CDS Spreads as Default Risk Indicators - Feb 2011
Transcript of CDS Spreads as Default Risk Indicators - Feb 2011
CDS Spreads as Default Risk Indicators
Robert J. Grossman, Group Managing
Director
Martin Hansen, Senior Director
February 2011
2/15/2011
2
www.fitchratings.com
Why Study CDS Spreads?
0
500
1,000
1,500
2,000
2,500
3,000
6/07 9/07 12/07 3/08 6/08 9/08 12/08 3/09 6/09 9/09 12/09 3/10 6/10
Monoline REIT Home Builder
Source: Fitch Ratings, Fitch Solutions.
Property-Sensitive Sectors — CDS Spreads
(bps)
• Pronounced volatility in CDS spreads during the crisis
• Spreads are one of the analytical tools used by Fitch’s credit analysts (e.g., identifying outliers)
2/15/2011
3
www.fitchratings.com
Agenda
• CDS Spreads as Default Risk Indicators
• Overview: CDS vs. Bonds
• Methodology
• US Experience During the Crisis
• Europe: Banks and Insurers
• Risk Management Implications
• Margining: A Closer Look
• Margining Practices
• CDS Leverage and Return Dynamics
• Systemic Implications
• Appendix: Select Research on Other Topics
• US Money Fund Exposure to European Banks
• ‘Burden-sharing’
• Samples Used in CDS Analysis
CDS Spreads as Default Risk Indicators
Overview: CDS vs. Bonds
Methodology
US Experience During the Crisis
Europe: Banks and Insurers
Risk Management Implications
2/15/2011
5
www.fitchratings.com
CDS Embed both Long and Short Positions
Credit Default Swaps — Overview
Protection Seller(“Long” Credit Risk)
Protection Buyer(“Short” Credit Risk)
CDS Premiums (Cost of Insuring
the Reference Entity)
Insurance Payout (Upon a Credit
Event of the Reference Entity)
● CDS are economically akin to insurance and provide an investment alternative to bonds.● The protection seller insures an underlying risk exposure (the reference entity) against a credit event,
such as bankruptcy or failure to pay. ● CDS credit events and bond defaults are generally aligned, but can differ (i.e. the conservatorship of
Fannie Mae and Freddie Mac triggered their CDS but did not result in bond impairments).● The protection buyer, in turn, must pay contractually-specified premiums.● However, unlike insurance, CDS contracts are actively traded in secondary markets.● CDS returns thus typically depend more on spread changes (and premium income) than a triggering of a
credit event. Spread tightening benefits the protection seller; spread widening benefits the protection buyer.
2/15/2011
6
www.fitchratings.com
CDS Spreads as Risk Indicators
• CDS spreads increasingly used for risk analysis
• ‘Market-implied’ view
• Updated frequently (e.g., daily)
• Applications: valuation, portfolio strategy, funding conditions
• Also provides an ‘alternative’ perspective of an entity’s creditworthiness
• E.g., Fitch uses in helping to identify credit ‘outliers‘
• Converting spreads to probabilities of default (PD)
• Assume 60% loss severity (i.e., 40% recovery rate)
• Annual CDS spread = 120 bps
• Thus, one-year probability of default = 2.0%
Probability of Default (1 yr) = CDS spread (annualized) / Loss Severity
2/15/2011
7
www.fitchratings.com
Background & Assumptions
• Data sourced from Fitch Solutions
• Monthly averages of daily (annualized) CDS spread observations
• Observations are excluded if based on low liquidity or if occurring after a credit event
• Fixed (rather than stochastic) recovery rate
• Higher loss severity assumption → lower implied PD
• Banks, insurance companies, homebuilders, REIT = 60% loss severity
• Monolines insurer = 80% loss severity
• Risk-neutrality
• Spreads do not embed risk premium beyond compensation for expected or average credit losses
• Of course, not necessarily the case in the ‘real world’
• Risk aversion
• Large downside of default events
• Concentration risk
2/15/2011
8
www.fitchratings.com
Overview of Findings
• CDS-implied PD for US sectors that experienced pronounced volatility during the crisis
• Analyzed PD trends during different phases of the crisis
• ‘Backtest’ of PD vs. subsequent realized defaults / credit events
Volatility in the Relative Risks of Sectors over the Crisis
June 2007 October 2008 March 2009 August 2010
Sector
CDS Spread (bps)
Implied PD (%) Sector
CDS spread (bps)
Implied PD (%) Sector
CDS spread (bps)
Implied PD (%) Sector
CDS spread (bps)
Implied PD (%)
HB 101 1.7 Monoline 1,656 20.7 Monoline 2,902 36.3 Monoline 1,527 19.1
REIT 40 0.7 REIT 607 10.1 REIT 1,081 18.0 REIT 275 4.6
Monoline 37 0.5 HB 447 7.5 Insurance 460 7.7 HB 249 4.2
Insurance 30 0.5 Bank 310 5.2 HB 313 5.2 Bank 161 2.7
Bank 26 0.4 Insurance 267 4.5 Bank 243 4.1 Insurance 151 2.5
HB − Homebuilder. Notes: The October 2008 and March 2009 periods illustrated above are intended to represent different points in time during the financial crisis and do not necessarily reflect peak CDS spreads for each sector. Figures above are annualized. Source: Fitch Ratings, Fitch Solutions.
