Transcript LVA, FVA, CVA, DVA impacts on derivatives management

LVA, FVA, CVA, DVA impacts on
derivatives management
Christophe MICHEL
Head of RCCAD Quantitative Research
AFGAP-PRMIA April 5th 2012
Contents
02
13
19
26
31
Introduction to LVA
Introduction to CVA
Impact of CSA on CVA
A New Pricing Framework
Centralized Risk Management
Introduction to LVA
Pricing principle
3
Discounting Forecasted Flows
To make a long story short, pricing is a question of forecasting future fixed, floating or conditional flows in a
given currency and discounting them consistently with a funding level in this currency (in order to avoid
arbitrage opportunities).
For a single flow we can formally write:
Time t value of the future flow
in a given currency
Vt f  B f t, T  Ft
Time t forecasted value of the future flow
in the given currency
Time t value of a unit of currency paid at T
According to a funding level
Collateral Impact
In case of collateral agreement, additional flows are to be taken into account in the valuation process.
Indeed, to be posted a collateral has to be funded and remunerated with the funding rate, on the other hand,
the collateral posted is remunerated with a given collateral rate described within the collateral agreement.
The additional flows are all differentials of interest generated by the difference between funding and
collateral rates applied to the collateral amount posted.
These additional flows have to be funded and their price is straightforward as soon as we know their
forecasted value.
Valuation of Collateral Impact
4
Collateral Impact Value: an Implicit Problem
Coming back on the case of a single flow, additional flows linked to a collateral agreement depend upon the
collateral amount paid at each period:
t3
t0
t1
t2
Collateral-linked flow paid at t4 equal to
the differential of interest of collateral
Remuneration posted at t3 :
t4
t5
t6
Future derivative value
Ct3  rf t3   rc t3  3
In general, the collateral amount to be posted is directly deduced from the value of the collateralized
derivative i.e. including additional interest flows, say Vc, which is different from the value of the unsecured
derivative (Vf).
Hence, the value of a collateralized derivative depends upon this value: we’re facing an implicit problem.
Valuation of Collateral Impact: the Standard Case
5
Collateral Impact Value: the Special Case of Bilateral Contracts
In case of bilateral collateral agreement, we have:
Ct   Vt c
In this special case, under model assumptions, one can show that the price of a collateralised contract is
obtained by discounting with the collateral remuneration rate instead of the funding rate.
For a single flow, we can formally write:
Time t value of the future flow in a given
Currency including collateral
additionaml flows
Vt c  Bc t, T  Ft
Time t forecasted value of the future flow
in the given currency
Time t value of a unit of currency paid at T
according to the collateral remuneration rate
Note that in this case, the value of the collateralized contract doesn’t depend any more upon the
funding rate.
Note also that if a given derivative pays flows in a given currency but is collateralized in another currency then
its collateralized value will depend upon FX-Term Changes.
Valuation of Collateral Impact: the General Case
General Collateral
Many other possibilities exist in practice. For instance the unilateral collateral agreement can be written as
follows:
Vt c if theMtMof thecollateralized portfoliois positive
C t   
0 else
In this case, the collateral impact remains implicit but cannot be explicitly solved like in the standard case.
Pricing Method
To treat the general case, one has to solve a high dimensional optimal control problem.
This is generally not do-able in practice but it is well approximated by Monte Carlo simulation method
Market data are diffused on simulated paths
On each node of the simulation the Mark-to-Market of each product included in the collateral agreement are approximated
The collateral amount to be posted on each node can then be deduced
Additional flows resulting from collateral agreement can then be evaluated on each node
The average of their present value approximate the LVA impact
Note that the collateral impact crucially depends upon the global portfolio within the collateral contract
and cannot be treated on a stand alone basis. A contribution to the global adjustment can be computed.
6
Before the crisis
7
.
Pre-crisis
Bor rates were market convention for discounting all trades, independently on the counterparty, CSA, etc…
Implicit assumptions
If the counterparty was uncollateralised, CA-CIB could lend/borrow (unsecured) @ Bor flat
If the counterparty was collateralised, it meant the interest paid on the collateral posted/received was Bor (which is consistent with
interest - mainly OIS as defined in the CSA)
Historically
Access to liquidity was taken as granted
EONIA, 3m Euribor 6m Euribor: historical values*
Basis were tight
6%
These assumptions were
verified
5%
4%
Eonia
3%
3m Euribor
6m Euribor
2%
Source: Bloomberg
1%
2000
2001
2002
2003
2004
2005
2006
2007
The IR pricing framework was a SINGLE CURVE one based on -Bor.
