This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1. JEAN-MARTIN AUSSANT DIRECTOR FIXED INCOME PRODUCT STRATEGY
PRMIA LONDON 18 MAY 2004
2.
Recent market environment
New market-implied techniques to manage credit risk
Introduction to the BDP (Barra Default Probability)
Practical Examples
Questions and answers (assuming I have the answers)
Discussion Outline
3.
Recent market environment
New market-implied techniques to manage credit risk
Introduction to the BDP (Barra Default Probability)
Practical Examples
Questions and answers
Discussion Outline
4.
Equity declines drove re-allocations to fixed income
Simultaneously government yields decreased to all time
lows
Credit default rates neared all time highs
Pension fund shortfalls (Focus on ALM)
Credit markets are increasingly complex
Universe of assets is expanding rapidly
Spread products are becoming more complicated
Limited headcount to cover expanding number of issues
Market Forces Change the Rules of Credit Investing
5. Currently Very Few Easy Opportunities
End of the bear credit market in 2003
Spreads have tightened to extreme levels
Lowest since 1998
Demand still high
Non-traditional investors
Source: S&P
6. Outperformance Is More Demanding Than Ever
Are we being correctly compensated?
Risk premium close to zero
How does a long-only investor win/outperform?
Spreads have nowhere to go
Move to
Lower-quality / higher-yielding
Find names with value still
Asset selection is key
7. One Default Can Negate Entire Portfolios Return
Credit market is strongly asymmetric
Earning the spread has become extremely difficult over the past
few years
Unprecedented market conditions with record downgrades and
defaults
Source: Lehman Brothers, 2002
8. Fundamental Analysis Alone Is Not Enough
Judgment of experienced analysts remains essential
However, judgment often impaired by questionable data
Nearly 1000 accounting re-statements in the last three years
(source: SEC)
Creative accounting
Whats required is a more efficient process to monitor, screen,
and select credit-risky investments
9. Gaining the Advantage in Credit Investing
To successfully manage credit, you need
Earlier, more accurate prediction of potential default
risk
Models that allow for the real- world uncertainty of financial
statements
Tools to make your credit analysis process more efficient
10.
Recent market environment
New market-implied techniques to manage credit risk
Introduction to the BDP (Barra Default Probability)
Practical Examples
Questions and answers
Discussion Outline
11. Market-Implied Measures Provide Additional Insight
12. Equity Market BDPs Go Beyond Traditional Models
BDP advantages
Incomplete Information framework assumes fundamental data may
be flawed
Use of Barras industry standard equity volatility forecast
Empirical study of historical leverage
Barras model signals significant uptrend in default risk months
earlier
13. Bond Market BIRs Lead Agency Ratings
Barra Implied Ratings take the bond markets perspective on
credit and match it to a best fit distribution of actual
ratings
Barra Implied Ratings typically can lead agency ratings by as
much as three months
Barras measures provide earlier warning to possible downgrade
14. Derivatives Market CDS Market a Leading Indicator
Credit Default Swap (CDS) rates often provide leading
indication of risk and value
CDS market is exploding: more than $4 trillion notional
outstanding and most big names actively traded (source: BBA)
Cash-CDS Basis History Merrill Lynch (5 Year USD)
15.
Recent market environment
New market-implied techniques to manage credit risk
Introduction to the BDP (Barra Default Probability)
Practical Examples
Questions and answers
Discussion Outline
16. Current Quantitative Default Models
Structural or Cause-and-Effect approach (Merton)
Default happens for a reason
Firm-specific information can be used advantageously
Reduced form approach
Default rates can be analysed statistically
Ad hoc, exogenously-given, default rate
17. Mertons Structural Model of Default
Default occurs at debt maturity if the firm value is below the
liabilities value
We thus need
A model of firm value process
Estimate of default point
Merton identified equity as being long a call option on the
firm value
Merton identified a bond as being short a put option on the
firm value
18. Mertons Structural Model of Default Payoff at maturity of
the bond Value of the Equity at maturity time T Value of the Bond
at maturity time T i.e. default free bond + short European put on V
@ K i.e. European call on V @ K
19. Mertons Structural Model of Default 0 D V 0 No Default
Probability of Default Default T
20. Reduced Form Models
Assume that default is totally unpredictable
Default comes unannounced
Based on a conditional default rate or intensity
Exogenously given
Fit well to market data including short credit spreads
Ad hoc, lack intuitive appeal
The picture:
?
