Everything you ever wanted to know about: CDOs. A Macquarie Forward Thinking insight

Transcription

1 Everything you ever wanted to know about: CDOs A Macquarie Forward Thinking insight

2 Introduction Why does the market like CDOs?: Simple, for the yield. Whether the economy is booming or in recession, whether equity markets are strong or weak, one thing remains the same: Investors need and demand income. With today s low inflation, returns on traditional term deposits and similar cash investments fail to meet investor needs. Investors have to look to alternative investments. This shift has led to a rise in popularity of higher yielding investments such as hybrid securities, property syndicates and more recently CDO s (Collateralised Debt Obligations). Income securities C o m p a r i s o n 12% 10% 8% 6% 4% 2% 0% A- BBB BBB- B- N/R Speculative o Investment grade grade or not rated ASX f listed income security yield as at 24/03/2004 source: Macquarie Group. Reset Preference Shares 12% 10% 8% 6% 4% 2% 0% A- BBB BBB- BB+ N/R Speculative grade Investment grade or not rated ASX listed reset preference shares yield as at 24/03/2004 source: Macquarie Group. f i x e d Hybrids 12% 10% 8% i n t e r e s t 6% 4% 2% 0% Capital Notes Convertible Notes Converting Preference Shares ASX listed convertible and hybrid securities yield as at 24/03/2004 source: Macquarie Group.

3 S&P rated CDO issuance has grown by 178%pa since 1996, now exceeding $567 billion This popularity stems from the ability to generate higher yield and their flexibility Although CDOs have been around in numbers since the mid- 1990s, the past few years has seen explosive growth in this asset class. Across Australia and New Zealand, 2003 saw the introduction of four listed issues of CDOs bringing Investment Grade CDOs to the retail investor for the first time, raising around A$480m between them. The market reacts quickly to meet demand, and demand has been very strong for one overriding reason: CDOs can offer a significantly higher level of yield at each level of credit rating than any other asset class of a similar credit rating. A large part of this booklet is devoted to the question of why this should be the case. The second factor driving demand is that CDOs offer a very high potential for customisation by the issuer. Return profiles, credit exposures and the currency denomination can be tailored to a level that was unthinkable a decade ago. Given the popularity of these issues, and the level of existing and untapped demand from investors for this type of higher yielding alternative investment, it is critical that advisers understand what CDOs are and how they can benefit their clients. CDOs, like managed funds, can involve hidden risks if not understood Like any financial instrument, CDOs are not without their risks. Poorly described or complex documentation can contain adverse terms that are not in the interest of investors. Their leveraged nature and great flexibility mean that they can be used to leverage the wrong sort of risk just as easily as the right sort. A poorly constructed CDO can experience a rapid series of downgrades in short order when the market turns against it. This means investors could be left with a choice between holding on to a higher-risk asset than they initially subscribed for, or selling on the secondary market at a loss. fortunately, however, they are not difficult to understand with the right insight from experts Fortunately, applying some basic common-sense guidelines, such as those outlined in this booklet, can help you assess each type of CDO. Most of the information you ll require is in the public domain. And the rest should be able to be obtained by asking the issuer of the document.

4 Insight 1: CDOs are similar to a typical corporate structure The basics What is a CDO (Collateralised Debt Obligation)? A CDO is a structured financial product whose returns are linked to the performance of a portfolio of debt obligations, such as corporate bonds. It is split into tranches, whereby the riskiest or lowest tranche, the equity tranche, receives the highest returns. Higher rated tranches offer protection against the risk of capital loss, but at proportionately diminishing returns. Portfolio of securities Highest ranking tranche(s) Lower risk, lower returns Mezzanine tranche(s) Lowest ranking (equity) tranche Retail investors Higher risk, higher returns An investment in such a CDO 1 will typically have the following elements: in which the business is investing in credit risk Quarterly or semi-annual interest payments Typically 5 years or thereabouts to maturity Interest level determined by: the running yield from the collateral, eg bank bills (which may or may not be swapped into a fixed return); plus a risk premium for taking on the risk of default on a portfolio of corporate debt securities. The higher the risk, the higher the premium. A rated portfolio typically comprising between 40 and 130 entities, each rated from AAA to BBB-. A portfolio diversified by nationality, industry and credit rating. 1 From this point on we use CDO to mean the types of CDO most commonly offered to Australian retail investors. These are more properly known as synthetic investment grade or investment grade cashflow CDOs.

