Sovereign Credit Default Swap Premia∗
Patrick Augustin†
‡
Stockholm School of Economics
This version: April 2012
Preliminary, comments welcome
Abstract
This paper provides a comprehensive overview of the young, but rapipdly growing literature on sovereign credit default swap premia, describes key statistical and stylized facts about
prices, the market, its players and related trading activities and attempts to raise some thoughtprovoking questions. We thereby hope to serve as a useful starting point for anyone interested
in the topic.
Keywords: Credit Default Swap Spreads, Default Risk, Descriptive Statistics, Literature Review,
Sovereign Debt, Term structure
JEL Classification: A31, D40, E02, F30, F34, G01, G12, G13, G14, G15, O57, Y10
∗
This version is preliminary and incomplete. Comments are welcome and appreciated. We would like to thank
René Kallestrup for valuable feedback. All remaining errors are our own.
†
Stockholm School of Economics, Finance Department, Sveavägen 65, 6th floor, Box 6501, SE-113 83 Stockholm,
Sweden. Email: Patrick.Augustin@hhs.se.
‡
1
Introduction
The late 1990s and early 2000s displayed a series of emerging market sovereign defaults. Nowadays,
the popular press is covered with news about developed countries in distress, in particular the
GIIPS countries: Greece, Italy, Ireland, Portugal and Spain. For most part of history, lenders to
national governments were unable to insure against sovereign default and had to trust creditors
that they honor their debt obligation in good faith. Our generation, on the other hand, has
witnessed the creation of insurance products to protect specifically against such contingencies.
These insurance products, so-called credit default swaps (henceforth CDS) have undergone two
decades of spectacular growth and, covering an underlying cash market of approximately 41 trillion
USD in 2011 represent an important part of the economy1 . Policy makers, regulators and investors
increasingly look at sovereign CDS prices to gauge the health of our system2 . While there have
been several treatments about the corporate CDS market, we feel that there is a void when it comes
to a comprehensive overview of the sovereign analogue3 . The purpose of this document is to fill
this gap.
More specifically, the goal of this paper is to provide an integrated overview of the key stylized
facts of the sovereign CDS market, its players, frictions, the determinants of prices and trading
related information. This is executed by reviewing the very young but growing literature, and
discussing mainly the publicly available data from the Depository Trust and Clearing Corporation
(DTCC), the Bank for International Settlements (BIS) and a historical dataset provided by Markit.
As the focus of this review is the synthetic fixed income market, we mainly refer in citations to
academic work on sovereign CDS spreads, but refer at times to academic work on sovereign cash
bonds4 .
We hope that this review provides a useful starting point for anyone interested in the topic and
that our exposition raises some thought-provoking questions. At the outset, we would like to note
some puzzling observations. Despite the huge press coverage of credit derivatives, they represent
1
The Economist global debt clock displays an estimate of current global public debt. Source: www.economist.com.
Hart and Zingales (Fall 2011) suggest the use of financial CDS to evaluate the health of banks.
3
See for example Stulz (2010) for an overview of corporate CDS and their relation to the financial crisis or Das
and Hanouna (2006).
4
As our goal is specifically the treatment of sovereign CDS spreads, we apologize to any authors who feel that
their work is relevant and should have been included in this review.
2
1
only a tiny fraction (4.5%) of the overall ”over-the-counter” derivatives market. Moreover, among
all traded credit default swaps, sovereign contracts make up roughly 9%. Although the overall size
of the CDS market is valued at 32 trillion USD, the net economic exposure is only 1.3 trillion5 .
Furthermore, why is it that contracts on emerging market economies tend to trade in high numbers
and small volumes, while developed economies tend to trade with a larger net economic exposure
per contract and have fewer contracts outstanding? Why are hedge funds so often blamed for
speculation in the sovereign CDS market, when in fact they represent only a small fraction of
all counterparties? Why is there such a surge in CDS trading on the US government in 2011?
What determines the shape of the sovereign yield curve when government term structures exhibit
similar patterns as those of corporations, but country-specific factors play relatively weaker roles in
explaining variation in CDS spread levels. And finally, why is it so important to avoid a trigger on
Greek CDS when the net reported economic exposure is below 4 billion at the end of 2011? These
facts and more are highlighted along the road.
The remainder of this document structures as follows. The second section explains the nature
and mechanism of CDS. Readers familiar with this may directly want to jump to section three,
which describes the market for sovereign CDS. Section four illustrates statistical stylized facts about
sovereign CDS prices. In section five, we focus on the determinants of these products and review
the literature. We follow up with a discussion on the relationship between sovereign CDS and cash
bonds, and CDS related frictions in section six. Finally, in section seven, we give an overview of the
trading in sovereign CDS markets based on the newly available data from the global data repository
DTCC, and provide some indication on how this relates to country specific fundamentals. Finally,
in section eight, we conclude.
2
What Are Credit Default Swaps
Engineered in the early nineties by JP Morgan to meet the growing demand to slice and dice credit
risk6 , credit default swaps represent the most simple (”vanilla”) instrument among the class of credit
5
The net economic exposure can be understood as an upper bound on insurance payments upon default. Actual
payments are expected to be lower, because recovery values of defaulted bonds are generally non-zero.
6
The exact year of creation is not clear. Tett (2009) refers to the first CDS in 1994, when J.P. Morgan off-loaded
its credit risk exposure on Exxon by paying a fee to the European Bank for Reconstruction and Development. More
2
derivatives. Nevertheless, they remain till this date highly controversial. While the proponents
would defend them as efficient vehicles to transfer and manage credit risk as well as means to
widen the investment opportunity set of return chasers, opponents denounce them as weapons
of mass destruction, time bombs7 , financial hydrogen bombs8 or speculative bets on government
default. For the sake of this review, let us stick with the factual definition of what they really are:
insurance contracts offering protection against the default of a referenced sovereign government or
corporation.
Technically speaking, a credit default swap is a fixed income derivative instrument, which
allows a protection buyer to purchase insurance against a contingent credit event on an underlying
reference entity by paying an annuity premium to the protection seller, generally referred to as
the Credit Default Swap spread. The premium is usually defined as a percentage on the notional
amount insured and can be paid in quarterly or bi-annual installments. Readers not familiar with
this particular notion of insurance, should simply bear in mind the example of car insurance. But
whereas for car insurance, the contract would be written on a specific vehicle, credit insurance
could be contracted on a given brand, think of a Volkswagen for the sake of the argument. Thus,
any Volkswagen involved in an accident would trigger a payout to the holder of the contract. In the
language of credit derivatives, you would purchase a CDS on a company, the reference entity, and
if that company fails to meet its obligations for any of a predetermined set of its debt obligations,
the payout would occur. More specifically, the contract comprises usually a specific class of the
firm’s capital structure, such as the senior unsecured or the junior debt portion of the company,
and references a particular amount of insured debt, the notional amount.
Failure to meet the debt obligations is labeled a credit event. Hence, a credit event triggers a
payment by the protection seller to the insuree equal to the difference between the notional principal
and the loss upon default of the underlying reference obligation, also called the mid-market value.
In general, the occurrence of a credit event must be documented by public notice and notified to the
investor by the protection buyer. Amid the class of qualifying credit events are Bankruptcy, Failure
importantly for this review, the author also references a deal between J.P. Morgan and Citibank asset management
on the credit risk of Belgian, Italian and Swedish government bonds around the same time (p.48).
7
See Warren Buffett,
Berkshire Hathaway Annual Report for 2002,
p.13,
on line at
http://www.berkshirehathaway.com/2002ar/2002ar.pdf
8
Felix Rohatyn, a Wall Street banker emplyed at Lazard Frères, quoted in Tett (2009).
3
to pay, Obligation default or acceleration, Repudiation or moratorium (for sovereign entities) and
Restructuring, and represent thus a broader definition of distress than the more general form of
Chapter 7 or Chapter 11 bankruptcy.
The settlement of insurance contracts comes in two forms, it may be either physically or cash
settled. In case of a cash settlement, the monetary exchange involves only the actual incurred losses
and the claimant holds on to the physical piece of paper providing him with a claim on the capital
structure of the sovereign balance sheet. On the other hand, if there is a physical settlement, the
claimant may hand any reference obligation referenced in the contractual agreement to the insurer
and obtains the full notional amount of the underlying contract in return. The insurer can then try
to maximize its resale value in the secondary market. This implies that the claimant literally holds
a Cheapest-to-Deliver option, as he may deliver the least valuable eligible reference obligation9 .
This option is particularly relevant in the case of a corporate restructuring, which is why the
restructuring credit event is the most critical. As a consequence, it has been adapted numerous
times and comes in multiple colors and associated contractual clauses. In particular, apart from
Full Restructuring, which allows the delivery of any obligation with a remaining maturity up to
30 years, a contract may simply omit this definition, in which case it is labeled No Restructuring,
or it may qualify as Modified Restructuring or even Modified Modified Restructuring. The former
definition is more prevalent in the U.S. and limits the remaining maturity of the deliverable reference
obligation to a maximum of 30 months, whereas the latter is predominant in Europe, but restricts
the maximum maturity to 60 months. Even after default, prices of bonds usually suffer from
wide market fluctuations, so the question remains as to how the precise value of the insurance
settlement is determined. Markets have converged to a practice where the mid-market value is
obtained through a dealer poll. Whether this pricing mechanism is efficient, remains unclear.
Three recent papers have taken a closer look at this issue for corporate CDS: Helwege et al. (2009),
Chernov et al. (2011) and Gupta and Sundaram (2012).
Credit default swaps are part of the over-the-counter market (OTC). Thus they are not trading on an organized exchange and remain to a large extent unregulated. The reason for this is a
9
See Ammer and Cai (2011) for evidence on sovereign CDS and Jankowitsch et al. (2008) for evidence on corporate
CDS. Ammer and Cai (2011) also study the short-run dynamics between sovereign CDS spreads and their cash
counterparts, in a similar fashion as Blanco et al. (2005) did for the corporate market.
4
provision inserted in 2000 in the Commodity Futures Modernization Act by Senator Phil Gramm,
who from 1995 to 2000 presided over the Senate Banking Committee, exempting CDS from regulation by the Commodity Futures Trading Commission (CFTC)10 . Nevertheless, guidance on the
legal and institutional details of CDS contracts is given by the International Swaps and Derivatives
Association (ISDA)11 . They also act as a non-voting secretary for the Credit Determination Committee, which deliberates over issues involving Reference Entities traded under Transaction Types
that relate to the relevant region, including Credit Events, CDS Auctions, Succession Events and
other issues. Moreover, ISDA has played a significant role in the growth of the CDS market by
providing a standardized contract in the 1998 (”the ISDA Master Agreement”), which was updated
in 2002, and supplemented with the 2003 Credit Derivatives Definitions (”The Definitions”) and
the July 2009 Supplement. The contractual details of these contracts are crucial, and as usual,
the devil lies in the details, as was recently proved in the restructuring case of Greek government
debt. European officials have heavily pushed towards a voluntary restructuring, which would not
be binding on all bondholders. In such a case, it would not be be likely that the agreed deal would
trigger payments under existing CDS contracts12 . The landscape for CDS has further altered with
the implementation of the CDS Big Bang and Small Bang protocols on 8 April and 20 June 2009
for the American and European CDS markets respectively. The biggest changes relate to the standardization of the coupon structure and the settlement process. Going forward, the Dodd-Frank
Act in the U.S. will likewise have a significant impact on the way these products are traded. The
Inter-continental Exchange (ICE), a subsidiary of the NYSE, is already recording growing market share in the clearing process and both academic and political voices call for a move towards
organized exchanges, more transparency and more orderly price dissemination13 .
Before moving on to the statistical features of the market for sovereign CDS, we want to point
out that we don’t describe their pricing mechanism. The interested reader may however refer to
Duffie (1999), Hull and White (2000a) and Lando (2004), who provide excellent descriptions of
these details.
10
See Nouriel Roubini and Stephen Mihm, Crisis Economics, pp.199.
See www.isda.org.
12
See Greek Sovereign Debt Q&A, October 31 2011, www.isda.org
13
Another growing clearing platform for CDS is provided by the CME.
11
5
3
The Market for Sovereign Credit Default Swaps
The last decade was accompanied with a spectacular growth in the market for credit default swaps,
nearing three-digit growth rates year by year. The Bank for International Settlements (BIS) publishes bi-annual statistics on the notional amounts outstanding and gross market values of OTC
derivatives14 , and also includes credit default swaps in their reporting since 2004. As can be seen
in Table (1), the market grew from approximately 6 trillion USD of notional amount outstanding
to a peak of 58 trillion USD in the second term of 2007. Subsequent surveys show a significant
drop in trading with figures down to 32 trillion USD in June 2011, which is mostly linked due to
a netting of outstanding positions as regulators’ increasing concerns about counterparty risk and
calls for higher transparency have led to portfolio compressions, but also partly due to the fact
that credit derivatives were central to the 2007-2009 financial crisis15 . While contracts written on
a single underlying reference made up for the bulk of the trading volume in 2004 with a market
share of 80%, multi-name instruments have increased more rapidly due to the practice of synthetic
securitization and represented roughly 44% of the market in 2011. While these numbers sound
impressive, the reader should bear in mind that credit default swaps still account for only a small
fraction of the overall OTC market, as is illustrated in Table (2). With a total trading volume
of roughly 32 trillion USD in June 2011, credit default swaps represent merely 4.58% of the total
amount of notional oustanding in all OTC derivatives, which has grown to 707.57 trillion USD.
The largest fraction comes from interest rate derivatives, which is estimated to have a size 553.88
trillion USD, or equivalently 78.28% of the market.
Nominal or notional amounts outstanding provide a measure of market size and a reference
from which contractual payments are determined in derivatives markets. However, such amounts
are generally not those truly at risk. The amounts at risk in derivatives contracts are a function of
the price level and/or volatility of the financial reference index used in the determination of contract
payments, the duration and liquidity of contracts, and the creditworthiness of counterparties. They
14
The notional amounts are likely to slightly underestimate the total value of the market, as until the end of 2011,
only 11 countries are reporting their OTC derivative statistics to the BIS. Belgium, Canada, France, Germany, Italy,
Japan, the Netherlands, Sweden, Switzerland, the United Kingdom and the United States. From December 2011,
Australia and Spain are expected to contribute to the semiannual survey, bringing to 13 the number of reporting
countries.
15
Portfolio compression refers to the process by which two counterparties maintain the same risk profile, but reduce
the number of contracts and gross notional value held by participants.
6
are also a function of whether an exchange of notional principal takes place between counterparties.
A measure commonly reported along the lines is gross market value, which provides a more accurate
measure of the scale of financial risk transfer taking place in derivatives markets. As is reported
in the BIS statistical notes, ”gross market values are defined as the sums of the absolute values of
all open contracts with either positive or negative replacement values evaluated at market prices
prevailing on the reporting date. Thus, the gross positive market value of a dealers outstanding
contracts is the sum of the replacement values of all contracts that are in a current gain position
to the reporter at current market prices (and therefore, if they were settled immediately, would
represent claims on counterparties). The gross negative market value is the sum of the values of
all contracts that have a negative value on the reporting date (i.e. those that are in a current loss
position and therefore, if they were settled immediately, would represent liabilities of the dealer to
its counterparties). The fact that these numbers are gross merely reflects the fact that contractual
positions of positive and negative values with the same counterparty are not netted out”. Taking
a look at the final column of Table (2), it is possible to see that the gross market value of credit
default swaps is 1.35 trillion USD during the first half of 2010, which is still a sizable market, but
this number is much less breathtaking than the total notional amount outstanding of 32 trillion
USD. Going a step further, gross credit exposure, which takes account of all legally enforceable
netting agreements, makes this number shrink further to a value of 375 billion USD, as can be seen
in the last column of Table (3).
