Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
16 July 2003
CONTENTS
n Section
Page
3.1 Moody’s ABS Rating Transition Study 2002 – Many Things to Write
42
Home About!
3.2 Moody’s CDO Rating Migration Study 1996-2002 – More Doom Than
50
Gloom!
1. European Structured
Credit Portfolios
General
HIGH insurer
Monoline
wrap
Single asset/manager
Few
assets portfolio
De-linked Insured Single
industry portfolio
Securitisations
Synthetic Securitisations
ABS/MBS /CMBS
THIRD Portfolio of
CLO/CBO
PARTY CLN insured
assets
Diversified
multiple asset
Future
portfolio
Flow Multiple assets
Operating Assets
LOW Single asset
Corp Bond
HIGH ORGINATOR/ISSUER
LOW
Ideal Securitised
Bond
Source: Merrill Lynch
n Insured Securitisations
Insurance can play different roles in securitisations – it can be applied to the
assets
in the securitised portfolio, or it can be applied to the securitisation bond in
order
to enhance its credit quality. Insurance can be provided on the bond level by both
general and specialised bond (monoline) insurance companies and on the
collateral level by both general and specialised (primary mortgage) insurers.
Through the insurance the risk of the assets or the bonds is shifted to the
insurance
provider. And the insurers’ credit substitutes the credit of the bond or the
assets.
It is prudent, though, to look through the insurance cover or wrap and determine
the risk, to which the insurance provider is exposed and the investor or issuer
could be exposed in case of insurers’ default.
The application of the ‘look through’ principal require understanding of the
process of insurance underwriting, the minimum credit level requirement for the
insurance to be applied, the general business of the insurer and the management of
the insurers’ risk portfolio and overall business. In that respect, the levels of
risk
should be different and analysed differently for
• Insurance providers on assets level (e.g. insurers specialising in mortgage
insurance who provide coverage on an individual or pool basis) where the
securitisation bond is backed by a pool of multiple assets insured on
individual or pool basis by one or several insurers;
• Insurance providers on a bond level through specialised bond (monoline)
insurers, whereby such insurers take only investment grade risk, manage their
risks on a portfolio basis and are tightly monitored by the rating agencies in
terms of individual and portfolio risks;
• General insurance providers, for whom bond insurance is just one of the
numerous lines of business.
Refer to important disclosures at the end of this report.
7
ABS/MBS/CMBS/CDO 101 – 16 July 2003
For investors, this is a way of shifting the exposure from the assets and the
originator away to another party, yet such exposure to the third party is only to
the
extent of the credit performance of the underlying bond or pool of assets. In
addition, depending on the type of insurance as discussed above, investors
exposure to the third party may be viewed as exposures to: a pool of assets, a
managed portfolio of risks or a general insurance business.
n Synthetic Securitisations
In the case of synthetic securitisation, the assets, usually used for credit
reference
only, remain on the balance sheet of the respective originator/issuer. Through a
number of mechanisms the originator seeks to shift the risk of those assets to
other
parties: through a credit default swap to the swap provider or through the purchase
of protection from the market, or a combination thereof. Raising financing is of
secondary importance, if any, for the seeker of credit protection.
The reference portfolio of assets is clearly defined but remains on the sponsor’s
balance sheet – the portfolio comprises residential or commercial mortgages,
unsecured or secured consumer and corporate loans, real estate. The risks
associated with this portfolio and subject to the transfer are defined through
‘credit
events’. If bonds are issued, in the case of partially funded or fully funded
structures, their credit performance is fully linked to the portfolio and its
credit
enhancement. In an effort to de-link the rating of the bonds from the rating of the
issuing bank, the note proceeds may be used to purchase a portfolio of government
bonds, mortgage bonds or MTNs, as collateral for the bonds.
Hence, the credit performance of the bonds depends on the credit performance of
the referenced portfolio, the enforcement of credit events, while their cash
performance is linked to the cash performance of the collateral portfolio. Unlike
conventional ABS or CLN structures, the amortization proceeds of the referenced
assets are not used to make payments under the structured bonds. Investors only
have synthetic exposure to the referenced portfolio, whilst debt service is met by
the yield of the collateral portfolio supplemented by insurance premiums paid by
the sponsor in return for credit protection for the referenced portfolio. The
purchased collateral is reduced to the extent necessary in order to compensate the
sponsor for losses incurred on the referenced portfolio. Ultimately, the collateral
is
liquidated to meet note holders’ principal payments.
Credit exposures are related to the reference portfolio, its credit quality and
credit
performance related to the definition of credit events and the determination of
their occurrence (where the trustee’s role is broader than usual). Others stem from
the collateral portfolio and its performance, mechanics to protect against its
market risk. Credit default swap counterparty is another element in the credit
picture. The originator plays a role by assuming the negative carry on the
collateral through regular payments made in the form of insurance premiums.
Investors’ levels and nature of credit exposure depends on the particular type of
synthetic securitisation. In the case of credit-linked bonds, investors’ exposure
is
to both the credit quality of the asset portfolio and the issuer’s credit standing.
In
the case of unfunded or partially funded structures, the exposure is to a
diversified
assets pool, highly leveraged for the latter. Investors are gaining exposure to the
credit risk as defined by the credit events of a diversified and enhanced pool,
while
the cash flows for debt service are derived from a non-related highly liquid, high
credit quality bond portfolio.
Hence, the need for a credit framework within which to position the securitisation
bonds. The proposed framework includes two dimensions to reflect the level of
credit (in)dependence of the securitisation bond from the originator and the third
parties. The credit risk of the asset itself is a reflection of the risk of the
originator
and the third party in combination with the credit enhancement as determined by
the rating agencies.
In the language of our schematic increasing credit dependence of the securitisation
bond from the originator and the third parties would be reflected in moving left on
the horizontal axis and up on the vertical axis, respectively. By determining the
level of credit dependence of each bond on the originator and the third parties, we
can position it in our securitisation credit plane:
• We believe the ideal starting point for full credit independence would be a
securitisation bond based on a multiple asset, well geographically and
industry diversified amortising portfolio with easily replaceable servicer. In
the low right corner of our credit plane credit dependence is de minimum.
Such a securitisaton bond should pretty well fit in any portfolio with minimal
surveillance.
• Moving left (backwards) on the horizontal axis, the securitisation bonds have
an increasing degree of originator risk, i.e. risk specific to a given company
and a given industry. At its extreme, future flow securitisation bonds offer
almost full company risk along with mitigated through structural
enhancements certain particularly unpalatable risks. In this case, credit
surveillance would require primarily focus on the originator. Intermediary
points will include operating assets securitisations based on portfolio of
businesses units or a single business.
• Moving up the vertical axis from the ideal starting point, securitisation
bonds
have an increasing degree of a third party risk to the point of extreme of,
say,
fully insured by a general insurer single asset transaction with some risk
mitigants on the underlying asset. Hence, surveillance is required for the
insurance company in question and the associated insurance industry segment.
Intermediary points would include on the insurance side: bonds based on a
broad portfolio of insured assets, monoline insured bonds, and on the
traditional ABS/ MBS side multiple industry portfolios, single industry
portfolios, decreasing number of assets in a pool down to a single asset/
single
manager situation.
• Moving away from the axes and towards the middle of the credit plane,
securitisation bonds would represent different combinations of originator –
third parties credit risks. Different level of monitoring of all the parties
involved would then be necessary.
The above three extreme points and the relevant intermediary points should be
used as points of reference when building and re-balancing the credit risk of a
securitisation bond portfolio.
2. Securitisation Fundamentals
❏ Legal segregation of the assets backing the ABS/MBS from the assets
originator
❏ Bankruptcy- remote issuing vehicle
❏ No investor recourse to assets originator in case of losses in the assets
pool
❏ Rating on senior notes exceeds rating of asset originator
❏ Off-balance sheet financing for asset originator
❏ Higher yield and more protections for investors
In most cases, the originator remains the servicer of the pool of assets, with a
back-up servicer identified at the outset of the transaction in case the originator
underperforms or is declared bankrupt during the term of the transaction. Cash
flows generated by the underlying assets are used to service the notes issued. The
credit quality of the debt is therefore, based in the first instance on the credit
quality of the legally segregated, ring-fenced pool of assets. However, the ratings
assigned to the asset-backed notes will take into account not just the credit
quality
of the underlying pool, but also the extent of internal or external credit support,
i.e.
the credit enhancement for the notes, as well as other structural and legal
features.
The assets backing the transaction are sold or transferred to the special purpose,
bankruptcy remote entity, issuing the asset-backed securities. The sale or transfer
is ‘absolute’, so that the creditors of the originator of the assets have no claim
against those assets, whereas the investors in the asset-backed securities have no
recourse to the originator of those assets in case of losses under the latter.
The benefits of securitisation to issuers is in the off balance sheet treatment
achieved, as well as the capital relief gained to the extent that the underlying
assets
attract regulatory capital charges. Another essential benefit is the
diversification
of funding sources. The cost of funding through securitisation is sometimes more
competitive for certain issuers.
Investors in asset-backed securities may be able to pick up yield over comparable
plain vanilla bonds. To a large degree, this incremental spread reflects the
relative
complexity of structures and – sometimes – the lower liquidity of securities in the
secondary marketplace.
A defined revenue stream is pledged to the bondholders for the purposes of the
securitisation deal. However, bondholders’ rights are not limited only to those
revenues. In case of default or bankruptcy of the originator, the bondholders have
a right to recourse and become general unsecured creditors to the originator.
Many of the future flow transactions are structured to mitigate certain sovereign
risks (e.g. currency risk) and allow a highly rated corporate or bank borrower from
a given low rated country to raise financing at more advantageous conditions than
the country’s sovereign ceiling would otherwise permit.
The ratings of the revenue-based bonds are generally in line with the unsecured
senior debt ratings of the respective originator. Credit support in the structure
is
achieved through a desired level of debt service coverage, and excess debt service
cash flows are remitted back to the originator or retained in case of adverse
events.
As in some other debt financing transactions, covenants play an important role to
protect investors.
Revenue-based securitisation is often used when the company faces restrictive
covenants on its other debt preventing it from pledging assets. The sale of future-
generated revenue circumvents such restrictions. In addition, in emerging markets
it helps accommodate higher credit quality companies located in lower credit
quality sovereigns, and allow the former to raise funds at more advantageous
terms than the sovereign ceiling would otherwise allow them. Project finance
bonds are yet another user of this securitisation structure.
through the so-called ‘secured loan structure’: investors are granted fixed and
floating charges over the assets of the company, which allow them to take over
those assets in case of default and determine whether to operate or liquidate them
to satisfy debt payments before potential creditors.
❏ Assets backing the transaction are the core operating assets of a company or
entity
❏ ‘Secured loan structure’ allowing investors to take over the company in case
of default on
the notes
❏ Operating risk
❏ Reliance on operating cash flows and liquidation value
❏ Widely applicable in acquisition, privatisation and principal finance
❏ Exposure to sectors which otherwise may not issue debt
❏ Risks of changes in the value of the assets and the timing of its
realisation
Credit ratings for operating asset securitisations do not need to be capped at the
ratings of the company, particularly, if the assets can be shown to generate income
reliably with little input from the operator and to preserve their value in case of
originator bankruptcy. Obviously, the credit quality of the underlying obligors
(i.e. paying customers of the company) is also an essential feature of the rating
analysis. In addition, some operating assets-based securitisations benefit from
reliance on payments from the government or quasi-public bodies.
From issuers’ perspective, operating asset securitisations can be employed as yet
another corporate finance tool especially by companies involved in acquisition and
principal finance, as well as companies involved in operations related to
government tendered service contracts.