5 the equity investor highly leveraged to business conditions and debtors protected by shareholders equity, at varying levels depending upon seniority. Risk and returns tranched into two or more tranches, with the highest rated tranche receiving the lowest rate of return, but being the last to experience losses: The equity (or lowest) tranche is risky, unrated. typically paying offering a gross yield over 20%pa, but likely to suffer losses as it has no protection. This tranche is typically the domain of hedge funds as it is not for the fainthearted. The final target returns, after expected losses, might be in the order of 15% pa. The higher tranches are typically rated between BB and AAA, depending upon the level of protection offered by the equity tranche (typically 3-7%). These tranches are completely protected against risk of loss unless the equity tranche and all lower rated tranches are completely wiped out. If defaults follow historic averages (for the types of portfolios typically used for CDO s) of 1-2% losses, these higher tranches should not experience any losses. More than one rated tranche may be issued. Highest rated tranches protected unless all equity and then mezzanine tranches are lost Mezzanine is protected unless entire equity tranche is lost Portfolio losses hit the lowest ranked tranche first Protection 9% 8% 7% 6% 5% 4% 3% 2% 1% AA and AAA note BBB and A notes Unrated (equity) notes

6 Why invest in CDOs? The diversification rationale for CDOs is to access an Australian fixed interest product, denominated in Australian dollars, where 90% of the risk is offshore. At least when it comes to income generating assets such as property, income securities, convertibles, hybrids, and resets Australian investors still have an overwhelming domestic bias. In the unhappy instance that the Australian economy reverses its recent form and underperforms the rest of the world, CDOs provide some useful balance in a portfolio. The investment rationale for CDOs is to exploit the spreads available on corporate debt, which are often in excess of the actual level of credit risk that is entailed in earning such spreads. This creates what is called an arbitrage opportunity. But why does this opportunity exist? To compensate holders of corporate debt for taking credit risk on any given security, the market spread ought to be at least: a) the expected loss rate of that security (i.e. default probability multiplied by the loss in event of default), plus b) a premium for uncertainty, plus c) a premium to cover transactions costs However the bulk of investment grade debt is held by high-grade debt fund managers, whose mandates often mean they are forced to sell if an investment grade bond is downgraded to below investment grade. This means they stand to realise an immediate capital loss. So we can add: d), the spread premium required for holding lower investment grade securities. Thanks to the operation of these four factors it has been observed that the long-run average spreads from BBB and A rated debt tend to be in excess of the loss rates on said debt. The investment rationale for investing in CDOs can therefore be summarised as: to make a leveraged arbitrage trade into the difference between lower investment grade bond yields and loss rates; and to make the final investment outcome more certain through diversification. By investing in an investment grade tranche of such a transaction, where the investor does not suffer the first loss, an investor is effectively giving up some of the upside on such a trade in return for greatly increased security and a less volatile return profile.

