as the world’s top rating agency, it had to reform itself, and structured finance showed how: rating wasn’t something done from an ivory tower, rating was done as a service for clients, and, in the case of structured finance, as a service for banks.

“We’re in the service business,” Clarkson said in an interview last year. “I don’t apologise for that.”


If Clarkson set the course for the personality shift in structured finance, the listing of Moody’s spurred it on. The agency was floated as a public company in 2000 – spun out of the financial publisher Dun & Bradstreet. As one former Moody’s staffer recalls: “The change was just precipitous. There was suddenly a concentration on profits. Management got stock options. It’s true there was a big personality shift in the company – lots of cozying up to clients went on.”

“We were lily white,” says Ann Rutledge of her time working in the Moody’s structured finance division in the mid-1990s. “But then the centre of gravity in Moody’s shifted. Moody’s went public.”

Soon there were stories of analysts going skydiving with clients; of structured finance experts and bankers on weekend getaways together; of golf outings and karaoke nights (Clarkson was known as a fan of the latter). Analysts at Moody’s now picked up their phones.

Alongside the new-found service culture came other, more fundamental, changes. “In the late ’90s, research and development at Moody’s slowed right down,” says Rutledge. “The banks caught up. And then they came to be the ones in the lead.” As the structured finance arms race accelerated, Moody’s and its peers became more passive participants. Where the firm had once been at the cutting edge of statistic analysis (analysing facts) and stochastic analysis (analysing probabilities), soon it found itself trying to keep pace with the latest engineering packages devised by the banks.

“Over time, disagreements with analysts were just smoothed out. And it was no longer the rating agency which always won,” says Raynes, who left Moody’s to set up his own credit analysis consultancy. “The banks won. Now the modelling is done by the street.”

The agencies were inundated with a huge volume of new structured finance deals that they were being asked to rate. At Moody’s, the flipside to the huge revenue growth was a high-pressure work environment. One analyst recalls rating a $1bn structured deal in 90 minutes. “People at the rating agencies used to say things like, ‘I can’t believe we got comfortable with that deal,’” says Raynes. “People talk about moral hazard at the banks, but the moral hazard for the rating agencies is extreme.”

Moody’s rejects the suggestion that it subordinated the integrity of its ratings to chasing a larger share of the market. It told the FT in May: “Moody’s has published extensively on our position that each possible business model brings with it potential conflicts. The real issue is how credit rating agencies manage them.

“The quality of our ratings opinions, commentary and analysis has been and continues to be our primary concern and commercial considerations are never a factor in the rating assigned to an issuer or transaction. Our ratings and research are our only products, and our reputation is our only capital.” Moody’s was invited to comment in detail on the specific points raised within this article, but declined.

Clarkson, according to several people at Moody’s, made clear to his analysts that accuracy was fundamental. There was no policy of subordinating ratings to market share. But, say some, there was a significant institutional fear of losing business. Though analysts weren’t strictly paid according to the amount of work they brought in, employees – who were often rewarded with stock options – had reason to see the company do well. Clarkson could not be reached for comment.


In the early days of the millennium, it was almost impossible for one of these bundles of bonds – known as CDOs or “collateralised debt obligations” – to get a triple-A rating from Moody’s if the collateral was entirely mortgages. The rating agency had a long-standing “diversity score”, which prevented securities with homogenous collateral pools from winning the highest rating. S&P didn’t have such a score and neither did Fitch. The result, even with service-oriented Clarkson at the head of the structured finance department, was a steadily dwindling number of mortgage CDOs rated at Moody’s.

As the mortgage CDO market continued to grow, Moody’s couldn’t continue as it had. In 2004, the diversity score was abolished – a decision approved by the Moody’s credit committee. The number of mortgage CDOs rated by Moody’s rocketed.

