The evolution of Asian CDOs

JAPAN OVERCOMES MARK-TO-MARKET FEARS

While Japan is generally viewed as a separate market from the rest of Asia by most financial institutions, the situation in the Japanese CDO market largely mirrors what is going on elsewhere in the Asia-Pacific region. Yoichiro Nakai, general manager of the structured products and sales department at Mizuho Bank in Tokyo, says the Japanese market is split roughly evenly between domestic and global CDOs. "Leveraged super-senior and short buckets are both very popular in Japan. Investors believe that volatility will continue, and some buyers want a way of escaping from exposure to US credit and getting more exposure to Europe," he says. Mizuho Bank, which has launched a number of new transactions in recent months as part of a series aimed at Japanese investors, reckons there are around 100 domestic buyers of CDOs in the market, including life insurers, megabanks, trust banks and pension funds.

Cash CDOs have been sold into Japan since the late 1990s, and synthetic structures for the past thee years or so. Dean Rostrom, Deutsche Bank's head of CDOs in Japan, says there has been a significant increase in the number of domestic CDO investors, and in the variety of structured products they use. "Japanese investors have held back to some extent while they study the accounting and regulatory issues associated with synthetics, including the effect of mark-to-market regulatory requirements and the potential impact on their income statements," he says. "But they are increasingly coming to terms with this, and many now accept the increased volatility, which is, after all, limited to the portion of their income that comes from derivatives."

Rostrom says Deutsche Bank continues to sell cash CDOs across the capital structure, from equity to triple-A tranches, but as with other Asian markets, there has been a growing reluctance on the part of investors to buy triple-As as spreads continue to tighten. "Some of the traditional triple-A buyers won't buy at less than Libor plus 30bp, which isn't possible right now," he says. Japanese cash CDO investors typically include banks, insurance companies, pension funds, endowments and corporates, with less interest from regional banks, which tend to be deterred by the longer time horizons of CDOs (typically from eight to 12 years) and the fact that they are structured in euros or US dollars.

Demand for synthetic mezzanine tranches in Japan has increased over the past couple of years, both in managed CDOs and bespoke structures, where a single tranche is issued to investors and the remainder managed by the arranging bank. "At this point, however, mezzanine tranches of synthetic CDOs don't offer as much value as previously, leading to increasing demand for outright equity tranches and principal-protected equity," says Rostrom. "The principal-protected equity gives an investor exposure to the valuable equity tranche, while protecting the principal investment from loss." This has been done across a variety of credits, including hedge funds, emerging markets and now synthetic corporate exposures. "In addition, the high-yield market offers value for synthetic investors," he notes.

Synthetic CDOs have proved popular with Japanese insurance companies, regional banks and corporates since they allow buyers to gain exposure to a global portfolio, but fund in yen. Synthetics can also limit the buyers' window of exposure to as little as seven years. "There is also a trend towards marketing principal-protected synthetic CDO products to high-net-worth individuals," says Rostrom. "There are high-net-worth individuals in Japan who don't have much way of getting yield - individuals who are accustomed to analysing sophisticated products and attracted to principal-protected products."

Other trends include long/short strategies in cash CDOs, for example, where an asset manager buys an ABS portfolio and then shorts corporate credits using CDS or indices. "The theory is that the ABS market looks attractive in the short and medium term, but CDOs are long-term products, and so if there is a housing slump in the US, for example, you have a short portfolio of corporate credits which are highly correlated to the long housing-related products and may mitigate the risk in the long portfolio." This kind of trade is becoming popular with big asset managers. "We're also seeing long/short strategies in the synthetic CDO market: you go long one tranche and short another," notes Rostrom. "The volatility in the corporate market has become very compressed and will almost certainly widen at some point in the future. This allows you to take advantage of that."

Demand for CDOs in Japan is mirrored in the current appeal for principal-protected products. Asset manager BlueBay, for example, which has also recently launched a $350 million emerging market CDO called Hemisphere, reports these have been a major driver of its business in Japan. So much so that the company opened a new representative office in Tokyo earlier this year. "We've traditionally had a strong presence in Japan, where many investors have invested in BlueBay funds," says Alex Khein, BlueBay's head of structured products and chief operating officer. "We distributed approximately $300 million of principal-protected structured products last year in Japan."

