MOODY'S
INVESTORS SERVICE
Turning the ratings
game upside down
TRADE
PROCESSING
Which STP vendors
are right for YOU?
Spanish
securitisation
THE BIG INTERVIEW
NEW REGS CHANGE
Pimco s Peter Bentley
on credit valuations
LANDSCAPE
PLUS:
SPECIAL FOCUS: Korea Ź Bond Liquidity Survey
P R O F I L E P R O F I L E
As the debate rages over the usefulness of credit ratings, Moody s unveils
a set of credit risk indicators derived from market movements. Will  market
implied ratings silence the agencies critics? Nikki Marmery investigates
Upgrading
the
rating process
Moody s MIR team: (from left) Simon Jiang, Chris Lam, Dan Russell, Robert Eckerstrom, David Munves, Njundu Sanneh and Tipanee Pipatanagul
nron s spectacular descent into the biggest corporate accounting practices. When Enron announced it had overstated given issuer s bond spreads to give it a bond implied rating. For types of credit risk. They identify default candidates: the agen-
bankruptcy ever seen in the US surprised everyone. Not profits by nearly $600m over five years on November 8, the stock example, the median credit spread for five-year B2 rated bonds cy s research shows that default rates are significantly higher for
Eleast the rating agencies that were still tipping the energy was at $8.41 and CDS spreads were at 133bp. By the time Moody s might be 377 basis points over Treasuries. If Acme Inc, a B2 issuers whose securities are trading with negative gaps compared
firm as an investment-grade credit four days before it filed for and S&P finally downgraded Enron to junk status on November rated credit, is trading at 168 basis points  the median spread with their Moody s rating. It might seem like something of a no-
bankruptcy protection on December 2, 2001. 28, its stock was worth little more than a dollar per share. of five-year Ba1 rated bonds  its bond implied rating would be brainer that securities the market takes a dim view of are more
Their collective failure to rate Enron s creditworthiness The moral of the story,  don t ignore the market , was a hard Ba1. As this is four notches above its actual rating, it is said to likely to default; but what is surprising is the degree to which it
adequately in the weeks leading up to its collapse earned the lesson for the rating agencies to learn. Five years on, one agency, have a ratings gap of +4. is true.
rating agencies the lasting enmity of US lawmakers, who have Moody s, has something to show for it. Ford Motor Company is a good example of an issuer s bonds Using a data set of 2,900 issuers, with 180,000 observations
been dying to get their hands on the right to regulate them Moody s, the oldest rating agency, and alongside Standard & trading below its Moody s rating: as of September 1, Ford s rat- gathered between January 1, 1999 and February 28, 2006, the
ever since.  The credit raters  despite their unique position to Poor s one of the two largest agencies by market share, has devel- ing was B2, while its bond implied rating was Caa1, giving it a one-year default rate for B2 rated issuers trading two notches
obtain information unavailable to other analysts  were no more oped a set of ratings indicators derived from market signals. These ratings gap of -2. In other words, Ford was then trading two below their Moody s rating was a massive 17.82%. That com-
astute and no quicker to act than others, said US senator Joe may be used as a counterpart to Moody s  normal ratings, which notches below  or cheap to  its rating. pares with a default rate of 3.61% for issuers trading flat to their
Lieberman, who chaired the governmental committee looking are based on analysts views of an issuer s creditworthiness. Moody s rating; or 0.59% for those trading two notches rich. In
into Enron s collapse. The indicators, dubbed  market implied ratings (MIR), high- Synthetic feature other words, if you held a portfolio of bonds that were trading
Yet the signs of the energy firm s imminent demise were there light discrepancies between an issuer s credit rating  in essence, To add extra dimensions to the implied ratings, Moody s applies two notches cheaper than the Moody s rating, you should expect
to be seen clearly in the markets several weeks before Enron s the rating agency s assessment of a company s financial situation the same analysis to an issuer s credit default swaps, using median nearly a fifth of them to default within a year.