Investors gain exposure to companies which otherwise may not issue debt and
face full credit exposure to the company in question. They are entitled to a full
priority claim against the company with the potential downside in the value of the
assets (decline in the market value of the assets) and the timing of the
realisation
of that value (length of period needed to realise those assets).
n Synthetic Securitisation
In the case of synthetic securitisation, the assets, usually used for credit
reference
only, remain on the balance sheet of the respective originator/issuer. Through a
number of mechanisms, the originator seeks to shift the risk of those assets to
other parties: through a credit default swap to the swap provider or through the
purchase of protection from the market, or a combination thereof. Raising
financing is of secondary importance, if any, for the seeker of credit protection.
The reference portfolio of assets is clearly defined but remains on the sponsor’s
balance sheet – the portfolio comprises residential or commercial mortgages,
unsecured or secured consumer and corporate loans, as well as real estate. The
risks associated with this portfolio and subject to the transfer are defined
through
‘credit events’. If bonds are issued, in the case of partially funded or fully
funded
structures, their credit performance is fully linked to the portfolio and its
credit
enhancement. In an effort to de-link the rating of the bonds from the rating of the
issuing bank, the note proceeds may be used to purchase a portfolio of government
bonds, mortgage bonds or MTNs, as collateral for the bonds.
Hence, the credit performance of the bonds depends on the credit performance of
the referenced portfolio and the enforcement of credit events, while their cash
performance is linked to the cash performance of the collateral portfolio. Unlike
conventional ABS or CLN structures, the amortization proceeds of the referenced
assets are not used to make payments under the structured bonds. Investors only
have synthetic exposure to the referenced portfolio, whilst debt service is met by
the yield of the collateral portfolio supplemented by insurance premiums paid by
the sponsor in return for credit protection for the referenced portfolio. The
purchased collateral is reduced to the extent necessary in order to compensate the
sponsor for losses incurred on the referenced portfolio. Ultimately, the collateral
is
liquidated to meet note holders’ principal payments.
Credit exposures are related to the reference portfolio, its credit quality and
credit
performance related to the definition of credit events and the determination of
their occurrence (where the trustee’s role is broader than usual). Others stem from
the collateral portfolio and its performance, mechanics to protect against its
market risk. Credit default swap counterparty is another element in the credit
picture. The originator plays a role by assuming the negative carry on the
collateral through regular payments made in the form of insurance premiums.
From issuer’s point of view, synthetic securitisation is a simpler and easier way
of
transferring risk especially when financing is not a primary goal, and can be
applied to much larger (than for traditional ABS) portfolios.
Investors’ exposure levels and the nature of credit exposure depend on the
particular type of synthetic securitisation. In the case of credit-linked bonds,
investors’ exposure is to both the credit quality of the asset portfolio and the
issuer’s credit standing. In the case of unfunded or partially funded structures,
the
exposure is to a diversified assets pool, highly leveraged for the latter.
Investors
are gaining exposure to the credit risk as defined by the credit events of a
diversified and enhanced pool, while the cash flows for debt service are derived
from a non-related highly liquid, high credit quality bond portfolio.
Originator
Originator
Assets
Assets
Servicer
Servicer Transfer of Asset Pool
(Sale or Assignment)
Credit
Credit SPV
SPV // Trust
Trust -- Issuer
Issuer
Swap
Swap
Enhancer
Enhancer
Liquidity Provider
Provider
Investors
Investors
Let us look at a very simple schematic diagram that outlines the main generic
structure of securitisation transaction (Slide 9).
The originator, a company or bank originates assets in the normal course of its
business and retains them on its balance sheet. It needs financing, and one way of
acquiring it, is by monetising the assets it has, selling them to another entity,
(an
SPV) established solely for the purpose of that financing. The SPV (the Issuer)
issues securitisation bonds to investors and applies the issuance proceeds to
purchase the assets from the asset originator. The SPV’s balance sheet now
consists of assets – the assets acquired from the originator (as they are no longer
on the originator’s balance sheet) and liabilities in the form of the bonds issued.
The SPV is a shell company, which holds the assets for the benefit of the bond
investors. Hence, the need for a company to look after those assets (the servicer).
In order to achieve desired credit quality of the bonds issued based on the assets,
there is a need for additional supports: credit, structural and legal enhancements.
These are put in place to mitigate the credit, legal, liquidity, interest rate,
currency
or other risks associated with the assets and the transaction. Hence, the roles of
credit enhancer, liquidity provider and swap counterparty. The securitisation
transaction is structured within a given legal framework in order to achieve the
legal separation of the assets from the originator.
The most senior tranches of a securitisation bonds, generally achieve AAA rating.
This is an essential aspect of securitisation. Securitisation allows originators
with
lower credit quality to issue AAA rated bonds. In other words, the originator
receives better funding irrespective of its own credit worthiness on a stand-alone
basis.
The legal separation of the assets from the originator, the credit and structural
enhancement in the securitisation structure altogether provide for a securitisation
bond for highest credit quality (rating).
❏ Assets Originator
➢ entity with funding needs and with assets
which can be used as collateral for ABS/MBS funding
❏ Issuer of ABS/ MBS
➢ entity specifically created for the purposes
of the securitisation - Special Purpose Vehicle (SPV) or
Special purpose Company (SPC)
❏ Outside Credit Enhancer
➢ entity providing credit enhancement through
insurance, guarantee or reserve account
❏ Servicer
➢ entity which collects and distributes the
cash flows from the assets
❏ Liquidity Provider
➢ entity that addresses the timing mismatches
between the collected cash flows from the assets and the
cash flows to be distributed under the
structured bonds
❏ Rating Agencies
➢ determine the credit strength of an ABS and
size the credit enhancement level necessary to achieve
that credit strength
The level of credit enhancement Finally, the credit enhancement level, or the
cushion sized to absorb the asset pool
losses, should correspond to the required or
expected credit rating under the asset-
corresponds to the level of back securities: the respective rating
requires different levels of credit
desired rating enhancement: highest for AAA, lower for A and
even lower for BBB or BB
security.
Expected loss approach Some of them apply the expected loss approach,
where they determine the
expected losses in the pool under various
scenarios. Under this approach, the
expected severity of loss to investors as well
as their frequency or probability of
occurrence is determined. The rating agencies
simulate the expected cash flows
that the pool could generate, determining the
potential losses that it could
accumulate. Some go further and link the
expected loss to the level of reduction
of the internal rate of return (IRR) of the
bond: the higher the IRR reduction, the
lower the bond rating.
Weak link approach Other rating agencies base their assessment on
the so-called weak link approach.
Under this methodology, the rating agencies
look at the confluence of different
entities and assets in the structure and
determine where the structure could ‘break’,
that is, the weakest link in the chain of
assets, counterparties and entities. The
final rating of the security cannot be higher
than the weakest link in the structure.
On that basis, the rating agency could
determine the probability of first dollar loss.
‘Probability of first dollar loss’ is another
rating approach, which can be
contrasted to the ‘expected loss’ approach.
Market convention differentiates
First dollar loss between the two by stating that the expected
loss approach involves a
consideration of both probability of default
and severity of loss on the liability side
of the securitisation, while first dollar loss
approach considers only the probability
of default, or other words, a missed dollar
payment is considered a default
regardless of the recovery (may be even equal
to 100%) that could follow.
It is also important to understand that more
often than not an asset-backed security
is rated by at least two rating agencies. Each
of them may have a different
approach and may focus on different factors or
weigh differently the same factors
to determine the performance under stress
scenarios and related expected loss.
The credit enhancement which the respective
asset-backed security carries is the
highest required by any one of the rating
agencies in order to achieve the desired
bond rating. In this respect, it is worth
investigating any split ratings that exist
especially on lower-rated tranches of the
securitisation bonds.
n Asset Classes
❏ Auto Loans ❏
Auto Leases
Rights)
ABS
ABS
REVOLVING
REVOLVING
PASSTHROUGH
PASSTHROUGH
HYBRID
HYBRID
Bullet/Cont.Am.
Bullet/Cont.Am.
Amortising
Amortising
Credit
Credit Cards
Cards
MBS
MBS
Floorplan
Floorplan
HELOCs
HELOCs HELs
HELs
Trade
Trade Receivables
Receivables Trade
Trade
Receivables
Receivables Auto
Auto Loans
Loans
CLO/CBOs
CLO/CBOs
Auto
Auto Leases
Leases
12
10 Controlled Amortisation
Bullet
Payments
8
0
1 2 3
4 5
Year
Bullet – One payment at the end of last year.
Controlled Amortisation – 12 equal monthly payments over the last year.
Generally, these revolving structures incorporate two periods. The first period is
the revolving period. During the revolving period the principal under the asset-
backed security remains outstanding in full and the investors receive only interest
payments. This reflects the cash flows generated by the assets. The revenue or
the yield generated by the assets is used to pay interest to investors (coupon) and
all the other expenses we talked about – servicing fee, LOC provider fees, losses.
The principal is used to purchase new receivables. This could continue for a
number of years, and in our example on slide 17, it continues for four years.
The question now is, once the revolving period is over, how do we apply
principal? One way of using principal is to deposit it into a reserve account, and
generally, we refer to this as principal accumulation period, an accumulation
period. We accumulate principal in a reserve account in order to pay the asset-
backed security at its maturity through a soft bullet payment. In other words, we
have a revolving/accumulation/soft bullet structure.
In other cases, instead of accumulating the principal in a reserve account, we can
start paying it out in regular instalments to investors. In other words, a
revolving
period is followed by controlled amortisation and investors receive their principal
back in several regular equal instalments. This is referred to as controlled
amortisation period. Alternatively, we have a revolving/controlled amortisation
structure.
n Early Amortisation Triggers
Essential features in these structures are the early amortisation events or
triggers.
During the revolving period, the principal is applied to buy new receivables. If
the
originator of those receivables goes bankrupt, there will be no receivables to
generate and there will be no receivables to be purchased with the principal
collected. The solution then is to start passing through the principal to
investors.
In other words, certain events (or the trigger event), ‘trigger’ the amortisation
of
the bonds on a pass through basis. All principal, as collected, is passed through
to
investors to pay down the asset-backed notes. The early amortisation event could
occur at any point in time during the revolving period, the accumulation period or
the controlled amortisation period.
Stylised Repayments
Schedule for Pass-Through Security
2000
1800
1600
Principal
1400
1200
Monthly 1000
Payments 800
Interest
600
400
200
0
1 2 3 4 5
6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
Time
n Securitisation Sponsors
❏ Banks
➢ assets on their balance sheet, risk transfers
❏ Companies
➢ trade receivables, exports
❏ Project Finance
➢ cash flow stream generated post completion
❏ Municipalities
➢ tax liens, social security contributions, parking tickets
➢ revenues from a specified entity - toll road, bridge, hospital, etc.
❏ Real estate developers
➢ commercial real estate (offices, shopping malls, hotels, etc.)
❏ Countries
➢ privatisation (PFI in the UK)
➢ export credits, external credits, etc.
Why securitise?
Exposure Management
Banks and companies have certain exposure limits
or credit lines, specified limits
Transfer of loans from credit
for their credit exposure to different entities.
By transferring loans off its balance
lines where exposure limits sheet, or transferring risk associated with some
of their loans and other assets, the
have been reached banks could free their credit lines and avoid
breaching their credit limits. They
can free up the lines for further lending to the
respective sectors or clients.
Removal of Illiquid Assets from Loan Book
Banks and companies can remove illiquid assets
from their balance sheets, as is
Increased liquidity the case with non-performing or sub-performing
loans (NPL), and sub-performing
real estate. Securitisation is often the solution
for banks burdened by NPL and an
acute need to sanitise their balance sheet.