This document is however about sovereign CDS, so we may wonder what these numbers look
like for insurance contracts written on government debt. Statistics on sovereign CDS have until
recently been much less detailed, but what is clear from Table (4) is that the amount of trading
in the sovereign CDS market with 2.91 trillion USD in 2011 represents approximately 9% of the
overall market for OTC credit derivatives. In contrast to corporate CDS, they are on the other
hand largely concentrated around single-name products, who reflect a trading volume of 2.75 trillion
USD, or respectively a fraction of 95%. While more detailed statistics for gross credit exposures
of sovereign CDS are not publicly available on the BIS website, we can get a noisy estimate if we
make the simplifying assumption that the fraction of gross credit exposure to total notional amount
traded is the same for sovereign reference entities and the overall market. Under this assumption,
7
the gross market value of all oustanding contracts is approximately 121 billion USD, while the total
net exposure reduces to roughly 33.6 billion USD. This is still an economically meaningful number,
but admittedly too small to justify interference in the market to avoid a payout trigger, unless it
is for other indirect reasons.
So who trades these products. Hedge funds are often blamed for their speculative one-sided bets,
but if we look at the counterparties involved in sovereign CDS trading in Table (5), we see that it is
to a large extent the reporting dealers, who operate in this market. Their reported trading volume
is approximately 1.8 trillion USD, reflecting a market share of 66.81%, followed second by Banks
and Security Firms with a market share of 21.54% or an underlying notional amount outstanding of
592 billion USD. Hedge funds are only third in line, with a total volume of 146 billion USD. While
this is no evidence, it suggests at least that sovereign CDS are used primarily for hedging purposes,
rather than for speculative bets16 . We should however be cautious with such fast conclusions, as
their trading volume has almost doubled from 2010 to 2011, which reflect arbitrage opportunities
linked to the European sovereign debt crisis, wheras banks and security firms have reduced their
exposure from 828 billion USD to 592 billion USD over the same time period.
In contrast to cash bonds, CDS have the attractive feature that they are constant maturity
products. This means that on every day, a market participant can buy an insurance contract with
an identical remaining maturity. Bonds, on the other hand, are issued only seldomly, and their
maturity decreases as time goes by. In addition, an investor has access to a full set of maturities,
as CDS trade usually from maturities of 6 months up to about 30 years. However, liquidity is
still mostly concentrated around five years17 . Table (6) shows that in 2011, the total volume of
notional amount outstanding with maturities above one and up to five years, was 23.19 trillion
USD, representing a total market share of 71.57%. This fraction has fluctuated closely around this
level since the BIS has started publishing their reports about CDS. The remaining contracts are
well balanced among maturities below one year, with a total volume 3.92 trillion USD or 12.11%,
or above five years, with a total volume of 5.29 trillion USD or 16.32%18 .
16
Bongaerts et al. (2011) also cite evidence that banks are primarily net buyers of (corporate) CDS, while insurance
companies and funds are mainly net sellers.
17
The following numbers are for the overall CDS market, and not specifically for sovereign CDS.
18
In fact, we should emphasize that for sovereign reference entities, trading is not as much concentraed in the 5-year
contract as it is the case for corporate reference entities. Evidence is also provided in Pan and Singleton (2008) and
Packer and Suthiphongchai (2003).
8
Last, but not least, Table (7) illustrates that the CDS market is mostly a transnational market.
81.70% trade with a counterparty based in a foreign country, in contrast to only 18.30% of all
contracts that are handled with a counterparty of the home country. The international nature of
the trading raises questions about the usefulness of unilateral policy decisions affecting the market,
such as the ban of naked CDS by Germany in May 2010.
4
Statistical Facts about Sovereign CDS
Beyond knowing the size of the market, which parties trade and with whom, a comprehensive
overview should also provide some stylized statistical facts about the probability distributions
of sovereign CDS. The following statistics are based on a sample of daily mid composite USD
denominated CDS quotes for 38 sovereign countries taken from Markit over the sample period 9
May 2003 until 19 August 2010, where all contracts contain the full restructuring clause. The
sample covers prices for a wide selection of the term structure, including 1, 2, 3, 5, 7 and 10-year
contracts and spans 5 geographical regions, including the Americas, Europe, Africa, the Middle
East and Asia.
A first set of summary statistics is provided in Table (8), where we have grouped the countries in
rating categories according to their average credit rating on external debt over the sample period.
The average term structure is always upward sloping, going from 15 basis points at the 1-year
horizon to 26 basis points at the ten-year maturity for AAA rated entities, and from 433 to 599
basis points for the B rated countries. A mean upward sloping term structure was also shown
for a set of three emerging countries in Pan and Singleton (2008). As the countries become less
creditworthy, the volatility of their CDS spreads jumps up as well. For example, the average
standard deviation of the 5-year CDS spread is 34 basis points for the most creditworthy country,
and 328 basis points for the least creditworthy. Moreover, the level of CDS spreads exhibit positive
skewness, with a value usually around 2, and they have very fat tails. But the excess kurtosis
generally decreases with the asset horizon. Finally, CDS spreads are also very persistent. Among
all statistics, the lowest and highest first-order autocorrelation coefficients are 0.9901 and 0.9979
respectively for daily observations.
9
While this measures provide a rough overview of the statistical stylized facts, a common theme
of interest among credit instruments is the slope of the term structure. A structural credit model
of Merton (1974) predicts that the credit curve is increasing for high-quality credit levels, humpshaped for intermediate credit quality and decreasing for low levels of creditworthiness. Lando
and Mortensen (2005) have confirmed these predictions for corporate CDS. For their analysis, they
rank their observations based on the average five-year spread level. We attempt to do the same
exercise here for our sovereign CDS sample. Hence for each country on each day, we look at the
5-year CDS spread and group the observations into ranges of 100 basis points and then report
within each category, the average spread, as well as the difference between the 10-year and the
1-year spread, the 10-year and the 3-year spread and the 3-year and the 1-year spread. The average
values are calculated across time and entities. The results, which are shown in Table (9), yield
similar conclusions as for corporate CDS. The overall slope, defined as the difference between the
10-year and 1-year spread, first increases as the credit quality weakens, and then decreases before
becoming negative at very high spread observations. Yet, if we look separately at the gap between
the 1 to 3-year, as well as the 3 to 10-year spread, it is evident that the former turns negative only
much later, confirming the humped shape curve for intermediate groups. Another way to look at
this is by computing the average slope levels in the same three segments, as well as the fraction of
negative values for sliding groups of 400 curves. This yields the outcome plotted in Figure (1), from
which we can draw the same conclusion. At first glance, such results make us feel comfortable that
statistical models of credit curves can be applied equally well to corporate and sovereign reference
contracts. From an empirical perspective, however, it raises the question about the determinants of
the shape of the sovereign credit curve, as country-specific factors seem to have little explanatory
power for the time-series variation of sovereign CDS, at least at higher sampling frequencies. This
contrasts with the predictions of a Merton (Merton 1974) type world. We will address this issue in
the following section.
10
5
Determinants of Sovereign CDS spreads
Up to this point, it remains to a large extent unclear what is driving variations in sovereigns CDS
spreads. Let us again go back to the classical Merton model to frame our thoughts. According
to this set-up, the credit spread of a company should be determined by the asset volatility of the
firm’s balance sheet. Or in other terms, characteristics specific to each firm should influence its
likelihood of default. To make the analogy with sovereigns, we would also intuitively expect the
fluctuations in default probabilities to be driven by country-specific fundamentals. Yet, it turns
out that a major fraction of the variation in sovereign CDS spreads is determined by a few global
variables, at least at higher trading frequencies. Moreover, most studies reference some sort of risk
originating in the United States. Where the camps of thought are divided, is whether these global
factors are related to fundamental macroeconomic risk, or rather to financial risk.
There is a long strand of literature, which has tried to understand bond spreads. Here, apart a
few minor exceptions, we will only focus on derivative contracts written on government debt. On
average, sovereign CDS spreads generally have a downward trend, with sharp spikes every time
there is a run-up in risk aversion due to a global risk-related event. Figure (2b) illustrates this
more clearly by plotting the average five-year CDS spread for the sample of 38 countries, which
we have discussed in the previous section. Indeed, the average spread quickly rises each time the
global economy suffers a major shock, such as changes in U.S. monetary policy following falling
growth rates, which have led traders to unwind their carry trades, major corporate bankruptcies
of the usual suspects Bear Stearns and Lehman Brothers, or when the European governments
underwent major political bailouts. Of course, it might be that the risk premium embedded in
spreads, compensating investors for unpredictable variations in future default rates, is driven by
one set of determinants, while expected losses are a function of others. But even for this point, the
literature is at odds. It might be useful to review the papers who have analyzed this question.
Sovereign CDS spreads comove significantly over time, as can be seen in graph (2a) for the 38
countries studied previously. Performing a principal component analysis on a sample of sovereign
CDS spreads generally reveals that the first principal component explains a very large fraction of
the variation, much higher than is the case for equities. The precise figure depends on sample
11
frequency and countries studied, but can be as high as 96% at the daily level or 64% at a monthly
decision interval as shown by Pan and Singleton (2008) and Longstaff et al. (2010) respectively19 .
This strong common component motivates the holy grail journey of what this common factor ought
to be.
Among those studies leaning more towards an explanation of global factors is for example
Longstaff et al. (2010). The authors carefully study monthly 5-year CDS of 26 countries over the
time period October 2000 to January 2010 and consider variables related to the local economy,
global financial variables, global risk premia, global investment flow measures and regional spreads.
They conclude that sovereign CDS can be explained to a large extent by U.S. equity, volatility,
and bond market risk premia. A tight link between sovereign CDS spreads is also reported by Pan
and Singleton (2008), who study the term-structure of daily CDS spreads of Korea, Mexico and
Turkey from March 2001 to August 2006. Their main interest lies in disentangling the risk neutral
default and recovery processes implicit in the term structure. With this goal in mind, they develop
a theoretical pricing model to decompose the spread into the part related to expected losses, and
the risk premium. Risk premia appear to comove strongly over time and are strongly related to
the CBOE VIX option volatility index, the spread between the 10-year return on US BB-rated
industrial corporate bonds and the 6-month US Treasury bill rate, and the volatility in the owncurrency options market. These findings corroborate the intuition of time-varying risk premia in
the sovereign CDS market.
Ang and Longstaff (2011) conduct a comparative analysis of CDS spreads written on sovereign
states in the United States and European countries. Rather than extracting the risk premium, they
extract the dependence of spreads on a common component and define this part as systemic risk.
The authors find evidence that this systemic risk component is influenced by global financial factors
and dismiss links with macroeconomic fundamentals. Related to the paper by Pan and Singleton
(2008) is the work of Zhang (2003), who develops a CDS pricing model and applies it to Argentina
to infer differences in expectations about expected recovery rates upon default and expected default
probabilities. In addition, the author also documents that the wedge between risk-adjusted and
historical default probabilities are associated with changes in the business cycle, both the U.S. and
19
Pan and Singleton (2008) study 3 countries. Augustin and Tédongap (2011) report that the first principal
component explains 78% of the daily variation for a set of 38 countries.
12
Argentine credit conditions as well as the overall local economy. Augustin and Tédongap (2011)
provide empirical evidence that the first two common components among a set of 38 countries
spanning a wide geographical region are strongly associated with expected consumption growth
and macroeconomic uncertainty in the United States. The authors study daily quotes over the time
period April 2003 to August 2010 and document that this link is robust against the influence of
global financial market variables such as the CBOE volatility index or the Variance Risk Premium.
In addition, they rationalize this new finding in a structural asset pricing model with recursive
preferences and a long-run risk economy where the default rate process is driven by the global longrun expectations of future consumption growth and macroeconomic uncertainty. Further evidence
of influence from the United States on sovereign CDS premia is provided by Dooley and Hutchison
(2009), who document how the subprime crisis channeled through to a sample of 14 geographically
dispersed countries, based on a series of both positive and negative real and financial news over
the sample period January 2007 to February 2009. While they document that a range of news
had statistically and economically significant effects, they find that in particular the Lehman event
and the expansion of Federal Reserve swap lines with the central banks of industrial and emerging
countries uniformly moved all country spreads (in the sample)20 . Finally, Wang and Moore (2012)
study the dynamic correlations between the sovereign CDS spreads of 38 emerging and developed
economies with the U.S. from January 2007 to December 2009. Results indicate that in particular
developed economies have become much more integrated with the United States since the Lehman
shock. This tighter link with the U.S. seems to be driven mainly by the U.S. interest rate channel.
A good transition between the ”pro-global” and ”pro-local” camps is provided by Remolona
et al. (2008). Based on the intuition that financial asset prices are driven by country-specific fundamentals and investor’s appetite for risk, Remolona et al. (2008) decompose monthly 5-year emerging
markets sovereign CDS spreads over the period January 2002 to May 2006 into a Market-based
measure of expected loss and a risk premium to analyze how each of these two elements is linked
to measures of country-specific risk and measure of global risk aversion/risk appetite. Fundamental variables include inflation, industrial production, GDP growth consensus forecasts, and foreign
20
On 13 October 2008, the Fed removed its USD swap limits to industrial countries, and on 29 October 2008, the
FOMC established swap lines with the central banks of Brazil, Mexico, Korea and Singapore for up to $30 billion
each.
13
exchange reserves. Proxies for global risk aversion are taken as the Tsatsaronis, Karamptatos
(2003) effective risk appetite indicator, the VIX and a Risk Tolerance Index by JP Morgan Chase.
They find empirical evidence that global risk aversion is the dominant determinant of sovereign
risk premium, while country-specific fundamentals and market liquidity matter more for sovereign
risk. Both components thus behave differently. Carlos Caceres and Segoviano (2010) analyze how
much of the rising spreads can be explained by shifts in global risk aversion, country-specific risks,
directly from worsening fundamentals, or indirectly from spillovers originating in other sovereigns.
The authors argue that the widening of spreads during the early period of the crisis was essentially
driven by changes in a self-computed measure of risk aversion, but later in the crisis, countryspecific factors identified by each countrys stock of public debt and budget deficit as a share of
GDP played the dominant role.
Carr and Wu (2007) present a joint valuation framework for sovereign CDS and currency options
written on the same economy and execute an empirical test for Mexico and Brazil over the time
period January 2, 2002 to March 2, 2005. They document that CDS spreads show strong positive
contemporaneous correlations with both the currency option implied volatility and the slope of the
implied volatility curve in moneyness21 . This analysis confirms their conjecture that economic or
political instability often leads to both worsened sovereign credit quality and aggravated currency
return volatility in a sovereign country. Interestingly, their results suggest that there are additional
systematic movements in the credit spreads that the estimated model fails to capture. Hui and Fong
(2011) reverses the analysis and documents information flow from the sovereign CDS market to the
dollar-yen currency option market during the sovereign debt crisis from September 2009 to August
2011. in a similar spirit, Pu (2012) exploits pricing differences between USD and EUR denominated
sovereign CDS spreads to show that the dual-currency spread significantly affects the bilateral
exchange rate for ten Eurozone countries from January 2008 to December 2010 and has predictive
power up to a perdio of ten days. Gray and Bodie (2007) apply the contingent claims analysis to
price sovereign credit risk by using the government’s balance sheet and exchange rate volatility as
inputs to their model and compare the results to observed CDS spreads. More recently, Plank (2010)
also suggest a structural model for sovereign credit risk based on macroeconomic fundamentals. In
21
More specifically, the authors refer to the delta-neutral straddle implied volatilities and the risk reversals.