7 What are the risks? The biggest risk of CDOs is that there are more defaults in the portfolio than the equity tranche protects against The primary risk of a CDO investment is that there are more defaults in the portfolio than the equity tranche can absorb. Listed CDOs often have a rating to guide investors as to the level of repayment risk associated with the investment. Rating agencies ratings for corporate bonds, CDO s and other credit issues imply a certain level of default risk, i.e. a BBB rating assigned to a corporate bond implies the same default risk as a BBB rating assigned to a CDO issue. However, CDOs are different to corporate bonds, so S&P s process is different. So how do the rating agencies rate CDO issues? To answer this question, consider a CDO-backed debenture offering investors a tranche of a CDO of 100 investment grade entities and with an Equity Tranche of 3.75%. The debenture is over the next highest tranche. S&P rate CDO tranches by considering the statistical probability of there being greater than 3.75% in total losses across the portfolio. In a portfolio of 100 entities, this would require around six defaults 2. While six defaults of a portfolio of 100 securities might seem like a small number, it is in fact multiple times the historic average loss rate, and substantially above even the peak rates seen during times of economic distress. and for Investment Grade tranches, this risk is very low The chart at right shows the Protection Amount 4.00% historic loss rate for a 3.50% randomised portfolio of 3.00% investment grade credits with 2.50% an average credit rating of 2.00% BBB+. In this simulation, the 1.50% Loss rate worst five year period since data 1.00% was first collected in 1981, 0.50% implies a loss rate of around 0.00% 1.95%, around half the buffer provided by the Equity Tranche. In other words, the next five years would need to experience twice the default rate than that seen over any period for the last 23 years for the tranche in the Historic 1 year default rate yr default rate 5 yrloss rate 2 The portfolio suffers losses when there is a credit event, typically a default on a loan payment. This default will lead to creditors achieving a partial recovery of the defaulted debt. Assuming that the average default is around 30% recovery rate (this is about what S&P assume for most CDOs), each default will result in a 0.7% loss on the portfolio (each company is 1% of the portfolio and recovery of 30% means loss of 70%). Therefore to suffer a 3.75% loss, there needs to be 6 defaults (5 defaults would result in 5 x 0.7% in losses, or 3.5%, which is less than the 3.75% buffer).

8 example given to suffer any losses. (Again, this chart assumes a 30% recovery rate. Default history data source: S&P. Chart: Macquarie Group). Of course, there s always the chance that a diversified portfolio of 40, 100 or even 130 names won t be a good proxy for the market as a whole. As a result, the agencies apply ratings a ratings penalty to portfolios that exhibit characteristics likely to lead to poor diversification, such as excessive concentration in a particular industry. Even within these fairly tight guidelines, though, there remains room for CDO tranches of the same rating to be of quite different quality. How reliable is this rating? The three main agencies which rate CDOs: Fitch, Moody s and Standard and Poor s 3, have developed clear and transparent processes that provide a high level of guidance to investors as to the default risk of a tranche. A BBB+ rating for a CDO implies similar default risk to a BBB+ rating for a corporate bond Rating agencies state that a BBB+ rating for a CDO implies similar default risk to a BBB+ rating for a corporate bond or any other rated security. That is the definition of a credit rating: the likelihood of experiencing a default. There is much debate about the best way to rate CDOs, and the agencies have come in for a fair bit of criticism from fund managers for supposed weaknesses in their methodology. However, their track record in general has been quite good i.e. credit ratings have generally been a good guide to relative default rates across the spectrum of corporate debt and structured credit securities. Although historic data can never conclusively predict future outcomes, several studies have shown the ratings agencies performance on structured credit issues is consistent with their performance on corporate bonds: as the evidence supports Rated tranches of CDOs have rarely defaulted. Of 2,715 issues S&P rated CDO tranches on issue in , only 10 defaulted, despite the extremely poor credit conditions existing in 1999 and Fitch (ratings agency) found that default rates on structured credit issues (including CDOs) to be lower than corporate issues. 3 We focus on S&P in this document as they appear to be most active in rating retail CDO issues in Australasia. The agencies use quite different methodologies, but there has been shown to be a very high degree of consistency between agencies you will seldom find a security that is rated very strong by one agency and junk by another. 4 Source: S&P Default Study

