CDO roundtable: Innovations in structured credit

Roundtable participants

NU, Euromoney There are more people buying CDOs than ever before in Europe. So, what is driving this interest?

JL, Solent Capital Put simply, lack of assets. There is lots of cheap liquidity at the moment and people are looking for assets to invest in.

DP, Henderson As well as lack of assets, there's lack of spread available. Anyone with a yield requirement over Libor or Euribor will find it hard to find assets that will meet their targets without going into the structured credit space.

AW, FSA And there is another driver: new entrants. The hedge funds who were less present a couple of years ago are now moving into all kinds of new products, including CDOs and negative basis trades, which has meant that certain people are buying things they were not buying in the past.

OM, JP Morgan  I think it is more fundamental. Real money investors are increasingly utilizing these products to help execute their broader ALM objectives. Much of the technology that's allowed CDOs risk/return to be tailored are being used more explicitly in asset allocation construction – creating customized payouts profiles that are not otherwise possible by investing in the asset class directly.

HL, SG CIB I would add one more thing: I believe that within the investor community there are more and more people with the technical capabilities to analyze these structures. More and more people are comfortable making these kinds of investments, and the dealers have developed a wide range of products to meet a variety of reverse enquiries, static or managed, be they single-tranche CDOs or CDO-squared or whatever.

JC, Euromoney How much do these new investors understand how these products work? And given the tight spread environment now, and given the leverage, what happens if spreads widen?

DP, Henderson The people we see when we sell these products understand them. But I do have concerns: one is an influx of real-money investors who do not understand them; another is investors who are using them as yield enhancement against a basic corporate benchmark. A lot of clients would not allow the use of leverage in their portfolios to enhance returns. If an asset manager went to a client with the request, "I'd like to borrow a lot of money against your portfolio to increase the returns you're achieving on it," many would refuse because it would be outside their mandate. Yet those asset managers are buying products with the same economic effect, but embedded within a credit-linked note. Real-money accounts which cannot buy credit default swaps should not be able to buy a synthetic CDO with those swaps embedded in them. I think these instruments do have good uses as investments, but their biggest use is the ALM work mentioned before, and where investors have a segregated fund allowed to invest in structured products.

OM, JP Morgan Remember, most of the risk being placed is double- and triple-A. It is true that seasoned buyers are moving down the credit spectrum, some even participating at the equity level by way of dedicated diversified investment programs, but much of the new volume of participants are entering at the upper end of the capital structure where sensitivity to defaults and spread widening is less.

HL, SG CIB I would add two things. First, the dealers have provided investors with sophisticated analytical tools. At SocGen we give single-tranche and CDO-squared prices to investors on deals we're doing. They can run simulations of spread-widening on our products. Second, dealers are designing structures that mitigate the impact of the spread widening, or that allow them to take advantage of it. Three years ago, perhaps, people bought paper because they were comfortable with the spread level at a particular rating. That is no longer the case. Investors really look in the portfolios and ask who the manager is, what their track record on synthetics looks like, what the rating and mark-to-market stability is and so on.

AW, FSA I have a slightly different point of view, in that I believe investors in the higher part of the capital structure are often asked to buy CDO products based solely on the triple-A rating from S&P and Moody's – this may be less true for investors in equity or lower-rated tranches, who are given more time to analyze their investments properly. In the past 12 months, we have often been asked to make a bid for the guarantee of a e200 million tranche of a CLO within one week, and often without even having received an offering memorandum. Even though we are familiar with the European and US leveraged loan markets, I believe that doing one's homework is extremely important. Reading the underlying documents, understanding the structures, analyzing and modelling stress scenarios on the cash flows, and knowing and understanding the manager's strategy are essential steps before investing in any CDO product, even if it takes more time than the market currently offers. At FSA we certainly go through all those steps before guaranteeing any CDO product.

HL, SG CIB On the ABS-side – including cash CLOs – there is so much demand that people are perhaps investing without looking closely enough at the deals. In synthetic CDOs it is a different story, with an investment process which is usually much longer.