14
this model, the probability of default reflected in the CDS spread depends on a country’s foreign
exchange reserves, as well as its exports and imports. Ismailescu and Kazemi (2010) consider
the possibility of rating changes affecting the sovereign CDS spreads of 22 emerging economies
over the time period January 2, 2001 to April 22, 200922 . Using classic event study analysis, the
authors conclude that the CDS of investment grade countries respond mainly to negative credit
rating announcements, while the spreads of speculative grade countries respond largely to positive
announcements. In addition, the information content of negative credit events is anticipated and
already reflected in CDS spreads by the time the credit rating change is announced. Alternatively,
CDS spreads do not fully anticipate upcoming positive credit rating events, which seem to contain
new information. Furthermore, the paper documents evidence of spillover effects arising mainly
from rating upgrades, with a one notch rise in the rating of an event country increasing the average
CDS spread of a non-event country by 1.18%. These effects are asymmetric, high grade countries
reacting stronger to rating upgrades and low-grade countries reacting stronger to rating downgrades.
Although controlling for previous rating downgrades reduces the magnitude of the spillover effect,
it is increased if countries have common a common creditor, which seems to be one channel of
transmission. However, competition effects arise if two countries have similar levels of trade flow
correlation with the U.S.
There is also a class papers documenting a link between sovereign credit risk implicit in the
CDS price and the health of the financial system. One such example is Acharya et al. (2011), who
illustrate the two-way feedback effect between sovereign risk and the financial sector, where government bailouts negatively impact the financial strength of the sovereign, which in turn reduces
the value of the government guarantees implicit in the financial institutions. As a consequence, the
CDS spreads of sovereign countries and financial companies co-move strongly once the government
has committed excessively to financial guarantees. Altman and Rijken (2011) propose a ”bottomup” approach to evaluate sovereign default probabilities by adapting the well-know credit-scoring
method to countries. This suggests that the local real economy significantly impacts default risk.
The outcomes are compared to implied CDS rankings provided by the financial market. Dieckmann and Plank (2011) study eighteen advanced western economies on a weekly basis during the
22
For early evidence on the relationship between credit ratings and pricing information, see Cossin and Jung (2005),
who show that ratings become more informative after a crisis event.
15
2007/2008 financial crisis23 . They find that both the state of a country’s financial system and of
the world financial system have strong explanatory power for the behavior of CDS spreads, while
the magnitude of this impact seems to depend on the importance of a country’s financial system
pre-crisis. In addition, they provide evidence that Economic and Monetary Union member countries exhibit higher sensitivities to the health of the financial system. While Acharya et al. (2011)
stress the two-way feedback effect, the authors here underscore the unilateral private-to-public risk
transfer through which market participants incorporate their expectations about financial industry
bailouts. Evidence that contingent liabilities of sovereigns arising from implicit or explicit guarantees of the banking system influence sovereign CDS premia is also provided in Kallestrup et al.
(2011). Also Kallestrup (2011) shows that there are strong interactions between sovereign credit
risk and macrofinancial risk indicators calculated based on bank balance sheet variables. Sgherri
and Zoli (2009) conduct an analysis of the co-movement between sovereign CDS and those of the
financial sector for ten European countries from January 1999 to April 2009. While they confirm
the dominance of a common time-varying factor, they find that concerns about the solvency of the
national banking systems have become increasingly important. Finally, Brutti and Sauré (2012)
show how financial shocks to Greece spill over to eleven other European economies. The magnitude
of contagion depends on the cross-country bank exposures to sovereign debt. All else being equal,
the transmission rate to the country with the greatest exposure to Greece (1.22 percent of GDP)
has been roughly 46 percent higher than the rate to the country with the least exposure (0.08
percent of GDP).
6
24 .
The relationship between sovereign CDS and bonds and Frictions
Duffie (1999) illustrates theoretically that the spread on a par floating rate note over a risk-free
benchmark should essentially be equal to the CDS spread25 . As such, the CDS-Bond basis, defined
as the difference between the CDS and bond spread, should be zero. However, empirically observed
23
The exact time period is January 2007 to 2010.
Further evidence on the relationship between sovereign and bank CDS is provided by Ejsing and Lemke (2011),
Alter and Schuler (2011) and Aktug and Vasconcellos (2010)
25
Other references are Lando (2004) and Hull and White (2000a).
24
16
imperfections of this theoretical arbitrage relationship has led researches to investigate whether the
CDS market is more informationally efficient than the cash bond market and whether it reflects
new information more quickly, or vice-versa. Alternatively, people have looked at the determinants
of the basis, which relates explicitly to limits of arbitrage and frictions in one of the two markets,
or both.
Regarding price discovery for the sovereign asset class, there has recently been a proliferation of
papers addressing the question of which market is more efficient because of the turbulences in the
sovereign debt market. Abstracting from technical details, these papers generally have the common
approach to test whether the two markets are co-integrated, that is whether they are characterized
by a long-run stationary relationship, and then to look at short-term deviations from equilibrium to
verify which market adapts to the other one. This is executed using the techniques of vector error
correction modeling, computing Hasbroucks’s and Gonazalo-Granger’s information measures, or by
analyzing statistical causality using Granger’s method. While for corporate reference entities, there
is more or less consent about the CDS market being more efficient, results for sovereign reference
entities are very mixed and ambiguous. These discrepancies are certainly related to the different
samples, time periods, sampling frequency and data sources. While some conclude in favor of the
bond market and others in favor of the CDS market, our interpretation of the literature is that
there is increasing price discovery in the credit derivative market. Arce et al. (2011) make the
sensible point that price discovery is state dependent and related to the relative liquidity in both
markets. As the CDS market has become more mature over time, this would also explain why its
relative informational efficiency has increased. A similar argument is also made in Ammer and Cai
(2011), who provide some evidence that the relatively more liquid market tends to lead the other,
as their measure of relative CDS price leadership is correlated positively with the ratio of the bond
bid-ask spread to the CDS bid-ask spread and negatively with the number of bonds outstanding.
Our further overall interpretation is that a positive basis is usually more persistent, due to the
difficulty of shorting bonds. Moreover, except for some minor exceptions, almost all studies focus
exclusively on the most liquid 5-year segment of the market. Without spending too much time on
the details, we provide a shopping list of the references in the following paragraph.
Palladini and Portes (2011) study 6 developed economies in Europe from 30 January 2004
17
through 11 March 2011 using CMA quote data and conclude that CDS prices lead in the price
discovery process26 . A similar conclusion is drawn by Varga (2009), who takes a special look at
Hungary from 6 February 2004 to 18 June 2008 using the CMA database27 . Using likewise the
CMA data, O’Kane (2012) studies cross-correlations and performs Granger causality tests for 6
European economies from 1 January 2008 to 1 September 201128 . He finds that for Greece and
Spain, Granger causality flows from the CDS to to the bond market, wile the results is reversed
for Italy and France. For Portugal and Ireland, there seems to be two-way causality. Studying 18
sovereign developed and emerging economies from January 2007 to March 2010, Coudert and Gex
(2011) conclude that bonds seem to lead for developed European economies, while the derivative
market wins the horse race in emerging economies29 . The authors also put forth a differential
liquidity argument and argue that the role of CDS has intensified during the crisis, as bond market
players with a bearish view will simply stay out of the markets, while CDS buyers will stay in
and purchase protection. This contrasts with Arce et al. (2011), who argue that the role of bond
markets has increased for European Economies during the financial crisis. They study 11 European
economies from January 2004 to September 2010 using CMA data30 . Fontana and Scheicher (2010)
look at 10 Euro area countries from January 2006 to June 2010 using again CMA data31 . In contrast
to most other studies, which use the 5-year spread, the authors study 10-year spreads. In line with
the previous mixed results, in half of the sample countries, price discovery takes place in the CDS
market and in the other half, price discovery is observed in the bond market. Further evidence is
provided by M. Kabir Hassan and Suk-Yu (2011), who study 7 emerging economies from January
2004 to October 2009 using data from JP Morgan Chase32 . They find that bond markets lead
for Brazil and the Philippines, CDS markets lead for Argentina, and there is no clear dominance
for Mexico, China, and South Africa. The additional contribution is to link pricing efficiency to
26
The 6 countries are Austria, Belgium, Greece, Ireland, Italy, and Portugal.
While Varga focuses on Hungary, some additional analysis is also provided for 14 emerging countries from 3
January 2005 to 30 May 2008. These countries are Brazil, Bulgaria, Czech Republic, South Africa, Estonia, Croatia,
Poland, Latvia, Lituania, Russia, Romania, Slovkia, Turkey and Ukraine.
28
The 6 countries are France, Greece, Italy, Ireland, Portugal and Spain.
29
See also Coudert and Gex (2010). The sample includes Austria, Belgium, Denmark, Finland, France, Netherlands,
Greece, Ireland, Italy, Portugal and Spain for the advanced European countries; Argentina, Brazil, Mexico, Lithuania,
Poland, Turkey and Philippines for the emerging countries; in addition the authors study 17 financial companies.
30
The 11 EMU countries are Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, the Netherlands,
Portugal and Spain.
31
Austria, Belgium, France, Germany, Greece, Ireland, Italy, Netherlands, Portugal and Spain
32
The countries are Argentina, Brazil, China, Colombia, Mexico, Philippines and South Africa.
27
18
financial integration. It is argued that sovereigns, for which bond markets lead the price discovery
process, are more financially integrated with the world economy.
Li and Huang (2011) witness an increasing contribution of sovereign CDS rates to credit risk
discovery based on a sample of 22 emerging countries using Thomson reuters data from 1 January
2004 to 31 July 200833 . C. Emre Alper and Gerard (2012), studying CDS premia in relation to
asset swap spreads, conclude in favor of price discovery in the derivative market given a sample
of 21 advanced economies from January 2008 to January 201134 . Their data comes both from
Datastream and Markit. Aktug and Bae (2009) study 30 emerging markets at a monthly sampling
frequency from January 2001 to November 2007 using data from Markit35 . They show that bond
markets lead CDS markets in general, but that they lag CDS spreads in some cases, and that the
markets have become more integrated over time. Support for the bond markets is also found in
Ammer and Cai (2011). The latter investigate 9 emerging economies from 26 February 2001 to 31
March 2005 using Markit data36 . Overall, they find that the bond market leads the CDS market
more often. In an early paper, Chan-Lau and Kim (2004) have a hard time concluding in favor
of any market using CreditTrade data for 8 emerging economies from 19 March 2001 to 29 may
2003.37 . Delis and Mylonidis (2011) look at Greece, Italy, Portugal and Spain using CMA data
from 9 July 2004 to 25 May 25 2010. Based on 10-year spreads, they argue that CDS almost
uniformly Granger cause bond spreads. Feedback causality is, however, detected during times of
intense financial and economic turbulences. Finally, additional elements of analysis can be found in
In et al. (2007), Boone et al. (2010), Li (2009), Bowe (2009), Carboni (2010) and Adler and Song
(2010).
As a conclusion, results are thus very inconclusive, but differential liquidity seems to matter.
Based on this intuition, Calice et al. (2011) investigate liquidity spillovers between the two markets
33
The sample is composed of Argentina, Brazil, Chile, China, Colombia, Czech Republic, Egypt, Hungary, Indonesia, Israel, Korea, Malaysia, Mexico, Morocco, Pakistan, Peru, Philippines, Poland, Russia, South Africa, Turkey,
Venezuela.
34
The 22 countries are Australia, Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy,
Japan, the Netherlands, Norway, Portugal, Spain, Sweden, the UK and the US.
35
They study Brazil, Bulgaria, Chile, China, Colombia, Croatia, Czech Republic, Dominican Republic, Ecuador,
Egypt, El Salvador, Hungary, Korea, Lebanon, Malaysia, Mexico, Morocco, Pakistan, Panama, Peru, Philippines,
Poland, Russia, South Africa, Thailand, Turkey, Ukraine, Venezuela and Vietnam.
36
Brazil, China, Colombia, Mexico, the Philippines, Russia, Turkey, Uruguay, and Venezuela.
37
The following countries are reviewed: Brazil, Bulgaria, Colombia, Mexico, the Philippines, Russia, Turkey, and
Venezuela.
19
and find substantial variation in the patterns of the transmission effect between maturities and
across countries. For several countries, including Greece, Ireland and Portugal the liquidity of the
sovereign CDS market has a substantial time varying influence on sovereign bond credit spreads. In
et al. (2007) study the volatility transmission among the two markets across emerging economies.
While the dynamics between the cash and derivative market raise interesting questions, another
angle to analyze is the determinants of the short-term deviations from the equilibrium relationship.
Differential liquidity may naturally explain the seemingly arbitrage opportunity. Early studies on
this topic assumed that the credit derivative market was perfectly liquid, as the underlying traded
instruments are by nature simple contractual agreements, in contrast to the cash market, where
hard money is actually exchanged when the asset is purchased. In this spirit, Longstaff et al. (2005)
postulated CDS spreads to be pure indicators of default risk in order to infer liquidity characteristics
of the bond market. Academic work on liquidity effects in the CDS market is very young, even
for corporate reference contracts, but the literature is rapidly growing. Two papers that study
liquidity characteristics in the synthetic credit derivative market for corporate reference entities are
for example Tang and Yan (2007) and Bongaerts et al. (2011). The former show that both liquidity
characteristics and liquidity risk explain about 20% of the time series variation in CDS spreads,
while the latter provide strong evidence for an expected liquidity premium earned by the protection
seller. The premium is however economically small. For sovereign CDS, Pan and Singleton (2008),
who study the term structure of recovery rates of Mexico, Brazil and Turkey, report anecdotal
evidence of liquidity effects from discussion with practitioners. Regarding the determinants of the
CDS-bond basis, a liquidity story is supported by Arce et al. (2011), who show that differential
liquidity between the bond and the CDS partially explains the difference. Specifically, differential
liquidity is proxied by the ratio of the bid-ask to the mid spread in the CDS market divided by
the bid-ask to the mid spread in the bond market. In addition, the authors show evidence for the
role of counterparty risk, which is the risk that a CDS counterparty may not be able to honor its
obligations from the insurance contract. This risk becomes especially important when the default
of the counterparty occurs at the same time as the default event of the underlying reference entity.
Arora et al. (2012) show that counterparty risk is significantly priced in the corporate market, but
that it has economically a small magnitude. On average, counterparty credit risk needs to increase
20
by 646 basis points to reduce the insurance price by 1 basis point. This small effect is likely the
result of collateralization (and maybe overcollateralization) agreements. In a similar spirit, Levy
(2009) also links the pricing discrepancies to both counterparty risk and liquidity. Kucuk (2010)
finds that the basis is strongly associated with various bond and CDS liquidity variables such as
(for bonds and CDS) the average sample bid-ask spread and percentage bid-ask spread, and (for
bonds) lagged trading volume, notional amount outstanding, age and time to maturity. Similar to
the debate on global versus local factors for sovereign CDS spreads, Fontana and Scheicher (2010)
relate the basis mainly to common factors.