Meanwhile, by 2006, Clarkson’s team, spread all over the globe, was delivering 40 per cent of annual income for Moody’s, eclipsing all the other parts of the business. Clarkson wasn’t involved in the day-to-day running of the structured finance unit – that was left to a network of deputies. Frederic Drevon was Clarkson’s prelate in Europe, working in cathedral-like offices in the City of London, and it was under his watch that Moody’s first came across the CPDO – a “constant proportion debt obligation”. The product had been designed by a crack team of credit experts at the Dutch bank ABN Amro and was called the “Holy Grail of structured finance” by analysts at Bear Stearns. It had been built with a view to achieving triple-A ratings, but also promised to pay investors a substantial return – more than 10 times what comparable triple-A instruments were offering. CPDOs were not mortgage-backed, but rather collections of bets on the creditworthiness of hundreds of European and US corporations.

Moody’s rated the first of the CPDOs in August 2006, after an unusually long analysis period. The rating it came up with was triple A – the closest thing you can get to risk-free. S&P also gave it a triple-A rating. About two weeks after those first ratings came out, Fitch, which was not hired to rate any CPDOs, said it couldn’t understand how they had been achieved: its own models had put CPDO bonds barely above junk grade. Fitch’s doubts – echoed by a few other researchers – did not steal ABN Amro’s thunder. It was the new bestseller, and received this plaudit from Citi’s research team: “Jokingly, we have started calling the product a Hydra. Almost every time we tried to kill it by subjecting it to severe stress, it seemed somehow to be able to recover par by maturity.”

At Moody’s, Drevon and his team had reason to be pleased. The CPDO was a new product in a new area of finance. By winning the first rating mandate, they had effectively guaranteed a big future income stream. As other banks rushed to bring their own versions of the CPDO to market, they naturally went to Moody’s and S&P for the ratings. CPDOs were reported to be the most lucrative individual instrument Moody’s had ever handled. “We’re working night and day to rate CPDOs,” said one Moody’s managing director to a conference in September that year.

Rating a new transaction, as an analyst, is a relatively simple procedure – but it can be time-consuming. From an ordinary desktop computer, you start the Moody’s rating software. A window opens in which you set the basic assumptions: duration of bond, payment, collateral details ... and then – click – the simulation is set running. Not once, but a million times, each time with a different outcome. It’s the average outcome from all those simulations that gives you a rating. “You’d run the program ... one million simulations at your desk... you might leave it going overnight, sometimes just come back the next morning,” explains an analyst.

Still, by the Christmas holidays, the rating team was burnt out. Most had gone home to their families. But one analyst kept working over the holidays, and on January 2 2007 sent the team an e-mail message: “Hi Guys! Happy New Year! I’ve spotted another bug in the CPDO code.” This in itself was nothing serious. Bugs in ratings are not uncommon; the Moody’s internal error-log lists numerous glitches and hiccups in the many codes the agency develops. Most of them are minor.

With CPDOs, however, there were quite a few more bugs than usual – which Moody’s attributed to the fact that it was a new product. What’s more, it wasn’t uncommon for the company’s client banks to be the ones spotting the errors. One bank recalls how it – and others – often phoned Moody’s to register problems: things that didn’t seem to stack up on their machines; tweaks or adjustments they wanted to insert. For the most part, they were just creases which needed ironing out.

The bug spotted over the Christmas holiday didn’t have a big impact on CPDO ratings. But another bug, spotted soon after, did. A single small error in the computer coding that Moody’s used to run its CPDO performance simulation had thrown the results way off. When the error was corrected, the likelihood of CPDO default increased significantly. CPDOs, it turned out, weren’t triple-A products at all. Preliminary results suggested the error could have increased the rating by as many as four notches.

A meeting of the rating committee was hastily arranged, involving some of the most senior managing directors in the company’s European structured finance division. Drevon, Clarkson’s deputy, was informed. The committee, however, did not disclose the error to investors or clients. The bug was corrected and the same model was then used to rate new CPDOs. But the new CPDOs still achieved triple-A ratings. The reason was that the committee made other changes to its code. A senior analyst proposed three alterations to the basic rating methodology. Two of them were adopted. The third was ditched because, as one document stated, “it did not help the rating”.