9 $A-denominated CDOs are generally thought of as Australian fixed interest investments Where does it fit in your client s portfolio? If a CDO is denominated in Australian dollars and the interest rate is determined with reference to Australian swap rates, we recommend classifying it in the Australian fixed interest asset class. Typically 60%-80% of the total return of such instruments will be generated with reference to Australian interest rates at that time. Therefore they can be expected to rise and fall with the Australian bond market. The remainder of the return will be generated through taking on international credit risk. In this sense the CDO is analagous to the bonds of an Australian corporate with a significant proportion of its business offshore. Compared to a domestic corporate bond, however, the CDO will be less sensitive to local business conditions but much more sensitive to global credit conditions. How do I access CDOs? Listed CDOs can be accessed via the primary market, usually by a float process common to most listed securities. Supply is generally limited to $60m-$200m per tranche, and much of the total volume is allocated prior to offer open date via a firm allocation. There may also be a public pool for those who did not get a firm allocation. After listing these securities can be bought and sold normally on the exchange. The purpose of this booklet Like any investment, without insight into how the investment works, there can be unexpected risks which is why Macquarie ensures its clients are educated on how to assess the quality of these new investments So far we have outlined the fundamentals behind a CDO. While conceptually the CDO is a pretty straightforward product, the underlying transactions that enable a CDO are highly complex and it s this complexity and the greatest benefit of CDOs their flexibility that can prove their greatest risk. W ithout prudent, professional structuring and advice, CDOs can contain hidden risks. This is the issue that some industry commentators have had with CDOs being offered to the retail market. Education and support is the key. It is critical to provide our clients with the necessary tools and information to enable them to understand what is being offered to them and the associated risks and benefits. Structured credit products like CDOs are likely to be here to stay. Investors seeking higher yields without extra risk will continue to seek out such rated investments, benefiting from the diversification offered by global corporate debt, while at the same time attaining regular income free of foreign exchange volatility. Macquarie is committed to continued innovation in retail investment products and to working with advisers to ensure they are armed for this type of next generation investment.

10 Insight 2: 7 habits of effective CDO traders Professional CDO traders look at a variety of elements of CDOs to assess value. These insights are equally applicable and accessible to advisers, if you know what questions to ask. Gary Vassallo, from Macquarie s Debt Markets team, has provided us with the following insights into the habits of successful CDO traders: Does the yield compensate for the credit rating? What is the Diversification Score? Are there any correlated pairs? Habit 1: Habit 2: Habit 3: The Obvious: Compare the Credit Rating and Risk Premium All of the following elements being equal, the higher the credit rating, the lower the risk premium. So if comparing a BBB rated 5yr CDO and a A- rated 5yr CDO both offering 8%, chances are that the A- rated CDO is better value. That s of course assuming the following checks don t uncover material differences between the two products. Look for high diversification Consider the geographic and industry diversification. As a general rule, there should not be more than 10% of a portfolio in any one industry. Geographic diversification needs to be considered relative to the dominance of the major economies, i.e. you can expect 50-60% of a portfolio to be US-domiciled, but it shouldn t be more than 80%. A mix of European and Asian companies provides further diversification of economy-specific risks. Similarly, a CDO with 100 companies in it is more diversified than one with 70 or 40 companies. There is a trade-off however between diversification and increasing the risk of correlation as there are only around 300 companies globally to choose from in creating CDOs. Ask the issuer of the CDO what the diversification score is for their portfolio. The higher the diversity score, the better. Reject related-party companies in the same portfolio Correlated companies in the portfolio increase the risks of one company s default causing another s.

11 What are the underlying collateral assets? What is the average credit spread or KMV? Habit 4: Habit 5: While S&P averages imply a very low probability of more than two or three defaults per 100 Investment Grade corporate debt securities over a five years period, S&P averages imply average levels of correlation between companies. All companies are linked in some way, even if indirectly. The trick with a CDO is to ensure that the portfolio is not too much more correlated than one would expect from randomly chosen companies. Most indicators of correlation, eg industry diversification, are policed closely by the ratings agencies. A few, however, are not. For example, a CDO portfolio that contains Ford Motor Company and The Hertz Corporation involves correlation risk as Ford owns 100% of Hertz. While it s not necessarily true that Ford s default would trigger Hertz s, it would certainly place Hertz s credit worthiness in question. Similarly, Visteon Corporation, while not owned by Ford, earns 70% of its revenues by selling parts to Ford, so Ford s demise would be likely to trigger considerable cashflow difficulties with Visteon. Professional CDO developers will consider correlation risk and so you should ask your issuer if they are aware of any correlated pairs within the portfolio. Understand underlying collateral As outlined earlier, CDO returns are based on returns from a risk free collateral account, plus a risk premium for the credit risk the investor takes on. The collateral account is typically highly rated, but the assets can be as varied as credit card receivables, residential mortgages or bank bills. If the collateral has to be liquidated at short notice to settle a credit event, the market value at which it can be liquidated may have changed somewhat. It is important to understand what the mark to market risk is on the collateral in a CDO, and who bears it. Again, ask the issuer. Cross-check quality of companies within the portfolio CDOs earn their risk premiums according to the credit risk of the portfolio and the structure of the CDO itself. The S&P rating provides a measure