Another potential source of friction, which can create a wedge between the CDS and bond
spread is the Cheapest-to-Deliver option embedded in CDS contracts. As we have already explained, upon the occurrence of a credit event, the holder of the insurance contract has the right
to deliver the cheapest among a set of defaulted debt obligations to his counterparty in return
for the insurance settlement. This option makes the contract riskier from the perspective of the
CDS seller, and should be relatively more important, the closer a country is to default. Ammer
and Cai (2011) provide a treatment of the sovereign case, while Jankowitsch et al. (2008) provide
positive evidence for corporate CDS. For sovereigns, some early analysis is also available in Singh
(2003). Fisher (2010) studies theoretically how heterogenuous investors with different beliefs about
the probability of default of a sovereign borrower affects the sovereign CDS-Bond basis. The results of the model depend on the assumption that a fixed proportion of investors is unable to sell
CDS. The latter constraint becomes binding in bad times, when a relatively bigger proportion of
investors is pessimistic and wants to sell of credit risk (buying CDS). But as the supply is limited,
this constraint creates a wedge between the bond and CDS prices and induces a positive basis.
Thus, the time-varying equilibrium basis depends on the time-varying proportion of pessimistic
investors. A second mechanism, that induces the (positive basis) in his model is the prospect of
lending fees from the bonds in bad times. As supply in the CDS market is restricted, some pessimistic investors, who want to take a short position in sovereign credit risk need to short-sell the
physical bond. This allows those, who hold on to the bond to charge higher fees by lending it out
to short-sellers. This fee raises bond prices, lowers spreads, and creates a positive basis. Finally,
Adler and Song (2010) provide evidence that short-selling costs were partly responsible for persist-
21
ing positive bases for sovereign reference entities. In addition, the authors provide a more general
theoretical pricing framework for the basis, which corrects for biases arising from acrrued spread
or coupon payments and bond prices away from their par value38 . In fact, the model illustrates
how accrued payments and bond prices below (above) par may mechanically introduce a negative
and positive (negative) basis respectively. It is exactly this latter effect, which may explain the
well documented basis smile. The average credit quality of a portfolio of bonds depends on the
statistical distribution of historical rating transitions. Hence, the joint consideration of price and
credit quality will determine the shape of the basis smile, which could very well be a ”frown”.
Another ”hot” topic about sovereign CDS is whether speculators manage to (negatively) impact
the market. In May 2010, BaFin, the German financial regulator went ahead with an outright
ban on naked sovereign CDS (and bond) positions, despite an official report failing to provide
conclusive evidence (See Criado et al. (2010).). The academic opinions don’t seem to agree on
this idea. Richard Portes (see Portes (2010)) has advocated a ban, arguing that naked CDS
buying (equivalent to shorting the bond) artificially drives up prices, thereby increasing a country’s
borrowing costs. This opinion seems to be shared in a study by Delatte et al. (2012). Darrell
Duffie on the other hand (Duffie (April 29 2010a, 2010b)) dismisses the argument and shows some
empirical evidence that there is no such evidence. He argues that a ban on naked sovereign CDS
trading may instead increase execution costs and lower the quality of price information. In contrast,
he argues that holders of both the derivative and the underlying cash bond have in fact no more
incentive to monitor the borrower, which may induce the borrower to invest less efficiently and
thus increases the probability of default. Bolton and Oehmke (2011) study this ”empty creditor”
problem theoretically for corporate CDS, while Subrahmanyam et al. (2012) provide empirical
evidence39 . An early theoretical treatment of the effect of the existence of CDS on the incentives of
sovereign borrowers was also made available by Goderis and Wagner (2011). More specifically, the
authors show that without insurance, the sovereign country has no incentive to optimally invest (or
exert effort) as there is the possibility of offering a debt restructuring in the bad states of the world,
38
The pricing framework considered in Adler and Song (2010) is a useful extension to the treatment in Duffie
(1999).
39
The concept of empty creditor refers to the separation of cash flow and voting rights through the use of credit
derivatives. A debtholder, whose exposure is insured with a CDS, keeps an economic interest in the firm’s claims,
but has no more risk alignment with other bondholders, who enjoy no such proection.
22
which the lender cannot credibly commit to reject. When the lender can purchase insurance on
the bond, the threat of rejecting the restructuring offer becomes credible and the debtor will have
to internalize more of the costs in the bad states. This provides incentive to invest more efficiently
ex-ante, which in turn reduces default. Strategic purchase by a (single) bondholder to increase
the bargaining power in debt renegotiation deal is thus beneficial from an ex-ante perspective.
However, in the presence of many bondholders, there may be coordination failures. This may
lead to a situation, where lenders buy more insurance than is socially optimal, which leads to an
increased probability of default.
Sambalaibat (2011) studies the issue of naked CDS theoretically and shows that the effect depends on the infrastructure of the insurance market. The overall outcome depends on the parameter
values, and naked CDS buyers may either induce over-investment, leading to lower borrowing costs
or to under-investment with higher borrowing costs. For a legal treatment of how credit derivatives
may affect the sovereign debt restructuring process, see Verdier (2004).
7
Trading in the Sovereign CDS Market
Very little is known about the trading patterns in the sovereign CDS market. Based on a threemonth sample of detailed transaction-level data over the time period May 1 to July 31 2010 generously received from a set of dealers, the Federal Reserve Bank of New York spoiled us a with a more
detailed picture of the CDS trading behavior in their staff report No. 51740 . Keeping in mind the
caveat that the data set stems from a period, where overall trading activity was below historical
average, the results indicate that CDS trade not that often despite their highly standardized nature.
For the 29,146 single name sovereign CDS transactions recorded for 74 reference entities, the most
actively traded reference entities traded on average 30 times per day, the less frequently traded 15
times per day on average and the infrequently traded reference entities traded only 2 times per day
on average. Another way to say this, is that a long tail of reference entities trade less than once a
day. In terms of size, the USD denominated trades were mostly traded in 5 million USD tickets,
with a median and mean trade of 10 and 16.74 million USD respectively. For EUR denominated
trades (which amounted to only 574 transactions), the mode was 10 million USD, with a median
40
See Chen et al. (2011).
23
trade of 5 million and a mean trade of 12.53 million. For matters of comparison, sovereign single
name CDS trades were on average twice as large as their corporate counterparts.
Although the number of trades seems rather low, market participation for sovereign reference
entities was rather active. On average, 50 market participants are reported to have traded at least
once a day, 200 traded at least once a week and 340 traded at least once a month. In addition,
dealers were more likely to be sellers of protection and the four most active dealers were involved
in 45% of the buying and selling of all CDS trades and in 50% of the overall notional amount.
Yet, according to regulatory definitions of the Department of Justice, the Herfindahl Hirshman
concentration index with levels ranging between 885 and 965 is considered to be low41 . Finally, the
market is confirmed to be highly standardized, as 92% of all single-name CDS contracts within the
sample had a fixed coupon and 97% had fixed quarterly payment dates42 .
Another more exhaustive source of information on CDS trading behavior is the the Depository Trust and Clearing Corporation (DTCC). In October 2008, the DTCC Trade Information
Warehouse started to publish detailed weekly reports on stocks and volumes in CDS trading. More
specifically, current and historical positions are shown on an aggregate level and for the 1,000 mostly
traded reference contracts. This move towards higher transparency is greatly welcome. Oehmke
and Zawadowski (2011) have used this data to study the determinants of corporate trading and
find that higher amounts of (corporate) CDS outstanding are associated with companies, which
have more assets on their balance sheet, more bonds outstanding, that are investment grade firms
or firms that have just recently lost their investment grade status, as well as with higher analysts’
forecast dispersion. Duffie (April 29 2010a) illustrates a few statistics on sovereign CDS from this
database in his testimony to the United House of Representatives, but we are not aware of any
formal treatment at this stage. Here, we provide a summarized overview of the main statistical
facts for government credit derivatives.
Tables (10) and (11) illustrate the average gross and net notional amounts outstanding in (mil41
Note that Giglio (2011) emphasizes the high concentration of the CDS market, thereby referring to industry
reports, which state that in 2006, the top 10 counterparties (all broker/dealers) accounted for about 89% of the total
protection sold. See footnote 8.
42
Using transactions and quote data from CreditTrade on 77 sovereign reference entities from January 1997 to
June 2003, Packer and Suthiphongchai (2003) emphasized the low trading volume by indicating that (in 2002), only
6% of all quotes resulted in transactions, with a strong concentration in a few reference entities. The top five names
accounted for more than 40% of all quotes. These countries were Brazil, Mexico, Japan, the Philippines and South
Africa.
24
lion) USD equivalents on all sovereign CDS contracts (excluding sovereign states in the US) among
the 1,000 mostly traded contracts over the time period 31 October 2008 through 11 November
201143 . In addition, there is information on the average number of contracts over that period, the
ratios of gross to net notional amount outstanding and the net notional amount outstanding to the
number of contracts. The countries are classified into five geographical regions following the practice of Markit: Americas, Asia Ex-Japan, Australia and New Zealand, Europe-Middle East-Africa
(EMEA) and Japan. While the weekly total gross notional volume recorded for sovereign countries
in the data repository is approximately 2.1 trillion USD, the net exposure, which accounts for offsetting effects between buyers and sellers and represents therefore a more economically meaningful
measure, reflects with 203.6 billion USD about 9.86% of that amount. Keeping in mind that this
number does not report the values for sovereign US states and other non-government supranational
bodies, this amount represents approximately 75% of the sovereign single name CDS outstanding
as reported by BIS in the first semester of 2011 (see Table (4)).
The total number over all country-specific averages of traded contracts is 147,258, the mean
ratio of gross to net notional amount is 10.74 and on average, the net credit exposure per contract
is 1.8 million USD. There is however a significant cross-sectional dispersion. Further analysis also
reveals that four of the five GIIPS countries rank among the top ten countries with the highest
amount of net notional exposure. Another striking feature we observe is that emerging economies
tend to have higher numbers of traded contracts, but smaller net exposures per contract on average,
while developed economies trade in bigger bulks, but have less contracts outstanding. The world’s
two biggest reference bond markets, the US and Germany, top the list with 5.16 and 6.66 million
USD per contract respectively. At the bottom of the list are the Philippines with an average of
360,000 USD per traded contract. Moreover, the top ten list of countries with the highest amount of
gross volume outstanding is dominated by emerging market economies, while developed distressed
economies cluster in the top ten list with the highest net notional exposure.
Tables (12) and (13) provide similar statistics at a yearly frequency since the data has been
made available by DTCC. From these numbers, it is clear to see that average weekly gross and net
volumes, as well as ratios of gross to net exposure have fluctuated only modestly, with the ratio
43
Tables (10) and (11) provide essentially the same information, but Table (10) provides a sort in descending order
for each column to help the exposition. The reader may find one or the other table more reader-friendly.
25
of gross to net exposure being roughly equal to a factor of 10 over time. The average number of
contracts reported in the warehouse has changed slightly more over time, which explains the bigger
fluctuations in the net notional exposure per contract, which has come down from 3.6 million USD
in 2008 to a mean 1.4 million in 2011.
Figures (3) and (4) provide some more intuition about how the CDS trading is related to
individual government characteristics for EMEA and Asian countries (ex-Japan). For these regions,
it is clear that within each year, there seems to be a statistical relationship between the net notional
amount of CDS outstanding and the nominal amount of government debt. This relationship appears
much weaker if debt levels are compared against gross exposures. Duffie (April 29 2010a, 2010b)
showed empirical evidence that the volume of CDS is unrelated to the level of spreads. This raises
once more interesting questions. We have seen that government characteristics and CDS volumes
are unrelated to to spreads, but characteristics are linked to volumes.
Table (14) separately reports statistics for the United States government and the sovereign US
states, which are registered in the DTCC database. In contrast to the statistics on other sovereign
governments, all yearly average values for CDS on US Treasuries exhibit a sharp increase in 2011.
In fact, the mean gross notional amount jumps by 90% from 2011 to 2011, going from 13.2 to
25 billion USD. That’s a huge surge. The same observation holds for net exposure, which rises
by 84% over the same time period, going from 2.4 to 4.5 billion USD, while the number of live
contracts increases from 479 million to approximately 1.1 billion contracts, which corresponds to
an increase of 120%. These spikes raise the provoking thought that investors may actually have
seriously considered the possibility of US government default during the lock-out period in summer
2011. In particular, as no such sharp changes are recorded for the individual states. There the
evidence is mixed. Gross exposures and number of traded contracts unilaterally increased, while
net notional exposure increased mainly for Illinois, while it dropped for instance for New York City,
Texas or Florida.
26
8
Conclusion
Sovereign credit risk is undoubtedly a topic of crucial importance. The supranational bailouts of
Greece, Ireland, Portugal and the vast amount of European Union summits ever since the start
of the sovereign debt crisis eliminate any doubt about the relevance of this topic44 . As we speak,
trying to avoid information on sovereign debt issues is like walking through a mine field in North
Korea.
This paper tries to provide a comprehensive overview of the young, but steadily growing literature on sovereign CDS, describes key statistical facts and features of prices, about the market,
its players and related trading activities and attempts to raise some thought-provoking questions.
We thereby hope to provide a good starting point for anyone interested in the topic. We had fun
writing this document, we hope you enjoyed it too.
44
We stopped counting at 16.
27
References
Acharya, V. V., Drechsler, I. and Schnabl, P. (2011). A pyrrhic victory? - bank bailouts and
sovereign credit risk, NBER Working Paper 17136 .
Adler, M. and Song, J. (2010). The behavior of emerging market sovereigns’ credit default swap
premiums and bond yield spreads, International Journal of Finance & Economics 15(1): 31–58.
Aktug, E. and Vasconcellos, G. (2010). Is sovereign risk related to the banking sector?, Working
paper, Lehigh University .
Aktug, Erdem, G. V. and Bae, Y. (2009). The dynamics of sovereign credit default swap and bond
markets: Empirical evidence from the 2001-2007 period, WP .
Alter, A. and Schuler, Y. S. (2011). Credit spread interdependencies of european states and banks
during the financial crisis, SSRN eLibrary .
Altman, E. I. and Rijken, H. A. (2011). Toward a bottom-up approach to assessing sovereign
default risk, Journal of Applied Corporate Finance 23(1): 20–31.
Ammer, J. and Cai, F. (2011). Sovereign cds and bond pricing dynamics in emerging markets: Does
the cheapest-to-deliver option matter?, Journal of International Financial Markets, Institutions
and Money 21(3): 369 – 387.
Ang, A. and Longstaff, F. A. (2011). Systemic sovereign credit risk: Lessons from the u.s. and
europe, Working Paper 16982, National Bureau of Economic Research.
Arce, O., Mayordomo, S. and Pena, J. I. (2011). Do sovereign cds and bond markets share the
same information to price credit risk? an empirical application to the european monetary union
case, Working Paper .
Arora, N., Gandhi, P. and Longstaff, F. A. (2012). Counterparty credit risk and the credit default
swap market, Journal of Financial Economics 103(2): 280 – 293.
Augustin, P. and Tédongap, R. (2011). Sovereign credit risk: Consumption strikes again, Working
Paper Stockholm School of Economics .
Blanco, R., Brennan, S. and Marsh, I. W. (2005). An empirical analysis of the dynamic relation
between investment-grade bonds and credit default swaps, The Journal of Finance 60(5): 2255–
2281.
Bolton, P. and Oehmke, M. (2011). Credit default swaps and the empty creditor problem, Review
of Financial Studies 24(8): 2617–2655.
Bongaerts, D., De Jong, F. and Driessen, J., J. (2011). Derivative pricing with liquidity risk: Theory
and evidence from the credit default swap market, The Journal of Finance 66(1): 203–240.