2/15/2011
9
www.fitchratings.com
Monolines: A Closer Look
0
500
1,000
1,500
2,000
2,500
3,000
3,500
6/07 9/07 12/07 3/08 6/08 9/08 12/08 3/09 6/09 9/09 12/09 3/10 6/10
Source: Fitch Ratings, Fitch Solutions.
Monolines — CDS Spreads and Implied PDs
(bps)
First Peak
Spread = 490 bps
Implied PD = 6.1%
Second Peak
Spread = 1,510 bps
Implied PD = 18.9%
Third Peak
Spread = 2,902 bps
Implied PD = 36.3%
• Spreads appeared to lead observable deterioration in credit fundamentals
• More than 60% of sample suffered a credit event / distress (i.e., CCC rating or lower)
2/15/2011
10
www.fitchratings.com
REIT and Homebuilders: A Closer Look
• Rise in implied PD, but no credit events in year following the peak
• PD for REITs increased by a multiple of 30x from trough to peak
0
200
400
600
800
1,000
1,200
6/07 9/07 12/07 3/08 6/08 9/08 12/08 3/09 6/09 9/09 12/09 3/10 6/10
REIT Homebuilder
Source: Fitch Ratings, Fitch Solutions.
REIT and Homebuilders — CDS Spreads and Implied PDs
(bps)
First Peak (Homebuilder)
Spread = 385 bps
Implied PD = 6.4%
First Peak (REIT)
Spread = 409 bps
Implied PD = 6.8%
Second Peak (REIT)
Spread = 1,154 bps
Implied PD = 19.2%
Second Peak (Homebuilder)
Spread = 481 bps
Implied PD = 8.0%
2/15/2011
11
www.fitchratings.com
U.S. Banks and Insurance Companies
0
100
200
300
400
500
600
6/07 9/07 12/07 3/08 6/08 9/08 12/08 3/09 6/09 9/09 12/09 3/10 6/10
Banks Insurance
Source: Fitch Ratings, Fitch Solutions.
Financial Services — CDS Spreads and Implied PDs
(bps)
First Peak (Banks)Spread = 427 bpsImplied PD = 7.1%
Second Peak (Banks)Spread = 247 bpsImplied PD = 4.1%
Peak (Insurance)Spread = 487 bpsImplied PD = 8.1%
• Only credit event (Washington Mutual) was coincident with 1st peak in implied PD
• Extraordinary external support (e.g. government assistance, acquisition)…
• …thus, senior debt obligations continued to perform, despite weakened condition
2/15/2011
12
www.fitchratings.com
U.S. Broker-Dealers
• Prior to crisis, spreads were relatively low for extended period
• As of October 2007, PD for sector was 1.1%...
• …however, several events of distress over ensuing 12-month period
• Highest CDS spread observed during period of study: Morgan Stanley (700 bps in October 2008)
0
100
200
300
400
500
600
700
800
1/06 4/06 7/06 10/06 1/07 4/07 7/07 10/07 1/08 4/08 7/08 10/08 1/09 4/09 7/09 10/09 1/10 4/10 7/10 10/10
Goldman Sachs Group Lehman Brothers Holdings Inc.Merrill Lynch & Co., Inc. Morgan StanleyBear Stearns Companies Inc.
Note: CDS spreads in text boxes are aggregated for the broker-dealer sector as a whole and calculated as the average of the spreads of
the individual entities.Sources: Fitch Ratings and Fitch Solutions.