JL, Solent Capital I agree. As cash CLOs are marketed, investors are asked to express an interest in order to ensure an allocation. Even then investors are being scaled back. Deals have been over-subscribed three, four and even seven times. Deals get tightened two or three times in the final week of pricing. That is a sign of an over-exuberant market, but it's precisely because people think they understand that asset class well that they're abandoning the normal sort of caution you would expect. Investors are still getting used to synthetics and so still do a lot more due diligence.

TM, UBS But at the same time, some of the innovations in the market – the long/short products for example – show that clients understand the risks very well and are looking for products to mitigate them.

HL, SG CIB Exactly. The long/short play in synthetic CDOs is for people who want to mitigate sensitivity to spread widening. That means they know that there's leverage in the product; they understand that the equity piece and a single-A piece don't have the same ratio when you compare the leverage and the spread and that you can play with those two pieces. In other words, investors now understand that they have a toolbox in terms of format, leverage and options. Occasionally a non-sophisticated buyer may buy a sophisticated product, but in general professional dealers deal with professional investors in a fairly heavily regulated market. I would say there is more scrutiny on the synthetic world than for many other asset classes.

VP, DWS We still tend to see the deal at least a month in advance, before it's launched, so we still get input into the structure. Synthetics are obviously different because they can be tailored to investors' demands, but even for standard structures our input is taken on board and we're very careful which tranches in which funds we buy.

JC, Euromoney And are CDOs unfairly singled out? After all, a lot of other things will be affected by rising rates and spread widening.

TM, UBS The interest-rate swap market has only existed in essentially a bull market in interest rates. So a lot of current practice may be challenged if rates go up meaningfully as in the 1970s. In the same way, what are people who have only come into structured credit during the tightening phase going to do when spreads go up?

JL, Solent Capital And why is the CDO market going to be worse affected than the underlying credit market? Some asset classes will be more stable, some more spread-sensitive.

JC, Euromoney There's more liquidity in the underlying.

OM, JP Morgan One can't ask for it both ways. If one wants tailoring on the way into an investment then one is getting exactly what one wants and is receiving unique value. Consequently, when it comes to exiting, liquidity will be less by definition. What is valuable to one investor may not be equally so to the next who will look to his/her own customization needs when assigning value to a structure.

HL, SG CIB And the dealers on synthetic CDOs are reducing the bid/offer. Spreads were wide when there was no correlation market but now this has developed and the CDS market is liquid – the indices are on a quarter basis point.

VP, DWS We're not going back to the situation of two to three years ago when we would find no bids for paper. In the triple-A pieces, if you are willing to weather the mark-to-market over long periods, then you have made money and I think that means that in the upper part of the capital structure spreads will not widen so much.

JL, Solent Capital That tension between investors able to take the mark-to-market fluctuations and those who are driven by them – such as leveraged hedge funds – creates interesting opportunities for people able to buy if the market becomes oversold.

TM, UBS Yes. Some clients go with mark-to-market and some take a more actuarial approach to risk and simply ask, "Am I getting my money back?" And credit is an unusual asset class, because it combines both those things: it has default risk, which is a long-term actuarial type of risk, and it also has the same short-term mark-to-market volatility risk that's normally associated with equities, rates or FX. That makes modelling credit assets difficult, but it also creates that tension between the long-term and short-term players.

Product development

NU, Euromoney Let's look at what's happening in the main asset classes. First, what about the CDO-squared sector?

OM, JP Morgan CDO-squareds by construction increase leverage and create steeper risk/return profiles. Some investors like the increased absolute return potential and accept the higher spread sensitivity. Many focus on the path-dependency of these instruments and like to analyze the value of that feature by studying the names in the inner and outer CDOs. But at the end of the day, CDO-squareds are just another way of expressing a view on the underlying asset class.

HL, SG CIB Also some investors like the fact that these securities contain more general risk on the market and less individual risk on each name, and so you have a more macro view with less sensitivity to unpredictable events such as fraud.