Chapter 11 filing, which left creditors holding some $16 billion and future outlook  and the market s view of that issuer  which five-year CDS spreads, and also examines equity implied ratings. MIR can also be used to predict potential ratings changes. An
of defaulted debt. is in effect the sum total of the expression of all bond, credit Determining these is more complex, and is based on Moody s issuer trading three notches below its Moody s rating is looking
On October 22 that year, Enron s stock price dropped 20% derivatives and equity investors views on that company. KMV EDF model, extracting credit risk information from an at about a 25% chance of downgrade over a one-year horizon,
to $20.65 per share, and five-year credit default swap (CDS) The concept is quite simple. Using data derived from all the issuer s equity price by assessing expected default frequency. according to MIR data from the same data set. And when you
spreads jumped 20% to 48 basis points, after the Securities & issuers it rates, Moody s has worked out an average bond spread By highlighting differences between an issuer s implied rating drill down further into specific ratings categories, the probability
Exchange Commission announced it was looking into the firm s for certain ratings categories. The MIR team then analyses a and its actual rating, the MIR team is able to flag up different of downgrade can increase further. The most extreme example
1 credit OCTOBER 2006 www.creditmag.com credit OCTOBER 2006 2
PHOTOGRAPHY: AMY FLETCHER
P R O F I L E P R O F I L E
Moody s rolled out a form of the product to its analysts in Still, all this leads to a clear conundrum. If credit risk infor-
2001. From the start, the agency used implied ratings as a guide mation derived from market movements is more accurate
The canary in the coalmine: How MIR acts as an early warning system
to the companies they rated, rather than factoring them in to than Moody s analysts assessments, then what s the point of
 We use it to look for credits that may be But credit risk officers are looking for their ratings judgments.  As one of the analysts said to me, analysts assessments of credit risk? Shouldn t the price of a
deteriorating so we can make our credit potential upgrades as well as downgrades, reports Munves,  it makes sure they ask the right questions in security be a function of its creditworthiness  not the other
risk officers aware of them, says one user he adds. If market implied ratings signal a the right way at the right time. way round? Isn t MIR an admission that the naysayers are right
of Moody s market implied ratings who credit is heading for an upgrade, the risk Three years later, Moody s rolled out MIR to customers. At  and Moody s ratings, along with those of the other agencies,
works in the internal credit risk depart- officer could free up the economic capital that point it became clear that MIR needed further research on are untimely and inaccurate?
ment of a major investment bank in New set aside against it for another area. its uses to help customers interpret the data, and the agency  They are different signals, affirms Russell.  They both serve
York.  We ve built our own system linking These types of clients  as opposed to turned to David Munves, a credit strategist with more than two a role and the market has spoken that it wants to continue to use
in all our exposures that alerts us of big investment managers looking for trad- decades of experience in the fixed-income groups of Lehman our ratings in a broad manner.
ratings gaps developing. ing ideas  might also be using a similar Brothers and Standard & Poor s, to lead the team. Serving as a benchmark for the implied ratings,  traditional
Watching ratings gaps evolve helps the tool developed by Riskmetrics Group,  I joined in December 2004, some months after it was rolled ratings are useful, because it s the discrepancy between the two
bank act faster in dumping bad credits, a risk management software firm spun out to customers, says Munves.  They wanted to know,  what s that delivers the signals.  We see them
explains this user.  If you look at some of out of JPMorgan in 1998. CreditGrades is the significance of this? Is it something I should act upon? Is that as complementary, says Russell.  We
the auto names in particular  say [auto parts provider] Dana Corp, an equity-based model for assessing the credit quality of publicly a high level of default risk? Without the research, no one really think we can play a unique role in
with a ratings gap of six  you can see the ultimate crash landing of traded companies. The model is used by Deutsche Bank, Goldman had the answers. doing rigorous research into mar-
that credit developing. Sachs and JPMorgan. It differs from Moody s market implied ratings ket signals relative to ratings so we
Dana Corp declared bankruptcy on March 3 this year. in that it assesses default probabilities unrelated to credit ratings. Team-building can move from instinct-based dis-
 It s like the canary in the coalmine: a sign that it s time to get out It also focuses purely on credit risk information derived from the Munves was joined by quant specialist Simon Jiang from another cussions and conclusions to more
of the credit, he says. equity markets. part of Moody s research team, and in November 2005 by assist- empirically based discussions of
ant vice-president Keith Gudhus and research associate Chris how credit ratings behave.