Companies that have accumulated large
amounts of real estate may want to focus on their
core businesses, but retain use of
their real estate assets without burdening their
balance sheets. They can transfer
those assets through securitisation by selling
real estate and leasing it back to their
own rise.
Transfer Uncertainties Related to Loan
Prepayments to Investors
Improvement of risk and asset- When finance companies and banks lend for
mortgages, they are exposed to the
liability management interest risk of those loans, expressed in their
prepayment behaviour. Normally,
when interest rates fall mortgage borrowers tend
to refinance their mortgages, i.e.
taking a new mortgage with a lower interest rate,
and using its proceeds to repay
in full the old mortgage loan. This creates
prepayment risk for the banks on their
balance sheets, and requires more active
management of their assets. By
transferring such assets away, the banks transfer
the related prepayments risk,
simplify, and improve the maturity management of
their balance sheet.
Achieve Better Pricing (Through Higher Debt
Rating)
With certain types of structures, like future
flow deals (say in the case of exports
Better pricing and longer term
from emerging markets), highly rated companies in
lower rated countries can raise
financing especially for cross- debt at better terms than the respective
sovereign ceiling would allow them to do
border securitisation on a straight corporate bond basis. This is
achieved by structuring a deal, where
the cash flows are generated by exports
denominated in hard currency and
captured offshore. The deal can be structured
with a rating above the sovereign
ceiling of that country, usually at the level of
the credit rating of the exporter on a
stand-alone basis, and the bond may have much
longer maturity that their
respective country or exporter would normally get
under straight bond funding.
AA
AA rated
rated
Senior
Senior AAA
AAA rated
rated
Debt
Debt
Senior
Senior Debt
Debt
(92)
(92)
(91.5)
(91.5)
Portfolio
Portfolio (100)
(100)
VS L+11bp
L+11bp
L+15bp
L+15bp
Yield:
Yield: L+35bp
L+35bp
Losses:6bp
Losses:6bp pa.
pa.
Junior
Junior Debt
Debt (7.5)
(7.5)
Tier
Tier 22
(4)
(4) L+40bp
L+40bp
L+60bp
L+60bp
Tier
Tier 22 (0.5)
(0.5) L+60
L+60
Equity
Equity
(4)
(4)
Equity
Equity (0.5)
(0.5)
Over-reliance on Securitisation
If a company heavily relies on securitisation, it will be determined to maintain
high levels of performance of the securitised assets at any cost. Performance of
any given asset pool securitisation bond below expectations would make it more
difficult for that company to access the securitisation market in the future. To
avoid that the originator may be willing to take some residual risk in the asset
pool
or to step in to support the asset pool performance. In other words, the originator
may indirectly allow for partial recourse, which is not really the stated purpose
of
securitisation and may provoke regulators’ objections.
Asset Risk Transfer Can Be Achieved Through Other Means
Finally, the transfer of risk associated with a given pool of assets could be
achieved without selling the assets, but simply by the transfer of that risk
through
credit default swaps in credit structures, or insuring that risk with an insurance
policy for an outside insurer or guarantor. We are referring here to securitisation
methods – synthetic and insured structures, different from the traditional
securitisation based on asset sale.
2.5. ABS/MBS
Investors and Their Considerations
It is important to
underline the broad range of investors in the securitisation
market and the
dominant role of banks in that context. The benefits of
securitisation
relative to other fixed income papers undoubtedly attract ever-
increasing number of
ABS investors.
n Investors in Non-US
ABS/MBS
20%
Iberia
4%
Germany /
Austria
Benelux 18%
Corporates SIV
14%
1% Pension Funds Insurance Co.s
France
10% 3% 12%
15%
Source: Merrill Lynch
Source: Merrill Lynch
by Investor Type
Scandinavia Rest of the World
Other
2%
2% Banks
5% UK
15%
18%
Insurance Co.s
18%
Ireland Germany /
Austria
1% 27%
Italy
12%
Pension Funds
6%
Iberia
Asset Managers
10%
57%
Corporates
Benelux
2%
8% France
SIV
17%
0%
Source: Merrill Lynch
Source: Merrill Lynch
n Investor Considerations
Diversification Tool
Buying into an asset or mortgage-backed security is definitely a diversification
tool when managing a broad portfolio of fixed income instruments. In addition,
asset-backed securities generally price wider than comparable corporate bonds,
which gives them attractive higher risk adjusted returns.
Rating Stability
Historically asset-backed securities ratings have been demonstrated to be generally
stable. There is lower rating volatility in the asset-backed bonds world than in
the
corporate or bank bonds world. This is to a large degree because these securities
already have built in credit and structural enhancements, as discussed earlier.
Rating agencies rating transition studies have long been a key investment
decision making tool. In this regard, every new study is anticipated with great
interest – even more so, in difficult years for the market. The catch, however, is
that even if the results presented by the studies are positive, investors, and
particularly those who have experienced problems, remain sceptical. And that is
natural – even when the averages are good, there are extremes with both
positive and negative signs, and the individual investor’s perception depends on
under which sign s/he happens to be.
Now that we better understand the data, let’s look at the study’s results and its
key
summary – the annual rating transition matrix, 1983 – 2002 (see Tables 3, 4 and
5). We particularly focus on the comparison between structured finance and
corporate finance rating transitions:
• Structured finance ratings are very stable with five of seven structured
finance
(and only one of seven corporate) broad rating categories experiencing no
rating change in more than 90% of the cases.
- Structured finance ratings are more stable in both investment and sub-
investment grade categories.
- While in the sub-investment grade category structured finance ratings
are
less likely than corporate ratings to be downgraded, they are
nonetheless
less likely to be upgraded – this may be a reflection of the going
concern
nature of a corporate entity and the liquidating nature of the assets
backing a structured finance rating.
• In addition, structured finance ratings are particularly less likely to be
downgraded than similarly rated corporates.
- Downgrade to upgrade ratio is much lower for structured finance than
for
corporate ratings historically and remained lower in 2002 despite the
significant increases in downgrades experienced by both sectors.
- Structured finance ratings, however, appear less likely to be upgraded
than corporate ratings judging by the upgrade ratios. We believe this
conclusion while correct based on the numbers presented, is somewhat
misleading when considering existing market practices. It is a well-
known fact that as the senior tranches of a structured finance security
amortize the credit strength of the junior tranches improves all other
conditions being equal – however, this is rarely reflected in a rating
upgrade for a number of reasons.
• Structured finance securities appear more likely (almost twice) to be
downgraded to Caa and below category from all broad rating categories
(except single B) in comparison to corporate finance ratings. This is a
somewhat worrisome observation. A better review of the instances of
structured finance ratings moving into the Caa and below category is needed
in order to derive the reasons behind such transition results.
- A possible explanation is that once a significant deterioration takes
place
in a securitized pool, a reversal is unlikely and the ratings are
likely to
continue to slide to the lowest possible category.
- The above finding, on the other hand, may contradict some assertions
that
structured finance ratings are stable because of intervention of
originators
to support troubled transactions, a limited instances of which have
been
observed in the past.
below
Aaa 98.90% 0.89% 0.13% 0.04% 0.00%
0.00% 0.03%
Aa 5.45% 91.46% 2.28% 0.63% 0.09%
0.03% 0.06%
A 1.13% 2.74% 93.54% 1.82% 0.52%
0.07% 0.18%
Baa 0.53% 0.65% 2.25% 90.40% 3.83%
1.26% 1.08%
Ba 0.14% 0.06% 0.78% 3.99% 86.33%
3.24% 5.46%
B 0.00% 0.06% 0.06% 0.46% 0.85%
88.95% 9.62%
Caa or below 0.00% 0.00% 0.00% 0.00% 0.17%
0.34% 99.49%
Source: Moody’s Investors Service
below
Aaa 89.93% 9.17% 1.00% 0.00% 0.00%
0.00% 0.00%
Aa 0.79% 89.66% 9.04% 0.37% 0.09%
0.02% 0.03%
A 0.05% 2.53% 90.68% 5.77% 0.70%
0.22% 0.04%
Baa 0.05% 0.28% 5.94% 86.95% 5.25%
1.12% 0.41%
Ba 0.01% 0.04% 0.61% 5.50% 82.59%
9.01% 2.23%
B 0.01% 0.06% 0.23% 0.61% 6.19%
81.22% 11.68%
Caa or below 0.00% 0.00% 0.00% 1.01% 2.57%
6.53% 89.88%
Source: Moody’s Investors Service
Table 5: Comparison of Aggregate Average Downgrade and Upgrade
Rates between Structured Finance and Corporate Finance, 1983-2002
(Broad-Rating-Based, Adjusted for Withdrawn Ratings)
Downgrade Upgrade Unchanged
Downgrade/Upgrade
Rate Rate Rate ratio
Structured Finance, 1983-2002 3.21% 2.70% 94.09%
1.2
Corporate Finance, 1983-2002 9.42% 4.14% 86.44%
2.3
Structured Finance, 2002 only 6.46% 1.41% 92.13%
4.6
Corporate Finance, 2002 only 14.42% 2.61% 82.97%
5.5
Source: Moody’s Investors Service
In our weekly commentary of the same title from Feb 20, 2003 we drew attention
to the latest Moody’s structured finance rating transition study capturing the
period 1983-2002. Here we expand our commentary to include several additional
aspects of the study.
We emphasised Moody’s conclusion about the higher rating stability of structured
finance as compared to corporate ratings.
• Structured finance better rating stability is confirmed over a horizon
longer
than one year. For example, triple-As retain their ratings in more than 95%
of
the cases over even a five-year period.
Reviewing the multi-year transition matrices we note the following:
• The incidence of ratings withdrawn for investment grade structured credit
ratings is much higher for the higher investment grade ratings than for
lower
investment grade and sub-investment grade by a factor of two or three. This
is
the exact opposite to corporate ratings – they have a much higher incidence
of
We could not find a similar data for the corporate bond world in order to derive
meaningful comparisons, maybe because the corporate world is more preoccupied
with defaults – that should be another point in favour of structured finance.
However, comparison of the average number of notches per rating move shows a
much more stable corporate sector (approximately 1.6) versus a more dynamic
structured finance sector (more than 2.6 notches in recent years and trending
upward).
This leads us to conclude that while structured finance ratings are more stable,
but once they start changing their rating changes tend to be much more dramatic
in either direction.
• The question arises then as to whether a given rating change is predicated on
a
preceding rating change of the same security – Moody’s calls this feature a
rating change momentum or path dependency. Structured finance ratings
appear to have stronger downgrade and upgrade momentum than corporates –
this is to some extent confirmed by the conditional annual rating transition
matrices.
When talking about rating dependency, it is worth noting that when one tranche in
a structured finance transaction experiences a rating change, there is a high
degree
of certainty that one or more tranches of the same transaction will experience a
rating change in the same direction. That is particularly notable in the sub-
investment grade tranches. This confirms what investors have always know
intuitively that there is a common denominator to the ratings of all tranches in a
given structured financing – that of the credit performance of the underlying pool,
among others.
Question arises, however, about the corporate ratings, given that in the corporate
world there are both issuer and issue specific ratings, and the range of issue
specific ratings are dependent on the issuer rating level and it changes, read
notching. In other words as corporate rating transitions track only issuer specific
ratings, the magnitude of rating change and particularly of contemporaneous rating
changes is not reflected in the corporate rating transition study, while it
affects,
apparently quite severely, the structured finance rating transition study.
Finally, as rating changes due to the ratings of Conseco/ Green Tree were reflected
in the structured finance study, we believe this fact distorts the rating
transition
conclusions made with regards to Baa and Ba – both in terms of general and
conditional (upgrade conditional on a downgrade) rating transitions.
