On a recent cloudy afternoon in April, at Merrill Lynch’s headquarters overlooking the New York harbor, a few dozen investment bankers piled into a conference room for egg rolls and sweet-and-sour pork.
Christopher Ricciardi, managing director of global structured credit products at Merrill Lynch and head of Wall Street’s biggest collateralized debt obligation (CDO) business, always orders celebratory “greasy food” for his team after they price a deal. That day, Merrill Lynch had sold a $330 million collateralized loan obligation called Marathon I.
The CDO group often eats fatty lunches these days and deals are priced almost weekly as Merrill Lynch solidifies its position as the world’s top issuer of CDOs. In 2004, the country’s third-largest investment bank priced $16.5 billion in deals, or almost 14% of the market total, according to Thomson Financial. Its competitors JPMorgan and Morgan Stanley trailed behind with 9% and 8% market shares respectively.
Merrill Lynch did not always rule the market. Before Ricciardi and his team joined, it was their former employer CSFB that consistently took first prize. Merrill could barely make the top 10. Almost immediately after Ricciardi switched jobs in April 2003, taking a good portion of his group with him, CSFB started sliding and Merrill soared. Since then, Merrill has led the league tables quarter after quarter, gradually increasing its lead over the competition, while CSFB has slipped to eighth position.
This year, with analysts expecting new CDO issuance around the world to rise 5–15% from last year’s $120 billion, Ricciardi expects Merrill to issue 30–40% more than it did in 2004. Perhaps more importantly, Merrill’s CDO revenues have jumped tenfold since Ricciardi and his team joined.
Although it may seem obvious that Merrill’s success in collateralized debt obligations has resulted directly from Ricciardi’s hire, the 36-year-old managing director says it’s all about the people he works with. “I want to resist the notion that it’s just me,” he says. “It’s very much a team effort.”
Ricciardi manages a staff of about 50 in New York who run origination, structuring and marketing, with additional desks in London, Tokyo and Hong Kong. The US origination desk is headed by Lars Norell, a Swede, and Harin De Silva, a Sri Lankan, while most of the sales effort falls under Douglas Mallach, head of structured product and derivative sales. Andy Phelps heads up the syndicate desk, Plamen Mitrikov is in charge of trust preferred securities (TruPS), Steve Kuppenheimer oversees collateralized loan obligations (CLOs) and Lang Gibson produces research.
“Chris is a great manager of those people,” says Daniel Cohen, founder and CEO of Cohen Brothers, a money management firm with $5.6 billion in assets that has worked with Ricciardi on structuring many deals over the years. Money managers are key players in the CDO market. They are the ones who first started issuing CDOs as a way to leverage the debt they held under management and bring in more assets. “To find someone who is great at looking at arbitrage and teaching people on his staff about it, that’s a great combination.”
A growth market
Collateralized debt obligations have been around since 1987, but annual issuance never exceeded $5 billion until 1996, which is when most analysts began tracking the market. In basic terms, CDOs are debt instruments that pool different types of collateral—loans, bonds, asset-backed securities, trust-preferred securities—and divvy them up into tranches of debt and equity with varying degrees of risk and return.
The early deals were usually $300–400 million CLOs backed by pools of high-yield leveraged loans, or collateralized bond obligations (CBOs) backed by high-yield bonds, with maturities of eight to 10 years. They were bought mainly by banks and insurance companies looking to bolster returns. CDOs offer a handsome return—these days of up to 240 basis points over comparable corporate bonds—because the market is complex, opaque and illiquid.
Initially, CDO supply came from “high-yield asset managers who sought to get money under management a lot faster”, says Lang Gibson, Merrill’s head of CDO research. Money managers would raise capital by issuing a CDO, and then use the money to buy a diverse mix of credit instruments.
High-yield CLOs and CBOs dominated the market until about 2001, when a corporate recession and a wave of defaults sent prices crashing. CDO issuance dropped from almost $68 billion to approximately $52 billion annually in 2001 and didn’t pick up again for three years. Investors sought safety in investment-grade securitizations—high-grade ABS CDOs—instead. Although high-yield CLOs are still popular, high-yield CBOs have all but disappeared.