U.S. Broker-Dealer CDS Spreads by Entity
(bps)
October 2007CDS spreads = 68 bps
Implied PD = 1.1%
March 2008CDS spreads = 274 bpsImplied PD = 4.6%
2/15/2011
13
www.fitchratings.com
Differing Experience for US vs. European Banks
• US banks experienced more pronounced distress at peak
• More recent convergence (e.g., as of Dec 2010…Europe = 146 bps versus US = 125 bps)
Banks - US vs. Europe
0
50
100
150
200
250
300
350
400
450
Jun-0
6
Dec
-06
Jun-0
7
Dec
-07
Jun-0
8
Dec
-08
Jun-0
9
Dec
-09
Jun-1
0
Dec
-10
(bps) US (25 of the largest North American banks, including major regionals)
EUR (25 largest banks, drawn from the global Top 50 by assets)
September 2008
US spread = 427 bps
Implied PD = 7.1%
versus
EUR spread = 127 bps
Implied PD = 2.1%
Peak spread (EUR) = 184 bps
Implied PD = 3.1%
Implied PD = 3.3%
2/15/2011
14
www.fitchratings.com
European Banks (By Country)
• Relatively low implied PD during initial stages of the credit crisis
• Changes in ‘ordinal’ ranking over time (e.g., each has been the widest at some point since March 2009)
CDS Spreads By Select Country (based on 25 largest EUR banks)
0
50
100
150
200
250
Jun-0
6
Dec
-06
Jun-0
7
Dec
-07
Jun-0
8
Dec
-08
Jun-0
9
Dec
-09
Jun-1
0
Dec
-10
(bps)
France (4 banks) Germany (3 banks) Italy (2 banks) Spain (2 banks)
June 2007 (average)
Spread = 8 bp
Implied PD = 0.1%
June 2010 (Spanish banks)
Spread = 233 bp
Implied PD = 3.9%
2/15/2011
15
www.fitchratings.com
European Banks (Credit Events)
• Avg. implied annual PD 1-year prior to Credit Events: 4.1%
• Avg. implied annual PD 3-months prior to Credit Events: 10.9%
Credit Events - Spreads and Implied PD for Prior 12 Month Period
0
500
1,000
1,500
2,000
2,500
t - 1
2 m
t - 1
1 m
t - 1
0 m
t - 9
m
t - 8
m
t - 7
m
t - 6
m
t - 5
m
t - 4
m
t - 3
m
t - 2
m
t - 1
m
CE (t
=0)
(bps)
0
5
10
15
20
25
30
35
40
(implied pd %)
Anglo Irish (Nov 2010) Bradford & Bingley (Jul 2009)Kaupthing (Oct 2008) Landsbanki (Oct 2008)Mean (four banks)
2/15/2011
16
www.fitchratings.com
European Insurance Companies
• No events of distress (e.g., default / downgrade to CCC or below) for insurers in this sample
Major European Insurers (sample of 17 entities)
0
100
200
300
400
500
600
Jun-0
6
Dec
-06
Jun-0
7
Dec
-07
Jun-0
8
Dec
-08
Jun-0
9
Dec
-09
Jun-1
0
Dec
-10
(bps)
March 2009 (peak)
Spread = 507 bps
Implied PD = 8.5%
June 2007
Spread = 12 bps
Implied PD = 0.2%
2/15/2011
17
www.fitchratings.com
Risk Management Implications
• CDS spreads can provide timely, market-based indicators of risk
• Particularly useful for valuation, active portfolio management, and assessing funding conditions
• In some cases, spreads lead observable credit deterioration (e.g., monolines during the crisis)
• Complement the informational content of other risk assessment tools (e.g., identifying outliers)
• Important to recognize the potential for ‘false positives’ and volatility
• For example, if used ‘peak’ spreads during crisis as a credit risk indicator, then either:
• Costly hedge
• Opportunity cost if sold off positions
• Additionally, both benefits and caveats in using spreads to derive PD estimates
• E.g., input to regulatory / economic capital calculations
• More ‘point-in-time’ view, but could undermine stability and reliability of results
• Procyclicality (higher PD → lower capital ratios → de-leveraging → further spread widening)
2/15/2011
18
www.fitchratings.com
Regulatory / Economic Capital Management (US example)
Volatility in CDS-Implied PD Could Drive Cyclicality in Basel Capital Charges (%)
Sector Period PD LGDa Reserves
(Expected Loss) Basel Capital Charge (Unexpected Loss)b
Expected Loss + Unexpected Loss
Trough 0.5 80 0.4 13.0 13.3 Peak 36.3 80 29.0 38.4 67.4
Monoline
Current 19.1 80 15.3 38.2 53.5 Trough 0.7 60 0.4 8.6 9.0 Peak 19.2 60 11.5 25.2 36.7
REIT
Current 4.6 60 2.7 15.6 18.3 Trough 1.7 60 1.0 11.7 12.7 Peak 8.0 60 4.8 18.9 23.7
Homebuilder
Current 4.2 60 2.5 15.0 17.5 Trough 0.4 60 0.3 8.8 9.1 Peak 7.1 60 4.3 21.4 25.7
Bank
Current 2.7 60 1.6 16.3 17.9 Trough 0.5 60 0.3 9.7 10.0 Peak 8.1 60 4.9 22.4 27.3
Insurance