Boone, L., Fransolet, L. and Willemann, S. (2010). Euro public debt and the markets: sovereign
fundamentals and cds market dynamics, Banque de France Financial Stability Review - Derivatives, Financial innovation and stability No 14: Barclays Capital.
Bowe, M., K. A. T. A. (2009). Information transmission and price discovery in emerging sovereign
credit risk markets, Working Paper .
28
Brutti, F. and Sauré, P. (2012). Transmission of sovereign risk in the euro crisis, Working Paper .
C. Emre Alper, L. F. and Gerard, M. (2012). Pricing of sovereign credit risk: Evidence from
advanced economies during the financial crisis, IMF Working Paper 12/24 .
Calice, G., Chen, J. and Williams, J. M. (2011). Liquidity spillovers in sovereign bond and cds
markets: An analysis of the eurozone sovereign debt crisis, SSRN Working Paper 1898596 .
Carboni, A. (2010). The sovereign credit default swap market: price discovery, volumes and links
with banks’ risk premia, Working Paper Bank of Italy .
Carlos Caceres, V. G. and Segoviano, M. (2010). Sovereign spreads: Global risk aversion, contagion
or fundamentals?, IMF Working Paper 10/120 .
Carr, P. and Wu, L. (2007). Theory and evidence on the dynamic interactions between sovereign
credit default swaps and currency options, Journal of Banking & Finance 31(8): 2383–2403.
Chan-Lau, J. A. and Kim, Y. S. (2004). Equity prices, credit default swaps, and bond spreads in
emerging markets, IMF Working Paper 04/27 .
Chen, K., Fleming, M., Jackson, J. and Sarkar, A. (2011). An analysis of cds transactions: Implications for public reporting, Federal Reserve Bank of New York Staff Report No. 517 .
Chernov, M., Gorbenko, A. S. and Mak, I. (2011). Cds auctions, Working Paper London School of
Economics .
Cossin, D. and Jung, G. (2005). Do major financial crises provide information on sovereign risk to
the rest of the world? a look at credit default swap markets, FAME Research Paper N.134 .
Coudert, V. and Gex, M. (2010). Credit default swap and bond markets: which leads the other?,
Banque de France Financial Stability Review - Derivatives, Financial innovation and stability .
Coudert, V. and Gex, M. (2011). The interactions between the credit default swap and the bond
markets in financial turmoil, Centre d’Etudes Prospectives Et d’Informations Internationales Document de Travail 2011/02 .
Criado, S., Degabriel, L., Lewandowska, M., Linden, S. and Ritter, P. (2010). Sovereign cds report,
Official Report DG COMP, DG ECFIN and DG MARKT .
URL: http://www.ft.com/intl/cdsreport
Das, S. and Hanouna, P. (2006). Survey of the literature: Credit default swap spreads, Journal of
Investment Management 4: 93–105.
Delatte, A.-L., Gex, M. and Lpez-Villavicencio, A. (2012). Has the cds market influenced the
borrowing cost of european countries during the sovereign crisis?, Journal of International Money
and Finance 31(3): 481 – 497. ¡ce:title¿Financial Stress in the Eurozone¡/ce:title¿.
Delis, M. D. and Mylonidis, N. (2011). The chicken or the egg? a note on the dynamic interrelation
between government bond spreads and credit default swaps, Finance Research Letters 8(3): 163
– 170.
Dieckmann, S. and Plank, T. (2011). Default risk of advanced economies: An empirical analysis of
credit default swaps during the financial crisis, Review of Finance .
29
Dooley, M. and Hutchison, M. (2009). Transmission of the u.s. subprime crisis to emerging markets:
Evidence on the decoupling-recoupling hypothesis, Journal of International Money and Finance
28(8): 1331 – 1349.
Duffie, D. (1999). Credit swap valuation, Financial Analysts Journal 55(1): 73–87.
Duffie, D. (2010b). Is there a case for banning short speculation in sovereign bond markets?, Banque
de France Financial Stability Review - Derivatives, Financial innovation and stability No. 14.
Duffie, D. (April 29 2010a). Credit default swaps on government debt: potential implications of
the greek debt crisis, Testimony to the United States House of Representatives, Subcommittee on
Capital Markets, Insurance, and Government Sponsored Enterprises .
Ejsing, J. W. and Lemke, W. (2011). The janus-headed salvation sovereign and bank credit risk
premia during 2008-09, Economics Letters 110: 28–31.
Fisher, N.-F. (2010). Explaining sovereign bond-cds arbitrage violations during the financial crisis
2008-09, Working Paper London School of Economics .
Fontana, A. and Scheicher, M. (2010). An analysis of euro area sovereign cds and their relation
with government bonds, SSRN eLibrary .
Giglio, S. (2011). Credit default swap spreads and systemic financial risk, Working Paper Harvard
University .
Goderis, B. and Wagner, W. (2011). Credit derivatives and sovereign debt crises, Unpublished
Paper, Tilburg University. .
Gray, Dale . F., R. C. M. and Bodie, Z. (2007). Contingent claims approach to measuring and
managing sovereign credit risk, Journal of Investment Management 5 4: 5–28.
Gupta, S. and Sundaram, R. K. (2012). Cds auctions and informative biases in cds recovery rates,
Working Paper New York University .
Hart, O. and Zingales, L. (Fall 2011). A new capital regulation for large financial institutions,
American Law and Economics Review 13(2): 453–490.
Helwege, J., Maurer, S., Sarkar, A. and Wang, H., Y. (2009). Credit default swap auctions and
price discovery., Journal of Fixed Income 19(2): 34 – 42.
Hui, C.-H. and Fong, T. (2011). Information flow between sovereign cds and dollar-yen currency
option markets in the sovereign debt crisis of 2009-2011, Working Papers 402011, Hong Kong
Institute for Monetary Research.
Hull, J. C. and White, A. (2000a). Valuing credit default swaps i: No counterparty default risk,
SSRN eLibrary .
In, F., Kang, B. U. and Kim, T. S. (2007). Sovereign credit default swaps, sovereign debt and
volatility transmission across emerging markets, Working Paper .
Ismailescu, I. and Kazemi, H. (2010). The reaction of emerging market credit default swap spreads
to sovereign credit rating changes, Journal of Banking and Finance 34(12): 2861 – 2873.
30
Jankowitsch, R., Pullirsch, R. and Veza, T. (2008). The delivery option in credit default swaps,
Journal of Banking and Finance 32(7): 1269 – 1285.
Kallestrup, R. (2011). Bank balance sheets and the dynamics of bank and sovereign credit spreads,
Working Paper Copenhagen Business School, .
Kallestrup, R., Lando, D. and Murgoci, A. (2011). Financial sector linkages and the dynamics of
bank and sovereign credit spreads, Working Paper Copenhagen Business School, .
Kucuk, U. N. (2010). Non-default component of sovereign emerging market yield spreads and its
determinants: Evidence from the credit default swap market., Journal of Fixed Income 19(4): 44
– 66.
Lando, D. (2004). Credit Risk Modeling, Princeton University Press.
Lando, D. and Mortensen, A. (2005). Revisiting the slope of the credit spread curve, Journal of
Investment Management, Vol. 3, No. 4, Fourth Quarter 2005 3(4): 1–27.
Levy, A. (2009). The cds bond basis spread in emerging markets: Liquidity and counterparty risk
effects, Working Paper .
Li, N. (2009). The price discovery process in credit derivative market: evidence from sovereign cds
market, American Journal of Finance and Accounting 1(4): 393 – 407.
Li, N. and Huang, A. Y. (2011). Price discovery between sovereign credit default swaps and bond
yield spreads of emerging markets, Journal of Emerging Market Finance 10(2): 197–225.
Longstaff, F. A., Mithal, S. and Neis, E. (2005). Corporate yield spreads: Default risk or liquidity?
new evidence from the credit default swap market, The Journal of Finance 60(5): 2213–2253.
Longstaff, F. A., Pan, J., Pedersen, L. H. and Singleton, K. J. (2010). How sovereign is sovereign
credit risk?, American Economic Journal: Macroeconomics (13658).
M. Kabir Hassan, G. M. N. and Suk-Yu, J. (2011). Credit default swaps and sovereign debt markets,
Working Paper .
Merton, R. C. (1974). On the pricing of corporate debt: The risk structure of interest rates, The
Journal of Finance 29(2, Papers and Proceedings of the Thirty-Second Annual Meeting of the
American Finance Association, New York, New York, December 28-30, 1973): 449–470.
Oehmke, M. and Zawadowski, A. (2011). Positioning in the cds market: Evidence from the dtcc
data, Working Paper Columbia University .
O’Kane, D. (2012). The link between eurozone sovereign debt and cds prices, Working Paper Edhec
Business School .
Packer, F. and Suthiphongchai, C. (2003). Sovereign credit default swaps, BIS Quarterly Review .
Palladini, G. and Portes, R. (2011). Sovereign cds and bond pricing dynamics in the euro-area,
Working Paper London Business School .
Pan, J. and Singleton, K. J. (2008). Default and recovery implicit in the term structure of sovereign
cds spreads, The Journal of Finance 63(5): 2345–2384.
31
Plank, T. J. (2010). Do macro-economic fundamentals price sovereign cds spreads of emerging
economies?, Working Paper .
Portes, R. (2010). Ban naked cds, Euro Intelligence .
Pu, X. (2012). The information content of dual-currency sovereign cds for exchange rate returns,
Working Paper .
Remolona, E., Scatigna, M. and Wu, E. (2008). The dynamic pricing of sovereign risk in emerging
markets: Fundamentals and risk aversion, Journal of Fixed Income 17(4): 57 – 71.
Sambalaibat, B. (2011). Credit default swaps and sovereign debt with moral hazard and debt
renegotiation, Working Paper Carnegie Mellon University .
Sgherri, S. and Zoli, E. (2009). Euro area sovereign risk during the crisis, IMF Working Papers
09/222, International Monetary Fund.
Singh, M. (2003). Are credit default swap spreads high in emerging markets? an alternative
methodology for proxying recovery value, IMF Working Paper 03/242 .
Stulz, R. M. (2010). Credit default swaps and the credit crisis, Journal of Economic Perspectives
24(1): 73–92.
Subrahmanyam, M. G., Tang, D. Y. and Wang, S. Q. (2012). Does the tail wag the dog? the effect
of credit default swaps on credit risk, SSRN eLibrary .
Tang, D. Y. and Yan, H. (2007). Liquidity, liquidity spillover, and credit default swap spreads,
Working Paper .
Tett, G. (2009). Fool’s Gold: How the Bold Dream of a Small Tribe at J.P. Morgan Was Corrupted
by Wall Street Greed and Unleashed a Catastrophe, Free Press.
Varga, L. (2009). The information content of hungarian sovereign cds spreads, Working Paper
Central Bank of Hungary .
Verdier, P.-H. (2004). Credit derivatives and the sovereign debt restructuring process, Working
Paper Harvard Law School .
Wang, P. and Moore, T. (2012). The integration of the credit default swap markets during the
us subprime crisis: Dynamic correlation analysis, Journal of International Financial Markets,
Institutions and Money 22(1): 1 – 15.
Zhang, F. X. (2003). What did the credit market expect of argentina default? evidence from default
swap data, SSRN eLibrary .
32
Table 1: Credit Default Swaps: Notional Amount outstanding (’000,000)
This table reports the total notional amount outstanding in million USD of credit default swaps, its growth (in %)
and the market share (in %) for the overall market, single-name and multi-name instruments. Data on total notional
amount outstanding are shown on a net basis, that is transactions between reporting dealers are reported only once.
Source: BIS.
Period
2004-H2
2005-H1
2005-H2
2006-H1
2006-H2
2007-H1
2007-H2
2008-H1
2008-H2
2009-H1
2009-H2
2010-H1
2010-H2
2011-H1
Total
6 395 744
10 211 378
13 908 285
20 352 307
28 650 265
42 580 546
58 243 721
57 402 760
41 882 674
36 098 169
32 692 683
30 260 923
29 897 578
32 409 444
Growth
0.00
59.66
36.20
46.33
40.77
48.62
36.78
-1.44
-27.04
-13.81
-9.43
-7.44
-1.20
8.40
Single-Name
5 116 765
7 310 289
10 432 038
13 873 445
17 879 280
24 239 478
32 486 272
33 412 115
25 739 929
24 165 086
21 917 050
18 493 632
18 144 580
18 104 619
Growth
–
42.87
42.70
32.99
28.87
35.57
34.02
2.85
-22.96
-6.12
-9.30
-15.62
-1.89
-0.22
33
Fraction
80.00
71.59
75.01
68.17
62.41
56.93
55.78
58.21
61.46
66.94
67.04
61.11
60.69
55.86
Multi-Name
1 278 979
2 901 090
3 476 247
6 478 863
10 770 985
18 341 068
25 757 449
23 990 644
16 142 745
11 933 083
10 775 633
11 767 291
11 752 998
14 304 825
Growth
–
126.83
19.83
86.38
66.25
70.28
40.44
-6.86
-32.71
-26.08
-9.70
9.20
-0.12
21.71
Fraction
20.00
28.41
24.99
31.83
37.59
43.07
44.22
41.79
38.54
33.06
32.96
38.89
39.31
44.14
34
Total contracts
Foreign exchange contracts
(Fraction)
Forwards and forex swaps
Currency swaps
Options
Interest rate contracts
(Fraction)
Forward rate agreements
Interest rate swaps
Options
Equity-linked contracts
(Fraction)
Forwards and swaps
Options
Commodity contracts
(Fraction)
Gold
Other commodities
Forwards and swaps
Options
Credit default swaps
(Fraction)
Single-name instruments
Multi-name instruments
of which index products
Unallocated
(Fraction)
2005-H1
282 665
31 081
11.00
15 801
8 236
7 045
204 795
72.45
13 973
163 749
27 072
4 551
1.61
1 086
3 465
2 940
1.04
288
2 652
1 748
904
10 211
3.61
7 310
2 901
–
29 086
10.29
Notional amounts outstanding (’000,000,000)
2006-H1 2007-H1 2008-H1 2009-H1 2010-H1
372 513
507 907
672 558
594 553
582 685
38 127
48 645
62 983
48 732
53 153
10.24
9.58
9.36
8.20
9.12
19 407
24 530
31 966
23 105
25 624
9 696
12 312
16 307
15 072
16 360
9 024
11 804
14 710
10 555
11 170
262 526
347 312
458 304
437 228
451 831
70.47
68.38
68.14
73.54
77.54
18 117
22 809
39 370
46 812
56 242
207 588
272 216
356 772
341 903
347 508
36 821
52 288
62 162
48 513
48 081
6 782
8 590
10 177
6 584
6 260
1.82
1.69
1.51
1.11
1.07
1 430
2 470
2 657
1 678
1 754
5 351
6 119
7 521
4 906
4 506
6 394
7 567
13 229
3 619
2 852
1.72
1.49
1.97
0.61
0.49
456
426
649
425
417
5 938
7 141
12 580
3 194
2 434
2 188
3 447
7 561
1 715
1 551
3 750
3 694
5 019
1 479
883
20 352
42 581
57 403
36 098
30 261
5.46
8.38
8.53
6.07
5.19
13 873
24 239
33 412
24 165
18 494
6 479
18 341
23 991
11 933
11 767
–
–
–
–
7 499
38 332
53 213
70 463
62 291
38 329
10.29
10.48
10.48
10.48
6.58
2011-H1
707 569
64 698
9.14
31 113
22 228
11 358
553 880
78.28
55 842
441 615
56 423
6 841
0.97
2 029
4 813
3 197
0.45
468
2 729
1 846
883
32 409
4.58
18 105
14 305
12 473
46 543
6.58
2007-H1
11 118
1 345
12.10
492
619
235
6 063
54.53
43
5 321
700
1 116
10.03
240
876
636
5.72
47
589
–
–
721
6.48
406
315
–
1 237
11.13
Gross market values (’000,000,000)
2008-H1 2009-H1 2010-H1
20 340
25 298
24 697
2 262
2 470
2 544
11.12
9.77
10.3
802
870
930
1 071
1 211
1 201
388
389
413
9 263
15 478
17 533
45.54
61.18
70.99
88
130
81
8 056
13 934
15 951
1 120
1 414
1 501
1 146
879
706
5.63
3.48
2.86
283
225
189
863
654
518
2 213
682
458
10.88
2.69
1.85
72
43
45
2 141
638
413
–
–
–
–
–
–
3 192
2 973
1 666
15.69
11.75
6.75
1 901
1 950
993
1 291
1 023
673
–
–
–
2 264
2 816
1 789
11.13
11.13
7.25
2011-H1
19 518
2 336
11.97
777
1 227
332
13 244
67.85
60
11 864
1 319
708
3.63
176
532
471
2.41
50
421
–
–
1 345
6.89
854
490
–
1 414
7.25
This table reports the total notional amounts outstanding in billion USD of Over-the-counter derivatives, their gross market values and the market share (in
%), broken down by risk category and type of instrument. Data on total notional amount outstanding are shown on a net basis, that is transactions between
reporting dealers are reported only once. Source: BIS.