Today the clear favorites are CDOs backed by asset-backed securities, which make up about 55% of the total CDO issuance. CLOs make up 30%, TruPS and the occasional CDO backed by emerging market debt or high-yield bonds make up 10%, and synthetic CDOs make up 5%.
Asset-backed securities are popular as collateral because they have historically suffered fewer defaults and downgrades than corporate bonds, says Gibson, even though they trade at wider spreads. At the same time, CDOs backed by ABS are generally safer than those backed by bonds because losses are filtered through the underlying securitization first.
Luckily for Merrill Lynch, Ricciardi and his team started using asset-backed securities to collateralize CDOs early and have ridden the wave all the way to the top. “Some of our success,” says Norell, “lies in the fact that we just happened to start working on and developing something that turned out to be very popular.”
Ricciardi started working on CDOs in 1997, while he was at Prudential Securities. Initially, the team struggled to be competitive because Prudential’s business in the collateral du jour—leveraged loans and emerging market debt—was meager. Prudential did, however, have a strong business in asset-backed securities. So when ABS CDOs began emerging, Ricciardi and his team were in a good spot to take advantage of the trend.
In 1999, they issued the first ABS CDO as it is defined today, backed by a diversified pool of residential mortgage-backed securities and home equity loans. “We thought ABS was excellent collateral for CDOs because asset-backed securities—particularly subordinate classes of, let’s say, home equity loans—trade at much wider levels than corporates of the same rating,” says Ricciardi. “It was our feeling that they trade at these wider levels not because they’re of worse credit quality but because of technical reasons.”
A year later, Ricciardi left Prudential for CSFB and took Harin De Silva with him. There they joined up with Norell, and the three helped CSFB become the largest issuer of CDOs. Then, in 2003, they headed for Merrill Lynch.
Areas to exploit
Up until 2003, Merrill Lynch barely had a presence in the CDO market, only pricing a couple of collateralized loan obligations a year. When Ricciardi joined he quickly ramped up the ABS and trust-preferred business and this year the firm plans to focus on further growing its CLO and synthetic CDO output.
ABS CDOs, backed mainly by residential mortgage-backed securities, have had a good run over the past few years as the housing market has rallied. ABS issuance rose from $27 billion in 2003 to $46 billion in 2004 alone, according to Standard & Poor’s. But some analysts worry about the consequences on ABS CDOs if the real-estate market stalls. “The biggest thing is always the leverage in the CDO market. When the general high-yield bond market got hit, CBOs backed by high-yield bonds got hit even more,” says Michiko Whetten, quantitative credit derivative analyst at Nomura Securities. “The same thing is concerning lots of market participants now, in that ABS CDOs have a very high exposure to the residential mortgage-backed securities market. So the fortunes of the ABS CDO market can go up and down with the housing market.”
Whetten says ABS CDO spreads may have tightened too much, and like other analysts she expects ABS CDO issuance to be flat to slightly lower this year. Ricciardi says he expects ABS CDO issuance at Merrill Lynch to rise marginally.
Some market participants expect CLOs to pick up some of the slack in the ABS CDO market. Traditionally, CLOs were backed by large syndicated leveraged loans. But as spreads in the sector have tightened, Merrill and other Wall Street firms have begun hunting for smaller credits with higher yields. For CLOs, this has meant a shift to the middle market.
Merrill Lynch has taken this one step further, issuing the first CLO backed mainly by mezzanine debt last December for the FriedbergMilstein Private Capital Fund. The CLO is worth $584 million. Mezzanine loans are more subordinate than the senior debt that usually makes up large syndicated loans, making the transaction riskier but higher-yielding than traditional CLOs. Although CLOs have been around for a long time, this type of innovation in that asset class is fairly new.
