0% found this document useful (0 votes)
8 views55 pages

Lecture 8 ARMI

This document discusses counterparty risk and how it affects the valuation of over-the-counter (OTC) derivatives. It introduces the concepts of credit valuation adjustment (CVA) and debt valuation adjustment (DVA) to account for counterparty credit risk. Specifically, it covers: 1) How counterparty risk introduces optionality into OTC contracts based on the sign of the position and whether the contract is in or out of the money at maturity. 2) A simple model for calculating CVA based on loss given default and the counterparty's probability of default. 3) How counterparty risk affects the Greeks like delta, gamma, and vega of a contract by making it nonlinear.

Uploaded by

Antonio Cobo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
8 views55 pages

Lecture 8 ARMI

This document discusses counterparty risk and how it affects the valuation of over-the-counter (OTC) derivatives. It introduces the concepts of credit valuation adjustment (CVA) and debt valuation adjustment (DVA) to account for counterparty credit risk. Specifically, it covers: 1) How counterparty risk introduces optionality into OTC contracts based on the sign of the position and whether the contract is in or out of the money at maturity. 2) A simple model for calculating CVA based on loss given default and the counterparty's probability of default. 3) How counterparty risk affects the Greeks like delta, gamma, and vega of a contract by making it nonlinear.

Uploaded by

Antonio Cobo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 55

Advanced Risk Management I

Lecture 9
Counterparty risk and XVAs
Learning objectives
• Counterparty risk and CVA/DVA
• Credit risk mitigation techniques and CSA
• FVA and the double counting issue
• Central Clearing Counterparty CCP and MVA
Counterparty Risk
A simple model
OTC vs Futures-style
• Over-the-Counter • Futures-style
• Bilateral relationship • Organized market
• Customized products • Standardized products
• Low basis risk • High basis risk
• Low liquidity • High liquidity
• Relevant risk • Relevant risks
– Market – Market
– Counterparty – Basis risk
Derivatives on OTC markets

• Most of financial derivative contracts, and


particularly those with retail counterparties are
traded on what is called Over-the-Counter (OTC)
market
• The OTC market allows the construction of
customized positions for hedging or investment
purposes
• The cost is illiquidity and credit risk (counterparty
risk)
The simplest example

• Consider a lineare OTC contract, i.e. forward,


determined at time 0.
• Remember that if we only focus on the risk of
price changes in the underlying asset we have
CF(t) = v(t,T)EQ[S(T) –F(0)]
= S(t) – v(t,T)F(0)
where F(0) is the forward price at time 0.
• Notice that the product is linear, meaning delta = 1
and the replicating portfolio is static.
Counterparty risk

• We make the assumption that: default occurs at time T,


which is the maturity of the contract: this is a simplifying
assumption that will be relaxed later one.
• We assume that if at maturity the marked-to- market value
of the derivative contract is positive for the counterparty
which goes in default, the other party is compelled to abide
by the contract and pay its obligation. On the other side, if
the value of the contract is negative for the counterparty in
default the other party has an exposure equal to that value,
with the same degree of seniority of the other liabilities.
This assumption corresponds to the reality of legal
provisions of counter party risk.
Long and short positions

• The value of the impact of counter party risk requires to


distinguish the sign, long or short of the position. This is
because counterparty risk is triggered by two events:
– Default of the counterparty
– The contract is “out-of-the-money” for the party in
dafault, that it the contract has negative value for the
counterparty in default.
• So, in case on the delivery date we have S(T) > F(0) the
contract is in-the-money for the party long in the contract.
If instead it is S(T) < F(0) the contract is in-the-money for
the short party. In the former case the long party in the
contract will be exposed to default risk, in the latter the
short one will.
Default and loss

• Denote A the long party of the contract and B the


short one.
• Let us introduce characteristic functions 1A and 1B
assuming value 1 if the party A or B is in a default
state and zero otherwise.
• Definiamo RRA e RRB i tassi di recupero delle due
controparti. Nello stesso modo definiamo le loss-
given-default LgdA = 1 – RRA e LgdB = 1 – RRB.
Risk of the long party

• The pay-off value of the forward contract must


take into account both its sign and its value in caso
of default of the relevant counterparty.
• From the viewpoint of the long end of the contract
we have
CFA(T) = max[S(T) – F(0),0](1 –1B) +
max[S(T) – F(0),0]RRB1B –
– max[F(0) –S(T),0] =
CF(T) – LgdB1Bmax[S(T) – F(0),0]
Risk of the short party