Current 2.5 60 1.5 16.0 17.6 aMonoline LGD is assumed to be higher than for the other sectors, consistent with the realized average final price from CDS auctions for monoline reference entities that have experienced a credit event. bThe Basel capital charges are based on the IRB formulae for each respective asset class and, for monolines, banks, and insurance companies incorporate the 25% upward adjustment in correlation values for financial institutions under the recent Basel III revisions. Capital charge calculations are based on an 8% total capital requirement, which is consistent with both the Basel II and Basel III calibrations (notes that Basel III capital calculations above do not reflect the capital conservation buffer of 2.5% that will be phased in over the next several years).
Source: Fitch Ratings, Basel Committee on Banking Supervision.
2/15/2011
19
www.fitchratings.com
Drivers of ‘Disconnects’ in Spreads vs. Fundamentals?
• Total return orientation of market participants (i.e., MTM value of CDS positions)
• Not necessarily reflective of longer-term horizon (e.g., one year) / fundamental credit risk
• CDS pricing can be driven by a number of factors not directly related to an entity’s
fundamental creditworthiness:
• Liquidity conditions
• Counterparty risk
• Risk aversion of market participants (i.e., ‘risk-neutrality’ assumption)
• Leverage inherent in CDS trading
“As the markets came under increasing strain on account of the financial turmoil, liquidity in the CDS markets also began to dry up, raising doubts as to their value as an indicator of risk and funding costs.”
European Central Bank, August 2009
Margining: A Closer Look
Margining Practices
CDS Leverage and Return Dynamics
Systemic Implications
2/15/2011
21
www.fitchratings.com
Margining: Some Definitions
• ‘Margin’ is security against a counterparty failing to perform on its obligations
• Bonds purchased on margin = amount by which bond’s market value exceeds loan to investor
• CDS = amount of collateral posted relative to market value (i.e., MTM) of the CDS contract.
• ‘Initial’ margin = size of “down payment” necessary to cover obligations under the CDS
contract
• Higher rates reduce both leverage and variability in return on capital (i.e., must commit more
‘capital’ against CDS exposure)
• Lower rates increase leverage / ROC variability
• ‘Re-margining’ = posting of variation margin to cover ∆ in MTM of the CDS contract
(i.e., restores margin relationship to initial rate)
2/15/2011
22
www.fitchratings.com
Determinants of Margin Rates
• Margin rates are typically higher for:
• Lower quality reference entities
• Riskier counterparties
• Longer-dated exposures
• CDS margin rates for protection buyers (short credit exposure) are typically lower
than for protection sellers (long credit exposure)
• Consistent with fundamental asymmetry in credit spread movements
• Typically managed on a portfolio basis
• Examples in this study focus on individual CDS positions (for illustrative purposes)
• ‘Re-margining’ also depends on valuation of underlying risk exposure
• Particular concern for CDS written on complex or illiquid assets (e.g., AIG)
2/15/2011
23
www.fitchratings.com
Margin Rates: Examples
Summary of Margin Rates for Investment-Grade Corporate Exposures (%)
Source CDS Bonds
IMF (April 2007) 1 0−3
IMF (August 2008) 5 8−12
Dealers (Estimate)a 5−7 10 Institutional Investors (Estimate) a 7 15 Garleanu/Pedersen 5 25
FINRAb 4−7 N.A.
Committee on the Global Financial System (June 2007)c N.A. 4−10
Committee on the Global Financial System (June 2009)c N.A. 10−20
aEstimates are based on Fitch discussions with a sample of dealers and institutional investors. The margin rates cited should be treated as general approximations of market practices and are not meant to reflect an industry standard, since practices tend to vary to some degree across institutions. bBased on FINRA rule 4240 margining grid for reference asset whose credit spread is between 0 bp and 300 bps and has a five-year maturity. cApplicable range of haircuts based on corporate bonds rated ‘BBB’ to ‘A’, with lower haircuts for prime and higher haircuts for unrated counterparties. Note that "haircuts" and "margin rates" are conceptually similar, but technically different. N.A. — Not applicable.