Table 8: Contemporaneous Rating Change Dependency across Tranches in the
Same Deal, 1983-2002
Downgraded
Upgraded Unchanged
Conditional on Another Tranche Being Downgraded 70.53%
0.60% 28.87%
Conditional on Another Tranche Being Upgraded 0.57%
66.62% 32.81%
Conditional on Another Tranche Sustaining No Rating Change 1.69%
2.29% 96.02%
Unconditional 4.15%
3.71% 92.14%
Source: Moody’s Investors Service
(Adjusted for Withdrawn Ratings, Refined-Rating Based, Unadjusted for the Number of
Notches Per Rating Move)
• Finally, while the conclusions from the rating transition study are generally
in
favour of structured finance ratings in comparison to the corporate ratings,
the
structured finance world is not homogeneous and comprises several distinctly
different sub-sectors. While they demonstrate similar rating stability levels
(excluding, naturally CDOs), they are distinctly different in terms of their
downgrade and upgrade frequencies and rating withdrawn frequency.
The tables below demonstrate the significant differences in terms of rating
stability and transition among the four major sub-sectors of the structured finance
market.
As becomes clear from the comparison among sub-sectors, CDO has the worst
track record in structured finance world. While this is true, investors should
consider the CDO rating transition study for better understanding of the CDO
sector performance – as we have stated on many occasions this is not a
homogeneous sector and it owes its poor name primarily to US high yield
arbitrage CDOs.
In addition, the different type of structures predominant in the CDO and other
sectors of the structured finance market, along with the different levels of stress
experienced by the underlying should be taken into consideration when
interpreting the data. We note the strong performance of the real estate based
securitisations, RMBS and CMBS sectors – there is something to be said about the
strength of the collateral! While the CDO sector tracked the performance shadows
the performance of the corporate sector, RMBS and CMBS mirror the real estate
one. However, we again emphasise the importance of the deal structures in the
respective sectors to counter or connive with the collateral performance.
Considering the performance of the sub-investment grade area, again there is
evidence of significant difference among sectors – CDOs and ABS being much
more volatile than CMBS and RMBS. Furthermore, the frequency of emerging
from the Caa and below is very low in the structured finance world – and much
higher in the corporate world – there seems to be little cure for terminal illness!
Refer to important disclosures at the end of this report.
49
ABS/MBS/CMBS/CDO 101 – 16 July 2003
Caa or
Aaa Aa A Baa Ba B
below
Aaa 97.43% 2.18% 0.32% 0.07% 0.00% 0.00%
0.00%
Aa 1.48% 92.65% 4.18% 1.28% 0.27% 0.07%
0.07%
A 0.55% 4.79% 90.96% 2.81% 0.68% 0.00%
0.21%
Baa 0.00% 0.24% 1.71% 88.20% 5.13% 2.44%
2.28%
Ba 0.00% 0.00% 0.20% 2.00% 85.60% 4.40%
7.80%
B 0.00% 0.00% 0.00% 0.00% 0.00% 77.51%
22.49%
Caa or below 0.00% 0.00% 0.00% 0.00% 0.00% 0.00%
100.00%
Source: Moody’s Investors Service
Caa or
Aaa Aa A Baa Ba B
below
Aaa 99.12% 0.70% 0.10% 0.04% 0.01% 0.00%
0.03%
Aa 6.00% 91.26% 2.04% 0.55% 0.06% 0.03%
0.06%
A 1.23% 2.41% 93.96% 1.65% 0.49% 0.08%
0.18%
Baa 0.62% 0.72% 2.35% 90.77% 3.61% 1.06%
0.87%
Ba 0.17% 0.07% 0.88% 4.32% 86.46% 3.04%
5.07%
B 0.00% 0.06% 0.06% 0.50% 0.95% 90.22%
8.20%
Caa or below 0.00% 0.00% 0.00% 0.00% 0.18% 0.37%
99.45%
Source: Moody’s Investors Service
• Moody’s has increased over time the number of CDO sectors – from 4 in the
2000 study, to 10 in 2001 to 12 at present – this reflects the market
development and data availability, and is akin in some respects to adding new
industries to the corporate sector.
As the data shows arbitrage cash flow CBOs have been consistently the worst
performers in the CDO world in terms of downgrades – say, half of all
downgrades, followed well behind by the arbitrage synthetic US$ and non-US$
transactions, and even further behind, by Balance Sheet Synthetic US$ and non-
US$ transactions. It is interesting, that while arbitrage cash flow CBOs have
performed fairly poorly across the board, the other two transaction types have had
a distinctly different performance: arbitrage synthetic non-US$ deals performed
worse than the US$ ones, while balance sheet synthetic non-US$ deals performed
better than the US$ ones. Overall, more than half of the CDO types have had a
fairly good performance, all considered and particularly taking into account the
most stressful conditions the corporate and other fixed income markets
experienced in the last several years.
Table 12: Moody’s CDO Downgrade by Sector for US and European Market (Tranches)
(% of total CDO downgrades from January 1, 2002 to December 31, 2002 (Row 2002)
and from January 1, 2001 to December 1, 2001 (Row 2001))
Deal Type/ Vintage 1995 1996 1997 1998 1999
2000 2001 2002 Total
ACF CBO 2001 3.45 14.94 14.37 13.22
1.15 47.13
2002 4.0 6.8 10.8 23.6
8.5 0.8 54.6
Arb Synth non-US$ 2001 0.57
2.87 10.34 13.79
2002 0.2
3.6 8.9 12.7
Arb Synth US$ 2001 1.15
3.45 4.6
2002 0.4
0.8 5.5 0.4 7.2
B/S Synthetic US$ 2001 1.15 10.92 2.3
4.6 0.57 19.54
2002 0.8 2.1 0.4
2.1 1.3 6.8
ACF CLO 2001 1.15 0.57 4.02
5.75
2002 1.1 1.5 2.8
0.4 5.7
ACF IG CBO US$ 2001
2002 0.2 0.2
0.6 3.6 4.7
B/S Synthetic non-US$ 2001 1.15
5.75 6.9
2002 0.2
2.5 1.7 4.5
Resecuritisations 2001
2002 0.2
0.4 0.6 1.3
Market Value 2001
2002 1.1
1.1
Emerging Markets 2001 1.15 0.57
1.72
2002 0.4 0.2 0.2
0.8
B/S Cash Flow US$ 2001 0.57
0.57
2002 0.4
0.4
ACF IG CBO non-US$ 2001
2002
0.2 0.2
Totals 2001 4.6 17.8 29.9 19
17.8 10.9 174
2002 0.4 5.3 9.1 16.1 26.8
19.1 22.7 0.4 471
A review of this table coupled with the experienced accumulated on the CDO
market would suggest that:
• It is incorrect to talk about the CDO market performance in general, as much
as it is incorrect to talk about the corporate market performance in general.
• Different credit performance may be suggesting that there is a potential
diversification benefit from investing in different CDO types. This is not,
however, the same as saying that the underlying CDO exposures present
Table 13: Corporate vs. CDO Transitions in 2002 and 2001, Probability for Downgrade
for Selected Credit Ratings (adjusted for withdrawn ratings)
Synth Arb Synth Arb
2002 ACF-CBO ACF-CLO (US$) (Non US$) Em. Mkts
All CDOs Corporates
AAA 19.4 1.0 31.8 55.0 0.0
10.7 8.2
Aa2 47.4 12.5 80.0 100.0 12.5
23.9 14.7
Baa2 44.3 4.3 66.7 66.7 16.7
26.4 24.1
Baa3 55.6 30.8 nm 85.7 12.5
35.9 21.6
Ba2 59.6 5.6 71.4 nm nm
30.6 33.6
Ba3 46.2 3.3 nm nm nm
26.1 27.7
widely accepted views, which have ultimately found their way into investor
and regulator stance towards CDOs.
• These numbers are actual and are not adjusted for the significantly different
numbers of securities in each rating category for corporates and CDOs, that’s
there is sample size error inevitably distorting the results. Said
differently, the
comparison is between ‘apples and oranges’ – you choose which is which.
It is a pity that Moody’s has not presented a similar analysis of the different
corporate sector – we are fairly certain that different corporate sectors downgrade
performance will look significantly different among themselves and against the
composite over a given period of time.
While the actual downgrade experience in given years is enlightening, it is longer-
term averages on which investment portfolios are structured. In that regard
Moody’s creates a weighted-average one-year transition matrix for CDOs and
theoretical one-year transition rate for Corporates (weighting each year’s one-year
corporate transition rates by the number of CDO ratings outstanding at each rating
level as of the beginning of that year) to allow for proper comparison. On such
basis the comparison is between ‘apples and apples’, and it looks increasingly
similar, as most apples do, except for those belonging to top brands.
Table 14: Average One-Year Downgrade Risk.
Comparison of 1-Year Average Rating Transition, 1996-2002, Probability of
Downgrade*
1yr Av ACF-CBO ACF-CLO Synth B/S (US$)** Em. Mkts
All CDOs Corporates***
AAA 7.0 0.4 8.8 0.0
5.2 6.0
Aa2 16.3 6.0 6.7 9.3
13.4 13.5
Baa2 20.5 3.6 26.7 3.7
14.3 17.1
Baa3 21.8 10.7 37.5 17.2
17.5 14.3
Ba2 36.6 10.5 31.3 9.1
20.4 26.2
Ba3 19.8 5.8 50.0 14.3
16.8 24.1
*Probability of downgrade is adjusted for withdrawn ratings
**For the period 1998-2002
***Figures are from ’theoretical’ corporate transition matrix, developed by Moody’s
Source: Moody’s, Credit Migration of CDO Notes, 1996-2002, for US and European
Transactions, April, 15, 2003
Third, near-term CDO sector performance trends, observed during the last
several years, look set to continue. That means that sub-sectors that under-
performed in the past should continue to do so, maybe to a lesser extent due to
deal de-leveraging and hopefully some improvement in corporate credit, while the
CDO sectors, which performed reasonably well in the past may continue to do so
with some exceptions, which may see slight deterioration.
An indicator of future performance could be the CDO watchlist. Below we
compare Moody’s CDO watchlists in early 2002 and early 2003. This watchlist
from early 2002 can be compared with the performance in 2002 shown earlier in
this report to determine its predictive power. The comparison seems to indicate a
fairly strong predictive power.
In the early 2002 watchlist the sectors ranked from worst-to-best in the following
order: ACF CBO, ACF CLO, B/S Synthetic USD, Arbitrage Synthetic USD, etc.
The actual performance in 2002 ranks the sectors from worst-to-best in the
following order: ACF CBO, Arbitrage Synthetic non-USD, Arbitrage Synthetic
US$, B/S synthetic US$, ACF CLO.
Judging by the watchlist from early 2003, the 2003 ranking of CDO sector by
performance in worst-to-best order could be as follows: ACF CBO, ACF IG CBO,
ACF CLO, Arbitrage Synthetic US$, Resecuritisation, etc. And again ACF CBO
are well ahead of the pack – a watchlist ratio of 4 to 1, at least – and will
continue
casting a long shadow on the CDO market.