TM, UBS If you think of regular CDOs as correlation exposure, CDO-squareds are amplified correlation exposure. So if you really want to take views on correlation there's no better way. Also, a lot of these deals come with cross-subordination. Investors perceive that as a value-added feature regular deals can't offer.

JL, Solent Capital What's interesting is that if you look at CDOs of CDOs in the cash world, it's exactly the same trade. You've got a different structure – cash flow structures – but the same risks: overlap, systemic risk, manager risk, concentration risk and so on.

DP, Henderson But with the cash CDO-squared you have 20 different managers, so you have a control issue

JL, Solent Capital Well if you really want to be exposed to systemic risk, isn't it better to have 20 managers than one?

HL, SG CIB Absolutely.

AW, FSA But why don't investors just take a leveraged position in various deals? Why do a CDO-squared?

JL, Solent Capital Because people don't have the management capabilities. Investors in CDOs are making a statement that they like the asset class and the yield, but they prefer somebody else to either select or manage that asset class for them.

VP, DWS You also have investors who want specific risks within the asset class. In our case we have the capabilities, but we have investors who want specific risks and that's why we pick CDOs for them.

AW, FSA The key issue in CDO-squareds is leverage – and the potential high loss severity associated with it – together with the value that the investor places on such leverage. Each investor must ask whether the price adequately remunerates the for the embedded leverage.

HL, SG CIB You cannot say now that people don't understand this. There have been articles by almost every dealer on the risk profile of these instruments.

JL, Solent Capital I don't know. I think there is still an issue with the way people talk, or think, about correlation – and that drives these trades. When dealers talk about correlation, they're talking about an explicit base correlation curve that is inferred from the traded indices and used as an input in a copula-based pricing model. When investors talk about correlation, generally they are referring to their perception of the actual historical or potential future correlation of spreads or defaults between assets. Those two are totally different. One is a second-order effect of supply and demand in the tranche market derived from the standard model used by the dealer community. The other is something you believe about defaults or spreads and affects a fundamental view about value. So those two different definitions allow investors to take different views, but dealers definitely do not price their deals based on their assumptions about implied or explicit correlations – real spread or default correlations.

AW, FSA That's the way investors price.

JL, Solent Capital It's the way investors price; it's not the way dealers price. If it were, then hedge funds would have a great time, but they don't: dealers price based on their arbitrage models.

TM, UBS One thing I would say about all the talk of whether CDOs and their derivatives are dangerous is that it ignores the kinds of things people do in other markets. For example, a significant portion of the Swiss investor base is happy to take a very concentrated portfolio of Swiss middle-market risk, but wouldn't touch a diversified portfolio. Or take emerging market paper. We see 30-year Turkish bonds that get placed in three hours, but try to get somebody to buy into triple-B corporate paper and they'll say, "Oh that's way too risky". I've sat down with Argentines who tell me a double-A CDO with fixed recovery rates is much more risky than buying Argentine debt. I used to think credit was about giving people rational, probabilistic pay-offs, but it's really about giving people what their taste buds say is good.

The changing face of CLOs

NU, Euromoney Let's move on to CLOs. The trend here seems to be lightly leveraged funds and reverse-enquiry driven deals? JL, Solent Capital Well, one interesting deal was the KKR transaction in which KKR set up a Reit and used the CLO market to finance themselves cheaply on a non-recourse basis. Obviously, you still have people doing classic deals and you have people who are effectively creating trading vehicles – the equivalent of what AXA do in the CDS market. So yes there are real developments in the CLO markets but it's a lot more sophisticated in the US than Europe.

OM, JP Morgan Yes, the asset manager history in the US is much greater given that the loan market has been established longer, so you can clearly see who adds value across economic cycles. The longer track records make investors more open to accepting more flexible structures for these asset managers, allowing them to express their views more freely – this includes shorting for instance, a key feature being added to the traditional cash vehicles.