Sources: Committee on the Global Financial System, “The Role of Margin Requirements and Haircuts in Procyclicality,” (2010), Bank for International Settlements; Garleanu, Nicolae and Lasse Heje Pedersen, “Margin-Based Asset Pricing and Deviations from the Law of One Price,” (2009), mimeo; International Monetary Fund, “Global Financial Stability Report: Financial Stress and Deleveraging,” (2008); Securities and Exchange Commission, “Proposed Rule Change by Financial Industry Regulatory Authority Pursuant to Rule 19b-4 under the Securities Exchange Act of 1934,” (2009).
2/15/2011
24
www.fitchratings.com
Margin Rates and Leverage / Return Dynamics
• Leverage magnifies 2.4% return on fully-funded cash position (if spreads tighten)…
• …but also amplifies the losses (in a spread widening scenario)
Margin Leverage, Exposure, and Return on Capital
Origination 50-bp Tightening 50-bp Widening
Credit Spread (bps) 100 50 150 Yield (%) 2.50 2.00 3.00
Bond (Noncallable) Initial Margin (i.e. Capital) ($) 1,000,000 1,000,000 1,000,000 Margin Rate (% of Notional) 10.0 10.0 10.0 Initial Exposure ($) 10,000,000 10,000,000 10,000,000 Borrowed Amount ($) 9,000,000 9,000,000 9,000,000 Bond Market Value ($) 10,000,000 10,236,783 9,769,445 Appreciation/Depreciation ($) 0 236,783 (230,555)
Return on Capital (%) 23.7 (23.1)
CDS Protection Seller Initial Margin (i.e. Capital) ($) 1,000,000 1,000,000 1,000,000 Margin Rate (% of Notional) 5.0 5.0 5.0 Initial Exposure ($) 20,000,000 20,000,000 20,000,000 CDS MTM ($) 0 491,994 (471,778)
Return on Capital (%) 49.2 (47.2)
The analysis above assumes: each instrument has a five-year tenor; both CDS and bond spread movements are identical; the CDS MTM is based on Treasury curve as of Aug. 4, 2010 and the on-the-run CDS contract; the market spread equals the deal spread when the contract is initiated, and as such, there is no upfront MTM payment; and the change in spreads is instantaneous and the returns therefore do not reflect contractual cash flows (e.g. CDS premiums or funding costs on the bond).
2/15/2011
25
www.fitchratings.com
Margin Rates and Leverage / Return Dynamics
• Investor’s return on capital ‘profile’ is highly sensitive to initial margin rates.
• Higher margin rates would substantially reduce return variability
(300)
(200)
(100)
0
100
200
300
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
Return on Capital (100-bp Initial Spread; 50-bp Tightening; Five-Year Maturity)
Return on Capital (100-bp Initial Spread; 50-bp Widening; Five-Year Maturity)
Margin Rate (%)
Margin Rates and Return on Capital (CDS Protection Seller)
Return on Capital (%)
(300)
(200)
(100)
0
100
200
300
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
Return on Capital (100-bp Initial Spread; 50-bp Tightening; Five-Year Maturity)
Return on Capital (100-bp Initial Spread; 50-bp Widening; Five-Year Maturity)
Margin Rate (%)
Margin Rates and Return on Capital (CDS Protection Seller)
Return on Capital (%)
(300)
(200)
(100)
0
100
200
300
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
Return on Capital (100-bp Initial Spread; 50-bp Tightening; Five-Year Maturity)
Return on Capital (100-bp Initial Spread; 50-bp Widening; Five-Year Maturity)
Margin Rate (%)
Margin Rates and Return on Capital (CDS Protection Seller)
Return on Capital (%)
(300)
(200)
(100)
0
100
200
300
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
Return on Capital (100-bp Initial Spread; 50-bp Tightening; Five-Year Maturity)
Return on Capital (100-bp Initial Spread; 50-bp Widening; Five-Year Maturity)
Margin Rate (%)
Margin Rates and Return on Capital (CDS Protection Seller)
Return on Capital (%)
2/15/2011
26
www.fitchratings.com
Re-Margining and Leverage / Return Dynamics
• Redeployment of gains → build-up of credit exposure during positive environments?