12 Are there any companies rated Speculative Grade by any agency? What are the recovery terms? Habit 6: Habit 7: terms of the default risk of the overall CDO, but the average credit spread of the portfolio can provide an indication of the credit risk of the underlying companies. This is not a perfect measure as other market factors can influence credit spreads of some companies (eg Ford usually trades at higher credit spreads than other BBBcompanies due to the volume of debt it has on issue). Probably the best quick check you can do to compare two similar CDOs is to compare the distribution of credit ratings. A concentration in the very lowest rating permitted for that portfolio is a sign something may be amiss. The poorer the ratings distribution, the more subordination the ratings agency will require. It is possible to construct a perfectly good portfolio out of BB securities, as long as the subordination is appropriately robust. Equity market prices also provide an insight into the default probability of a corporate credit. Proprietary systems such as Moody s KMV, which estimate alternative credit ratings based on equity price data, are becoming popular as a way of vetting portfolios for excessive risk. Look at the lowest rated names The lower the credit rating, the higher the probability of default. Therefore it s worth taking a closer look at the five or ten1 riskiest names in the portfolio those on negative watch and/or with the lowest ratings. Are you aware of any special risks which may not be captured by the credit rating? Do they dovetail well with your other investments, or are you doubling up on risk? Look for favourable recovery As discussed earlier, in the event of a default, the CDO issuer will seek to recover some of the defaulted debt from the defaulting company. CDOs vary in terms of the process for this recovery. Some will seek bids from the market for up to 45 days after default, and choose the highest bid. Others will wait up to 120 days, which, on average, will mean higher recoveries as the market has more information about the defaulting company and the probability of recoveries. One further innovation is the

13 guaranteeing of a certain recovery rate, eg 30c in the dollar, giving the investor certainty about the impact of any default. Ask the issuer whether they are guaranteeing a recovery rate, and if not, what is the process for determining recoveries.

14 Insight 3: What your clients need to know about CDOs In short, CDObacked bonds will pay their coupons and return 100% of the capital unless there are too many defaults in the portfolio For an Investment Grade CDO, the probability of this occurring is very low What returns will I get? Despite the complexity of the underlying instrument, the end investment is relatively simple to understand: A 5 year CDO-backed bond, such as the ones issued in Australia and NZ recently will pay its coupon for 5 years unless there are more defaults than the CDO tranche is protected against. That is, the BBB+ rated example used above, which had protection against around 6 defaults, will pay 8%pa for 5 years and return the investors capital in full, unless there are 7 or more defaults. If there are 7 or more defaults, the investor will lose some or all of their principal. In order for a tranche to be rated BBB+, the risk of 6 or more defaults will have to be low. But beyond that, the loss of capital is accelerated. This is the trade-off faced by the CDO investor. The probability of that number of defaults should be compared against historic evidence of default rates for investment grade securities. In fact, it is a simple process to show the historic default rate for a portfolio with the same credit ratings as those used in the CDO. For our BBB+ example, the following graph shows the loss rate that would have occurred on the whole portfolio. Note that the closest the BBB+ tranche ever came to a loss was the period in which losses would have been around 2.2%, just over half of the buffer. CDO tranche buffer vs historic losses Cu 4.00% mu lati 3.50% Generator Protective buffer Bonds (equity buffer tranche) ve De 3.00% fau 5yr cumulative loss rate 2.50% lt/ implied by historic default Lo 2.00% rates ss Ra 1.50% tes 1.00% 0.50% 0.00% As new investments, those that know the risks to avoid What are my risks? The major risk is that there are more defaults than the buffer can withstand, as outlined above. Breaking that risk down, investors need to be aware of any additional risk (beyond the market averages implied by the above graph) caused by the factors we outlined in Insight 2, namely: Are there any correlated pairs in the portfolio that will increase the probability of default within the portfolio?