The FT revealed the CPDO rating error at Moody’s in May. The agency commissioned an external investigation, led by law firm Sullivan & Cromwell. In July, the findings were released. Staff had “engaged in conduct contrary to Moody’s Code of Professional Conduct”, said the agency. “Specifically, some committee members considered factors inappropriate to the rating process when reviewing CPDO ratings following the discovery of the model error.” Moody’s instigated a company-wide review of methodology and modelling as a result, and began disciplinary proceedings against staff.

Around the same time that Drevon’s Paris team discovered the rating error in CPDOs, another Moody’s team – a dozen or so junior analysts and directors in the US – spotted a different potential problem in the systems at Moody’s. This problem wasn’t a coding bug.

From January 2007, the Moody’s US residential mortgage bond team began tracking a disturbing rise in the number of subprime mortgages going delinquent. The delinquency rate is the first of three key measures that rating agencies use to assess the soundness of a mortgage-backed bond. The second key measure shows the number of people delinquent for more than 90 days and the third shows the number of foreclosures. The trio form the danger-alert system on any mortgage bond: if homeowners miss a payment, they make it into the first category. If they miss three payments in a row, they fall into the ignominy of the second. After that, they lose their home; the mortgage loan defaults. Green, amber, red.

What shocked the Moody’s analysts about the delinquency rate they were watching – the green lights – was that it had moved up very fast. And it was happening not just in isolated regions but all over the US. More worrying still: in some cases, mortgage loans were jumping the lights and moving straight from green into red. This was not in their models – nor anyone else’s.

The agency’s chief economist, Mark Zandi, had been warning of a US housing downturn for some time. In May 2006, he wrote that the housing environment “feels increasingly ripe for some type of financial event”. But Moody’s triple-A rated bonds were thought safe – as were those highly rated by other agencies. According to a report in March 2007, the risks of the defaults in subprime mortgage bond pools climbing further up the structured finance chain were “mild to moderate”. Nothing needed be done unless the second warning – the amber lights – lit up.

Pretty soon they did. The same people who missed one mortgage payment had now missed three. Outwardly, the rating agencies were sanguine. “Over the past several years, Moody’s rating changes in the second half of the year have greatly outnumbered actions in the first half,” said Moody’s, trying to explain the large number of downgrades it seemed to be making.

But behind the scenes, Moody’s was already taking steps which would drastically revise its outlook.

At the end of July, the company decided it needed to update its rating methodology. The review was publicly announced on August 2. Five days later, Clarkson was appointed chief executive of the company, replacing Raymond McDaniel. Meanwhile, the new delinquency assumptions were calibrated, and the analysts typed the revised data into their machines. When they came back to their desks, they realised, one bond at a time, the severity of the mortgage crisis. On August 16, on a mild, rainy day in New York, Moody’s released the results of its revised methodology. In one fell swoop, it downgraded 691 mortgage bonds. The two biggest other rating agencies – Fitch and S&P – were issuing unprecedented downgrade notices too.

The action was the first in a series of surprises for the credit markets. In each of the succeeding weeks, it seemed, Moody’s and the other rating agencies had more bonds to downgrade. And each set of downgrades was a convulsive shock. In the final few months of 2007, Moody’s downgraded more bonds than it had over the previous 19 years combined. Panic gripped trading floors. Titanic structured vehicles, created by banks to warehouse their “riskless” mortgage bonds, became untouchable for short-term investors. As a result, two big German banks revealed that they were within a whisker of collapse, and virtually overnight all the world’s banks stopped lending to one another.


“We were preparing for a rainstorm and it was a tsunami,” said Clarkson earlier this year. Moody’s insists there was no way it could have foreseen the onset of the credit crisis. But were any of the agencies looking? Moody’s hadn’t updated its basic statistical assumptions about the US mortgage market since 2002.

Some analysts believe Moody’s and the other rating agencies could have acted sooner. There were debates in 2006 among the firm’s mortgage-security rating teams about the quality of the collateral they were allowing to be rated. Too many American mortgages seemed prone to fraud; the risk of a house-price collapse was looming; subprime lending was exploding across the US. E-mails which the SEC recovered from the rating agencies – S&P, Moody’s and Fitch – paint a grim picture. One senior director at an unnamed rating agency wrote to a colleague: “I have been thinking about this for much of the night. We do not have the resources to support what we are doing now ... we need staff to keep up with what is going on in subprime and mortgage performance in general NOW.” Another wrote: “Doing a complete inventory of our criteria and documenting all of the areas where it is out of date or inaccurate would appear to be a huge job.”