Table 15: Moody’s CDO Downgrade Watchlist by Sector for US Market and European
Market As of March 18, 2003 for Row 2003 and as of February 8, 2002 for Row 2002
Deal Type/ Vintage 1995 1996 1997 1998 1999
2000 2001 2002 Total
ACF CBO 2002 4 12 20 36
8 80
2003 2 8 27
37 16 90
Arb Synth non-US$ 2002
2 2
2003
4 2 6
Arb Synth US$ 2002
7 7
2003
1 17 1 19
B/S Synthetic US$ 2002 6
2 8
2003 1
4 5
ACF CLO 2002 2 6 6
14
2003 1 2 6
9 2 20
ACF IG CBO US$ 2002
2003 2
13 6 21
B/S Synthetic non-US$ 2002
2003
1 1 2
Resecuritisations 2002
2003 10
5 15
Market Value 2002
2003 4
2 6
Emerging Markets 2002 2 1
3
2003
B/S Cash Flow non-US$ 2002
2003
3 3
B/S Cash Flow US$ 2002 2
2
2003
3
ACF IG CBO non-US$ 2002
2003
Totals 2002 2 6 19 32 38
8 11 116
2003 1 4 14 44
74 46 4 187
Source: Moody’s, Credit Migration of CDO Notes, 1996-2002, for US and European
Transactions, April, 15, 2003 and
Credit Migration of CDO Notes, 1996-2001, February 27, 2002
In conclusion, CDO market performance apparently owes its bad name to the
arbitrage cash flow CBO sector, which in fact represents 36.56% of all CDO
tranches rated by Moody’s. It is unfortunate that one third, the bad third, of the
market has such a strong negative bearing on the rest of it in the eyes of European
investors. It may be time to re-examine the current attitude and adopt a stance
much more relevant to the market realities.
4. Assessing Subordinated
Tranches in ABS Capital Structure
• Likelihood of servicer
bankruptcy/replacement and its effects on pool
performance.
• Effects of payment rates on the
accumulation of losses and excess spread
variability.
• Break-even and sensitivity analysis, etc.
These aspects are more important to
subordinated tranches investors than to the
most senior tranches investors given the lower
loss cushion protection they have
and the higher likelihood of negative
consequences of performance volatility. We
would discuss those particular aspects in more
detail below.
n Waterfall Priorities
ABS capital structure determines the
prioritisation of cash flows among different
tranches. The senior tranche (usually rated
AAA) is the one, which usually
receives its payments of interest and
principal first – that assures that any cash
collections first service the senior tranche.
As for the subordinated tranches, one
should understand how the interest and
principal cash payments are distributed,
for example:
• Whether interest is paid pari passu or
sequentially for the senior and
mezzanine tranche;
• Whether principal is paid pro rata or
sequentially for the senior and
subordinated tranches;
• Whether interest payments for the junior
tranches can be subordinated to
principal payments for the senior
tranches; and
• Whether there is some kind of a trigger,
which changes the priority of interest
or principal distribution during the life
of the transaction.
As a rule of thumb, with the risk of stating
the obvious, it is best for a
subordinated tranche to receive its interest
on a pari passu basis and its principal
on a pro rata basis.
Spread traps can be unique to a In addition, the series may have certain
specific excess spread capture
series, reducing the amount mechanisms, which require accumulation of
excess spreads in a designated reserve
available to other series account for the benefit of that series or its
most junior tranche under certain
circumstances. Consequently, the excess spread
from that series available to other
series in the master trust would be reduced. In
most cases of master trust
structures any residual excess spread is
released to the originator. One exception
is the aircraft master trust structure, where
in some cases the excess spread is used
to ‘turbo’ the principal repayment of the
senior bonds outstanding.
• The location of principal from non-
amortising to amortising series, or from
non-accumulating to accumulating series.
Utilisation of Principal Collections
Another key point in a master trust structure
is the allocation of principal to repay
the series, or put differently, the type of
amortisation. The principal can be:
• Accumulated in a special account to
ensure a bullet payment at expected
maturity of the notes (the so-called
accumulation structure meant to establish
a soft bullet), for example: the CARDS
Master Trust; or
• Passed-through to investors up to a specified
amount (controlled or regulated
amortisation structure), for example:
Arran One MT, or fully (rapid
amortisation structure, usually a result
of the occurrence of a trigger event in
all master trusts).
In this regard, it is the level of principal
payment rate that determines the speed of
accumulation of the principal to achieve the
soft bullet structure or the speed of
amortisation in case of rapid amortisation. In
case of controlled amortisation the
required principal payment rate necessary to
meet the required controlled principal
payment amount is usually much lower than the
actual pool principal payment
rate.
Principal can be re-directed to In a typical credit card master trust structure
the revolving period is followed by
support series in their an amortisation or an accumulation period. In a
mortgage master trust structure the
accumulation phase revolving period may be interrupted in order to
accumulate collected principal and
prepayments, and resume upon payment of the
soft bullet. Under normal
circumstances, in a master trust there are
series in a revolving period and in an
accumulation/amortisation period. Hence,
principal from the revolving series may
be re-directed to support series in their
amortisation or accumulation periods.
Such redistribution mechanism will not work if
all series are accumulating or
amortising simultaneously.
With regards to the amortisation or
accumulation period, a corollary question is
related to the speed of amortisation or
accumulation. The payment rate becomes
important as it determines the likelihood of
extension affecting in a domino
fashion the senior and the junior tranches.
Principal Outstanding
Cumulative Losses
100%
10%
8%
80%
12% 8%
16%
60%
6%
40%
4%
20%
2%
0%
0%
0 5 10 15
20 25 30
Oustanding Principal
Cumulative Losses
100%
20%
6% 12%
18%
80%
16%
60%
12%
40%
8%
20%
4%
0%
0%
0 5 10 15
20 25 30
Principal Outstanding
Controlled Accumulation
100%
10%
0% Reserve Account
80%
8%
5% Reserve Account
60%
6%
40%
4%
20%
2%
0%
0%
0 5 10
15 20 25 30
Indeed trust excess spread returned to 6.5% during January 2001. Of greater
concern would have been a gradual erosion in excess spread from 6.6% to 1.4%
over a period of a few months, which would suggest an erosion of portfolio credit
quality, diminishing the credit enhancement available beneath the subordinated
pieces.
n In Conclusion
ABS capital structure allows investors to assume different risks and achieve
different awards by taking a view on the performance of a specific asset pool and
its originator/ servicer. The lower the investor moves down the capital structure,
the more intensive the credit analysis becomes and a more sophisticated cash flow
review is required. The a priori understanding of the collateral pool and its
expected performance, and the subsequent close monitoring of its actual
performance are more time consuming. However, the rewards, as evidenced by
the yield pick-up when moving down the credit levels, would compensate
investors for that.
n Basic Structure
n Seller Certificate
The seller certificate’s primary purpose is to
absorb the daily fluctuations in the
amount of receivables in the trust.
Fluctuations in the amount of receivables are
due to many factors including seasonal effects,
returns of merchandise, and the
reversal of fraudulent charges. As the amount
of receivables outstanding declines
(increases) the principal amount of the seller
certificate also declines (increases).
The minimum amount of the seller certificates
for most Master Trusts is generally
around 4-7% of the principal receivables
transferred to the trust.
n Investor Certificates
Investor certificates are the Each series of investor certificates may have
several classes. The senior, Class A,
notes sold to the market is rated AAA while the junior classes, the B
and Class C, are generally rated A
and BBB, respectively. Prior to the
introduction of BBB pieces, issuers used cash
collateral accounts (“CCA”) or collateral
investment amounts (“CIA”). Though all
three methods work to provide the same credit
enhancement, they differ in both
cash flow mechanics and marketability. For the
purposes of our discussion, we
will only focus on the Class C structure which
has grown to be the European norm
over the past year through transactions by
Barclays, Royal Bank of Scotland, and
MBNA.
Part of a Master Trust’s appeal for issuers is
that it can issue multiple series,
thereby spreading the cost across various
transactions. Series issued by the same
trust can have different characteristics (e.g.,
coupons, principal amounts, and
maturities) despite representing a claim on a
common pool of assets. For investors
this raises questions as the manner and extent
to which cash flows can be allocated
amongst the series.
Ex pected Legal
Maturity Maturity
100%
Rev olv ing
Period Accumulation
Period
0%
1 11 21 31
41 51
n Credit Protection
As previously mentioned the credit enhancement
for most credit cards ABS
consists of the subordination of cash flows
and, in certain circumstances, reserve
cash accounts. The subordinated cash flows
represent excess spread and junior
classes.
n Seller/Servicer Support
Despite the legal separation between
seller/servicer’s and the transaction, most
have a strong interest to have their
transactions perform successfully. Reasons
range from reputational risk to the
desire not to compromise future access to the
ABS capital markets. This results in an
unwritten form of credit enhancement
comes in the guise of “moral” support
from the transactions’s seller/servicer. This
consists of providing “support” to credit
their credit card ABS, which was not
required under the legal documentation.
Possible methods include over-
collateralising investor certificates or
substituting assets of a superior credit
quality into the pool.
Though not an explicit form of Though this is yet to publicly occur in
European Credit Card ABS market, a US
enhancement, seller “support” example is First Union who subordinated
its right to receive a servicing fee to
is an important protection ABS holders. For example, Mercantile
discounted principal receivables which
increased the trust’s portfolio yield.
Similarly, Banc One (First Chicago) added
better performing seasoned accounts to
its trust.
historically been hardest for subordinated tranches to pass the test given their
relatively small sizes. By issuing all tranches within a series as notes, CCCIT by-
passes this issue. In past other issuers have used a second trust to package
certificates into notes, but primarily at the BBB level.
Application to MBS – One of the most important innovations in the securitisation
of mortgages is the application of the master trust structure, typically associated
with credit card ABS, to residential MBS. Pioneered in the UK, it quickly gained
favor with the largest mortgage originators and securitisers, leading to the
establishment of large securitisation programs with regular multi-currency and
multi-tranche issuance.
n Property-Type Mix
The UK and Continent also differ by types of
property available in the real estate
investment market. The UK market remains the
most transparent market, and
long lease structures have facilitated historic
property lending in the retail and
office sectors. Office properties represent just
under 40% of UK commercial
property stock, and retail properties a further
40%. Industrial property accounts
for much of the remaining 20% of stock. In
contrast, Continental commercial
property markets consist of 50 to 70% office
properties, less than 10% retail
properties (excluding Ireland), and less than 5%
industrial properties. The balance
of Continental properties consists of multi-
family/housing association residential
real estate.
n Supply and
Demand
The degree
of supply in a property market further influences the amplitude of the
market
cycle. In general, prime retail properties with limited competition (due to
Limited competition through
restrictive
planning regimes or limited space) have exhibited lower volatility than
restrictive planning regimes office
properties. Due to historically buoyant consumer spending and low price
results in European retail rental inflation,
retail properties are less subject to speculative development than office
growth that flattens, rather properties,
which have a greater bias towards oversupply and the related property
than falls, in a downturn market
cycle.
Again,
whereas the office property cycle exhibits positive rental growth at the
peak and
negative rental growth at the trough, the retail property cycle shows
instead
moderately positive rental growth at the peak, and flat rental growth at the
trough.
n
Illustrating the Cycles in Europe
European
national markets follow different cycles not only on a national level, but
also on a
property type level. This in itself is a diversification opportunity for a
Munich Munich
Milan Milan
Berlin Berlin
Amsterdam,
Amsterdam,
Dublin Dublin
Dublin
Dublin
Rental Rents
Rental Rents Rental
Rents
growth falling
Madrid,
slowing Berlin
slowing London slowing
London
Frankfurt Frankfur
Retail Retail
Madrid
Office Office
Paris
Milan
London Dublin
Refer to
important disclosures at the end of this report.