In CDOs structured from trust-preferred securities (TruPS), however, innovation has always been a part of the game. Trust-preferred securities are hybrid instruments—part equity and part debt—which small financial institutions issue to fund growth or repurchase stock without diluting shareholder value. Issuers of TruPS receive a tax break from the government, and for the past five years small banks and insurance companies have been issuing them in order to put them in CDOs.
Initially, TruPS CDOs were used as a way for small banks and insurance companies to raise money on Wall Street, giving them access to the type of market they couldn’t usually afford. By pooling their financing needs, borrowers could split issuance fees. At the same time, investors were given exposure to a market that was usually overlooked.
The first CDO backed by TruPS was issued by First Union in 1998, and since then a handful of specialized money managers like Cohen Brothers have sprung up, zooming in on regional and local financial institutions. They help underwriters source collateral, negotiate with the borrowers and analyze the credits. Managers also give investors the comfort that someone other than the underwriter and borrower is looking out for their interests, because money managers want to see their assets perform well.
These days, TruPS CDOs are being extended to more asset classes. Earlier this year Merrill Lynch was the first to issue a CDO backed by a pool of real-estate investment trust (Reit) TruPS: the $729 million Taberna Preferred Funding vehicle managed by Cohen Brothers. The deal had 10 investment-grade tranches, the most senior of which yielded 47 basis points over Libor.
“It’s quite a process,” says Ricciardi, who says it took about two years from when the idea was first hatched to execution. “It’s a new product. You have to work with the rating agencies, educate investors, educate the Reits so they’ll issue the product.” It took a while to convince the Reits to borrow through a vehicle they had never heard of before, and it took some effort to convince rating agencies to take a look at small institutions usually below their radar screens.
Ricciardi says he expects Reit trust preferreds to grow rapidly this year and next at least, although this asset class demonstrates one of the drawbacks of CDOs: there has to be underlying collateral to back them, and supply is limited. “It’s not clear how much more capital Reits need. It is a more limited audience than banks. There are 10,000 banks, but there are only two or three hundred Reits,” says Ricciardi. “At some point we’re going to run out of guys to do this for.”
This problem exists in other asset classes as well. “In ABS, the availability of assets has been a sticking point,” says Norell. “There’s a finite amount of them issued.” Only about $10–12 billion in subordinate classes of home equity loans are issued annually, he says, putting a natural cap on CDO issuance.
But as Norell points out, the TruPS technology can be extended to any type of borrower who is too small to tap capital markets directly. And to mitigate any slowdown caused by lack of collateral, Wall Street has set its sights on synthetic CDOs, backed by the large and growing supply of credit default swaps.
At the end of the day, the CDO market is all about tweaking existing structures to fit new kinds of borrowers and asset classes, in the never-ending search for collateral. “The ideas are easy to come up with,” says Ricciardi. “Once you have CDOs, people ask, ‘what else can I do with CDOs?’ It’s the execution that’s difficult.”
One potentially very lucrative asset class that has proved tough to crack is the municipal bond market, because it has been too expensive to offer worthwhile arbitrage opportunities. Ricciardi says market participants have been talking about muni CDOs for years, and once the conditions are right Wall Street will be ready to act. Merrill is a top municipal bond issuer and secondary market trader. “Munis will be a big area of innovation,” he says. “It’s such a big market with high credit quality and a lot of diversity.”
Wall Street is also trying to find ways to collateralize commodities. But De Silva, co-head of US origination and structuring and the synthetic CDO expert on the team, says Merrill will only issue commodity CDOs if the firm can find a way to make it worthwhile. “If you try to use a new asset class, rather than doing just a one-off deal, you try to create a transaction that is repeatable,” says De Silva. Given the time and effort involved in structuring CDO deals from scratch, it would make little sense to do otherwise.
One of the problems with commodity CDOs is identifying investors who want to buy protection on commodity price movements five years ahead. Commodity trading desks who might be candidates typically look out six to 12 months, says De Silva. A commodity CDO would likely have to be a synthetic product, given that the market already revolves around derivatives such as futures and options, says Ricciardi.
Even though the Merrill team says it hasn’t devoted much time to commodity CDOs, the group is always thinking about other asset classes that can be used to issue new vehicles.