15 Is the portfolio well diversified? Are there any Speculative Grade securities in the portfolio, (using either S&P or Moodys ratings)? Are any of the reference securities on Negative Watch (i.e. likely to be downgraded soon) 5? Are the recovery terms either long-dated (eg 120 days after default) or fixed (eg 30% fixed recovery)? What due diligence has been done on the underlying CDO documentation? Does the return compensate for the risk? will benefit from higher returns without additional risk This is the ultimate question for the investor. On the Australian fixed interest market as at 24 March 2004, the following yield /rating combinations were available: Income securities 12% 10% 8% 6% 4% 2% 0% A- BBB BBB- B- N/R Speculative Investment grade grade or not rated ASX listed income security yield as at 24/03/2004 source: Macquarie Group. 5 S&P assume that any rating on negative watch is one notch lower in credit quality, so having some of these issues in a portfolio is not necessarily bad.

16 Insight 4: For a more technical look at CDOs, read on Inside the black box In this section we look at how a CDO actually works, including the various parties involved in bringing them to market. The five steps in creating a CDO-backed bond such as those issued recently in Australia and New Zealand are: 1. Determine the structure of your CDO 2. Choosing the portfolio of companies 3. Creating a portfolio of credit default swaps 4. Tranching 5. Adding tranche s spread to collateral account Step 1: Determine the structure of your CDO Start by setting a structure to suit your risk/ return requirements As outlined in Insight 2, there are numerous structural choices to be made. How many companies should be included, of what credit ratings, from what countries and industries, at what credit rating and yield, and so on. The choice on structure depends upon the risk/ return requirements of the investors. Retail CDOs typically target a credit rating of BBB to AA-, with yields 1%-3.5% above the risk free rate, and use a portfolio of companies. Step 2: Choosing the portfolio of companies then select the portfolio of companies, avoiding the traps from Insight 2; and enter into a portfolio of credit default swaps to earn the risk premium. Choosing the portfolio is a trade-off between earning enough premium ( credit spread ) for taking on the credit risk of each company, and avoiding the traps identified in Insight 2, eg correlated pairs. Step 3: Creating a portfolio of credit default swaps (CDS) Now you need to purchase the credit risk for your portfolio, by entering into a CDS over the entire portfolio.

17 What is a Credit Default Swap? A CDS is a standardised contract between an investor that buys credit risk in exchange for regular payments from the seller. This process is akin to insurance: the buyer is providing the seller with insurance against credit risk in exchange for premiums. These are common transactions occurring daily around the world. In the first six months of 2003 $2.69 trillion worth of derivatives transactions took place (this figure includes CDS, baskets and portfolio transactions). The main sellers are lenders such as banks, and the buyers are institutional investors such as insurance companies and fund managers. To illustrate, consider an imaginary CDS on Ford, with the following terms: Amount: $1,000,000 Spread: 100 basis points Term: 5 years Credit Default Swap illustration Credit Protection Seller 1% of $1m ($10,000) per annum (if no default) Protection (pay up to $1,000,000 if Ford senior unsecured debt defaults) Credit Protection Seller Note that to earn the 1% per annum income stream, the Credit Risk Buyer doesn t need to put up any capital. As it is somewhat unlikely that Ford will default on its debts in the next five years, this may seem like a good deal. However, as the potential liability is up to $1m, Buyer will have to have a very high credit rating to enter into this sort of transaction. Alternatively, this may be achieved by a lien or other form of custodial arrangement whereby part of another pool of assets is legally separated from Buyer s other assets in order to protect the legal rights of Seller in the event of a default. The process that unfolds if Ford experiences a credit event can be broken down as follows: 1. Ford does something to trigger the credit default swap. Typically this credit event will be either: being declared bankrupt; failing to make an interest payment on a relevant debt security, or; undergoing a restructuring of its debt obligations. 2. A specified amount of time elapses to allow the affected debt to begin trading normally in the secondary market for distressed debt. 3. Once the specified time has elapsed, one of two things takes place. Typically, a valuation process takes place according to the terms of the CDO, eg the terms may specify that the security be priced with five reputable dealers and the highest of the five prices becomes the valuation price for purposes of the swap. 4. This valuation therefore determines the recovery rate on the default. For example, if the distressed Ford debt is trading at 70 cents in the dollar, the implied recovery rate is 30%. 5. This recovery rate is deducted from the notional value of the swap to determine the cash settlement amount. For example, in this example the recovery rate is 30% and the notional amount is $1m, the seller must pay the buyer $700, Settlement must take place within a few days of the valuation date.