75
ABS/MBS/CMBS/CDO 101 – 16 July 2003
•Monument
(>50 loans)
•Dolerite
•SCIP
•GECO
(Synthetic)
•Wuertthyp EU (Synthetic)
Portfolio
•Dutch Dream (Synthetic)
•Europa (Synthetic)
•Nymphenburg (Synthetic)
Other Loan
•Eurohypo 2001-1 (Synthetic)
Portfolios •Real
Value One (Synthetic)
•Paternoster
(<50>20 loans)
•Acres
CMBS Property
•ELOC
1 to 8, and 11
Conduits & Large
•Windermere
Loan Portfolios
(<20 loans)
•ELOC
9 and 10
Sale-Leaseback
•IMSER
•Telereal
Tenant
•Canary Wharf
Single
•Broadgate
•Trafford Centre
Property •La
Defense
Single-Borrower •ELOC
12
•Pan-
European Industrial Properties
Multi
•France Industrial Properties
Property
•HOTELoC
•CIT
n Transaction Motivation
Not only do the risk profiles differ among the three exposure types, but the
motivation behind the transactions and, often, the ongoing servicer commitment
also vary. We differentiate, by originator, among:
Commercial Banks
Diversified multi-borrower portfolios generally represent part of the commercial
property book of a financial institution seeking to transfer risk from its balance
sheet to the capital markets.
Developers
Developers often employ large-loan “property exposure” CMBS in order to
achieve lower financing costs than in the traditional bank lending market.
Corporates and Governments
Corporates and governments often become the sole and long-term tenants of
properties within a “tenant-type” CMBS transaction. In these cases, such
insititutions have sought to divest of non-core assets, restructure the balance
sheet,
or have desired a source of long-term financing.
n Debt-Service Coverage
The second most important indicator of commercial
property performance is cash
DSCR is generally more flow. The first year net cash flow figure can be
derived for any property
investment and represents the EBITDA of the
property (or Net Operating Income,
important than LTV
“NOI”, in real estate parlance) less capital
expenses (e.g. tenant improvements).
This figure is then compared with the required
debt-service under the loan to
arrive at debt-service coverage (“DSCR”) and
interest coverage (“ICR”) measures.
Of note, Moody’s DSCR often differs from the
published underwriter’s. In
Components of rental income
calculating the DSCR, Moody’s assumes a standard
annuity amortisation,
differ by jurisdiction and regardless of the terms of the loan, as well as
uses its own forecast of future
impact coverage ratios and interest and capitalisation rates in order to
arrive at the debt-service figures.
valuation
It is important to inquire as to the components of
NOI and rental income, as they
are not standard across Europe (see the discussion
of Lease Terms in the chapter
on Legal Aspects of European Commercial Property).
n Loan-to-Value Ratio
In addition to interest and debt service coverage,
the other key credit metric is the
loan-to-value ratio (“LTV"). The LTV measures the
property value relative to the
debt financing. Though an important factor in
determining recovery values upon a
loan default, the LTV is subject to varying
degrees of error depending on the:
• Quality of the appraisal in the related
country.
• Elapse of time since the most recent or
initial appraisal.
• Relative illiquidity of the commercial real
estate market.
• Presence (or lack) of comparable transaction
data.
Furthermore, “exit” or “balloon” LTVs do not
estimate future increases, or more
importantly, decreases in property value. Although
some properties are re-
appraised annually, in general, the LTV is not a
mark-to-market measure. Exit
LTV is often not marked-to- LTV therefore indicates only that principal
amortisation is occurring, and loans
are being re-paid. Though important credit
indicators, these are largely captured
market already in DSCR and, as time progresses, speak
little about the saleability of a
property upon the default of a loan (the primary
purpose of the LTV ratio being an
estimate of recovery value upon default).
Investors must therefore take their own
view as to the likelihood of a significant decline
in realisable value.
Lease terms range from Commercial leases across Europe vary as to the
balance of property rights and
costs allocated between the tenant and landlord.
Lease terms and tenant rights
landlord-friendly in the UK and also vary between office or industrial tenants,
and retail tenants, the latter of which
Ireland to tenant-focused in the may have greater rights in some jurisdictions.
All leases other than UK and Irish
southern European countries leases are generally indexed to inflation, in
the form of a proportion of the cost of
living index in the respective country. In any
case, the basic factors determining
the strength of a lease include:
• Term;
• Rental rate;
• Rent review;
• Landlord obligations and tenant rights; and
• Tenant quality.
Lease terms range from considerably landlord-
friendly in the UK and Ireland, to
more tenant-focused as one moves south across
the Continent. The landlord-
friendliness of a selection of European lease
terms are arranged from top to
bottom, generally most to least favourable.
n Country comparison
The table below summarized concisely several
structural elements (including lease
terms, tenant and landlord rights, enforcement
of security, and planning
permissions) of the commercial mortgage markets
in France; Germany; Italy; the
Netherlands; Spain; Sweden; the UK and Ireland
(see Table 22). Other aspects,
which need to be examined include: standard
lease terms, tenant rights, subletting
and assignment, planning permissions, appraisal
and valuation, security and
enforcement. This is a subject to a much more
detailed report.
Full Landlord
Tenant
Valuation
Typical Freq. of Statutory Basis of Indexa- Stamp Tax Recovery
Recovery insures, but Local VAT
Right to
and
Standard Lease Rent Right to Rent tion of for Property
From Excl. recovers Property
payable on Assign/ Termin.
Re-
valuation Lease Length Review Renewal Review Rent
Transfer Tenant Structural from tenant Landlord Tenant
Taxes Rent Sub-Let Early
UK Net, OMV,
Yes 15 years 5 years Yes Open None 4%
Yes No Yes Minimal Structural Tenant
17.50% Yes/Yes ** Only at
and 15)
US Gross or
Yes 5-10 years None (step No (in COLI*** or Annual Varies by
Yes, in Yes Yes Depends Depends Depends
0% Usually Only at
Refer to important disclosures at the end of this report.
Net, OMV,
(three rent usual) practice, 2-5% fixed state/
county many (unless on whether on whether on
whether prohibited break
term)
Germany Net,
“Long- Yes 5-10 years Rare No COLI Annual
3.50% No Yes Yes Structural Normal
Landlord, 16% No/Yes ** No
term”
value,
repairs maint. but (where
Infreq
Costs reimbursed parties opt
by tenant to tax)
Netherlands Net,
Yes 5-10 years 5 or 10 No (except Cost of Annual 6%
No Yes No Structural Minor Real estate
19% Yes/Yes ** No
(market
55%, to tax)
review*)
tenant
45%)
Italy Gross,
Yes 6+6 years 6 years Yes 75% of Annual 10%
No Yes Yes Structural Normal Tenant
20% Usually Only at
Historical COLI
(individual) repairs maint.
(some prohibited break
cost/
Costs landlords clause
Independent,
are exempt)
Infreq
Sweden DCF,
Yes 3-5 years 3-5 years Yes (in 75-100% of Annual 3%
No Not usual No Structural Internal Landlord
25% (where NA NA
Historical (end of most COLI
repair and repairs and (1% of landlord
cost/
lease) cases) (above
normal unusual assessed opts to tax)
Independent initial
maint. wear value)
lease*)
Spain Gross,
OMV, No 3, 5 or 10 3-5 years No COLI Annual 6%
No Yes Not Usual Structural Normal Landlord
16% Usually Only at
Independent, only)
Infreq
France Gross,
Yes 9 years 3 years Yes INSEE 3 yearly, 4.8%
(slight No Yes Yes Structural Normal
Negotiable 19.60% Usually At break
7. Basics of Synthetic
Collateralised Debt Obligation
A broader range of asset types One of the other key advantages of synthetic CLOs
– and one that is arguably the
can be used as reference primary driver behind growth of European bank
synthetic CLOs – is that such
collateral structures allow for on balance sheet credit
hedging of an asset pool that may
otherwise be unsuitable for securitisation, or
funding off-balance sheet. This
would include unfunded assets (guarantees,
derivative positions, etc), loans with
restrictions on assignment or loans from
jurisdictions with legal or other obstacles
on transferability. Indeed, many synthetic CLOs
are referenced to a portfolio of
multi-jurisdictional loans.
Transaction execution is As a matter of course there is much less
transaction documentation involved in
simpler European synthetic CLO deals compared to the true
sale variety – it is enough to
mention the absence of an asset sale and transfer
agreement. Synthetic CLOs
allow banks to avoid the onerous process of
combing through each underlying
loan document and unraveling assignment clauses
or other restrictions in order to
sell the asset. As a result, transaction
execution is that much simpler.
E x a m p le o f fu n d in g c osts
F u lly F u n d ed
(pa ) assu m in g 5% L IB O R P artially F u n d ed
S u p er S en io r C red it
S e n io r N o tes
D efa u lt S w a p
(9 0 )
(8 7 )
L+25
10 bps
L +27
M ezza n in e N o te s
M e zza n in e N o tes
(4 )
(4 )
L+60
L +60
J u n io r N o tes (3 ) L + 1 5 0
J u n io r N o tes (3 ) L + 1 5 0
R eta in ed E q u ity (3 )
R eta in ed E q u ity (3 )
Swap
‘Super Senior’
Counterparty
Unfunded Originator’s
Portion Reference
Loan
Portfolio
Pay-out of
realised loss
Sale of
under ‘Credit
Event’
Collateral
Asset Backed
Funded Portion
Credit Default
Swap Bankruptcy Remote
Interest
Agreement
Issuing SPV Noteholders
Commercial Bank &
Reference Loans
Collateral
Swap Premiums
Principal
Interest Payments due
to Noteholders
to Noteholders
Investors
Reference Credit
Occurrence of
PayOut under‘Credit Event’
‘Credit Event’
Credit
Default Swap Premiums
Chart 38: Key Risk Considerations in Bank CLOs – Conventional Versus Synthetic
Originator /
Risks
Reference Pool Sponsor Bank
Structural Market
Servicer
Associated with
Credit Quality Credit Risks
Considerations Considerations
Considerations
Collateral
• Credit enhancement as a
performance • Competence in counterparty (if
applicable)
through vs bullet, call
different scenarios
assets • Effectiveness of
risks
• Availability of back up
• Protection against
deterioration in value or
saleability of collateral
• Role of supporting parties (like
Text in italics denote considerations that are unique to the respective structure
type. Source: Merrill Lynch
n Some
Differences Between Synthetic and Conventional CLOs
We have
identified several key investment differences between a typical synthetic
CLO and a
conventional, true sale CLO:
Bullet
Redemption, Liquid Collateral
The
availability of readily saleable collateral in synthetic CLOs allows for bullet
redemption
of the notes in most cases, which in itself is investor positive. A
powerful
feature here is that full note redemption – assuming a default neutral
scenario –
will be immediately realizable under certain acceleration trigger events
(sponsor
bankruptcy for instance). By contrast, in cash based transactions the
length of an
early amortization payout window will be determined by underlying
pool payment
rates.
Of course,
full note redemption in this instance will depend on the value and
saleability
of the collateral in the open market and/or on the ability of the put or
repo
counterparty to meet its obligations.
Potential
Exposure to Collateral Credit Risks
The credit
quality of the collateral determines the credit quality of the synthetic
CLO notes.
From an
investment perspective, the collateral presents less credit concerns when
it is
comprised of selected government securities. But a number of transactions
are backed
against non-government securities including most often Pfandbriefe.
Refer to
important disclosures at the end of this report.
96
ABS/MBS/CMBS/CDO 101 – 16 July 2003
The triple-A ratings on these non-zero risk weighted securities allow for the
senior
notes of synthetic CLOs to achieve similarly high ratings. Yet non-government
collateral is likely to be more credit volatile than true ‘risk free’ paper. For
example, it is far from certain whether Pfandbriefe ratings can withstand a
significant deterioration of the respective bank’s rating, or indeed a change in
the
legal or regulatory framework supporting these instruments2. A downgrade of the
collateral may result in a downgrade of the notes.
Less Scope for Servicing and Structure Related Risks
Adequate servicing of the asset pool (administering collections, remitting
payments, etc.) in order to meet debt servicing under the capital structure is an
important consideration in cash flow CLOs. The servicing aspect of a CLO is less
of an issue in a synthetic structure:
• In case of the outright insolvency of the seller/servicer, there is no need to
transfer the servicing function. Using the collateral pool, as described
above,
investors can be taken out of the transaction immediately. Hence, the risks
inherent in cash based structures that a proficient substitute servicer cannot
be
found are of no relevance to synthetic CLOs.