During and post crisis
8
.
During and post crisis
Liquidity crisis
Drying up of lending/borrowing between banks
Creditworthiness of banks questioned
Pre-crisis funding assumptions no longer hold
Unsecured funding much more expensive
Interest paid on collateral (usually OIS) significantly diverged from Bor rates
3M EONIA/Euribor spread (LHS)
5Y 3m EONIA/Euribor basis (LHS)
5Y 3m/6m Euribor basis (RHS)
EONIA, 3m Euribor 6m Euribor: historical values*
6%
5%
200
25
150
20
100
15
50
10
4%
3%
2%
0
1%
0%
2007
2008
Eonia
2009
2010
3m Euribor
2011
2012
6m Euribor
2013
-50
2007
5
0
2008
2009
2010
2011
2012
2013
Growing importance of Credit Support Annex (CSA)
Liquidity & credit risk issues are more and more actively managed by banks.
The main trend is a clear shift towards growing use of CSA as collateralization remains among the most
widely used methods to mitigate counterparty credit risk in the OTC derivatives market.
Growth of value of total collateral (USD billions)
Source: ISDA Margin Survey 2011
Growth of collateral agreements
9
Credit Support Annex (1/3)
10
Each CSA determines the collateralization terms between counterparties: bilateral or
unilateral, type of collateral, currency, haircut, threshold, minimum transfer amount and all
other details are stipulated in the CSA. The amount of margin posted and the margin
interest – and consequently the valuation of the structure – will depend on the CSA terms.
CSA determines the range of assets that may be posted as a
collateral – cash in different currencies, government bonds or
corporate or mortgage backed securities
The threshold level will determine how much of the exposure is
collateralized. Margin transfers will be made only if the minimum
transfer amount is exceeded.
Bilateral CSA collateral profile
The choice of collateral currency will alter the expected
return as the cash funding terms are tied to the corresponding
currency’s overnight rate (need of cross currency swap if the
collateral currency is not the same as the deal currency)
Counterparty posts collateral and receives interest
+
For cash-only CSA’s, the funding curve corresponds to the
specified collateral interest rate
PV
Bank posts collateral and receives interest
Credit Support Annex (2/3)
Description of the CSA
Margin transfer form
Unilateral – only one way transfers – CA CIB posts a collateral but the counterparty dos not (required by some
supranational entities )
Bilateral – symmetric transfer terms
Margin call frequency
Daily margin call is frequency required (becoming a market standard). Longer than daily margin call frequency can
be practical for markets and assets that are not volatile
Threshold
This is the exposure amount below which collateral is not required (the threshold represents an amount of
uncollateralized exposure). A threshold of zero implies that any exposure is collateralized
Minimum transfer amount
The smallest amount of collateral that can be transferred.
It is used to avoid the workload associated with a frequent transfer of insignificant amounts of collateral
11
Credit Support Annex (3/3)
12
What is a standard CSA ? Pay attention to the detail when negotiating CSA terms
Standard CSA terms
Bilateral margin transfer form: symmetric
transfer terms for both parties
Non standard agreements
No CSA
Daily margin call frequency (weekly can be
also accepted as standard)
Unilateral CSA (can be unilateral in our favour;
often supranational institutions require an
unilateral CSA in their favour)
No threshold – up to 5M threshold is considered
as reasonable
CSA with very high threshold – 50-100M
threshold will impact pricing
Cash and G7 bonds (haircuts – 0%-2% on short
term maturities and 5-10% on longer term
maturities)
CSA with rating triggers (example: CA-CIB
needed to post an independent amount to EIB
due to the S&P rating action)
In EUR or in USD, remunerated at EONIA
FLAT or Fed Funds FLAT
Sub-optimal CSAs – negative spreads on
cash collateral, securities received that can’t be
re-hypothecated
Introduction to CVA
CVA Basics – Definition
14
To be or not to be paid ?
The Credit Value Adjustment appears in the pricing framework when a credit risk is taken into account.