EQUITY-BASED CDOS TARGET ASIA

With Asian investors keen to achieve extra yield through the equity and lower-rated tranches of CDOs, the launch of a new product by Credit Suisse First Boston (CSFB) is likely to receive close attention in the region.

CSFB launched the first ever collateralised equity debt obligation (Cedo) transaction in May, a structure that is essentially a CDO using equity rather than credit as collateral, which allows the bank to offer considerable spread pick-up compared with traditional credit-based CDOs. At the same time the structure maintains high ratings across tranches. CSFB is currently marketing a second Cedo of up to $500 million, to be issued in early December, targeting both European and Asian investors. Jean-Manuel Dersy, head of equity derivatives sales in Europe for CSFB, says it has received interest from Korea, Japan, Taiwan and Singapore.

Dersy says the key elements of the Cedo are that the underlying assets are an equity default swap (EDS) instead of a credit default swap. The underlying assets are usually highly customised, because equity derivatives technology is used to customise the EDS. And there are two underlying portfolios, one long and one short. Another key factor is the way in which these two portfolios of names are selected. This is done by comparing credit and equity through CSFB's proprietary Credit Underlying Securities Pricing (Cusp) model, which allows it to derive implied equity volatilities from credit and vice versa.

While other banks have already applied EDS to credit - Daiwa Securities last year with a CDO based on EDS and aimed at Japanese investors, and JPMorgan with its 2004 Odysseus transaction which was based on a mixed portfolio of CDS and EDS - CSFB says its EUR229 million issue is the first transaction to achieve both an enhanced credit rating (of triple-A for the senior tranches) and deliver spread pick-up over traditional credit-based CDOs.

Lionel Fournier, co-head of CSFB's equity derivatives structuring group, says the concept of the Cedo came about after the bank identified relative-value differences between credit and equity. "Sometimes volatilities in the equity markets are not in line with credit spreads, and so you can structure a deal which will exploit the relative value between credit and equity," he says. "That's where the idea of the Cedo came from initially, and we then started to work with rating agencies to structure such a deal."

The obvious solution was to use EDS techniques to define the distressed equity risk: in other words, the risk that a stock will drop by a certain amount. "It's not exactly a credit event," says Fournier, "but you can argue that if the stock drops by, say, 60%, then the company's cashflow is stressed. This is to some extent similar to a credit event, although you can't set an exact figure for when it actually becomes the equivalent of a credit event."

One potential problem with these products is that as they are more difficult to price than CDOs of CDS, rating agencies have - initially at least - been cautious when rating them. As a result, there has been in reality a one-notch difference between ratings applied in theory and practice - meaning dealers have found it difficult to get a sizeable tranche of a CDO based on EDS with anything higher than an A rating. The Zest transaction, for example, contained A2-rated Japanese government bonds that helped it achieve a most senior tranche with an A3 rating - higher proved impossible, says the bank.

CSFB's solution enabled it to offer highly rated senior notes and spread pick-up by using a long/short structure with two portfolios. First there is the risk portfolio: a static portfolio of single-stock EDS where the investor is long risk. Then there is an insurance portfolio of short positions. In the event that a risk event affects the stocks in the risk portfolio, the impact should be offset by the insurance portfolio. "This gives a relatively market-neutral structure which is much more stable than pure long positions," says Fournier. "Obviously the rating agencies give you credit for doing this, which allows us to achieve the high rating on the senior tranche."

The relative value between credit and equity - based on models that identify inefficiencies between the pricing of each one - is used to create the composition of the two portfolios. "Both portfolios consist of names where there is a discrepancy between the volatility we observe in the equity markets and volatility we derive from the credit markets," says Fournier. "The risk portfolio comprises high-volatility names for a comparatively high credit rating, and in the insurance portfolio will be low-volatility names for a comparatively low credit rating. This allows us to offer spreads considerably higher than the normal structured CDO market normally pays."

In late October, spreads on the triple-A Cedo tranche were around Libor plus 60bp, compared with around Libor plus 50bp on a credit-based synthetic CDO; Libor plus 100bp for the double-A Cedo tranche, compared with Libor plus 65-75bp for a normal synthetic; and Libor plus 200bp for the single-A tranche of Cedo, compared with Libor plus 140-150bp. These spreads have come in since the Cedo was launched in May this year, as market conditions have changed.