• Any breach of servicer duties or warranties can be easily ‘reversed’ as no
sale
/transfer of assets would have taken place. Having synthetic – as opposed to
cash – exposure to the reference pool also means that any breach of structural
criteria (such as eligibility conditions) can be quickly ‘reversed’. A
reversal
in this instance amounts simply to canceling the default protection covering
those particular assets.
• As there is no cash flowing from the reference pool, risks associated with
cash
transfer, commingling or set-off inherent in true sale CLOs are not relevant
for synthetic structures.
Synthetic CLOs also have less scope for other structural cash flow related risks
such as basis or currency mismatches. In multi-currency denominated deals, for
instance, foreign exchange risk can be assumed by the protection buyer. In cash
structures, any collections denominated in currencies other than the issuing
currency will need to be hedged, of course, exposing investors to counterparty
risk.
Potential Exposure to Sponsor Bank Risk
The exposure to the sponsor bank depends on the roles it performs in a synthetic
CLO structure. Such exposure could include and be mitigated in the following
ways:
• The risks to the payment obligations of the sponsor bank as credit protection
buyer can be fully mitigated in synthetic CLOs, as noted earlier in the
report,
through – for example – an arrangement to deposit the payment well before
the payment date.
• In addition to being the buyer of credit protection, the sponsor bank may also
be a repo or hedging counterparty (vis-à-vis collateral mark-to-market risks)
in the transaction. The risks associated with the sponsor bank being a repo or
hedging counterparty based on its short-term rating can be mitigated through
appropriate downgrade language. While not a constraint to rating the
synthetic CLO notes, any deterioration in the bank’s credit quality will
require
remedial action, failing which the collateral is liquidated and the notes paid
down.
2
In its report titled ‘German Pfandbriefe, Moody’s Analytical Approach’ (June
1996), the rating agency stated that “the probability of default for Pfandbriefe
can
not be isolated from the creditworthiness of the issuing entity”.
Refer to important disclosures at the end of this report.
97
ABS/MBS/CMBS/CDO 101 – 16 July 2003
• Clearly, investors are directly exposed to the sponsor bank’s credit in any
synthetic structure where the collateral pool consists of bonds credit-linked
to
the bank. Where the collateral comprises highly rated non-government
collateral that is ‘related’ – though not technically ‘linked’ – to the
sponsor
bank (to include, in our opinion, Pfandbriefe), the ability to pay down the
notes under seller credit deterioration or insolvency may not be entirely
assured either. Such collateral is likely to be sensitive to the performance
and/or credit profile of the bank (a hedge provided by a suitably rated third
party should mitigate this risk). In a number of synthetic CLOs, note holders
may ultimately have to take physical delivery of the collateral in lieu of
redemption if a market sale or repurchase cannot be effected3 (examples of
such deals include Globe 2000 and Cast 2000).
In conclusion, many synthetic CLOs are structured in such a way that there is
multiple reliance on the sponsor bank. That is, there may be many ‘layers’ of
bank risk in the transaction – for instance, credit dependence given the nature of
the collateral, reliance on the bank as a hedging and/or repo counterparty, etc.
The
more such layers, the greater the ultimate linkage to the bank, in our view, even
if
such risks are mitigated at each level. True de-linked cash flow CLOs do not
typically have exposure to the sponsor bank to the extent that many synthetic
transactions may do.
Loss Determination and Settlement
Synthetic CLOs have more complex settlement and valuation issues related to the
credit events in comparison to the cash based structures, where loss
crystallization
and impact on the deal’s liabilities are relatively straightforward:
• On the one hand, credit events and payout amounts in synthetic CLOs are pre-
defined, whereas in cash or ‘true sale’ based structures any form of
underlying non-payment or default adversely affects investors. This feature
of synthetic CLOs is clearly investor positive.
• But on the other hand, credit events that are too broadly defined may trigger
premature and more severe loss for noteholders. In transactions where the
credit events are not clearly described, investors would have to rely on a
third
party for the interpretation and validation of a covered event.
The credit events, therefore, need to be carefully examined. In our view, synthetic
CLOs with broad credit event definitions and complex workout procedures prove
a challenge to rating agency analysis – default probabilities and loss severity
assumptions in such deals are far from being an exact science. Structures where
the loss payout is fixed from the outset mitigate risk of loss volatility, but we
realize that only very few deals are structured this way.
Greater Reliance on the Trustee
There is a greater reliance on trustees in being able to protect the interests of
noteholders in synthetic CLOs. Among other duties, the trustee in synthetic CLOs
will be required to be vigilant vis-à-vis:
• Monitoring and selling the collateral as required.
• Verifying the determination and allocation of losses.
But there is one important exception – under a seller bankruptcy (as mentioned),
the trustee’s performance in a conventional structure becomes crucial. Continued
deal servicing post seller insolvency, by contrast, is of no relevance in synthetic
transactions.
3
In our analysis, we do not take into consideration the potential benefits of being
delivered collateral in lieu of payment. Such benefits include the replacement of
notes by lower risk weighted paper.
Refer to important disclosures at the end of this report.
98
ABS/MBS/CMBS/CDO 101 – 16 July 2003
More recently, the question cash or synthetic is equally applicable to the way the
exposures in the CDO portfolio are created through:
• CDS: selling protection on given names (CDS); buying protection (which is
the equivalent of cash bond shorting) is also possible in some cases.
• Cash: buying exposure to given names in the form of bond purchases (short-
selling of bonds has not been allowed, to our knowledge), or
• Hybrid: a combination of the two above above.
To these two broad categories of assets, one can also add total rate of return
swaps, bi-variate risk positions, etc., which we believe will be of limited use in
future deals, as have been in past ones.
In this regard, we believe that the key question is not so much how the reference
pool is created (we will address this issue later), but rather
1. What is its credit quality? and
2. What level of diversification can be achieved and maintained?
Given that CDS exist primarily on investment grade corporates, in order to
improve the diversification of a CDS pool, cash bonds are a welcome addition. In
fact, cash exposures can take the form of bonds, ABS, convertibles (stripped), etc.
We note that particularly ABS, in its broadest meaning (that is, all asset
classes),
can add to the portfolio exposures to credit sectors not easily available through
CDS or other cash alternatives. Such sectors include consumer loan portfolios as
in the traditional ABS/RMBS, corporate portfolios on names, where CDS are
scarce (high yield) or non-existent (mezzanine loans, SME bank loans, CMBS,
etc.). Such a portfolio should benefit from broader defensive diversification and
higher rating stability, judging by recent studies of rating transitions of
different
fixed income instruments performed by the rating agencies. These studies also
emphasize the difference in transition ratios and in spread volatility for ABS and
corporates under the same economic conditions of duress experienced in the last
few years – this can be viewed as evidence of lower correlations between the two
sectors.
In addition, a hybrid managed CDO allows CDO manager to explore possible
relative value opportunities that may emerge between CDS and cash assets to the
benefit of investors in the respective CDO or achieve better returns through more
efficient leverage through CDS. Overall, a hybrid structure should allow for the
optimization of asset allocation given CDO manager’s skills and sector
knowledge.
One should not forget that regardless of how a given deal is structured it is,
after
all, mainly about credit. A portfolio of credits – cash bonds or reference entities
–
require similar analysis in terms of credit quality, diversity, single credit
exposure,
total spread, default and recovery expectations, etc. However, the way one arrives
at them may require somewhat different analysis.
n One/Multiple CDS – One/Multiple CDS Counterparties
In its early days synthetic securitisation usually involved the transfer of the
risk of
the entire portfolio through a single basket credit-default swap to one
counterparty, often the SPV. A possible variation, in the case of partially funded
synthetic CDOs for example, would be two separate risk transfers: one CDS
directly with the super-senior protection provider and one with the SPV.
While these structures are applicable to static and dynamic arbitrage synthetic
CDOs, the choice depends on the level of management anticipated in a given
transaction.
For static deal one CDS counterparty is the rule, as could be with lightly managed
portfolios, where the SPV can trade with or through the CDS counterparty. As the
level of management and manager discretion increases, the number of
counterparties to the SPV can increase into multiple counterparties.
Hence, the simplicity and ease of execution of the static balance sheet structures
are replaced with higher documentary complexity of actively managed CDS
structures to allow for an active and more efficient portfolio management. While
the need for competitive pricing for trading and settlement of CDS requires
several CDS counterparties, an excessive number of counterparties used may
inadvertently create the additional problems of documentation complexity and
management of counterparty risk. Hedging and offsetting trades with two
counterparties creates additional difficulties in determining the credit risk of
the
position.
So the question is not so much how many CDS there are in the deal, but rather:
1. How many swap counterparties- one or multiple - are used? and
2. How is counterparty risk managed?
3. Is it necessary to have multiple conterparties, i.e. is the portfolio expected
to
be actively managed?
Given that a CDS is an agreement between a protection seller and a protection
buyer and the related payment flows going in both directions, the buyer and the
seller have credit exposure to each other, that is – counterparty risk. A CDS
counterparty default has implications for:
• Regular payments of protection premium by the protection buyer. A mitigant
to the protection buyer counterparty risk is the requirement to make
protection
payments in advance, i.e. at the beginning of the protection period.
• Event-related, one-time protection payment due from the protection seller.
Such risk can be addressed through a requirement of a minimum rating trigger
of the protection seller. A trigger breach requires a substitution of the
protection seller with a higher rated one, a guarantee of the protection
seller’s
obligation by an appropriately rated third party, or posting of collateral by
the
protection seller (the level of collateral depends on the level of deviation
from
the minimum required rating).
• CDS MTM valuation in case of termination and funding of the termination
payment (repayment at par) through a liquidity line or through a senior
position in the deal’s cash flow allocation. MTM should not be eroding the
subordination levels in the deal.
n Liquidity
In typical cash CDO structure, given the pass-through nature of the cash flows
emanating from the asset pool and directed to servicing the CDO liabilities,
liquidity is usually needed to cover temporary cash flow shortfalls usually due to
timing mismatches. In a synthetic CDO structure and particularly managed
synthetic CDOs, particular needs for liquidity may or may not arise. So instead of
an assumption that there is a liquidity facility in the deal, the questions instead
should be:
1. Is there a need for a liquidity facility in a given synthetic CDO structure?
and
2. How is the liquidity facility sized and procured?
Particular need for liquidity in a synthetic CDO are created by:
• Shorting CDS,
• Trading losses that are crystallized by termination,
• Credit events requiring cash settlement or bond delivery (bond must be
purchased).
Furthermore, liquidity may play an active part in the execution of portfolio
management strategy and specifically asset allocation changes.
Certain trading strategies, for example uncovered shorts, the liquidity needs may
be substantial and can be satisfied only through a dedicated liquidity facility.
The
size and cost of such facility are of key importance, as high liquidity cost may
reduce the benefits of some savings achieved on the liability side of the synthetic
CDO through the low cost super-senior tranche hedging.
Another aspect of liquidity management is the shortfall of Classes B and C interest
payments. A potential solution is introducing pikeable interest for these classes
in
the form of the capitalization of the missed interest for two consecutive periods
for
the single-A rated tranche and indefinitely for the BBB-rated tranche.
n Loan-to-Value Ratios
Loan-to-value (LTV) averages and distribution within a pool are also important.
The LTV ratio is considered the determinant of the borrower’s willingness to pay,
reflecting the portion of the borrower’s equity in the property. LTVs are, for many
reasons, not directly comparable across geographic borders; therefore the LTV
profile must be examined against benchmark performing pools in each jurisdiction
to determine whether pool LTVs are more or less aggressive.