Table 2: OTC Derivatives: Notional Amount outstanding (’000,000,000)
Table 3: Exposures of Credit Default Swaps: Notional Amount Outstanding (’000,000)
This table reports the total notional amount outstanding in million USD of credit default swaps as well as the gross
and net market values. Data on total notional amount outstanding are shown on a net basis, that is transactions
between reporting dealers are reported only once. Source: BIS.
Notional Amounts outstanding
Gross Market Values
Gross Credit Exposure
2005-H1
10 211 378
187 932
–
2006-H1
20 352 307
294 332
–
2007-H1
42 580 546
720 725
–
2008-H1
57 402 759
3 191 868
–
2009-H1
36 098 169
2 973 449
–
2010-H1
30 260 923
1 665 944
–
2011-H1
32 409 444
1 344 731
374 525
Table 4: Sovereign Credit Default Swaps: Notional Amount outstanding (’000,000)
This table reports the total notional amount outstanding in million USD of Sovereign credit default swaps and the
market share (in %) by type of product. Data on total notional amount outstanding are shown on a net basis, that
is transactions between reporting dealers are reported only once. Source: BIS.
Period
2004-H2
2005-H1
2005-H2
2006-H1
2006-H2
2007-H1
2007-H2
2008-H1
2008-H2
2009-H1
2009-H2
2010-H1
2010-H2
2011-H1
Total
6 395 744
10 211 378
13 908 285
20 352 307
28 650 265
42 580 546
58 243 721
57 402 760
41 882 674
36 098 169
32 692 683
30 260 923
29 897 578
32 409 444
Sov.Single-Name
––
––
1 258 139 ( 9.05 %)
640 971 ( 3.15 %)
1 100 990 ( 3.84 %)
1 489 655 ( 3.50 %)
1 837 873 ( 3.16 %)
2 177 364 ( 3.79 %)
1 650 743 ( 3.94 %)
1 761 083 ( 4.88 %)
1 943 005 ( 5.94 %)
2 392 580 ( 7.91 %)
2 542 331 ( 8.50 %)
2 749 259 ( 8.48 %)
Sov.Multi-Name
––
––
––
––
––
––
––
––
––
––
––
––
––
158 516 (0.49 %)
Sov.All
––
––
––
––
––
––
––
––
––
––
––
––
––
2 907 775 (8.97 %)
Table 5: Counterparties of Single-Name Sovereign Credit Default Swaps
This table reports the total notional amount outstanding in million USD of Single-Name sovereign credit default swaps
by type of counterparty as well as their relative market share (in %). Data on total notional amount outstanding are
shown on a net basis, that is transactions between reporting dealers are reported only once. Central Counterparty
(CCP) is defined as an entity that interposes itself between counterparties to contracts traded in one or more financial
markets, becoming the buyer to every seller and the seller to every buyer. Both contracts post-novation are captured.
Insurance and Financial Guaranty firms include reinsurance and pension funds. Source: BIS.
All Counterparties
Reporting Dealers
Other Financial Institutions
Central Counterparties
Banks and Security Firms
Insurance and Financial Guaranty Firms
SPVs, SPCs and SPEs
Hedge Funds
Other Financial Customers
Non-Financial Institutions
2006-H1
640 971
301 320
(47.01)
292 907
(45.7)
–
168 174
2 110
712
18 386
103 525
46 745
(7.29)
2007-H1
1 489 655
781 382
(52.45)
644 922
(43.29)
–
216 677
5 293
2 375
27 603
392 976
63 347
(4.25)
35
2008-H1
2 177 364
1 123 078
(51.58)
985 117
(45.24)
–
569 346
8 356
323
13 519
393 573
69 169
(3.18)
2009-H1
1 761 083
877 496
(49.83)
813 241
(46.18)
–
647 135
4 371
418
9 437
151 880
70 357
(4,00)
2010-H1
2 392 580
1 320 039
(55.17)
1 017 122
(42.51)
828 023
8 858
7 905
88 519
83 817
55 429
(2.32)
2011-H1
2 749 259
1 836 889
(66.81)
891 483
(32.43)
2 075
592 194
14 688
18 106
145 178
119 241
20 885
(0.76)
Table 6: Maturity Structure of Credit Default Swaps: Notional Amount Outstanding (’000,000)
This table reports the total notional amount outstanding in million USD of credit default swaps by remaining maturity
as well as their respective market shares. Data on total notional amount outstanding are shown on a net basis, that
is transactions between reporting dealers are reported only once. The remaining contract maturity is determined by
the difference between the reporting date and the expiry date of the contract and not by the date of execution of the
deal. Source: BIS.
Period
2004-H2
2005-H1
2005-H2
2006-H1
2006-H2
2007-H1
2007-H2
2008-H1
2008-H2
2009-H1
2009-H2
2010-H1
2010-H2
2011-H1
Total
6 395 744
10 211 378
13 908 285
20 352 307
28 650 265
42 580 546
58 243 721
57 402 760
41 882 674
36 098 169
32 692 683
30 260 923
29 897 578
32 409 444
1y or less
491 270
670 627
861 871
1 573 542
2 336 401
2 866 737
3 376 160
3 968 090
2 978 334
3 843 679
3 432 429
3 328 461
3 181 610
3 924 650
Fraction
7.68
6.57
6.20
7.73
8.15
6.73
5.80
6.91
7.11
10.65
10.50
11.00
10.64
12.11
Over 1y up to 5y
4 659 521
7 139 210
9 821 293
13 019 148
16 876 503
24 353 126
36 037 715
36 923 426
26 724 773
23 191 972
21 307 787
20 786 997
21 480 952
23 194 893
Fraction
72.85
69.91
70.61
63.97
58.91
57.19
61.87
64.32
63.81
64.25
65.18
68.69
71.85
71.57
Over 5 years
1 243 778
2 400 398
3 225 106
5 759 336
9 437 359
15 360 681
18 829 848
16 511 244
12 179 538
9 062 498
7 952 441
6 145 442
5 235 000
5 289 922
Fraction
19.45
23.51
23.19
28.30
32.94
36.07
32.33
28.76
29.08
25.11
24.32
20.31
17.51
16.32
Table 7: Credit Default Swaps by location of counterparty
This table reports the total notional amount outstanding in million USD of credit default swaps by location of
counterparty as of June 2011, as well as their respective market shares. Data on total notional amount outstanding
are shown on a net basis, that is transactions between reporting dealers are reported only once. The notional
amounts outstanding is allocated to one of the locations listed in the table on an ultimate risk basis, ie according
to the nationality of the counterparty. Home country means country of incorporation of the reporters head office.
Based on the data reported by 10 countries. Source: BIS.
June 2011
Total
Total
32 409 444
Home Country
5 931 892
36
Fraction
18.30
Abroad
26 477 552
Fraction
81.70
37
AAA
Mean
Median
Minimum
Maximum
Volatility
Skewness
Kurtosis
AC1
A
Mean
Median
Minimum
Maximum
Volatility
Skewness
Kurtosis
AC1
BB
Mean
Median
Minimum
Maximum
Volatility
Skewness
Kurtosis
AC1
1
15
2
0
304
26
2
6
0.9940
1
49
12
1
1235
69
2
6
0.9967
1
110
78
15
822
92
2
7
0.9979
2
17
2
0
301
28
2
6
0.9949
2
55
18
2
1172
72
2
5
0.9970
2
157
115
7
845
111
1
4
0.9977
3
19
3
1
293
30
2
5
0.9970
3
61
24
3
1127
74
2
5
0.9971
3
196
146
1
903
125
1
3
0.9976
5
23
4
1
274
34
2
5
0.9975
5
71
34
5
1015
77
2
5
0.9974
5
255
197
1
1032
138
1
3
0.9974
7
24
5
2
270
34
2
4
0.9975
7
76
41
5
952
75
2
5
0.9973
7
281
225
74
1036
138
1
3
0.9972
10
26
7
2
266
33
2
4
0.9974
10
81
49
6
893
73
2
5
0.9972
10
305
250
72
1039
138
1
3
0.9971
category. AC1 is the first-order autocorrelation coefficient. Source: Markit
AA
Mean
Median
Minimum
Maximum
Volatility
Skewness
Kurtosis
AC1
BBB
Mean
Median
Minimum
Maximum
Volatility
Skewness
Kurtosis
AC1
B
Mean
Median
Minimum
Maximum
Volatility
Skewness
Kurtosis
AC1
1
24
3
0
557
36
2
6
0.9969
1
77
33
3
1190
105
3
9
0.9976
1
433
328
19
3654
371
2
7
0.9901
2
27
4
1
536
38
2
6
0.9972
2
95
54
3
1100
106
2
8
0.9974
2
484
403
34
3504
362
2
6
0.9916
3
31
5
1
499
40
2
5
0.9975
3
109
74
5
1110
105
2
7
0.9970
3
517
455
55
3400
350
2
6
0.9926
5
38
8
2
461
44
2
5
0.9978
5
132
106
8
1106
103
2
7
0.9967
5
564
517
117
3234
328
2
6
0.9927
7
41
12
2
433
44
2
4
0.9978
7
143
123
11
1096
99
2
6
0.9966
7
574
538
140
3111
303
2
7
0.9889
10
45
16
3
410
43
2
4
0.9977
10
155
139
14
1081
96
2
6
0.9964
10
599
562
186
3053
286
2
6
0.9928
by taking the mean. Similarly, the standard deviation is calculated as the standard deviation of the data series aggregated at each date within a given rating
Mean (Median) spread is calculated as the historical mean (median) spread, where at each date, all observations within a given rating category are aggregated
each date to each country. The equally weighted historical average is then rounded to the nearest integer, which is used as the final rating categorization. The
CDS prices are mid composite quotes and USD denominated. Rating classification is achieved by assigning an integer value ranging from 1 (AAA) to 21 (C) at
The table reports summary statistics for the CDS term structure of 38 sovereign countries over the sample period May 9th, 2003 until August 19th, 2010. All
Table 8: Summary Statistics
Table 9: Slope of the CDS Spread Curve
Summary statistics by groups of the 5-year spread level. For each group and maturity, the average spread level and
slope has been calculcated across time and entities. The groups are sorted according to the 5-year spread level.
Source: Markit
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Group
(101-200 bps)
(201-300 bps)
(301-400 bps)
(401-500 bps)
(501-600 bps)
(601-700 bps)
(701-800 bps)
(801-900 bps)
(901-1000 bps)
(1001-1100 bps)
(1101-1200 bps)
(1201-1300 bps)
(1301-1400 bps)
(1401-1500 bps)
(1501-3234 bps)
1y
66
139
209
273
362
503
617
622
724
1 008
1 075
1 207
1 347
1 357
2 369
Average CDS premium (bps)
2y
3y
5y
7y
88
107
143
160
172
201
245
263
256
295
347
367
334
383
447
468
433
484
542
557
567
603
642
633
693
725
745
713
759
812
850
846
845
894
940
932
1 051
1 052 1 043
1 002
1 142
1 149 1 145
1 095
1 256
1 257 1 251
1 195
1 382
1 376 1 363
1 309
1 410
1 425 1 444
1 395
2 362
2 334 2 281
2 213
38
10y
177
279
386
490
572
661
741
840
919
964
1 052
1 148
1 260
1 349
2 146
Average Slope (bps)
10y-1y 3y-1y 10y-3y
111
41
69
141
62
79
176
86
91
217
111
106
210
122
88
158
100
58
125
108
16
218
189
29
195
170
25
-44
44
-87
-23
74
-98
-60
50
-109
-87
29
-116
-8
67
-75
-223
-35
-188
#curves
14 012
5 995
3 711
2 444
1 137
645
346
230
148
96
70
48
33
69
155
#entities
35
34
27
24
21
17
10
8
6
4
2
1
1
1
1
Table 10: Trade Information Warehouse Data
This table reports the average gross (Column 3) and net (Column 4) notional amount (in million USD) on CDS contracts
outstanding in USD equivalents of the sovereign reference entities among the 1,000 mostly traded contracts over the time
period 31 October 2008 through 11 November 2011 (excluding sovereign US states). All numbers are reported in million USD.
Column 5 indicates the average number of contracts live in the Depository Trust & Clearing Corporation’s (DTCC) Trade
Information Warehouse (Warehouse) over the same time period. Column 6 reports the ratio of gross to net notional amount
outstanding and column 7 is the average ratio of net notional amount outstanding to number of contracts outstanding. The
notional values are represented as US dollar equivalents using the prevailing foreign exchange rates. The final row labeled Total
reports the sum over all rows for column 3, 4 and 5, and the average for columns 6 and 7. The values are sorted in descending
order.
(1)
Country
Italy
Turkey
Brazil
Spain
Mexico
Russia
Germany
Greece
France
Philippines
Korea
Portugal
Hungary
Argentina
Venezuela
Ukraine
UK
Austria
SouthAfrica
Indonesia
Ireland
China
Belgium
Colombia
Poland
Japan
Kazakhstan
Peru
Malaysia
Thailand
Bulgaria
Netherlands
Romania
United States
Sweden
Finland
Denmark
Australia
Czech Rep.
Slovak Rep.
Latvia
Iceland
Vietnam
Israel
Panama
Croatia
Qatar
Norway
Lithuania
Slovenia
Chile
Egypt
Estonia
NewZealand
SaudiaArabia
Lebanon
Tunisia
HongKong
Total
(2)
Gross Notional
232 651
152 667
147 673
111 477
103 758
103 099
70 774
66 072
64 073
63 228
57 731
55 638
54 333
50 781
49 909
46 002
41 957
40 389
37 797
34 006
34 003
33 142
32 153
29 762
28 989
28 935
21 914
21 441
18 575
18 148
17 412
15 908
15 732
14 574
14 560
11 385
10 718
10 499
9 070
8 881
8 241
8 186
7 855
7 550
6 912
6 730
5 864
5 275
5 149
3 994
3 974
3 205
2 968
2 700
2 188
2 027
1 972
1 705
2 066 309
(3)
Country
Italy
Spain
Germany
Brazil
France
UK
Austria
Portugal
Greece
Mexico
Turkey
Belgium
Russia
Japan
Ireland
Korea
China
Hungary
Sweden
United States
Netherlands
Philippines
Indonesia
Denmark
SouthAfrica
Poland
Colombia
Finland
Argentina
Australia
Venezuela
Peru
Ukraine
Romania
Bulgaria
Malaysia
Thailand
Kazakhstan
Czech Rep.