18 Step 4: Tranching your portfolio Next, tailor the risk/ return to your target outcome The great benefit of tranching is customisation. If a CDO issuer has a client which requires $100m of AArated debt and another which requires $45m of BB+ rated debt, this can generally be arranged often within a few weeks. To illustrate this benefit, lets return to our example portfolio from above. A portfolio default swap such as this is not particularly interesting as a standalone investment instrument. It will provide you with a reasonable return, but a good chance of experiencing defaults in two or three of your 100 companies, and therefore some losses to counterbalance your returns. However, through the mechanism of tranching one can create a number of securities of different risk - return profiles out of a single portfolio. To continue our example above, imagine our portfolio default swap above is paying a yield of 1% for protection over $100m of assets. by splitting the risk premium income payable and the credit risk exposure into tranches Tranching diagram This means a total income stream of $1m per annum is available. It is possible to split the $100m into tranches, with the lowest rated tranches having the highest yields and vice versa. The following diagram shows how the $100m of risk and $1m pa of income might be split up. each with progressively greater credit risk and progressively higher income. Losses cascade up in the structure, W ie the bottom rated tranche experiences the first loss and no other tranches lose income or capital unless the bottom tranche has been 100% wiped out Super senior tranche. This has virtually no chance of default, statistically speaking. It pays little or no spread. AAA tranche. 2% of total. Will not experience losses unless losses occur several times worse than worse case historical levels. Pays a spread of 0.4%. AA tranche. 2% of total. Will not experience losses unless portfolio loses more than 6%, or more than 4 times historic levels. If losses are above 8%, will be wiped out. Pays a spread of 1.2%. Income cascades down, ie the bottom tranche doesn t receive income until all other tranches have been paid BBB tranche. 2% of total. Will not experience any loss if defaults remain within historic range, but will be affected if losses rise above 4% and will be wiped out if losses rise above 6%. Pays 3% interest. Equity tranche, also known as first loss piece. Accepts the full impact of the first 4% of losses to the portfolio, and will be totally wiped out if loss rate reaches 4%. If the portfolio experiences the historical average of 1.3% losses, will experience a 32.5% capital loss over 5 years. Pays 25% interest. Not for the fainthearted. Historic loss range of 0 2%

19 The income that might be earned on the various tranches is as follows (numbers are purely illustrative real-life CDOs vary widely): Class Rating Spread Total Return* % of total A Super senior ~0 6.0% 90** B AAA 0.4% 6.4% 4 C A 2.5% 8.5% 2 D Equity 15% 21.0% 4 *assuming risk free rate = 6% ** notional The highest rated tranche is not sold to the public. It is so highly rated that it is effectively risk free, but also has little potential to pay interest. The effect of the super senior tranche being notional is that the entire spread on the CDO transaction, and the risk, is concentrated into maybe 10% of the total notional size. For this reason, CDOs are often referred to as leveraged transactions, even though they are not geared in the traditional sense. Step 5: Combining the payoff from the CDO portfolio with the risk-free rate This final step simply involves adding your chosen tranche to the risk-free return from a collateral account. What this means in practice is that your investment capital is placed into a collateral account, typically backed by AAA rated securities and therefore paying the risk free rate of return. If the collateral account is paying say 6% for 5 years fixed, and the tranche you ve chosen is paying a 2.5% spread, the total return will be 8.5%pa. Floating rate securities are a little different. In this case the investor receives the collateral rate (typically at or close to the 90 day bank bill rate) plus a spread of 2.5%. If the yield curve is positively sloped, i.e. long term interest rates are higher than short term rates, then total return will initially be higher for a fixed rate security than for floating rate. For example, if the long end of the curve is paying 6% as above, and the 90 day rate is 5.5%, then the initial yield will be 8.0%. But the investor will benefit if short term rates rise, boosting the yield from the collateral account.