A key element of the LTV ratio is the determination of property value. In the UK
and Ireland, for example, property value is based on appraised values by
professional surveyors. In France, some properties are valued simply on the basis
of market values of properties in the surrounding area. In Germany, lenders are
required by law to value property based on ‘sustainable’ rent. In Sweden, policies
differ from lender to lender, but generally use tax valuations, which are performed
every six years but are indexed annually, based on recent sales. In the
Netherlands,
the concept of loan-to-value is instead sized to loan-to-foreclosure value,
resulting
in a value between 80%-90% of the market value (not dissimilar to the practice of
German lenders).
n Income Ratios
Price-to-income or net income-to-mortgage payment determines the borrower’s
ability to pay debt service out of current income. Price-to-income ratios are less
commonly used by non-US mortgage lenders, where debt service as a proportion
of monthly income is a more powerful determinant of the approved loan amount.
n Mortgage Products
The type of mortgage products – such as annuity mortgages and life insurance
mortgages – and the effects of the mortgage type on debt servicing and debt
recoveries also factor into an investor’s consideration. Redraw and equity-release
loans, for example, are available in a number of markets, including Australia, the
UK and Ireland, and increase the leverage of a given borrower. In the Netherlands,
savings and investment-linked loans are the norm, while in Italy mortgage loans
may be linked to family businesses.
n Tax Implications
Tax implications relating to mortgage debt must be considered. In several
jurisdictions, tax advantages drive the structure of the mortgage product offered,
as well as the manner in which debt is repaid. Tax regimes have a heavy influence
on payment behavior and loan structure. For example, in the Netherlands,
mortgages are used as a tax-planning instrument and are linked to savings
accounts, investment portfolios or life insurance contracts where the related
investment returns are generally not taxed. Hence interest-only loans are very
common, and loans outstanding do not decrease, as rapidly over time as in the US,
despite being similar in that it is largely a fixed rate market. Another example
occurs in Australia where mortgages receive a tax deduction on principal repaid
but not on interest, thus providing an incentive for early repayment.
n Pool Seasoning
The seasoning, or ageing, of a mortgage is associated with the building up of the
borrower’s equity in the acquired property, which has a strong influence on the
borrower’s motivation to continue servicing the mortgage or to abandon it. A
pool’s ageing is also associated with a loss curve; mortgage loss curves tend to be
front-loaded, in that losses occur early in the life of the mortgage pool and
decrease and stabilise later in its life.
n Obligor Profile
The characteristics of the obligors in the pool – whether they are sub-prime or
non-conforming, salaried or self-employed, private or social housing – are
important; some characteristics are more negative, others are more positive. For
sub-prime and non-conforming mortgage pools, although not an established sector
outside the UK and Australia, performance varies dramatically from that of prime
mortgage pools. Generally, these differences include a higher expectation of
defaults, and prepayments are typically faster and are driven by a change in
borrower credit quality, due to a change in circumstances rather than interest
rates,
which are the dominant factor for prepayment in prime mortgage pools. In contrast
to pools with riskier obligors, some pools carry a large proportion of high-quality
obligors, such as civil-servant mortgage loans. These borrowers are considered
higher quality because they are less likely to become unemployed.
(exceeding 60% LTV), but are secured by first-ranking mortgages. Is the mortgage
for a tenant-owner or a single family? In Sweden, tenant-owner loans are secured
not by a lien on the property, but by a perpetual right to occupy the property. For
these loans it is important to scrutinize the financial condition of the
cooperative
and the credit quality of the borrower. In Italy, the distinction between fondiario
and ipotecario is critical in that fondiario loans usually achieve beneficial
treatment during the enforcement process.
n Geographic Concentration
The geographic concentration of mortgage loans is related to the demographic and
economic characteristics of the region. For instance, Italian mortgages for
properties in the south and the islands have historically exhibited higher default
rates than those for northern and central Italy. Hong Kong Island housing has
traditionally been more resilient to price downturns than Kowloon and the New
Territories. In Portugal and Sweden, the population is heavily concentrated in one
or two large cities, so mortgage pools may be subject to housing market dynamics
that are less well diversified than in other markets.
n Recovery Value
Recovery value and the time it takes to recover a property should not be
overlooked. Property values could be reduced by market value declines,
foreclosure costs and carrying costs from delinquency to foreclosure. Recovery
times vary widely across markets; in the Netherlands, foreclosure to recovery
proceedings can take as little as three months, while in Italy the process can take
up to 10 years.
n ‘Set-Off’ Risk
‘Set-off’ is the risk associated with the borrower’s ability to offset mortgage
debt
against a deposit held with the bank originator of the mortgage. In Germany and
the Netherlands, set-off is allowed between savings or insurance contracts and
mortgage debt in the event that the originating bank defaults. Structural steps
taken to reduce or eliminate this risk should be examined. That said, set-off is
only
an issue for cash transactions. Synthetic transfer (which is typical in Germany)
obviates the need for any additional legal provisions.
n Substitution
Substitution is a structural provision, common in master trust and revolving
transactions, allowing for the addition of new mortgages to the original pool.
Eligibility criteria should be examined carefully to minimize the potential for
pool
deterioration over time.
n Insurance Policies
Insurance policies may be present in a number of forms. Private mortgage
guarantee insurance is used almost exclusively in Australia, on both a pool and
individual loan basis. Private mortgage guarantee insurance is also used, although
not as broadly, in France, the UK, Ireland and Hong Kong. State mortgage
insurance is available in Belgium and the Netherlands. In jurisdictions such as
Germany, Spain and Portugal, lenders may require life insurance policies in
conjunction with certain mortgages, for example, to assure balloon payments on
interest-only mortgages.
n Consumer Protection
Consumer protection laws vary by jurisdiction. In Sweden, over indebtedness
legislation excludes mortgage payments, in that it leaves the borrower responsible
for the debt even if otherwise over indebted; in France, similar legislation
includes
mortgage debt, as well as other consumer debt.
• Prepayments and their effect on floating rate MBS (mainly weighted average
life, or WAL) and fixed rate MBS (duration and convexity).
• Prepayment assumptions used in credit enhancement modeling and MBS
pricing.
• Clear understanding of the differences between expected maturity and legal
maturity for pass-through structures, for example, pass-through structures are
rated by legal, not expected, maturity.
• Investment characteristics of specific MBS tranches – for example, senior,
subordinated, interest only (IO) payments or principal only (PO) payments; in
the case of IO, this includes the effects of prepayments, resets on the fixed
rate loans, excess spread, and WANM (weighted average net mortgage rate).
• Syndicate composition and commitment to secondary market making
associated with the expected liquidity of the bonds.
• Finding an appropriate benchmark for pricing disparate asset classes, all
falling under the label of MBS, e.g. performing versus non-performing loans,
originator experience, geographic distribution and loan seasoning. As default
curves vary by jurisdiction (e.g. defaults peak, on average, two years from
origination for Italian mortgages, and four years from origination for UK
mortgages), similar seasoning for different jurisdictions may have different
credit implications.
n Guarantor-Dependent Ratings
In a number of jurisdictions, insurance is provided by either state or private
insurers to encourage mortgage lending. In Australia, for example, it is a
widespread practice for the mortgage loans in securitised pools to be insured on an
individual basis and on a pool basis. Private insurers also play a role in the
Irish,
French and UK mortgage markets. State guaranteed mortgages are often present in
mortgage pools in the Netherlands and Belgium. These additional parties within
the mortgage chain emphasizes the need for a clear understanding of the role of
mortgage insurers in the credit performance, credit rating, and rating volatility
for
such MBS.
Australian mortgage pools are generally insured by pool policies (up to a defined
limit) or individual loan policies. As such, any credit-related aspects of their
performance defer to the performance of the respective insurance companies under
the insurance contracts. Changes in the process of mortgage origination, mortgage
product features, and the consequences of increased competition have specific
credit effects on mortgage performance, the effects of which are dampened or
absorbed by the respective mortgage insurers. In this regard, it is important to
follow market changes that could affect the motivation of mortgage insurers to
honor claims or increase their ability to reject claims.
Investors’ exposure to the mortgage insurers is to the extent claims are submitted
and are expected to be honored by the insurance companies. As the credit ratings
of MBS depend on the insurance coverage of the underlying mortgage pools, MBS
ratings may be correlated with the ratings of the insurance providers for a
specific
MBS pool. The rating of the insurance providers reflects their ability to pay
claims
under the granted insurance policies. Hence changes in the ratings of the claim-
paying ability of insurers could lead to changes in the credit quality of the
insured
mortgage pools and the related ratings of the respective MBS.
So, comparing the LTV of the mortgage pools of different MBS, all other things
being equal, should give investors a good indication of their respective default
probabilities. If only the LTVs were calculated the same way in all MBS!
Unfortunately, they are not. And in order to use them in a meaningful way in
comparing different MBS, investors should make proper adjustments.
Analyst Certification
We, Alexander
Batchvarov, Jenna Collins and William Davies, hereby certify that
the views each of us
has expressed in this research report accurately reflect each of
our respective
personal views about the subject securities and issuers. We also
certify that no part
of our respective compensation was, is, or will be, directly or
indirectly, related to
the specific recommendations or view expressed in this
research report.
Copyright 2003 Merrill Lynch, Pierce, Fenner & Smith Incorporated (MLPF&S). All
rights reserved. Any unauthorized use or disclosure is prohibited. This report has
been
prepared and issued by MLPF&S and/or one of its affiliates and has been approved
for publication in the United Kingdom by Merrill Lynch, Pierce, Fenner & Smith
Limited,
which is regulated by the FSA; has been considered and distributed in Australia by
Merrill Lynch Equities (Australia) Limited (ACN 006 276 795), a licensed securities
dealer
under the Australian Corporations Law; is distributed in Hong Kong by Merrill Lynch
(Asia Pacific) Ltd, which is regulated by the Hong Kong SFC; and is distributed in
Singapore by Merrill Lynch International Bank Ltd (Merchant Bank) and Merrill Lynch
(Singapore) Pte Ltd, which are regulated by the Monetary Authority of Singapore.
The
information herein was obtained from various sources; we do not guarantee its
accuracy or completeness. Additional information available.
Neither the information nor any opinion expressed constitutes an offer, or an
invitation to make an offer, to buy or sell any securities or any options, futures
or other
derivatives related to such securities ("related investments"). MLPF&S, its
affiliates, directors, officers, employees and employee benefit programs may have a
long or short
position in any securities of this issuer(s) or in related investments. MLPF&S or
its affiliates may from time to time perform investment banking or other services
for, or solicit
investment banking or other business from, any entity mentioned in this report.
This research report is prepared for general circulation and is circulated for
general information only. It does not have regard to the specific investment
objectives,
financial situation and the particular needs of any specific person who may receive
this report. Investors should seek financial advice regarding the appropriateness
of
investing in any securities or investment strategies discussed or recommended in
this report and should understand that statements regarding future prospects may
not be
realized. Investors should note that income from such securities, if any, may
fluctuate and that each security’s price or value may rise or fall. Accordingly,
investors may
receive back less than originally invested. Past performance is not necessarily a
guide to future performance.
The bonds of the company are traded over-the-counter. Retail sales and/or
distribution of this report may be made only in states where these securities are
exempt from
registration or have been qualified for sale. MLPF&S usually makes a market in the
bonds of this company.
Foreign currency rates of exchange may adversely affect the value, price or
income of any security or related investment mentioned in this report. In addition,
investors in
securities such as ADRs, whose values are influenced by the currency of the
underlying security, effectively assume currency risk.
Refer to important
disclosures at the end of this report.
122