Lam. Gudhus s background was in trading, having worked in But is it necessary for market
is B rated issuers, for whom there is a higher than 50% chance of to the market? MIR helps investors avoid opportunities to buy BNP Paribas loan syndication and trading department, and implied ratings to function rela-
downgrade over the next year for all issuers trading three notches expensive paper, says Munves. before that at the Gelber Group, where he traded corporate tive to Moody s ratings? Could not
or more below their Moody s rating. It also helps traders pick out the  biggest trading signals from debt. Lam researches the MIR database and contributes to the default probability be derived from
A third application of the tool is relative value analysis. By a mass of information. Autos traders for example might know monthly comments; he was previously at private equity firm CAI absolute price move-
comparing relative implied ratings moves, the data can signal the ins and outs of Ford and GM spread movements; but with Managers in New York where he specialised in buyouts, restruc- ments, irrespective
which bonds are likely to rise or fall against the broad market 75 rated names in the sector, they are likely to have less intuitive turings and acquisitions. of rating gaps?
in the coming year  effectively delivering buy and sell signals expertise on the less liquid names. This makes the tool particu- Vice-president Robert Eckerstrom, who with Gudhus writes One credit
for investors benchmarked against indices. The data shows, for larly valuable for supervisors and risk managers. market-orientated research and researches long-term projects, analyst at an
example, that bonds trading with a ratings gap of -3 are more  How well does a boss know what s going on? asks Munves. joined in July. He previously worked for the Government of Sin- investment
than 50% likely to see their bond implied rating rise over the  They need all the help they can get to stay on top of 300 400 gapore Investment Corporation, where he was an interest rate bank in
next 12 months  indicating an outperformance of the broad names. With MIR, it s very easy to get those names uploaded, portfolio manager and co-managed a credit portfolio.
market. Conversely 65% of bonds trading three gaps rich to their look at ratings gaps and get email alerts when ratings gaps appear. Moody s also hired two emerging markets specialists earlier this
Moody s rating should expect to see their bond implied rating It helps managers keep an eye on large number situations, which year: economists Tipanee Pipatanagul, who monitors Asian and
decrease over the coming year, indicating underperformance is how Moody s analysts use it. Information overload is killing emerging European economies, and Njundu Sanneh, who looks
versus the market. people. MIR picks out what s important. at MIR-related credit market trends in regional markets. Pipatan-
That s a reflection of the typical cycle of the market, says David agul previously worked as an economist at the US Treasury, and
Munves, the managing director of credit strategy research at Gestation period Sanneh transferred from Moody s Credit Trends service, where he
Moody s in New York, who leads the MIR team.  Fund manag- The roots of Moody s market implied ratings were growing long provided emerging markets commentary.
ers will rotate out of expensive names and into cheap names; it s before accounting scandals such as Enron s so clearly demon- The newly formed team each saw the potential for the product
the traditional pattern of issues being oversold, stabilising and strated their value.  Since time immemorial, Moody s has gotten from the perspective of their diverse backgrounds.  Putting my
coming back. calls from customers saying,  you ve got this thing rated like a B3, trading hat on, says Gudhus,  I thought,  this would really help
Bonds with the positive ratings gaps are the ones investors but it s trading like a B1. What s going on?  says Dan Russell, me with my trading ideas. 
want to avoid in order to outperform the index.  Holding a the managing director responsible for new business initiatives at
portfolio of bonds here, you ll have a lot that will lose value Moody s in New York.