• Re-margining losses → negative cycle of losses / distressed selling / counterparty risk?
Re-Margining and Redeploying Investor Capital
Origination Spreads Tighten by 50 bps Spreads Widen by 50 bps
Credit Spread (bps) 100 50 150 Yield (%) 2.50 2.00 3.00
Bond (Noncallable)
Initial Margin (i.e. Capital) ($) 1,000,000 1,000,000 1,000,000 Margin Rate (% of Notional) 10.0 10.0 10.0 Appreciation/Depreciation ($) 0 236,783 (230,555)
New Margin Amount ($) 1,236,783 769,445 New Exposure (Market Value) ($) 10,000,000 12,367,830 7,694,450
CDS protection seller Initial margin (i.e. Capital) 1,000,000 1,000,000 1,000,000
Margin rate (% of Notional) 5.0 5.0 5.0 CDS MTM ($) 0 491,994 (471,778)
New Margin Amount ($) 1,491,994 528,222 New Exposure (Notional) ($) 20,000,000 29,839,880 10,564,440
Note: This analysis assumes that investors close out the original position and immediately redeploy the initial capital plus gains realized (in the case of spreads tightening) or initial capital less the loss realized (in the case of spreads widening). In the case of CDS, it is also assumed that investor capital is redeployed in a new CDS contract possibly referencing a different entity where the market spread is very close or equal to the contract spread so as to limit the impact on the analysis of any theoretical transfer of the MTM amount between investor and counterparty at the time the swap is entered into; also not considered in this analysis is any impact related to the transfer of accrued income.
2/15/2011
27
www.fitchratings.com
Changes in Margin Rates During Stress
• From dealer’s perspective → buffer to weather further spread / counterparty erosion
• For margined investors → amplifies de-leveraging and liquidity pressures
• Appears to pose more of a risk for leveraged bond investors
• Margin rates may be reset at relatively short intervals (i.e., term of securities financing)
• For bilateral CDS, fixed initial margin rate for life of the contract…
• …however, can (in effect) increase if counterparty’s financial condition weakens (i.e., lowering of
margining threshold, or unsecured credit exposure available)
Increasing Margin Rates Under Stress: A Shock to the System?
Bond (Noncallable) Stable Margin Rate Dynamic Margin Rate
Spread Widening (bps) 50 bps 50 bps Initial Margin (i.e. Capital) ($) 1,000,000 1,000,000 MTM loss ($) (230,555) (230,555) Residual margin (i.e. Capital) after MTM Loss ($) 769,445 769,445 Margin Rate (%) 10.0 20.0 Exposure (Market Value) ($) 7,694,450 3,847,225 Net Bonds Sold to Meet Margin Call ($) (2,074,895) (5,922,220)
2/15/2011
28
www.fitchratings.com
Systemic Implications
• Three potential sources of pro-cyclicality:
• Low initial margin rates (i.e., investors can build large positions)
• Re-margining (i.e., investment losses and gains can be magnified)
• Changes in margin rates under stress (increase in collateral buffers might inadvertently trigger
liquidity pressure)
• Even cash investors that do not employ leverage can be exposed to pricing volatility
stemming from reliance by some investors on margin financing
• CDS spread volatility (which might partly be explained by margin rates) could create
negative feedback loop for some issuers
• If financial reforms / movement to CCP were to result in higher margin requirements,
CDS leverage would likely be reduced
• …less spread volatility?