20 It is this collateral account that will be called upon in the event there are more defaults than the tranche you have invested in can withstand. For example, if your tranche has protection against 6 defaults, if the 7 th default occurs, the counterparty of the credit default swap for the defaulting company will require payment, and that payment will come from the collateral account.

21 Insight 5: Jargon buster Parties to the transaction Arranger, Arranging Bank Issuer Swap Counterparty Custodian How the parties come together The Arranging Bank is the one that sets up legal structure necessary to issue the CDO and is responsible for putting together the portfolio, getting the CDO rated and marketing the offer. The Issuer is the entity that issues the CDO. Invariably, the issuer is separate from the arranging bank. The most common structure is a Special Purpose Vehicle, or SPV, a thinly capitalised, standalone company with the sole purpose of issuing structured credit notes. The SPV is typically not a subsidiary of the Arranging Bank or on its balance sheet. The SPV has no active decision making power it simply carries out the terms of the CDO. The CDO is in effect selling protection on behalf of the investors, so it needs a Swap Counterparty to buy protection from it. The Arranging Bank often acts as a Swap Counterparty, though it could just as easily be an institutional investor. The Custodian holds assets in trusts for the CDO, pays out interest and capital as per the terms of the CDO payment waterfall, and assists in the valuation process when credit events occur. The administrative aspect of a CDO can be quite complicated, especially when multiple currencies and time zones are involved, and may be shared between one or more administrative specialists, possibly including the Arranging Bank. Having identified some of the parties, the structure can now be laid out graphically. It will aid clarity if we add some sample numbers to the transaction. In this example, we have assumed the transaction size is $100m, the risk free rate (measured by the 5 year swap rate) is 5%, and the spread on the CDO is 3%. Interaction of the parties to the CDO transaction Investors Capital Capital Capital + 8% income Protection Issuer Income Income stream Swap counterparty Custodian (holds collateral on trust)

22 Checklist Things to ask your CDO provider: 1. What is the return? 2. What is the credit rating? 3. What due diligence has been done on the underlying documentation? 4. What are the recovery terms? 5. Are there any related parties in the portfolio? 6. What diversification does this portfolio provide versus other assets I already have a large exposure to? 7. What form of underlying collateral is backing the transaction? 8. What are the lowest rated names in the portfolio? Are any speculative grade? Macquarie Equities Limited ABN (MEL), trading as Macquarie Financial Services, Participant of Australian Stock Exchange group, Australian Financial Services Licence No: , Level 18, 20 Bond Street, Sydney, NSW This educational booklet has been prepared by MEL for general information purposes only and is provided for the use of licensed financial advisers only. To the extent permitted by law, MEL accepts no responsibility for errors or misstatements, negligent or otherwise. The information may be based on assumptions or market conditions and may change without notice. This booklet is based on information obtained from sources believed to be reliable but MEL does not make any representation or warranty that it is accurate, complete or up to date nor does it accept any obligation to correct or update the information or opinions in it. No part of the information is to be construed as a solicitation to make a financial investment. This information is confidential and must not be copied (either in whole or part) and the recipient may not distribute it to other persons. MEL is the owner of the copyright in material in this booklet unless otherwise specified. MEL accepts no liability whatsoever for any direct, indirect, consequential or other loss arising from any use of this booklet and / or further communication in relation to this booklet.