 A lot of the time people have intuition, but
against the benchmark, he says. Towards the beginning of this decade, various factors conspired
It s in this relative value analysis that MIR is at its most useful. to make the idea of extracting credit risk from price information
empirical data doesn t back it up. MIR
It means a trader is able to back up his intuition with cold, hard more viable: the wider use of the Merton approach to analys-
helps investors avoid opportunities to buy
facts as to whether a credit really is as cheap as it  feels .  A lot of ing default risk from equity market information; the emergence
the time people have intuition, but empirical data doesn t back it of CDS data, enabling the use of risk signals from this market;
expensive paper
up. How many notches exactly is it cheap? It looks like it s trad- and the availability of traded levels for corporate bonds from the
David Munves, Moody s
ing cheaply, but has it become more expensive over time relative Trace reporting engine.
3 credit OCTOBER 2006 www.creditmag.com
CORBIS
P R O F I L E P R O F I L E
have happened had an investor bought and sold bonds based
on patterns of behaviour revealed by MIR data in the context Case study: MIR raises alarm on Philippine debt
of the market environment of the time. For example, the paper
demonstrates that in 2000  a  horrific year for the corporate Emerging markets is one of the new project areas Moody s MIR
bond markets, says Munves  you were better off buying expen- is looking into, after the addition of economists Tipanee Pipatan-
sive bonds (those trading rich to their Moody s rating): the  +2 agul and Njundu Sanneh earlier this year. One early example of
portfolio returned 6.64%. In every year after that, buying rich the team s work in this area is a recent study on the Philippines.
bonds proved to be a losing strategy. And in the best year for Financial markets in the country enjoyed a strong rally over the
corporate bonds  2003   it was a real losing proposition : the summer on the back of the government s improving fiscal situ-
 +2 portfolio returned -19.68%. ation, expectations of an end to US and local rate hikes and the
return of foreign inflows to the country.
 Super implied ratings Against such a backdrop, the Philippines Composite Equity
The team is also looking into what gives more accurate indica- index rose 13.4% between June 14 and August 16, after a sell-off
tors: implied ratings derived from bond spreads, CDS spreads in spring brought it crashing down 20.2% from its record high of
or equities. Early indications are that the CDS market is the 2,589 on May 8. The Philippine peso rallied 4% against the dollar
most efficient, probably because it s a liquid two-way market, over the same period.
says Munves. Once the analysis is complete, the team hopes to Credit investors looking for opportunities in emerging mar-
determine a  super implied rating derived from an optimum kets might be buoyed by such market optimism and see solid
weighting of all three indicators. investment potential in the country. But a warning signal from
Other long-term research studies include a project looking at Moody s MIR prompted the agency to advise investors to  curb
past leveraged buyouts and working out whether bond and equity their enthusiasm on the sovereign.
implied ratings can be used to identify risky LBO situations. It noted on August 18 that bond implied ratings for Philippine
As such, MIR is looking at significant expansion. It s tripled issuers rose by one notch between the end of June and August
the number of firms using MIR to  north of 500 over the past 15 to a rating of Ba2, taking them to two notches above the sov-
year, says Russell.  We ll probably double staff size before the ereign s actual Moody s rating of B1. As issuers with bond implied
London thinks so:  To say the market is implying a particular Delphi is an example of why an investor would want to keep year is over, and expand in London. We expect rapid growth. ratings gaps of two notches tend to underperform over a 12-
rating is egocentric: it breaks everything down as though the rat- both sets of ratings in mind. Earlier in this decade, bond implied Which leads us to one extreme scenario: what happens if month view, Moody s cautioned investors on Philippine debt.
ing was the common language of the credit. What the market is ratings of the then Baa2 rated US auto parts manufacturer were everyone started to buy into Moody s market implied ratings? Qualitative factors back up that cautious view, says Pipatanagul.