DP, Henderson We manage leveraged loans internally because we want to have access to that asset class without going to external managers. There are a lot of real-money asset managers – big institutions – who are not prepared to make that commitment in terms of resource and therefore are happy to go and outsource. For them, buying into the lightly levered funds is an easier sell internally than saying, "I'd like to buy some CLO equity".

An interesting development in the leveraged loan market is that we are finally breaking into the UK pension market, which traditionally has not had exposure to leveraged loans. It is a good product for that client base, but there are too few asset managers who have the access to the consultant base and who are able to market their leveraged loan expertise to that particular style of client.

TM, UBS And effectively you are getting a whole new class of investors buying the equity pieces of lightly levered loan funds who would not touch a regular equity piece of a CLO. You get the fixed-income department as opposed to the alternative asset department, so they can usually write much bigger tickets, which, as Dominic was alluding to, is broadening the client base.

VP, DWS For us leveraged loans were the asset class of choice for quite a large portion of our CDO investment. I should say though that I am slightly worried about the leverage of the underlying and also about the lack of good correlation data – neither of which I think are fully priced in.

NU, Euromoney Let's talk about the CDO of ABS sector. The cash mezzanine model is almost dead, at least in Europe. Do we see the market going entirely synthetic? JL, Solent Capital Well, we need documentation which people can agree on. There are lots of banks offering CDS of ABS, but I don't think there's common documentation between even two banks. Before we can get to truly synthetic ABS deals, we need a CDS market that works.

AW, FSA I think CDS on single-name ABS is the future and at FSA we are very excited about such a market developing. And I agree that a CDS market is the key to the development of CDOs of ABS. But for the CDS market participants must agree on credit events and other technical issues – such as physical delivery of the defaulted asset. I think it will be a year or two before we see a market where investors can easily exit positions and manage their risks in single-name ABS as they do in other markets, such as single-name corporates.

HL, SG CIB Just one point. If you want to do a synthetic CDO of ABS – single-tranche anyway – then you need liquidity and that's only there in the triple-A. We did the first CDO of triple-A ABS but then the question is, "how do you manage the fund?". There's no doubt about demand, it's massive; but to deliver, you need to source the assets and you need to have a solution in place for the funding. On CDS of ABS, I would be interested to know who is really interested in buying protection on triple-A ABS and at what price? Because the only ones I've met are hedge funds and that is not enough for a market.

JL, Solent Capital Well if I want protection on an asset class I need mark-to-market protection, which means being able to trade between counterparts. For that to work I need documentation that works. It's no different than in the CDS market where people, slowly but surely, have worked out that they need common reference obligations. And going back to the question about CDO of ABS, I think one of the reasons that you've seen the death of euro ABS is because the arbitrage has been worked through, at least for now. And the same thing is probably true of the US mezzanine ABS market. Whenever that happens the risk for investors is that managers and originators try to enhance yield by including more high yielding assets – aircraft leases for example. But I agree that the development of CDS on ABS is the most exciting thing to happen to ABS for a long time and it will change the way ABS works. It won't change the CLO market, but it'll change the ABS market.

AW, FSA I would like to respond to Hubert about who in the market is interested in buying protection on triple-A ABS assets. My answer is: many more investors than just hedge funds. A lot of negative basis trade desks at various financial institutions have bought protection from FSA (and other monolines) over the last three years on triple-A ABS assets, particularly CLOs. Managers of investment portfolios have also bought protection on their triple-A assets from FSA. Conversely, FSA would gladly be a buyer of protection on triple-A ABS assets, in order either to manage its overall leverage or as an alternative to first-loss reinsurance. I would love a real market to develop so that I could pick up my phone and say, "I would like to buy protection on a 10% junior AAA tranche of this ABS, so that FSA can guarantee the entire AAA tranche of that ABS, but can manage its risk exposure/appetite by hedging itself on the first 10%."