• For the short end of the contract, the default event


is relevant only in the hypothesis that the contract
ends in-the-money.
• From the viewpoint of the counterparty
CFB(T) = max[F(0) – S(T),0](1 –1A) +
max[F(0) – S(T),0]RRA1A +
– max[S(T) – F(0),0] =
– CF(T) – LgdA1Amax[F(0) – S(T),0]
Counterparty risk

• Counterparty risk corresponds to a short position


is options.
• The option is of the call type for the long endo of
the contract and of the put type for the other end
of the contract.
• Exercise of the option is contingent on two events
– The value of the underlying asset at time T
– Default event of the relevant counterparty
Contract evaluation

• The value of the product from the point of view of


the long end of the contract will be given by
CFA(t) = S(t) – v(t,T)F(0) –
EQ[v(t,T)LgdB1Bmax[S(T) – F(0),0]]
• From the viewpoint of the short end of the
contract we will then have
CFB(T) = – S(t) + v(t,T)F(0) –
EQ[v(t,T)LgdA1Amax[F(0) – S(T),0]]
Credit Valuation Adjustment (CVA)

• Counterparty risk is represented by


EQ[v(t,T)Lgdi1imax[(S(T) – F(0)),0]]
with i = A, B and  = 1(–1) for call (put) options
• Counterparty risk is made by
– Interest rate risk
– Market risk of the underlying
– Credit risk of the counterparty
– Recovery risk
• All these factors may be correlated.
A simple model

• In what follows we will assume that

– Interest rate is independent of the other risk


factors
– Default risk of the counterparty is not
dependent on other risk factors.
CVA valuation
• The value of counterparty risk is then

v(t,T)EQ[Lgdi1i] EQ[max[(S(T) – F(0)),0]]

with i = A,B. Assume Lgdi to be given and constant. Then,

EQ[Lgdi1i] = Lgdi EQ[1i] = Lgdi (1 – Fi),

• In case of independence, then

CVA = v(t,T) Lgdi (1 – Fi) EQ[max[(S(T) – F(0)),0]


CVA and hybrid derivatives

• CVA are technically «hybrid derivatives», that is


derivatives that are subject to different risk
sources, typically market and credit risk.
• Contingent CDS: CDS that upon a credit event
give the pay-off of a credit derivative
• Risky swaps: swap contracts that may terminate
before maturity if there is a credit event of a name.
Effects of counterparty risk

• Effect 1: ruling out counterparty risk leads to


undervaluation of the overall exposure to credit risk
• Effect 2: if one does not consider counterparty risk, he
comes out with the wrong price, and the wrong hedge.
• Effect 3: counterparty risk makes linear product non
linear, so that changes in volatility may affect the value
of the contract even though it is linear and one would
not expect any effect.
Greek letters

• The sensitivity of the contract to small changes in


the underlying is no more that of a linear contract.
We get in fact

A = 1 – ELBN()
B = – 1 + ELAN(– )
with

 =(ln(S(t)/F(0))+(r + ½ 2)(T – t))/[(T – t)1/2 ]


Gamma and Vega

• The second order effect of finite changes in the


underlying is now given by
– n()/ [S(t)(T – t)1/2]
• Changes in volatility affect the value of the position
through a vega effect
– S(t)n()/ [(T – t)1/2 ]
 =(ln(S(t)/F(0))+(r – ½ 2)(T – t))/[(T – t)1/2 ]
An example

• Forward contract
– Notional 1 million
– Volatility 20%
– Maturity 1 year
• Counterparty
– Loss given default (Lgd): 100%
– 1 year default probability: 5%
• Counterparty risk at the origin of the contract, for
both the long and the short end of the contract:
3983
Long position
500000

400000

300000

200000

100000

0
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5

-100000

-200000

-300000

-400000

-500000
Long position delta
1,01

0,99

0,98

0,97

0,96

0,95

0,94
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5
30000
Counterparty risk (long)

25000

20000

15000

10000

5000

0
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5
Short position
500000

400000

300000

200000

100000

0
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5

-100000

-200000

-300000

-400000

-500000
Short position delta
-0,94
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5