Appendix: Select Research on Other Topics
US Money Fund Exposure to European Banks
‘Burden-sharing’
Samples Used in CDS Analysis
2/15/2011
30
www.fitchratings.com
MMFs Reduced Exposure Prior To Recent VolatilityMMF Exposure to European Banks
(as % of total assets under management for ten largest MMF)
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
2H06
1H07
2H07
1H08
2H08
1H09
2H09
1H10
2H10
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
Ireland SpainPortugal ItalyEurope total (right axis) UK, France, Germany (right axis)
Source: U.S. Money Market Funds: Recent Trends in Exposure to European Banks
(Fitch Macro Credit Research, December 10, 2010)
2/15/2011
31
www.fitchratings.com
Regional Variation in MMF Exposure to Banks
Ireland Italy Portugal Spain UK France Germany Benelux Swiss Nordic
Europe
(total) Japan Oceania
North
America
2H06 0.4% 1.2% 0.0% 0.4% 8.8% 7.5% 6.2% 4.6% 2.1% 2.0% 34.2% 2.6% 1.1% 11.4%
1H07 0.3% 0.7% 0.0% 0.3% 8.1% 7.0% 5.9% 5.2% 4.0% 1.9% 33.9% 2.1% 1.5% 10.7%
2H07 0.6% 0.7% 0.1% 1.3% 10.4% 6.1% 3.5% 6.2% 3.4% 2.6% 35.1% 1.7% 1.5% 14.1%
1H08 1.5% 2.3% 0.0% 2.2% 8.3% 8.1% 2.3% 5.0% 1.9% 2.4% 34.9% 1.1% 2.6% 10.0%
2H08 0.5% 2.7% 0.0% 3.2% 8.4% 9.8% 1.8% 4.7% 1.9% 3.0% 36.0% 0.8% 3.1% 14.5%
1H09 0.0% 3.0% 0.5% 3.0% 8.3% 13.5% 3.1% 5.6% 1.9% 3.8% 42.7% 3.5% 3.2% 10.7%
2H09 0.4% 3.2% 0.3% 2.9% 8.1% 14.4% 4.4% 6.9% 0.9% 4.6% 46.1% 4.6% 5.3% 8.2%
1H10 0.0% 1.9% 0.0% 1.7% 7.0% 10.8% 4.6% 6.0% 1.2% 4.6% 37.5% 4.0% 4.8% 8.5%
2H10 0.0% 1.8% 0.0% 1.3% 6.4% 11.9% 4.3% 6.6% 1.5% 4.9% 38.7% 4.9% 5.4% 7.0%
MMF Exposure to Bank CDs and CP
(as % of total MMF assets under management)
Note: “Europe (total)” column in table above does not necessarily equal the exact sum of the European countries listed because of (1) rounding errors and (2) the exclusion of countries to which the MMFs sampled were not highly exposed
over the observation period.
2/15/2011
32
www.fitchratings.com
Changes in the Largest Single-Name Bank Exposures
As of 2H 2007…
% of Total
MMF assets As of 2H 2010…
% of Total
MMF assets
Citibank 3.6% BNP Paribas 3.1%
Barclays 3.4% Rabobank 3.0%
JP Morgan Chase 2.7% Credit Agricole 2.8%
HBOS 1.9% ING 2.5%
Societe Generale 1.9% Natixis 1.9%
Bank of America 1.9% Westpac 1.9%
UBS 1.7% Societe Generale 1.7%
ABN AMRO Bank 1.7% RBS 1.5%
RBS 1.5% Nordea Bank 1.3%
Fortis Bank 1.4% HSBC 1.3%
% of exposure represented by Certificates of Deposits (CD):
• Citibank (as of 2H 2007): 8%
• BNP Paribas (as of 2H 2010): 91%
2/15/2011
33
www.fitchratings.com
Burden-sharing: External Support vs. Hybrids / Sub Debt
0
200
400
600
800
1000
1200
Global Public Sector Capital
Support
Bank Risk Capital If Equity
Conversion
(EUR billion)
Hybrids
360
Subordinated
Debt
660
1,1101,020
Note: figures in chart are in (€1,000,000)
Source: Burden Sharing: Who Pays Next Time? (Fitch Macro Credit Research,
May 10, 2010)
2/15/2011
34
www.fitchratings.com
Appendix: Sample for US CDS Analysis
Bank Insurance Monoline REIT Home Builder Bank of America Corporation Aetna Inc. Ambac Financial Group, Inc. AMB Property LP Centex Corporation Bank of New York Co Inc. AFLAC Incorporated Assured Guaranty Corp. Archstone-Smith Operating
Trust D R Horton Inc.
Bank One Corporation Allstate Corporation FGIC Corporation Avalon Bay Communities Lennar Corporation BB&T Corporation American Financial Group, Inc. Financial Security Assurance
Inc. BRE Properties, Inc. M.D.C. Holdings, Inc.
Canadian Imperial Bank of Commerce
American International Group Inc.
MBIA, Inc. Camden Property Trust NVR Inc
Capital One Financial Corporation
Anthem Insurance Companies Inc.
MGIC Investment Corporation Developers Diversified Realty Corporation
Pulte Homes, Inc.