OM, JP Morgan Given that the cash European ABS CDO market has been arbitraged away I find it ironic that all of a sudden we can find value in synthetic ABS, precisely because of the fact that we can't fully fabricate the raw material behind it without a developed market. The arbitrage for synthetic ABS is coming from cheap funding – a bank that is not fully costing the total package of funding, credit and market risk. Synthetic ABS structures only break-up the value of the package into its components – they don't change the combined value – so the basic arbitrage equation is not changed. Separately, the US ABS CDO market is not really an ABS market, it's a home equity market, so it's like a layman's version of the CMO market, without all the bells and whistles.

NU, Euromoney Let's move on to CFOs of private equity participations and hedge fund participations. Does anyone around the table have any high hopes for this sector?

JL, Solent Capital No, I think liquidity's too easy at the moment. It's been easy to raise capital for private equity and hedge funds and raising too much capital counts against you. So unless you have people demanding principal-protected structures to get investment into private equity or hedge funds, it's an odd thing to want to do.

OM, JP Morgan We will see one or two transactions this year. The drivers will be to secure term financing against deals that historically were short-term bank-financed or done in higher spread markets that will be called and refinanced at todays better levels.

JC, Euromoney One innovation was commitment calls. The rating agencies are saying a lot of those fund of private equity deals were coming along.

OM, JP Morgan Commitment calls on capital better align the usage of cash in these structures with the timing needs of private equity investing which makes practical sense – so I would expect to see more of that feature in those deals. More interesting than the private equity topic, however, is the hedge fund-of-funds market because it is entering a phase of scalability. This is part of a bigger trend in that a lot of the technology that has permitted the growth of structured credit – the development of derivatives and investable indices – is developing in the hedge fund-of-fund space. This will help dealers to deliver scaled customization to clients by allowing them to hedge and intermediate these risks more systematically as they do with structured credit. So you will see the availability of single-tranches with different degrees of leverage, and all types of overlayson HFOFs develop.

DP, Henderson The key there is having an investable index.

NU, Euromoney Just linger on the cash side for a second there. Anne, do you buy those deals?

AW, FSA No. And we're not currently looking at them. To be able to guarantee private equity CDOs would require serious research on the underlying collateral. We haven't yet done the work, and FSA does not enter any new line of business without having first thoroughly researched and understood its components.

DP, Henderson Which is one of the concerns about the product in the first place, to the extent that someone as sophisticated as FSA isn't prepared to look at it.

OM, JP Morgan I'm not even sure people are really placing the equity from these deals. Again, I think banks are just providing leverage to portfolios as part of their prime brokerage programs which is difficult to compete with on a price basis. The investors in these portfolios may want to raise term finance through a rated, capital-markets-targeted structure if they do not want to be exposed to the risk that the short-term financing afforded by prime brokerage gets called away – a risk that exists when you least want it, that is when the portfolio's market value declines.

JC, Euromoney Valeriy, do you have any CFO paper?

VP, DWS No, we don't, and for the same reasons as FSA.

NU, Euromoney Okay, we'll move to equity default obligations – or equity default swaps in CDOs. A year and a half on from when they first appeared, what are the trends?

OM, JP Morgan The analogy here is the convertible bond market: are they fixed-income or are they equity? And the answer is both. But as with any hybrid it is difficult to identify a person at an investing organization that has the combined skills in one place. That said it does present an opportunity for those investors who do organize around this market fragmentation – in fact convertible arbitrage hedge funds were created for, and very successful at, exploiting this similar cross-market opportunity.

HL, SG CIB There have been static and managed deals but neither has been a massive success. I think there is some interest but it's still a limited number of investors.

OM, JP Morgan EDS have not gained broad acceptance as an asset class in their own right so scalable development of structures around them is unlikely as structures provide for second-order customization around an asset class. Technicians will play between those spaces and use EDS as tools in portfolio strategies, but EDS won't be part of any core asset gathering strategy.

TM, UBS There's also the issue of the skew and term structure of equity volatility. You could well get paid more for shorter maturity exposure with a relatively deep out-of-the-money strike, so why get paid less in five years than you would at one year or less? And in terms of liquidity, equity default swaps push the boundary. If things don't really trade in five years and if they don't really trade at 50% much less 70% out-of-the money strikes, do you want to use them if you are a serious equity options player? Also there were plenty of names in 2002 that did go down 70% and didn't trigger a credit event.

