Академический Документы
Профессиональный Документы
Культура Документы
Robert J. Grossman, Group Managing Director Martin Hansen, Senior Director February 2011
3,000 2,500 2,000 1,500 1,000 500 0 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
Pronounced volatility in CDS spreads during the crisis Spreads are one of the analytical tools used by Fitchs credit analysts (e.g., identifying outliers)
www.fitchratings.com
2/15/2011
Agenda
CDS Spreads as Default Risk Indicators
Overview: CDS vs. Bonds Methodology US Experience During the Crisis Europe: Banks and Insurers Risk Management Implications
www.fitchratings.com
2/15/2011
www.fitchratings.com
2/15/2011
www.fitchratings.com
2/15/2011
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
www.fitchratings.com
2/15/2011
Overview of Findings
Volatility in the Relative Risks of Sectors over the Crisis
June 2007 October 2008 March 2009 August 2010 CDS CDS CDS CDS Spread Implied spread Implied spread Implied spread Implied Sector (bps) PD (%) Sector (bps) PD (%) Sector (bps) PD (%) Sector (bps) 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.
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
www.fitchratings.com
2/15/2011
6/08
6/09
6/10
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)
www.fitchratings.com
2/15/2011
1,200
Second Peak (REIT) Spread = 1,154 bps Implied PD = 19.2% Second Peak (Homebuilder) Spread = 481 bps Implied PD = 8.0%
1,000
800
First Peak (Homebuilder) Spread = 385 bps Implied PD = 6.4%
600
400
First Peak (REIT) Spread = 409 bps Implied PD = 6.8%
200
0 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
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
www.fitchratings.com
2/15/2011
10
600 500 400 300 200 100 0 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. Second Peak (Banks) Spread = 247 bps Implied PD = 4.1% First Peak (Banks) Spread = 427 bps Implied PD = 7.1% Peak (Insurance) Spread = 487 bps Implied 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
www.fitchratings.com
2/15/2011
11
U.S. Broker-Dealers
U.S. Broker-Dealer CDS Spreads by Entity
Goldman Sachs Group Merrill Lynch & Co., Inc. Bear Stearns Companies Inc. Lehman Brothers Holdings Inc. Morgan Stanley
(bps)
March 2008 CDS spreads = 274 bps Implied PD = 4.6% October 2007 CDS spreads = 68 bps Implied PD = 1.1%
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)
www.fitchratings.com
2/15/2011
12
De c06
De c07
De c09
US banks experienced more pronounced distress at peak More recent convergence (e.g., as of Dec 2010Europe = 146 bps versus US = 125 bps)
www.fitchratings.com
De c10
2/15/2011
De c08
Ju n07
Ju n06
Ju n09
Ju n10
Ju n08
13
200
50
De c06
De c07
De c08
De c09
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)
www.fitchratings.com
De c10
2/15/2011
Ju n08
Ju n09
Ju n06
Ju n07
Ju n10
14
m t1
12
11
10
t-
t-
t-
t-
t-
t-
t-
t-
Avg. implied annual PD 1-year prior to Credit Events: 4.1% Avg. implied annual PD 3-months prior to Credit Events: 10.9%
www.fitchratings.com
2/15/2011
t-
t-
CE
t-
(t= 0)
15
De c07
De c08
De c09
De c06
No events of distress (e.g., default / downgrade to CCC or below) for insurers in this sample
www.fitchratings.com
De c10
2/15/2011
Ju n06
Ju n07
Ju n08
Ju n09
Ju n10
16
For example, if used peak spreads during crisis as a credit risk indicator, then either:
Costly hedge Opportunity cost if sold off positions
www.fitchratings.com
2/15/2011
17
Monoline 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.
www.fitchratings.com
2/15/2011
18
CDS pricing can be driven by a number of factors not directly related to an entitys 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.
www.fitchratings.com
2/15/2011
19
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)
www.fitchratings.com
2/15/2011
21
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
www.fitchratings.com
2/15/2011
22
Estimates 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).