Citigroup Inc. Assurant Inc. Radian Group Inc. Duke Realty LP Ryland Group, Inc. Fifth Third Bancorp AXA Financial Inc. The PMI Group, Inc. Equitable Resources, Inc. Toll Brothers, Inc. JPMorgan Chase & Co. Berkshire Hathaway Inc. Equity One Inc KeyCorp Chubb Corporation Federal Realty Investment
Trust
Marshall & Ilsley Corp CIGNA Corporation First Industrial LP Mellon Financial Corporation First American Corporation Health Care Property
Investors, Inc.
National Bank of Canada Genworth Financial Inc. Health Care REIT, Inc. National City Corp Hartford Life, Inc. Healthcare Realty Trust
Incorporated
PNC Financial Services Group, Inc
Horace Mann Educators Corp Highwoods Realty LP
Popular, Inc. John Hancock Financial Services Inc.
Hospitality Properties Trust
Royal Bank of Canada Liberty Mutual Insurance Companies
HRPT Properties Trust
Sovereign Bancorp, Inc. Lincoln National Corporation iStar Financial Inc. State Street Corporation Manufacturers Life Insurance
Company Kimco Realty Corporation
SunTrust Banks Markel Corporation Liberty Property LP Toronto Dominion Bank Marsh & McLennan Companies,
Inc. Mack-Cali Realty Corporation
U.S. Bancorp Massachusetts Mutual Life Insurance Company
National Retail Properties Inc
Wachovia Corporation MetLife, Inc. New Plan Excel Realty Trust, Inc. (acquired by Centro in 2007)
Washington Mutual Incorporated Nationwide Financial Services, Inc.
ProLogis
Wells Fargo & Co New York Life Insurance Company
Regency Centers, L.P.
Pacific Life Insurance Co United Dominion Realty Trust, Inc.
Principal Financial Group Vornado Realty Trust Prudential Financial Inc. Washington Real Estate
Investment Trust
Safeco Corporation Weingarten Realty Investors The Progressive Corporation The St. Paul Travelers Group of
Companies, Inc.
Torchmark Corporation Unitrin, Inc
2/15/2011
35
www.fitchratings.com
Appendix: Sample for EUR CDS Analysis
Bank Insurance
ABN AMRO Bank NV Aegon N.V.
Banca Intesa SpA Allianz AG
Banco Bilbao Vizcaya Argentaria SA Assicurazioni Generali SpA
Banco Santander Central Hispano Aviva Plc
Bank of Scotland AXA
Barclays Bank plc Hannover Rueckversicherungs-AG
Bayerische Landesbank Girozentrale ING Groep NV
BNP Paribas Legal & General Group plc
Commerzbank AG Munich Re
Credit Agricole SA Old Mutual Plc
Credit Suisse Group Prudential plc
Danske Bank AS Royal & Sun Alliance Insurance plc
Deutsche Bank AG SCOR
Fortis SA/NV Standard Life Assurance Company
HSBC Holding plc Swiss Life Holding
ING Bank NV Swiss Reinsurance Co
Lloyds TSB Group plc Zurich Financial Services AG
Natixis
National Westminster Bank plc
Nordea Bank AB
Rabobank Nederland
Royal Bank of Scotland plc
Societe Generale
UBS AG
Unicredito Italiano SpA
2/15/2011
36
www.fitchratings.com
Disclaimer
Fitch Ratings’ credit ratings rely on factual information received from issuers and other sources.
Fitch Ratings cannot ensure that all such information will be accurate and complete. Further, ratings
are inherently forward-looking, embody assumptions and predictions that by their nature cannot be
verified as facts, and can be affected by future events or conditions that were not anticipated at the
time a rating was issued or affirmed.
The information in this presentation is provided “as is” without any representation or warranty.
A Fitch Ratings credit rating is an opinion as to the creditworthiness of a security and does not
address the risk of loss due to risks other than credit risk, unless such risk is specifically mentioned.
A Fitch Ratings report is not a substitute for information provided to investors by the issuer and its
agents in connection with a sale of securities.
Ratings may be changed or withdrawn at any time for any reason in the sole discretion of
Fitch Ratings. The agency does not provide investment advice of any sort. Ratings are not
a recommendation to buy, sell, or hold any security.
ALL FITCH CREDIT RATINGS ARE SUBJECT TO CERTAIN LIMITATIONS AND DISCLAIMERS. PLEASE READ THESE
LIMITATIONS AND DISCLAIMERS AND THE TERMS OF USE OF SUCH RATINGS AT WWW.FITCHRATINGS.COM.
New York One State Street Plaza New York, NY 10004
London 30 North Colonnade Canary WharfLondon E14 5GN