DP, Henderson That is why people structured CDOs with equity default swaps embedded within secondary default baskets. But then you've got the issue of – in a managed trade – how does the manager start extracting some of these names and what's the hit on the subordination when you start trying to remove equity or credit default swaps and replace them with other names? There's a lot more work to be done here before most people round this table get comfortable with these structures.

JC, Euromoney Would you use equity default swaps in a portfolio?

DP, Henderson No.

JL, Solent Capital Why not use short-dated equity puts? If you want to find a correlation between default and equity, find your correlation and tailor your put option to fit that. I'm not sure it makes a huge amount of sense within the confines of the CDO.

Mitigating exposures

HL, SG CIB What about long/short? This is one big thing we've seen in the past year with structures that address concerns about potential spread widening. And when I say long/short, I don't mean managers being able to go short in a portfolio, that's a different story. I mean different exposures with different sensitivities, for example simultaneously going long an equity piece and short the senior piece to achieve positive carry with a global position which is spread neutral.

TM, UBS We've definitely seen a proliferation of trades constructed using constant maturity credit default swaps – CMS trades, but referenced to credit spreads instead of interest-rate swaps – trying to create spread-duration neutrality. And it's not just the most sophisticated people who are looking for these products. Anyone can say, "Come on, look at credit spreads right now; you've got to be joking if you think I'm going to go long credit spreads at 10-years with levels this tight, and you've got to be deranged if you think I'm going to go leveraged long credit spreads at these kinds of levels". And so ordinary people are asking for these kinds of products and you could also think of the whole structured credit market as pausing for breath. There are a lot of people realizing they have very leveraged credit books in one form or another, and they're thinking, "My God, what do I do next".

DP, Henderson CMCDS are a good way of getting people comfortable with the risk of spreads widening, but then the issue is that people see the cost of having that CMCDS protection against spread widening and don't believe that it's compensating them enough. So they think spreads are going wider but they're not prepared to give up some of the running yield from the CDO structure to compensate them or to benefit if the spreads start to widen.

AW, FSA The question of whether or not to buy a product that protects against spread widening depends on how much value each investor puts on the mark-to-market volatility of its book. Different investors have different answers. Being a buy and hold entity and therefore having its mark-to-market always converging back to zero, FSA may have less incentive to purchase such products at current prices.

DP, Henderson  Exactly, and as you find the more marginal buyers coming in who aren't buy and hold, who are increasingly concerned about mark-to-market, that's going to start reducing the number of people you can sell to. But you're right, it's down to the relative-value players who have to put the capital to work, or the institutions and the real-money accounts who are starting to get involved, but then get panicked when they realize what would happen to the value of these notes if spreads were to go wider.

HL, SG CIB There is also an issue with certain groups of investors being forced by regulators or their prime brokers into investing in products that are less volatile in terms of mark-to-market. This, and the fact that tight spreads make people wonder what happens if they widen, creates a natural need among investors for this kind of product.

DP, Henderson The shift in the regulatory environment will certainly push people to look at that. The FSA has recently come out with new regulations for insurance companies in the UK. With their new ability to go short tranches and to mix and match the way they invest in tranches, it's going to be important to a lot more people to consider the capital they need to hold against their risk positions and their liabilities. I think that as an industry there are a lot more opportunities coming forward driven by the regulatory changes.

Moving into the middle market

NU, Euromoney Oldrich, you mentioned middle-market CDOs.

OM, JP Morgan On the cash side I see more coming. We've seen a German Mittelstand deal that took the much talked about "originate for sale model" to its extreme where corporates were offered mezzanine finance from the SPV directly and simultaneously financed in the capital markets. SME mezzanine finance is being taken pan-European and provides hard-to-get funding for small corporates. Investors like it because it offers diversification against their existing holdings with an asset class they could not otherwise source directly.