Formally, one can explicit this risk by multiplying each payment flow by the following function:
1 if theflow is actuallypaid at T
FT a flow paid at T becomesFT  
0 if theflow isn' t paid at all
In case of a counterparty default at time T, the key question is to measure the exposure.
A priori, the exposure is the sum of all positive mark to market of each transaction remaining at time T with
this counterparty.
In case of netting agreement with the counterparty, the exposure becomes the sum of all netted positive
mark to market of each set of transactions within each netting agreement contract.
A classical formula
A classical CVA formula can be expressed as the amount of discounted future Expected Losses
CVA  LGD *
Maturity

t 0
Loss Given Default
PDt 1,t * EPEt * DFt
Default Probability
Exposure at Default (Discounted)
CVA Basics – Key Ingredients
15
Expected Positive Exposure (EPE)
Calculated via simulations process (Monte Carlo…)
Computation including netting and collateral agreements
Involves only the Positive Exposures in case of Counterparty Default
Definition of Exposure linked to the mark to market of transaction
Evaluated Contingent on the default of the counterparty
–
including right way / wrong way risks
CSA or break clause have a huge impact on EPE
Market CDS Curve
Default Probability
Implied from CDS spreads (market-implied) or,
Historical default probabilities
Loss Given Default / Recovery Rate
Market Implied (where possible) : LGDMarket
Internal Recovery measure : LGDInternal
CVA Basics – Key Ingredients
16
Risk parameters:
Peak Exposure: Maximum of the MtM of the transaction over its lifespan, given a high confidence level (VaR
95%)
Loan Equivalent: Average EPE over time (measure to determine risk equivalent in terms of loan for economic
capital calculation)
Tail Credit Risk (CVaR): average of 5% maximum potential losses
10y EUR IRS CA-CIB Receives Fixed 2.3%
CVA Basics – Interpretation
CVA as an Accounting Provision
Measure of Expected Loss
Use of Historical Default Probabilities
Inactive Risk Management (Acceptance to ‘carry’ the Credit Risk)
No Market Hedges in Place (IR, FX or Credit)
Credit Risk remains in the individual traders books
CVA as a Mark to Market Adjustment
Market Cost of Dynamically Hedging Counterparty Risk
Use of Market Implied Default Probabilities (implied from CDS)
Active Risk Management Strategy in place
Local Sensitivities
Jump to Default Risks
Concentrations / Wrong Way risks
Market Hedges in place (IR, FX and Credit)
Credit Risk transferred to a centralised CVA desk
17
CVA Basics – Summary
CVA is a fair market adjustment to the derivatives portfolio
CVA charges are there to offset losses in the global CVA portfolio due to new trades
Incremental impact on the portfolio
CVA P&L is flat provided the cash transfers take place
CVA hedges are initiated in order to offset future CVA P&L volatility for CA-CIB
Due to portfolio MTM movements (IR, FX, Credit Hedges)
Due to counterparty CDS Spread Changes (Credit Hedges)
At inception: CVA P&L is flat, Hedge P&L is flat:
1. CDS spreads increase: CVA increases (Loss) versus Hedge P&L (Gain)
2. CDS spreads decrease: CVA decrease (Gain) versus Hedge P&L (Loss)
3. CDS spreads stay static: Negative carry on Hedge (Loss) is offset against positive carry of CVA (Gain)
CVA Credit Hedges also provide
Jump to default risk management
Basel III capital reductions
Debit Value Adjustment (DVA) is the CVA seen from the counterparty. To take it into account allows symmetrical
views on the shared portfolio.