www.fitchratings.com
2/15/2011
23
Bond (Noncallable)
Initial Margin (i.e. Capital) ($) Margin Rate (% of Notional) Initial Exposure ($) Borrowed Amount ($) Bond Market Value ($) Appreciation/Depreciation ($) Return on Capital (%)
The analysis above assumes: each instrument has a five-year tenor; both CDS and bond spread movements are identical; the CDS MTM is based on Treasury curve as of Aug. 4, 2010 and the on-the-run CDS contract; the market spread equals the deal spread when the contract is initiated, and as such, there is no upfront MTM payment; and the change in spreads is instantaneous and the returns therefore do not reflect contractual cash flows (e.g. CDS premiums or funding costs on the bond).
www.fitchratings.com
2/15/2011
24
300
Return on Capital (%)
Investors return on capital profile is highly sensitive to initial margin rates. Higher margin rates would substantially reduce return variability
www.fitchratings.com
2/15/2011
25
Bond (Noncallable)
Initial Margin (i.e. Capital) ($) Margin Rate (% of Notional) Appreciation/Depreciation ($) New Margin Amount ($) New Exposure (Market Value) ($)
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.
Redeployment of gains build-up of credit exposure during positive environments? Re-margining losses negative cycle of losses / distressed selling / counterparty risk?
www.fitchratings.com
2/15/2011
26
From dealers 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 counterpartys financial condition weakens (i.e., lowering of margining threshold, or unsecured credit exposure available)
www.fitchratings.com
2/15/2011
27
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?
www.fitchratings.com
2/15/2011
28
2H 06
1H 10
1H 07
2H 07
1H 08
2H 08
Source: U.S. Money Market Funds: Recent Trends in Exposure to European Banks (Fitch Macro Credit Research, December 10, 2010)
www.fitchratings.com
2/15/2011
2H 10
1H 09
2H 09
30
MMF Exposure to Bank CDs and CP (as % of total MMF assets under management)
2H06 1H07 2H07 1H08 2H08 1H09 2H09 1H10 2H10 Ireland 0.4% 0.3% 0.6% 1.5% 0.5% 0.0% 0.4% 0.0% 0.0% Italy 1.2% 0.7% 0.7% 2.3% 2.7% 3.0% 3.2% 1.9% 1.8% Portugal 0.0% 0.0% 0.1% 0.0% 0.0% 0.5% 0.3% 0.0% 0.0% Spain 0.4% 0.3% 1.3% 2.2% 3.2% 3.0% 2.9% 1.7% 1.3% UK 8.8% 8.1% 10.4% 8.3% 8.4% 8.3% 8.1% 7.0% 6.4% France Germany Benelux 7.5% 6.2% 4.6% 7.0% 5.9% 5.2% 6.1% 3.5% 6.2% 8.1% 2.3% 5.0% 9.8% 1.8% 4.7% 13.5% 3.1% 5.6% 14.4% 4.4% 6.9% 10.8% 4.6% 6.0% 11.9% 4.3% 6.6% Swiss 2.1% 4.0% 3.4% 1.9% 1.9% 1.9% 0.9% 1.2% 1.5% Nordic 2.0% 1.9% 2.6% 2.4% 3.0% 3.8% 4.6% 4.6% 4.9% Europe (total) 34.2% 33.9% 35.1% 34.9% 36.0% 42.7% 46.1% 37.5% 38.7% North Japan Oceania America 2.6% 1.1% 11.4% 2.1% 1.5% 10.7% 1.7% 1.5% 14.1% 1.1% 2.6% 10.0% 0.8% 3.1% 14.5% 3.5% 3.2% 10.7% 4.6% 5.3% 8.2% 4.0% 4.8% 8.5% 4.9% 5.4% 7.0%
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.
www.fitchratings.com
2/15/2011
31
As of 2H 2010
BNP Paribas Rabobank Credit Agricole ING Natixis Westpac Societe Generale RBS Nordea Bank HSBC
% of exposure represented by Certificates of Deposits (CD): Citibank (as of 2H 2007): 8% BNP Paribas (as of 2H 2010): 91%
www.fitchratings.com
2/15/2011
32
Note: figures in chart are in (1,000,000) Source: Burden Sharing: Who Pays Next Time? (Fitch Macro Credit Research, May 10, 2010)
www.fitchratings.com
2/15/2011
33
www.fitchratings.com
2/15/2011
34
www.fitchratings.com
2/15/2011
35
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.
www.fitchratings.com
2/15/2011
36
New York
One State Street Plaza New York, NY 10004
London
30 North Colonnade Canary Wharf London E14 5GN