On that recent April afternoon, after pricing the Marathon deal and finishing his Chinese lunch, Ricciardi led a few members of his team to a windowless conference room to discuss a secret project they had been working on for two years. Their mission was to use a new type of asset to back an existing category of CDOs. This type of deal had never been done before, and collateral had to be sourced, issuers persuaded and rating agencies brought on board. While the deal they discussed was already quite advanced and will likely be launched in the third quarter of this year, many others they have considered in the past end up not being feasible.
“You never know when a product is going to take off,” says Ricciardi. “You have to have the intersection of demand from investors, and demand by the people on the supply side,” he says. Synthetic ABS CDOs, for example, had several false starts before finally sparking investor demand last year.
Ironically, while one challenge is bringing investors on board, the other is keeping the masses away. “We need to continuously innovate,” says Norell. “We don’t necessarily want everything to be too simple and universally accepted by everyone, because then the relative increase in yield to investors who do want to take the time to understand complex product is going to be difficult to find.”
Norell points out that collateralized debt obligations carry handsome yields compared with corporate bonds of the same rating because fewer people understand them. By the same token, ABS CDOs yield more than CLOs because people don’t understand the mortgage-backed market as well as loans.
“If you polled a hundred capital markets participants about IBM, most people are going to have some opinion of it. But if you went to the same group and asked them about an Ameriquest or Countrywide triple-B mortgage security, very few people are going to have any idea of how to analyze it,” says Norell. This complexity is what adds yield to ABS CDO products, he says, because demand is limited to a small pool of very specialized investors.
In too deep
Is there a danger that the market is so opaque that even those who invest in it are taking bigger risks than they should? Nomura’s Whetten says some investors who have entered the market recently might be getting in over their heads. “The people who jump into the market last tend to be the ones with the least knowledge,” she says. “People who invested and got hit in 2001 and 2002 said, ‘we just didn’t know what we were buying.’ It’s possible that something on a smaller scale could happen because everyone is buying CDOs with a me-too attitude.” She adds that many investors with little understanding of structured finance have simply, and dangerously, applied their knowledge of corporate bonds to asset-backed securities.
However, Norell argues that CDO investors are savvier than most. In 2003, he says, when the ABS market widened due to concerns about an economic slowdown and possible deflation, CDOs didn’t suffer as much because investors seemed to have “seen through the hyperbole and determined that it didn’t affect them”.
One undeniable source of risk, which helps contribute to the yield on CDOs, is the lack of a real secondary market. David Weeks, head of secondary trading at Merrill Lynch, says secondary market volumes in cash CDOs were under $30 billion in 2004. “Generally speaking we advise our investors that the products we sell them are buy-and-hold securities,” says Norell. “Liquidity is tricky, and the rule seems to be that it’s never there when you really need it,” he says, explaining that when a transaction performs poorly investors have a difficult time selling it.
But again, the same principal applies: if the market were to become too liquid, returns for investors might start to diminish, says Norell.
In this vein, many Wall Street players would like to keep the CDO market as opaque as possible to keep yields juicy. Ricciardi, however, has been spearheading an effort to increase transparency. As co-chairman of the Bond Market Association’s CDO committee, he has set up the CDO Library, a website which carries pricing, structuring and modeling information on all deals. The website is meant more as an inter-dealer tool than an informational service for individual investors, although “qualified institutional buyers”—investors with more than $100 million in assets—can access it. The library has been a mixed success, and not all Wall Street players always post complete information in a timely fashion.
Ricciardi argues it might be premature to demand full transparency on CDOs when even the much older ABS market still has far to go. Other market participants agree that the CDO market is now far more liquid and transparent than it was five years ago.
For now, it seems, investors are more interested in returns than transparency or simplicity. Yields on triple-A CDO tranches have already tightened substantially from last year, and buyers continue to pile on. As long as structured finance investors remain hungry for CDOs, Merrill Lynch’s CDO group will be eating a lot of egg rolls.