AW, FSA I agree, SMEs are a very interesting asset class. But SMEs in Germany are not the same as SMEs in France and are not the same as SMEs in the UK. So again, investors, as we do, should take the time to understand each transaction and associated collateral.

JL, Solent Capital But at least the asset class is a new asset class. A lot of the other things we're talking about are just variations on a theme – ways of squeezing more juice out of what's already there.

HL, SG CIB Yes and perhaps the next step is to build an index of SMEs in different countries.

JC, Euromoney What about the inclusion of emerging market corporate debt in CDOs?

AW, FSA To get increased spread people are including larger buckets of emerging-market corporate debt in what used to be sovereign-only deals. I'm not sure anyone knows enough about the insolvency regimes in all the countries. More importantly, when those transactions are structured synthetically, I'm not sure how to assess the restructuring risk of, say, a single-B oil company in Russia. Who can make that kind of assessment over a number of countries in one CDO?

JL, Solent Capital I don't think this should stop someone investing, it just changes the price. Frankly, if you were doing that trade I don't know why you'd want to be in the triple A. You'd probably want to be in the equity, because if you're going to increase the level of risk and finance it cheaply, the best place to be is the bottom of the pile.

OM, JP Morgan I think it comes back to the asset manager question. If you have an asset manager you're comfortable with, that has proven trading experience in the underlying asset class, then I think you give them the latitude to rely on their judgement. To me it's like any service provider – either you trust their advise or you don't.

JL, Solent Capital The real question about emerging-market CDOs is, "Is it a good vehicle for taking a view on emerging markets?" Given the assumptions made about correlation by the agencies, you don't get a particularly levered basket. If you really like emerging markets, shouldn't you go and invest in a leveraged fund that can trade quickly in and out of a whole variety of instruments in lots of countries? I don't think you want to be in a leveraged vehicle which says, "These are the terms on which I invest and these are the terms I'm stuck with", because there's huge systemic risk in the emerging markets.

DP, Henderson Also you have to look at the manager and the underlying names. There are a lot of synthetics these days that claim to be diversified through industry sector but which in fact simply concentrate country risk through multiple exposures to government-sponsored entities – South Korean names being a classic example.

NU, Euromoney Do you buy emerging market structures, Valeriy?

VP, DWS Currently not. We would ask, "Does an emerging-market bond fit into a CDO structure?" Well, they are similar to high-yield bonds and those do not work when the cycle goes into extremes – both are very volatile, low-recovery assets and you need more flexibility than the rigid CDO structure gives you. Also, CDOs of emerging-market securities are extremely illiquid which we do not like.

JC, Euromoney What about constant proportion portfolio insurance?

HL , SG CIB These are just a different way to package credit risk. It's just part of the toolbox. But I agree that it is a trend and that it is bringing in new investors into the credit market who perhaps would have viewed other products as being too risky.

TM, UBS Obviously CPPI has been around in equities and the fund-of-funds market for a long time and as Hubert was saying, it's about time we got our act together in the credit world and started offering these products. It broadens the pool of investors who can look at structured credit products, because many investors in the insurance sector as well as retail understand CPPI but have never been comfortable with CDOs. Also CPPI makes people look at credit in pure market-value terms rather than in rating terms. The traditional credit buyer says, "Give me a rating and a spread and I'll play". Equity investors or fund-of-fund investors are much more interested in proper volatility – the earlier Jonathan definition: not correlation of defaults, but correlation of paths that assets take over time. And if you can look at credit as an asset class that has a path over time and look at the dynamics of credit spreads, you attract investors who don't have time to learn about ratings and don't want to engage with Moody's and S&P and who just want credit as a market value diversifier. There's a nice story to tell about credit spreads and how they interact with rates and how they interact with equities which attracts clients who are not traditional credit investors. And the last twist on CPPI is in the banking community that can use CPPI products because they can use a credit line to UBS, in which they probably have unused capacity, get an attractively low capital charge and get an interesting investment.