In the case of the multibillion-dollar CDO market, Moody’s denies that it was aware of serious problems with its methodology. “[We] change methodologies on a regular basis to enhance them and to reflect recent events and our best estimates of the future,” says the firm. In the end, as long as the banks were churning out CDO deals – and in early 2007, they seemed to step up a gear – the rating agencies continued to rate them. In the words of one senior executive at the firm, it was “like a suction pump”. A highly placed figure from Clarkson’s former team says: “If you speak to anyone individually at the rating agencies then of course they’ll say they’re not to blame for the crisis. I don’t think any individuals are losing sleep over it.”

Moody’s is not a rated company. If it was, its outlook might be bleak. In its latest results, filed in August, revenues at the company were reported to have halved from the same quarter a year earlier. US lawmakers are already exploring the role the rating agencies played in the lead-up to the credit crisis. Influential senators have called for fines to be imposed in the event that any wrongdoing is proved. In a July report, the SEC demanded far greater transparency from the rating agencies, requiring them to disclose much more information to the market about the securities they rate. Staff were to be prohibited from receiving gifts from their banking clients with a value greater than $25. In Europe, Charlie McCreevy, the European Union’s internal markets commissioner, said it was “time to end the rot at the heart of the structured finance rating process” and called for “mandatory, well-targeted and robust internal governance reforms”.

The rating agencies say they are co-operating fully with the authorities. But then there are the civil cases. Only last month, the Abu Dhabi Commercial Bank filed a lawsuit against a number of Wall Street banks as well as Moody’s and S&P, alleging that they inaccurately rated securities that the bank had been unfairly sold. The bank’s management say they expect some of the biggest sovereign wealth funds in the Middle East to join them in the lawsuit.

Moody’s used to be headquartered at Church Street in Manhattan, on a prime piece of downtown real estate a block south of Ground Zero. Looming above the entrance to the squat building was a bas-relief of polished bronze that John Moody had installed when the office opened in 1951. In hammered-out letters a foot high each, the inscription read: “Credit: Man’s Confidence in Man”. Below that was a quotation from the antebellum-era US senator Daniel Webster: “Credit is the vital air of the system of modern commerce. It has done more – a thousand times more – to enrich nations than all the mines in the world.”

The frieze has now been melted down. Shortly before Clarkson was appointed chief executive, the firm moved to more glamorous headquarters in the reconstructed World Trade Center Tower 7. In the place of the old Moody’s HQ, a monument to the property market will stand: America’s tallest residential tower.

In the meantime, with trust eroded on Wall Street and in financial centres around the world, governments are bailing out the banks that bought toxic assets that had been given the A-OK by the ratings agencies. It’s unclear, as yet, what the ramifications for ratings will be. Stricter oversight will be enforced. Moody’s will have to distance itself from its clients. Clarkson has already resigned. He has been replaced by Michel Madelain – formerly the head of corporate and sovereign rating, a core part of the Moody’s of old. But lawmakers may not have the appetite to go after the rating agencies. The world’s financial markets have credit rating hard-wired into them.

There is perhaps a middle ground. Egan-Jones is a small, newly registered agency which, unlike Moody’s, S&P and Fitch, earns its income from bond investors, not issuers. Ratings are paid for by subscription, and not by fees from the companies and banks trying to sell the deals in the first place. But going to an investor-pays model is probably too big a change to ask for more broadly. American and European market regulators seem happier to push for a much-reformed status quo.

The agencies, meanwhile, need to start earning back trust. “Rating is a religious process,” says Sylvain Raynes. “Once you’ve lost the confidence of investors, it’s gone. You can’t get that faith back. The rating agencies have lost their influence, just like the Church did 500 years ago. People don’t have faith in ratings any more and the scale of this crisis is such that I’m not sure they ever will again.”

Sam Jones is a reporter for FT Alphaville

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't copy articles from FT.com and redistribute by email or post to the web.