-0,95

-0,96

-0,97

-0,98

-0,99

-1

-1,01
Counterparty risk (short)
30000

25000

20000

15000

10000

5000

0
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5
Counterparty risk and vol. (long)
30000

25000

20000

0,1
0,2
15000
0,3
0,4

10000

5000

0
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5
Counterparty risk and vol. (short)
30000

25000

20000

0,1
0,2
15000
0,3
0,4

10000

5000

0
0,5 0,6 0,7 0,8 0,9 1 1,1 1,2 1,3 1,4 1,5
CVA and DVA
CVA and DVA
• In recent years the new CVA risk factor have been
added to the replicating portfolio.
• On top of that, for accounting purposes the new
concept of DVA was also included. The DVA is
the adjustment to be added to the risk-free price of
the product to allow for the dependence of the
own party.
• The price of a forward contract is then
Risk free product – CVA + DVA
Again in the simplest model
• A first sight application of the CVA/DVA pricing would be to
extend the price in our example to
CF(t) = S(t) – v(t,T)F(0) (risk free product)
– LgdB(1-FB)v(t,T) EQ[max[S(T) – F(0),0] (CVA)
+LgdA(1-FA) v(t,T) EQ[max[F(0) – S(T),0] (DVA)
…but this is not correct, because one should consider the first that
would default between A and B.
• In principle, this is only correct, when default of A and B are
independent.
Default before maturity
• Assume now that default may occur before maturity, for
example by a time .
• The value of exposure for the long position is now
max[S() – v( ,T)F(0), 0 ]
and for the short position
max[v( ,T)F(0) – S(), 0 ]
• The value of exposure is given by a sequence of options
that will be multiplied times the value of the default
probability of the counterparty in the sub-periods.
Counterparty risk

• Partition the lifetime of the contract in a grid of dates {t1,t2,


…tn}
• Denote Qj(ti) the survival probability of counterparty j = A,
B beyond time ti.
• Compute
[FB(ti -1) – FB(ti) ]Call(S(ti), ti; v(ti ,T)F(0), ti )
[FA(ti -1) – FA(ti) ] Put(S(ti), ti; v(ti ,T)F(0), ti )
respectively for long and short positions
• Aggregate the values obtained in this way from 1 through n.
CVA/DVA
Counterparty risk in swap contracts

• In a swap cotnract both the legs are exposed to


counterparty risk.
• In the event of default of one of the two parties the other
takes a loss equal to the marked to market value of the
contract, equal to the net value of the cash-flows.
• Remember that the net value of the swap contract is
positive for the long end of the contract if the swap rate on
the day of default of the contract is greater than the rate on
the origin of the contract.
Swap counterparty risk exposure

• Assume the set of dates at which swap payments are made


be {t1, t2,…, tn} and default of the counterparty that
receives fixed payments (B) took place between time t j-1
and tj. In this case, the loss for the party paying fixed is
given by

Lgd B  vt , ti 1 max sr t j , t n  k ,0 


n -1

i j

where sr is the swap rate at time tj and k is the swap rate at


the origin of the contract. Notice that this is the payoff of
a payer swaption (a call option on a swap).
Swap counterparty risk exposure
• Assume the set of dates at which swap payments are made
be {t1, t2,…, tn} and default of the counterparty that pays
fixed payments (A) took place between time tj-1 and tj. In
this case, the loss for the party receiving fixed is given by

Lgd A  vt , ti 1 maxk  sr t j , t n ,0 


n -1

i j

where sr is the swap rate at time tj and k is the swap rate at


the origin of the contract. Notice that this is the payoff of
a receiver swaption (a put option on a swap).
Wrong way risk
• A more general approach is to account for
dependence between the two main events
under consideration
– Exercise of the option
– Default of the counterparty
• Copula functions can be used to describe
the dependence structure between the two
events above.
Credit risk: long party
Vulnerable Call Swaptions: Financial Institution Paying Fixed

0,012

0,01

0,008 Independence
Perfect positive dependence

0,006

0,004

0,002

0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Swap credit risk: Baa
Fixed-Payer
Correl. 5years 10years 15years 20years 25years 30years
Rho
0 0,0259% 0,0686% 0,0935% 0,1141% 0,1385% 0,1678%
0,25 0,0536% 0,1448% 0,2178% 0,2574% 0,3056% 0,3658%
0,5 0,0813% 0,2210% 0,3420% 0,4007% 0,4726% 0,5637%
0,75 0,1090% 0,2971% 0,4663% 0,5440% 0,6397% 0,7617%
1 0,1367% 0,3733% 0,5905% 0,6873% 0,8068% 0,9597%