Slovak Rep.
Iceland
Norway
Israel
Slovenia
Egypt
Latvia
Lithuania
Panama
Croatia
Vietnam
HongKong
Chile
NewZealand
Qatar
SaudiaArabia
Lebanon
Estonia
Tunisia
Total
(4)
Net Notional
23 287
14 846
13 665
13 307
12 687
7 740
7 120
7 102
6 926
6 845
5 748
5 642
4 973
4 709
4 594
4 054
3 952
3 702
2 907
2 804
2 762
2 677
2 261
2 239
2 190
2 144
2 108
2 031
2 004
1 979
1 953
1 806
1 697
1 255
1 213
1 199
1 136
1 086
1 002
941
903
881
808
805
726
726
710
699
663
602
586
539
517
514
483
459
457
289
203 661
Source: http://www.dtcc.com
39
(5)
Country
Brazil
Turkey
Mexico
Russia
Philippines
Korea
Italy
Argentina
Venezuela
Spain
Hungary
Indonesia
SouthAfrica
Ukraine
China
Greece
Colombia
France
Japan
UK
Poland
Portugal
Thailand
Malaysia
Peru
Germany
Kazakhstan
Bulgaria
Ireland
Austria
Romania
Belgium
Vietnam
Iceland
Latvia
Australia
Panama
Croatia
Israel
Sweden
Czech Rep.
Netherlands
Qatar
Egypt
Slovak Rep.
Denmark
Lithuania
United States
Finland
Chile
Estonia
Slovenia
Lebanon
Tunisia
NewZealand
SaudiaArabia
Norway
HongKong
Total
(6)
# Contracts
10 955
10 087
8 760
7 401
7 364
6 156
6 132
5 413
5 033
4 606
4 603
4 294
4 168
3 764
3 341
3 215
3 107
2 736
2 693
2 683
2 627
2 519
2 437
2 291
2 212
2 051
1 938
1 771
1 749
1 733
1 638
1 533
1 145
1 127
1 020
963
957
918
831
781
756
737
730
718
689
617
598
544
443
420
391
334
324
300
297
264
220
125
147 258
(7)
Country
Germany
United States
HongKong
France
Finland
Austria
Norway
Italy
Netherlands
Sweden
Belgium
Denmark
Spain
UK
Portugal
Ireland
Slovenia
Greece
Australia
SaudiaArabia
Japan
NewZealand
Lebanon
Slovak Rep.
Czech Rep.
Chile
Brazil
Lithuania
China
Estonia
Egypt
Israel
Tunisia
Poland
Peru
Hungary
Iceland
Mexico
Romania
Panama
Croatia
Latvia
Qatar
Bulgaria
Colombia
Russia
Korea
Turkey
Kazakhstan
Indonesia
SouthAfrica
Vietnam
Malaysia
Thailand
Ukraine
Venezuela
Argentina
Philippines
Average
(8)
Net/Contract
6.66
5.16
4.69
4.64
4.58
4.11
4.01
3.8
3.75
3.72
3.68
3.63
3.22
2.89
2.82
2.63
2.41
2.15
2.05
1.83
1.75
1.74
1.42
1.37
1.32
1.29
1.21
1.19
1.18
1.17
1.01
0.97
0.97
0.82
0.82
0.8
0.8
0.78
0.77
0.73
0.72
0.71
0.7
0.69
0.68
0.67
0.66
0.57
0.56
0.53
0.53
0.53
0.52
0.47
0.45
0.39
0.37
0.36
1.8
Table 11: Trade Information Warehouse Data
This table reports the average gross (Column 3) and net (Column 4) notional amount (in million USD) on CDS contracts
outstanding in USD equivalents of the sovereign reference entities among the 1,000 mostly traded contracts over the time
period 31 October 2008 through 11 November 2011. All numbers are reported in million USD. Column 5 indicates the average
number of contracts live in the Depository Trust & Clearing Corporation’s (DTCC) Trade Information Warehouse (Warehouse)
over the same time period. Column 6 reports the ratio of gross to net notional amount outstanding and column 7 is the average
ratio of net notional amount outstanding to number of contracts outstanding. The notional values are represented as US dollar
equivalents using the prevailing foreign exchange rates. The final row labeled Total reports the sum over all rows for column 3,
4 and 5, and the average for columns 6 and 7. The values are reported in descending order according to the net notional amount
outstanding. The countries are grouped in five regions: Americas, Asia ex-Japan, Australia and New Zealand, Europe/Middle
East and Africa (EMEA) and Japan.
(1)
Country
Italy
Spain
Germany
Brazil
France
UnitedKingdom
Austria
Portugal
Greece
Mexico
Turkey
Belgium
RussianFederation
Japan
Ireland
Korea
China
Hungary
Sweden
United States
Netherlands
Philippines
Indonesia
Denmark
SouthAfrica
Poland
Colombia
Finland
Argentina
Australia
Venezuela
Peru
Ukraine
Romania
Bulgaria
Malaysia
Thailand
Kazakhstan
CzechRepublic
SlovakRepublic
Iceland
Norway
Israel
Slovenia
Egypt
Latvia
Lithuania
Panama
Croatia
Vietnam
HongKong
Chile
NewZealand
Qatar
SaudiaArabia
Lebanon
Estonia
Tunisia
Total/Average(∗)
(2)
DC Region
EMEA
EMEA
EMEA
Americas
EMEA
EMEA
EMEA
EMEA
EMEA
Americas
EMEA
EMEA
EMEA
Japan
EMEA
Asia Ex-Japan
Asia Ex-Japan
EMEA
EMEA
EMEA
EMEA
Asia Ex-Japan
Asia Ex-Japan
EMEA
EMEA
EMEA
Americas
EMEA
Americas
Australia NZ
Americas
Americas
EMEA
EMEA
EMEA
Asia Ex-Japan
Asia Ex-Japan
EMEA
EMEA
EMEA
EMEA
EMEA
EMEA
EMEA
EMEA
EMEA
EMEA
Americas
EMEA
Asia Ex-Japan
Asia Ex-Japan
Americas
Australia NZ
EMEA
EMEA
EMEA
EMEA
EMEA
(3)
Gross Notional
232 651
111 477
70 774
147 673
64 073
41 957
40 389
55 638
66 072
103 758
152 667
32 153
103 099
28 935
34 003
57 731
33 142
54 333
14 560
14 574
15 908
63 228
34 006
10 718
37 797
28 989
29 762
11 385
50 781
10 499
49 909
21 441
46 002
15 732
17 412
18 575
18 148
21 914
9 070
8 881
8 186
5 275
7 550
3 994
3 205
8 241
5 149
6 912
6 730
7 855
1 705
3 974
2 700
5 864
2 188
2 027
2 968
1 972
2 066 309
(4)
Net Notional
23 287
14 846
13 665
13 307
12 687
7 740
7 120
7 102
6 926
6 845
5 748
5 642
4 973
4 709
4 594
4 054
3 952
3 702
2 907
2 804
2 762
2 677
2 261
2 239
2 190
2 144
2 108
2 031
2 004
1 979
1 953
1 806
1 697
1 255
1 213
1 199
1 136
1 086
1 002
941
903
881
808
805
726
726
710
699
663
602
586
539
517
514
483
459
457
289
203 661
Source: http://www.dtcc.com
40
(5)
# Contracts
6 132
4 606
2 051
10 955
2 736
2 683
1 733
2 519
3 215
8 760
10 087
1 533
7 401
2 693
1 749
6 156
3 341
4 603
781
544
737
7 364
4 294
617
4 168
2 627
3 107
443
5 413
963
5 033
2 212
3 764
1 638
1 771
2 291
2 437
1 938
756
689
1 127
220
831
334
718
1 020
598
957
918
1 145
125
420
297
730
264
324
391
300
147 258
(6)
Gross/Net
9.99
7.51
5.18
11.10
5.05
5.42
5.67
7.83
9.54
15.16
26.56
5.70
20.73
6.14
7.40
14.24
8.39
14.68
5.01
5.20
5.76
23.62
15.04
4.79
17.26
13.52
14.12
5.60
25.33
5.30
25.56
11.87
27.11
12.53
14.35
15.49
15.98
20.18
9.05
9.44
9.06
5.99
9.35
4.96
4.41
11.35
7.25
9.89
10.14
13.04
2.91
7.37
5.22
11.40
4.53
4.41
6.50
6.82
10.74∗
(7)
Net/Contract
3.80
3.22
6.66
1.21
4.64
2.89
4.11
2.82
2.15
0.78
0.57
3.68
0.67
1.75
2.63
0.66
1.18
0.80
3.72
5.16
3.75
0.36
0.53
3.63
0.53
0.82
0.68
4.58
0.37
2.05
0.39
0.82
0.45
0.77
0.69
0.52
0.47
0.56
1.32
1.37
0.80
4.01
0.97
2.41
1.01
0.71
1.19
0.73
0.72
0.53
4.69
1.29
1.74
0.70
1.83
1.42
1.17
0.97
1.80∗
41
2008
.
48 945
.
17 169
12 964
136 964
15 442
2 897
19 814
28 208
4 326
4 929
4 866
.
.
2 240
3 736
22 091
38 051
35 990
.
33 875
8 864
30 813
17 589
5 149
153 732
7 296
22 479
50 311
6 440
.
3 302
16 178
75 514
4 904
.
2 313
6 475
19 707
66 345
17 152
25 066
4 033
11 839
105 708
.
5 581
.
31 681
64 649
6 245
16 620
.
179 469
59 556
4 537
12 843
46 073
5 997
30 539
Panel A: Gross Notional (’000,000)
2009
2010
2011
2 493
4 032
6 683
45 810
51 213
56 383
2 722
8 025
19 634
30 413
44 938
51 203
21 035
30 670
50 583
123 478
148 786
176 463
15 647
17 676
19 535
3 262
4 134
4 825
24 200
33 615
45 583
27 278
29 422
33 344
5 775
6 880
8 136
7 389
10 047
10 691
7 995
10 966
14 742
.
5 649
7 264
.
2 236
4 174
2 762
3 280
2 984
6 545
13 011
16 591
37 464
62 020
105 638
49 896
74 005
97 639
48 128
78 189
78 552
.
1 705
.
43 582
57 916
66 628
8 839
7 974
7 544
31 737
34 112
37 142
24 300
37 482
44 400
6 115
7 225
10 069
187 481
238 376
293 888
12 034
29 874
51 688
23 375
22 225
19 745
56 473
59 615
58 449
7 359
8 724
9 053
1 675
1 949
2 127
4 476
5 394
6 006
18 174
18 961
19 063
88 742
106 557
123 461
10 125
17 825
22 533
1 659
2 393
3 086
2 959
5 435
8 251
6 375
6 792
7 761
18 658
19 760
26 984
67 258
63 381
57 767
22 602
30 165
37 352
41 192
62 951
69 833
4 789
5 540
7 855
13 878
16 542
17 700
101 719
103 458
103 749
.
.
2 188
7 666
9 561
10 145
2 945
4 138
4 943
34 253
37 776
43 140
77 492
111 699
159 851
11 093
15 493
19 132
17 873
19 266
17 456
1 932
1 841
2 141
156 877
151 600
143 699
46 198
46 450
42 535
8 716
13 156
25 020
21 323
47 368
65 251
42 703
50 787
57 968
6 246
8 281
9 582
27 732
32 192
38 228
Source: http://www.dtcc.com
Country
AbuDhabi
Argentina
Australia
Austria
Belgium
Brazil
Bulgaria
Chile
China
Colombia
Croatia
CzechRepublic
Denmark
Dubai
Egypt
Estonia
Finland
France
Germany
Greece
HongKong
Hungary
Iceland
Indonesia
Ireland
Israel
Italy
Japan
Kazakhstan
Korea
Latvia
Lebanon
Lithuania
Malaysia
Mexico
Netherlands
NewZealand
Norway
Panama
Peru
Philippines
Poland
Portugal
Qatar
Romania
RussianFederation
SaudiaArabia
SlovakRepublic
Slovenia
SouthAfrica
Spain
Sweden
Thailand
Tunisia
Turkey
Ukraine
United States
UnitedKingdom
Venezuela
Vietnam
Average
Average
4 403
50 588
10 127
35 931
28 813
146 423
17 075
3 780
30 803
29 563
6 279
8 264
9 642
6 457
3 205
2 816
9 971
56 803
64 898
60 215
1 705
50 500
8 305
33 451
30 943
7 140
218 369
25 223
21 956
56 212
7 894
1 917
4 795
18 094
98 568
13 847
2 379
4 740
6 851
21 277
63 688
26 818
49 760
5 554
14 990
103 659
2 188
8 238
4 009
36 713
103 423
12 991
17 804
1 971
157 911
48 685
12 857
36 696
49 383
7 527
32 376
2008
.
2 815
.
4 883
3 871
11 160
1 717
670
2 099
2 442
680
1 139
1 706
.
.
549
1 413
5 533
9 905
7 459
.
4 388
1 193
2 194
4 485
595
17 388
1 741
2 471
5 326
891
.
709
1 485
4 529
1 588
.
651
708
2 158
2 903
2 083
5 030
456
1 561
7 188
.
1 017
.
2 484
13 895
1 829
1 187
.
5 816
2 777
1 296
2 368
2 175
627
3 305
(14)). The values are reported in descending order according to the country name.
Panel B: Net Notional (’000,000)
2009
2010
2011
Average
468
834
1 379
894
1 817
1 755
2 353
2 185
438
1 427
3 862
1 909
6 933
8 325
6 364
6 626
4 836
5 670
6 893
5 317
10 204
13 811
16 730
12 976
1 331
1 167
1 031
1 311
482
548
568
567
1 938
3 723
6 919
3 670
2 152
1 983
2 136
2 178
611
674
709
668
1 145
946
874
1 026
1 887
2 330
2 646
2 142
.
519
545
532
.
518
935
726
466
456
428
475
1 648
2 272
2 314
1 912
7 587
11 744
21 123
11 497
10 936
13 988
17 192
13 005
7 959
7 533
4 911
6 966
.
586
.
586
4 052
3 569
3 316
3 831
976
815
864
962
1 860
1 974
3 075
2 276
4 741
4 868
4 123
4 554
576
662
1 289
781
20 972
25 306
24 762
22 107
2 061
4 843
8 206
4 213
1 271
855
867
1 366
3 720
3 772
4 517
4 334
792
720
623
757
412
441
482
445
785
703
632
707
1 149
1 016
1 416
1 266
5 769
6 727
8 690
6 429
2 524
3 117
2 852
2 520
481
495
544
507
684
975
1 035
836
646
649
818
705
1 705
1 612
2 083
1 889
2 569
2 506
2 957
2 734
1 936
2 237
2 288
2 136
6 659
8 355
6 554
6 649
341
434
821
513
1 377
1 189
1 130
1 314
5 951
4 110
4 415
5 416
.
.
483
483
1 042
888
870
954
687
815
918
807
2 079
2 106
2 357
2 256
12 095
15 087
17 932
14 752
2 842
3 086
2 987
2 686
1 051
1 036
1 341
1 154
281
268
317
289
5 396
5 809
6 067
5 772
2 062
1 541
1 242
1 905
1 987
2 441
4 478
2 550
3 361
9 185
12 174
6 772
1 630
1 884
2 363
2 013
423
602
806
614
2 818
3 145
3 674
3 239
2008
.