JC, Euromoney In the equity markets they have a lot of problems with the path-dependent nature of CPPI.

HL, SG CIB The key is documentation. Now I think dealers and investors understand that they have to look at the way deals are worded. Is it a note? Is it a fund? What is it referencing exactly? Is it an explicit formula or not? TM, UBS Yes, and what products should go in a CPPI? They are path-dependent and if you're short volatility, you're short knock-outs in certain situations. So long credit exposure, or credit duration, is a bad choice for CPPI products, because it is so closely correlated with big trends up and down. With a CPPI product you have to do more homework and design something that is less prone to trending behaviour.

OM, JP Morgan What I find fascinating about CPPI on synthetics is that it should be a cheaper form of delivery of CPPI, because people delta hedging the underlyings behind the tranches have real-time transparency of when the knock-out is triggered. They're not looking to a separate market for a valuation. In effect, two risk-management strategies are combined into one.

JC, Euromoney Dominic, do you offer CPPI products?

DP, Henderson Yes. CPPI is a product that we did last quarter of last year and we're looking at another product for this year as well.

JC, Euromoney And do you notice any difference in the types of investors who purchase your CPPI products versus your CDO notes?

DP, Henderson Well, it depends on the strategy that you embed in the CPPI product in the first place. Depending on the strategy, CDOs and CPPI products can be complementary, as one can balance the other in terms of spread movement. To the extent that they're both positive carry, you can significantly reduce the volatility of the return from your portfolios by marrying the two together.

JC, Euromoney One issue with all these products is risk management. So what are the largest headaches?

OM, JP Morgan I think for investors one problem is that there are a lot of tools that let them analyze individual transactions in great detail but almost nothing public that lets them do scenario analysis on an entire portfolio. This makes it very hard to select transactions which combine to satisfy a particular portfolio risk/return strategy. This links into my opening point on how constructing asset allocation strategies is driving the market. We recognized this a while a go and have been offering our clients structured credit portfolio risk management tools working across all types of asset classes and different tranche levels for 18 months

DP, Henderson As a manager, the key is to be able to analyze the positions that we have across all of the portfolios that we're running; we need to identify the concentrations of risk within those portfolios. There is a trade off as a manager between having too many names – more than you can analyze and monitor actively – against constructing the same portfolio over and over again and selling it out to the same investors who don't feel that you're diversifying their portfolio. So it's a question of optimizing those two aspects. As a multi-asset manager, Henderson benefits from being able to target different asset classes. We're not always in the synthetic space; we do have the ability to look at loans; we did a property securitization last year.

JC, Euromoney What about investors?

VP, DWS The worrying short-term risk is the mark-to-market volatility of the portfolios and the interaction between the different asset classes in the more complex portfolios.

The longer-term worry is potential credit losses would be. I think spreads will anticipate any crisis well in advance, so risk management is managing spread duration, asset structure and so on. In CDOs of the same manager the key is the overlaps. We have a system that tracks individual loans, and managers don't make it easy for us. Trustee reports have a different description and code for the same leveraged loan and if they bought it three times from different dealers they'll have three identifiers. Those overlaps worry me.

AW, FSA We do not look at a transaction on the basis of whether or not it fits an overall portfolio risk/return strategy. We analyze each transaction on its own merits and each transaction must have a zero loss probability. But monitoring a large portfolio is complicated. We have more than 100 guarantees of CLOs and tracking those is not straightforward. We have a surveillance department entirely dedicated to the monitoring of those transactions and the underlying collateral. Notwithstanding that, we also have an overall portfolio strategy to monitor the proportion, as a percentage of our overall book, of corporate risk versus ABS risk, versus US municipal risk and so on.

HL, SG CIB From a dealer's point of view it's slightly different. I think that with the growing complexity of the products, you need to have technology that is scaleable and resilient and able to cope not just with current volumes and structures but with the increasing hybridization of products. And there's a link between this and with the second big issue, which is operational risk.

NU, Euromoney Well that's where we will have to leave it for this time. So thank you all for participating.