Fixed-Receiver
Correl. 5years 10years 15years 20years 25years 30years
Rho
0 0,0040% 0,0150% 0,0355% 0,0429% 0,0491% 0,0556%
0,25 0,0030% 0,0113% 0,0266% 0,0322% 0,0368% 0,0417%
0,5 0,0020% 0,0075% 0,0177% 0,0215% 0,0245% 0,0278%
0,75 0,0010% 0,0038% 0,0089% 0,0107% 0,0123% 0,0139%
1 0,0000% 0,0000% 0,0000% 0,0000% 0,0000% 0,0000%
Swap – CVA + DVA

N 1
 N -1 
NPV t0 , t N   Lgd B  EQ n   vt n , ti 1 maxRt n , t N   k ,0  FA n  1  FA n  
n 1  i n 
N 1
 N -1 
 Lgd A  EQ n   vt n , ti 1 maxk  Rt n , t N ,0  FB n  1  FB n 
n 1  i n 
Credit Risk Mitigation
CRM: theory
• In principle one can think of different techniques
to mitigate counterparty risk
1. Margin deposit at the origin of the contract
2. Position evaluation at daily on weekly period and
requirement of the payment of a collateral.
3. Netting agreement so that in case of default the net
exposure between the counterparty is liquidated.
From CVA to CSA
• According to the so-called ISDA Agreement the
credit mitigating techniques used apply netting
and the Credit Support Annex (CSA) requiring
periodic marking-to-market of the exposures.
• Before the crisis, it is not clear how much these
techniques were diffused in the market before
the crisis.
CSA pricing
• The frontier of pricing is about the impact of collateral.
• If collateral is posted every day, then the evaluation must
be made using daily discounting (EONIA) and it is risk-
free (apart from loss occurring overnight).
• Money for collateral must be collected by issuing debt.
The cost is then the spread between the funding interest
rate and EONIA.
• It is more so if “rehypothecation” of collateral is not
allowed.
Netting and rehypothecation
• Netting and rehypothecation are similar
• Netting:
– Counter party A posts collateral on the net exposure
with party B
• Rehypothecation:
– Counter party B is free to use the collateral posted by
counterparty A. So, for example it can use it to post
collateral with counter party C.
Funding Valuation Adjustment
(FVA)
Funding Benefits and Funding Costs
Unsecured Derivative Contracts
• Now consider an unsecured derivative contract
with a corporate client that is reflected on a mirror
contract (back-to-back) with a banking counter
party under the CSA rules.
• In this case, in principle we could compute
CF(t) = CFRF(t) – CVA + DVA – FVA
• Since FVA = FCA – FBA we may also write
CF(t) = CFRF(t) – CVA + DVA – FCA + FBA
FBA/DVA the double counting issue
The discussion on FVA
• According to traders, FVA must be included in the price of the
derivative because it is charged to the derivative desk by the
treasurer.
• According to academic the FVA should not be included in the
price of unsecured derivative contract because according to the
textbook the price should only reflect the hedge on the market, and
not the price accruing to bondholders and stockholders.
• According to accountants, the FVA should not be included because
the same contract would have different prices depending on the
parties.
• The industry, for the big names, includes CVA, DVA and FVA
(possibly allowing for double counting).
Counter party risk after Lehman
• Dodd-Frank (US) and EMIR (European Market
Infrastructure Regulation)
• Simple derivative contract: must be negotiated and
cleared in Central Clearing Counterparty (CCP).
• Complex derivative contracts: can be traded
between counter parties, but with the use of
collateral.
Margin Valuation Adjustment
(MVA)
• In CCP trades require an initial margin. In other
terms, while under CSA a linear contract that is
worth zero at origin does not require any initial
collateral, in CCP trades both parties must post a
collateral.
• Initial margin must be funded issuing debt. This is
called MVA. Collateral plays the role of what is
called «variation margin» in futures market.
• MVA is not re-hypothecable.
CCP Risk
• The CCP is providing guarantee to the
contracts by a system of funds that include:
• The initial margin posted by buyers and
sellers
• The guarantee fund that is posted by the
banks participating in the CCP
• The equity endowment of the CCP.

You might also like