UBS brings transparency to murky CDOs


Published on:

Issuer: Diversified Global Securities Limited (UBS Principal Finance)Type of deal: Cashflow arbitrage CDO of CDOsAmount: $236.95 millionDate: December 13 2001Underwriter: Société Générale

It sometimes seems that the world of finance is nothing but a series of opaque ruses designed to exemplify the expression mutton dressed up as lamb. However, after 18 months of bursting internet bubbles, crises at former corporate darlings such as Xerox, Lucent Technologies and Enron, and a belated admission from the Federal Reserve that a nine-month-old recession undermines its hopes for a soft landing, a deal comes along to redress some of the balance.

It belongs to a product segment that epitomizes the mystery of finance - collateralized debt obligations (CDOs). This deal, $236.95 million of securities issued by UBS Principal Finance, is from an even more mysterious sub-sector - it's a CDO of a pool of CDOs.

But unlike its brethren, whether straight collateralized debt obligations or CDOs of CDOs, this deal is totally transparent and also static. In other words, investors were able to check out both the CDOs in the collateral pool and the high-yield assets in the underlying CDOs first. Also, they can check it on-line, and have a guarantee that the managers of the fund - UBS Principal Finance - cannot, defaults aside, tinker about with the portfolio.

Most CDOs, by contrast, are actively managed. They have a chequered history, achieving notoriety last year, primarily as the asset class that forced American Express to take a charge of $826 million in July. Some of that was basic high-yield debt exposure but most was CDO deterioration. Bank One, American General, Torchmark and Lincoln National each announced smaller CDO losses of up to $75 million. The hit to hedge funds and other financial firms not required to report losses is estimated to be in the billions.

In part they were victims of the mid-1990s bond-market hubris: the conviction that a need for yield equated to an understanding of risk. Spreads were at their tightest for many years back in 1996 and 1997. European currency convergence had all but eliminated anything near double-digit coupons in eurozone sovereign debt and Russia's inaugural Eurobond traded in the low triple digits over US treasuries.

That was the time when we were told how sophisticated bond investors had become - how they could distinguish between individual credits around the globe. High-yield bonds were back in favour in the US and were a slowly growing market in Europe. Emerging-market high-yield paper was all the rage.

When the Asian crisis hit in the second half of 1997, many of these so-called sophisticated investors hunting for yield proved to be largely hot money subject to a herd mentality, a picture reinforced by the Russian meltdown nearly a year later. Higher-yielding paper was dumped in favour of US treasuries and other high-grade securities, and the US junk-bond market entered a two-year funk culminating near the end of 2000 with average spreads more than 1,000 basis points over treasuries and an annual default rate approaching 10%.

That filtered through to CDOs last year, and was at times made worse when a manager of a CDO portfolio chose to try to trade out of the crisis. High-yield bonds are illiquid at the best of times, and, says John Niblo, a director in structured credit arbitrage at UBS Principal Finance: "any time there is a problem in the CDO market, the bid price is based upon a worst case scenario. That liquidity premium eats into potential returns and so makes it very difficult for a manager to improve performance by trading out."

Even worse, says his colleague, executive director and head of structured credit arbitrage Michael Barnes, there are instances where "a manager has allowed a portfolio to go down in quality, versus trading to improve overall quality, to protect the current cash flow to equity holders. In other words, the portfolio is being managed for its smallest segment." On average only 10% of a CDO is equity, and this offers substantially higher returns than subordinated debt, usually in the region of 20% to 25%. Often some of the equity is retained by the portfolio manager, offering the temptation to protect the company - and the manager's job - at the expense of the majority of debt investors.

Some CDOs of CDOs are designed to take advantage of underperforming regular CDOs, with the leaders in the market being such firms as Zais and Triton. They, too, are actively managed deals. The UBS deal is different. "Our focus is on value, not management," says Barnes. "And the value is in choosing the right assets to start with which take advantage of the difference between the BB rating of the underlyings and their actual spread." As a result UBS takes a 5bp service fee and no management fee.

It's the first such deal where all the underlying securities are sub-investment grade, which made potential investors wary. "The usual approach in CDOs of CDOs has been for the managers to rely on investors to trust their judgement on what goes in the portfolio," says Lisa Underwood, head of the CDO group at underwriter Société Générale. "The UBS deal, however, is fully transparent from the start." So investors know the composition of the deal (for example, 90.66% of it is CDO securities, the rest Reits - asset-backed and commercial mortgage securitization deals), how much is emerging markets paper ($46.7 million), all the ratings, whose deals - by manager and underwriter - are in the portfolio, and what percentage they constitute.

They can also check out the vintages of CDOs in the portfolio. It's no surprise that there are no developed markets CDOs in there from 1997 or 1998, two years where a large number of poor deals were issued. In fact, 42% of the total comes from 2001, 25.11% from 2000. "It wasn't planned that way," says Niblo. "We were looking for the best assets, and these were just the right ones."

Being solely BB also made it a difficult deal to put together. "BB tranches of CDOs are difficult to place," says Underwood. "Most investors buy the investment-grade or equity tranches. The BB tranches are small, say $15 million per deal, and usually illiquid." And there were a lot of poorly performing tranches out there, making a tough task even tougher.

"We spent a month fending off bonds," says Barnes. "We had people trying to sell us their poor-performing or deteriorating credit."

It took four months for UBS to find and acquire the securities it wanted. It warehoused them, waiting until it had 100% of the assets before beginning to market the deal: often managers will start the marketing process at 60% or 80%, increasing the ramp-up risk of not being able to find suitable assets in time.

Two other factors hindered the deal: the terrorist attacks of September 11, and the growing crisis in Latin America. "We only have about $1.5 million of Argentine Sovereign credit in there, out of approximately $10 billion in underlying assets and were able to tell investors that very quickly," says Barnes. "But the World Trade Centre attacks probably put us back about a month." The UBS team is already hunting for new assets for new deals but has already achieved one of its main aims. "At least we're not a first-time manager any more," says Barnes. "We heard that a lot while marketing, investors saying that we should only come back once we were onto our second deal. Although some of them did approach us later as the story got out."

This deal ought to go a long way in opening up the opaque world of CDOs. It's not attempting to get loans or bonds off the balance sheet, nor is it trying to raise money for the issuer, as an increasing number of straight CDOs do. And, says Barnes: "We're not in it to gather assets under management. That's why it's a static pool. We're in it to make money for us and our investors."

The static versus managed CDOs debate will be a long running one, but transparency is a must.