17.4
.
3.5
3.3
12.3
9.0
4.3
9.4
11.6
6.4
4.3
2.9
.
.
4.1
2.6
4.0
3.8
4.8
.
7.7
7.4
14.0
3.9
8.6
8.8
4.2
9.1
9.4
7.2
.
4.7
10.9
16.7
3.1
.
3.6
9.1
9.1
22.9
8.2
5.0
8.8
7.6
14.7
.
5.5
.
12.8
4.7
3.4
14.0
.
30.9
21.4
3.5
5.4
21.2
9.6
8.8
Panel C: Ratio Gross/Net
2009
2010
2011
Average
5.3
4.8
4.8
5.0
25.2
29.2
24.0
23.9
6.2
5.6
5.1
5.6
4.4
5.4
8.0
5.3
4.3
5.4
7.3
5.1
12.1
10.8
10.5
11.4
11.8
15.1
19.0
13.7
6.8
7.5
8.5
6.8
12.5
9.0
6.6
9.4
12.7
14.8
15.6
13.7
9.5
10.2
11.5
9.4
6.5
10.6
12.2
8.4
4.2
4.7
5.6
4.3
.
10.9
13.3
12.1
.
4.3
4.5
4.4
5.9
7.2
7.0
6.0
4.0
5.7
7.2
4.9
4.9
5.3
5.0
4.8
4.6
5.3
5.7
4.8
6.0
10.4
16.0
9.3
.
2.9
.
2.9
10.8
16.2
20.1
13.7
9.1
9.8
8.7
8.7
17.1
17.3
12.1
15.1
5.1
7.7
10.8
6.9
10.6
10.9
7.8
9.5
8.9
9.4
11.9
9.8
5.8
6.2
6.3
5.6
18.4
26.0
22.8
19.1
15.2
15.8
12.9
13.3
9.3
12.1
14.5
10.8
4.1
4.4
4.4
4.3
5.7
7.7
9.5
6.9
15.8
18.7
13.5
14.7
15.4
15.8
14.2
15.5
4.0
5.7
7.9
5.2
3.4
4.8
5.7
4.7
4.3
5.6
8.0
5.4
9.9
10.5
9.5
9.7
10.9
12.3
13.0
11.3
26.2
25.3
19.5
23.5
11.7
13.5
16.3
12.4
6.2
7.5
10.7
7.3
14.0
12.8
9.6
11.3
10.1
13.9
15.7
11.8
17.1
25.2
23.5
20.1
.
.
4.5
4.5
7.4
10.8
11.7
8.8
4.3
5.1
5.4
4.9
16.5
17.9
18.3
16.4
6.4
7.4
8.9
6.8
3.9
5.0
6.4
4.7
17.0
18.6
13.0
15.7
6.9
6.9
6.7
6.8
29.1
26.1
23.7
27.4
22.4
30.1
34.2
27.1
4.4
5.4
5.6
4.7
6.3
5.2
5.4
5.6
26.2
27.0
24.5
24.7
14.8
13.8
11.9
12.5
10.0
11.2
11.6
10.5
ends on 11 November 2011. The overall statistics in the final row labeled Average include information for the U.S. states, which are reported in a separate table (Table
outstanding. The notional values are represented as US dollar equivalents using the prevailing foreign exchange rates. The sample period starts on 31 October 2008 and
among the 1,000 mostly traded contracts as of 11th November 2011. All numbers are reported in million USD. Panel C indicates the ratio of gross to net notional amount
This table reports the average yearly gross (Panel A) and net (Panel B) notional amount on CDS contracts outstanding in USD equivalents of the sovereign reference entities
Table 12: Trade Information Warehouse Data
42
2008
.
5 011
.
686
437
11 097
1 515
302
1 736
3 168
588
415
236
.
.
304
92
547
760
1 131
.
3 126
1 209
3 973
660
546
3 390
295
2 097
4 612
857
.
423
1 929
6 560
157
.
58
895
2 168
8 037
1 589
797
416
1 294
7 483
.
451
.
3 639
2 008
276
2 061
.
13 325
5 402
124
433
4 781
866
2 279
Panel A: # Contracts
2009
2010
2011
427
866
1 510
4 916
5 478
5 990
270
684
1 847
1 338
1 905
2 196
793
1 441
2 715
9 818
11 097
12 074
1 618
1 782
1 985
359
432
498
2 177
3 458
4 869
3 112
3 024
3 186
791
937
1 109
649
823
870
399
498
1 087
.
691
941
.
424
1 012
375
424
389
242
503
675
1 005
2 490
5 464
1 185
2 097
3 254
1 701
3 944
4 524
.
125
.
3 857
4 788
5 542
1 111
1 149
1 105
3 967
4 314
4 713
1 000
1 858
2 704
665
743
1 183
4 122
6 401
8 685
655
2 840
5 354
2 143
1 854
1 769
5 630
6 590
6 561
972
1 068
1 053
244
305
349
541
612
682
2 195
2 367
2 386
8 061
8 979
9 748
413
818
1 134
145
256
348
90
228
387
896
924
1 078
2 031
1 969
2 715
8 014
7 378
6 464
2 136
2 725
3 285
1 446
2 834
3 732
525
638
1 139
1 463
1 672
1 869
7 352
7 394
7 447
.
.
264
618
734
765
274
328
407
3 916
4 122
4 620
2 429
4 699
7 533
594
795
1 083
2 267
2 569
2 553
269
283
332
11 088
9 580
8 880
3 958
3 587
3 420
241
479
1 054
881
3 120
4 699
4 464
5 093
5 669
900
1 206
1 413
1 994
2 273
2 736
Source: http://www.dtcc.com
Country
AbuDhabi
Argentina
Australia
Austria
Belgium
Brazil
Bulgaria
Chile
China
Colombia
Croatia
CzechRepublic
Denmark
Dubai
Egypt
Estonia
Finland
France
Germany
Greece
HongKong
Hungary
Iceland
Indonesia
Ireland
Israel
Italy
Japan
Kazakhstan
Korea
Latvia
Lebanon
Lithuania
Malaysia
Mexico
Netherlands
NewZealand
Norway
Panama
Peru
Philippines
Poland
Portugal
Qatar
Romania
RussianFederation
SaudiaArabia
SlovakRepublic
Slovenia
SouthAfrica
Spain
Sweden
Thailand
Tunisia
Turkey
Ukraine
United States
UnitedKingdom
Venezuela
Vietnam
Average
Total
934
5 349
934
1 531
1 346
11 021
1 725
398
3 060
3 123
856
689
555
816
718
373
378
2 377
1 824
2 825
125
4 328
1 143
4 242
1 556
784
5 649
2 286
1 966
5 848
987
299
565
2 219
8 337
630
250
191
948
2 221
7 473
2 434
2 202
680
1 574
7 419
264
642
336
4 074
4 167
687
2 363
295
10 718
4 092
475
2 283
5 002
1 096
2 330
Panel B: Ratio Net Notional/Contracts (’000,000)
2008
2009
2010
2011
Total
.
1.1
1.0
0.9
1.0
0.6
0.4
0.3
0.4
0.4
.
1.6
2.1
2.1
1.9
7.1
5.2
4.4
2.9
4.9
8.9
6.1
3.9
2.5
5.4
1.0
1.0
1.2
1.4
1.2
1.1
0.8
0.7
0.5
0.8
2.2
1.3
1.3
1.1
1.5
1.2
0.9
1.1
1.4
1.1
0.8
0.7
0.7
0.7
0.7
1.2
0.8
0.7
0.6
0.8
2.7
1.8
1.1
1.0
1.7
7.2
4.7
4.7
2.4
4.8
.
.
0.8
0.6
0.7
.
.
1.2
0.9
1.1
1.8
1.2
1.1
1.1
1.3
15.3
6.8
4.5
3.4
7.5
10.1
7.6
4.7
3.9
6.6
13.0
9.2
6.7
5.3
8.6
6.6
4.7
1.9
1.1
3.6
.
.
4.7
.
4.7
1.4
1.1
0.7
0.6
0.9
1.0
0.9
0.7
0.8
0.8
0.6
0.5
0.5
0.7
0.5
6.8
4.7
2.6
1.5
3.9
1.1
0.9
0.9
1.1
1.0
5.1
5.1
4.0
2.9
4.3
5.9
3.1
1.7
1.5
3.1
1.2
0.6
0.5
0.5
0.7
1.2
0.7
0.6
0.7
0.8
1.0
0.8
0.7
0.6
0.8
.
1.7
1.4
1.4
1.5
1.7
1.5
1.1
0.9
1.3
0.8
0.5
0.4
0.6
0.6
0.7
0.7
0.7
0.9
0.8
10.1
6.1
3.8
2.5
5.6
.
3.3
1.9
1.6
2.3
11.2
7.6
4.3
2.7
6.5
0.8
0.7
0.7
0.8
0.7
1.0
0.8
0.8
0.8
0.9
0.4
0.3
0.3
0.5
0.4
1.3
0.9
0.8
0.7
0.9
6.3
4.6
2.9
1.8
3.9
1.1
0.6
0.7
0.7
0.8
1.2
0.9
0.7
0.6
0.9
1.0
0.8
0.6
0.6
0.7
.
.
.
1.8
1.8
2.3
1.7
1.2
1.1
1.6
.
2.5
2.5
2.3
2.4
0.7
0.5
0.5
0.5
0.6
6.9
5.0
3.2
2.4
4.4
6.6
4.8
3.9
2.8
4.5
0.6
0.5
0.4
0.5
0.5
.
1.0
0.9
1.0
1.0
0.4
0.5
0.6
0.7
0.6
0.5
0.5
0.4
0.4
0.5
10.4
8.2
5.1
4.2
7.0
5.5
3.8
2.9
2.6
3.7
0.5
0.4
0.4
0.4
0.4
0.7
0.5
0.5
0.6
0.6
3.6
2.5
1.8
1.4
2.2
include information for the U.S. states, which are reported in a separate table. The values are reported in descending order according to the country name.
prevailing foreign exchange rates. The sample period starts on 31 October 2008 and ends on 11 November 2011. The overall statistics in the final row labeled Average
sovereign reference entities among the 1,000 mostly traded contracts as of 11th November 2011. The notional values are represented as US dollar equivalents using the
A) and the average ratio (Panel B) of net notional amount on CDS contracts outstanding in million USD equivalents to the average yearly number of contracts of the
This table reports the average yearly number of contracts live in the Depository Trust & Clearing Corporation’s (DTCC) Trade Information Warehouse (Warehouse) (Panel
Table 13: Trade Information Warehouse Data
43
2008
.
.
.
.
.
.
.
.
4 537
Panel A: Gross Notional (’000,000)
2009
2010
2011
Average
4 253
7 928
9 981
7 387
2 688
3 855
4 636
3 726
.
2 061
3 088
2 575
.
1 682
1 707
1 695
1 989
3 075
4 358
3 141
2 333
2 922
3 318
2 857
3 004
4 423
5 655
4 361
1 937
2 363
2 719
2 340
8 716
13 156
25 020
12 857
Source: http://www.dtcc.com
Country
California
Florida
Illinois
Masschusetts
NewJersey
NewYork
NewYorkCity
Texas
United States
2008
.
.
.
.
.
.
.
.
1 296
Panel B: Net Notional (’000,000)
2009
2010
2011
Average
698
885
986
856
298
296
267
287
.
329
608
468
.
171
174
172
516
440
475
477
282
348
437
355
756
528
386
557
394
221
186
267
1 987
2 441
4 478
2 550
2008
.
.
.
.
.
.
.
.
3.5
Panel D: Ratio Gross/Net
2009
2010
2011
Total
6.1
9.0
10.1
8.4
9.0
13.0
17.4
13.2
.
6.3
5.1
5.7
.
9.9
9.8
9.8
3.9
7.0
9.2
6.7
8.3
8.4
7.6
8.1
4.0
8.4
14.7
9.0
4.9
10.7
14.6
10.1
4.4
5.4
5.6
4.7
ends on 11 November 2011. The values are reported in descending order according to the sovereign’s name.
2008
.
.
.
.
.
.
.
.
124
Panel
2009
232
161
.
.
120
104
171
81
241
C: # Contracts
2010
2011
599
989
274
318
152
403
112
116
215
401
194
266
262
356
96
155
479
1 054
Average
607
251
277
114
245
188
263
111
475
(Warehouse). The notional values are represented as US dollar equivalents using the prevailing foreign exchange rates. The sample period starts on 31 October 2008 and
amount outstanding and Panel D lists the average yearly number of contracts live in the Depository Trust & Clearing Corporation’s (DTCC) Trade Information Warehouse
entities among the 1,000 mostly traded contracts as of 11th November 2011. All numbers are reported in million USD. Panel C indicates the ratio of gross to net notional
This table reports the average yearly gross (Panel A) and net (Panel B) notional amount on CDS contracts outstanding in USD equivalents of the U.S. sovereign reference
Table 14: Trade Information Warehouse Data
44
(a)
windows of 400 curves. Source: Markit.
(b)
up to the right-most point representing curves 71,801 to 72,200. Figure 1b shows the average slope level in basis points as a function of CDS level, in sliding
against the average CDS premium among the 400 curves with the lowest 5-year spread level. The next point does the same for curves 2 to 401, and so on,
10y-1y. The 72,200 curves are sorted by the 5-year premium. Source: Markit. For instance, the left-most points plots the fraction of decreasing CDS pairs
Graph (1a) plots the fraction of decreasing CDS pairs as a function of the CDS level in sliding windows of 400 basis points for the segments 3y-1y, 10y-3y and
Figure 1: The Slope of the Credit Spread Curve
45
Figure 2: Co-Movement of Sovereign CDS Spreads and Risk Aversion
0
05/2003
200
400
600
800
1000
1200
02/2004
11/2004
08/2005
06/2006
03/2007
Time
12/2007
5-year Sovereign CDS Spreads (Source: Markit)
(a)
09/2008
06/2009
04/2010
inspired by the illustration in Pan and Singleton (2008) Source: Markit.
(b)
19th, 2010. Graph (2b) plots the historical average 5-year CDS spread of the 38 countries in the sample over the same time period. The second plot was
Graph (2a) illustrates the historical 5-year CDS spread for 38 countries spanning five geographical regions over the time period May 9th, 2003 until August
Spread
Figure 3: Trade Information Warehouse Data
The following graphs plot the average yearly net notional amount of CDS outstanding (in million USD) by the yearly
amount of Government debt (in million USD) for the EMEA and Asia ex-Japan countries among the 1,000 mostly
traded reference entities. The graphs are separated by year. Graph (3a), (3b) and (3c) refer to 2008, 2009 and 2010
respectively. The notional values are represented as US dollar equivalents using the prevailing foreign exchange rates.
Source: www.dtcc.com.
(a)
(b)
(c)
46
Figure 4: Trade Information Warehouse Data
The following graphs plot the average yearly gross notional amount of CDS outstanding (in million USD) by the
yearly amount of Government debt (in million USD) for the EMEA and Asia ex-Japan countries among the 1,000
mostly traded reference entities. The graphs are separated by year. Graph (4a), (4b) and (4c) refer to 2008, 2009 and
2010 respectively. The notional values are represented as US dollar equivalents using the prevailing foreign exchange
rates. Source: www.dtcc.com.
(a)
(b)
(c)
47