doing is implying a default probability. regularly trading two notches higher  at the level of an A3 Would the market end up trading off highly leveraged market In particular the government s exceptionally high public-sector
But there s another, more compelling argument in favour of credit. By February 2002, the bond implied rating sank a notch signals? Bonds trading cheap to their Moody s rating being sold debt makes the country highly vulnerable to shocks.  Although
Moody s market implied ratings: stability. Precisely because they below its Moody s rating to Baa3, before swinging back to the off, sending the spreads wider, and making them even cheaper Moody s recognises that the country s strengthened external
don t change with every news-related market movement, they Moody s rating and below again later that year. The Moody s to their Moody s rating  thus heightening the signal to sell? payments position provides a buffer to transitory shocks or
serve as a more stable indication of longer-term risk, making rating had remained unchanged throughout.  Markets often Bonds trading rich being bought, thus tightening their spreads policy mis-steps, this is not enough to significantly reduce the
them just as valuable to investors, argue Munves and Russell. overshoot and come back to the Moody s rating, says Russell. and exacerbating their positive ratings gap to Moody s rating? country s debt ratios, she says.  Even assuming a best-case sce-
Market-based metrics give you  more refined signals , says  The question is: how do you use the two together to maximise Russell, however, isn t concerned about this nightmare vision nario for fiscal reform this year, the ratio of national government
Munves,  but there s an offset, and the offset here is volatility. the value of both? of the future.  The good news about markets is it s hard to find and non-financial public sector debt to revenue will likely stand
Markets move faster, they re more volatile, and they re wrong That, of course, was before Delphi s default in October 2005. 10 people who agree on anything, he says.  Even if everyone at around 400% at the end of 2006, a level that is well above that
on occasion. As it stumbled headlong down the path to default, the Moody s did [use market implied ratings], they would interpret the out- of similarly rated countries.
Around 90% of Moody s market implied ratings, for example, rating lagged the market indicators  but only just. put differently. Political risk ahead of congressional elections in May 2007 is
change in the course of one year, and some 76% reverse that Now the MIR team is focusing on new projects. A portfolio A credit risk manager, for example, focuses on the default rate also an issue, she notes.
move during the next 12 months. By comparison around 20% of paper is in the pipeline, the result of a back-testing exercise to of entities, which for a portfolio of 100 B2 bonds with -2 rat-
Moody s ratings change throughout a year, with a reversal rate examine what happened to portfolios of bonds grouped by their ings gaps would be 17.42%. As a result he would likely buy
of just 1%. ratings gap over the past five years. The results reveal what would protection in the CDS market against these names. However
an active portfolio manager would take a different view:  He
would assume he is a superior name-picker, explains Munves,
Beware of the dog: next year s likely defaults and downgrades
 so would be able to hold a portfolio of cheap bonds and have
Issuer Sector Market implied rating Probability of default Probability of downgrade a default rate much lower than the market average. He would
1 HCA Healthcare Caa1(Ba2) 21.5% (1.8%) 48.4% (16.1%) likely sell CDS protection. Thus the names for which protec-
2 Ford Motor Company Autos Caa1(B2) 17.4% (5.5%) 35.4% (15.7%) tion would be bought or sold would vary according to the views
3 Abitibi-Consolidated Pulp & paper Caa1 (B1) 10.6% (2.7%) 35.9% (18.8%) of the participants.
4 Beazer Homes USA Construction B3 (Ba1) 7.1% (0.8%) 47.6% (17.5%) This underlines the point that Munves and Russell keep com-
5 Fairfax Financial Holdings Financial services Caa2 (Ba3) 2.9% (1.3%) 42.6% (18.7%) ing back to:  What we are saying is that MIR data is an initial
screening tool, says Munves.  People then have to make their
N.B. Figures in RED are  market implied ; figures in BLACK are according to Moody s actual rating
investment or risk decisions, as always.
5 credit OCTOBER 2006 www.creditmag.com credit OCTOBER 2006 6
CORBIS







T

i i i i i i i

i i i
i i i i i
żÿ
i i i i i i i
i 
żÿ
i i i i i i
i i
żÿ
i i i i i
żÿ
i i i i i i i
żÿ
i i i i
żÿ żÿ
i i
żÿ żÿ
i i i i