21. Model Comparison
Based on a model definition of default
Intuitive, appealing
The default time is often (implicitly) predictable
Hard to fit to empirical data
Based on an exogenously given default rate
Ad hoc
The default time is always totally unpredictable
Easy to fit to empirical data
Structural / Cause and Effect
What we want: a hybrid model
Incorporate the best features of structural and reduced
form
Avoid their pitfalls
Reduced Form
22. The Barra Default Probability (BDP) Model
A genuine hybrid of cause-and-effect (structural) and
reduced-form models (compensator approach)
Based on a default time that is not predictable
Makes use of all publicly available liability statements and
equity market data
Assumes investors have incomplete information
Calibrates easily to short credit spreads
Intuitive and appealing
23. Barra Default Probability Model Intuition 0 V 0 T
Distribution of possible default boundary levels Paths of Asset
Value Process Expected level of default barrier Width represents
uncertainty in the default barrier level Time Asset Value
24. Default Barrier Scaled Beta Distribution Mean = current
debt Standard deviation, calibrated or user-configured
25. BDP Model Uncertainty Can Be Varied Barra Default
Probability Model Variant 1 Variant 2 Variant 3
26. BDP Model A Firm Becomes Distressed 9/17/2001 9/10/2001
Credit term structure steepens and short-term spreads increase
27. BDP Model Subtlety Healthy Firm 4/15/2002 4/10/2002 15%
drop in equity Credit term structure steepens but short-term
spreads barely move
28. Testing the Model ROC Curves
Radar Operators in WWII: Plane (or flock of birds)?
Medical Diagnosis: Is this persons Thyroid OK?
Astronomy: Is this a Planet?
Marketing Analysis: Will this household buy insurance?
Credit Risk: Will this name default ?
Non-Event Event MODEL FORECAST Event (+) Non-Event (-) True
Negative False Negative False Positive True Positive REAL
WORLD
29. ROC Curves Merton Comparison Merton BDP Random Method
30. ROC Curves First Passage Comparison First Pass BDP
New market-implied techniques to manage credit risk
Introduction to the BDP (Barra Default Probability)
Practical Examples
Questions and answers
Discussion Outline
33. BIR Good Complement to Agency Ratings ZOOM ZOOM
34. BDP Outlier Identification
Inspecting like-credits in a new way can sometimes turn up
opportunities or threats
USD BBB Consumer Cyclicals Average BDP = 0.20% Toys R Us BDP =
4.97% Source: Barra Credit
35. BDP Warning of Toys R Us Downgrade?
Bond implied ratings moved in October as well
A few days later spreads widened, and months later Toys R Us was
downgraded to below investment grade (junk) Source: Barra
Credit
36. BIR Mandate Restrictions
Early warnings of Potential Downgrades can allow managers to
exit worrying names before the flood
The Bond Market was pricing in concerns back in September when the
Barra Implied Rating for Parmalat dropped to Sub-IG Source: Barra
Credit
37. BDP Early Warning
The equity market was also signalling concerns for
Parmalat
The BDP moved well into HY levels before December Source: Barra
Credit Investment Grade (approximately) . .
38. Capital Structure Arbitrage
Differing views from two markets on the capital structure point
out interesting opportunities
FedExs announcement of its planned acquisition of Kinkos triggered
concern implied by the equity market but not reflected in bond
spreads Source: Barra Credit
39. Market-Implied Measures Offer More Insight Source: Barra
Credit