18
Impact of CSA on CVA
CVA impact of collateral on stand alone transaction (IRS)
The Credit Support Annex is a key tool for credit risk mitigation
EPE computation includes netting and collateral agreements
EPE is calculated up to the threshold
Above threshold, mark to market drift is calculated on margin call period + collateral lag period (10days)
CSA features are built into EPE simulations and consequently have an impact on CVA
Collateral features
Trade description
Start date:
10 Apr 2012
Unilateral vs Bilateral
Maturity:
10Y
Threshold
Notional:
EUR 100M
Frequency
CA-CIB receives:
2.30% (SA, act/360)
Currency
CA-CIB pays:
6M Euribor (SA, 30/360
Minimum Transfer Amount (MTA)
Counterpart:
XXX
5Y cds:
143bps
Internal rating:
B
20
CVA impact of collateral on stand alone transaction (IRS)
21
Risk exposure is maximal for CA-CIB
No collateral agreement in place
EPE
Credit Risk
Loan equivalent
10,000,000
Peak exposure:
9 508 071
9,000,000
Market CVA:
137 124 (1.49 bps p.a)
7,000,000
Historical CVA:
9 942
8,000,000
6,000,000
(0.11 bps p.a)
5,000,000
4,000,000
3,000,000
2,000,000
Market CVA ~ hedging cost of loan lquivalent on cds market
1,000,000
0
12
20
CSA agreement in place
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
Sharp decrease of risk profile due to CSA risk mitigation
CSA in place
Threshold: 0
Bilateral
MTA: 0
1,600,000
Frequency: daily
Currency: EUR
1,400,000
EPE
Credit risk
Loan equivalent
1,200,000
1,000,000
Peak exposure:
1 378 356
Market CVA:
27 948 (0.30 bps p.a)
Historical CVA:
1 850
800,000
600,000
(0.02 bps p.a)
400,000
200,000
0
12
20
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
CVA impact of collateral on stand alone transaction (IRS)
Weekly frequency: drift is computed on one week +
10 days collateral lag
CSA agreement in place
CSA in place
Threshold: 0
Bilateral
MTA: 0
Frequency: weekly
EPE
Currency: EUR
22
Credit risk
Loan equivalent
1,800,000
1,600,000
1,400,000
1,200,000
Peak exposure:
1 704 987
Market CVA:
33 938
(0.37 bps p.a)
Historical CVA:
2 252
(0.02 bps p.a)
1,000,000
800,000
600,000
400,000
200,000
0
12
20
CSA agreement in place
CSA in place
Threshold: 0
Bilateral
MTA: 0
1,600,000
Frequency: daily
Currency: EUR
1,400,000
13
20
14
20
15
20
EPE
16
20
17
20
Credit risk
18
20
19
20
20
20
21
20
Loan equivalent
1,200,000
1,000,000
Peak exposure:
1 378 356
Market CVA:
27 948 (0.30 bps p.a)
Historical CVA:
1 850
800,000
600,000
(0.02 bps p.a)
400,000
200,000
0
12
20
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
CVA impact of collateral on stand alone transaction (IRS)
23
CSA agreement in place
EPE
CSA in place
Threshold: 500K
Bilateral
MTA: 0
1,800,000
Frequency: daily
Currency: EUR
1,600,000
Credit risk
Loan equivalent
2,000,000
1,400,000
1,200,000
Peak exposure:
1 878 356
1,000,000
800,000
Market CVA:
43 346 (0.47 bps p.a)
Historical CVA:
3 263 (0.024 bps p.a)
600,000
400,000
200,000
0
12
20
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
CSA agreement in place
CSA in place
Threshold: 0
Bilateral
MTA: 0
1,600,000
Frequency: daily
Currency: EUR
1,400,000
EPE
Credit risk
Loan equivalent
1,200,000
1,000,000
Peak exposure:
1 378 356
800,000
Market CVA:
27 948 (0.30 bps p.a)
600,000
Historical CVA:
1 850
400,000
(0.020 bps p.a)
200,000
0
12
20
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
CVA impact of collateral on stand alone transaction (CRS)
Collateral features
Trade description
Start date:
10 Apr 2012
Maturity:
10Y
Unilateral vs Bilateral
Notional:
EUR 100M
Threshold
EURUSD:
1.3091
Frequency
CA-CIB pays:
6M Euribor (SA, act/360)
Currency
CA-CIB receives:
USD 2.25% (SA, act/360)
Minimum Transfer Amount (MTA)
Notional exchange:
Beg and end
Counterparty
Name:
XXX
5Y cds:
143bps
Internal rating:
B
24
CVA impact of collateral on stand alone transaction (CRS)
25
Risk exposure is maximal for CA-CIB
No collateral agreement in place
EPE
Credit Risk
Loan equivalent
100,000,000
90,000,000
Peak exposure:
94 738 968
Market CVA:
1 718 224
(18bps p.a)
Historical CVA:
192 376
(2 bps p.a)
80,000,000
70,000,000
60,000,000
50,000,000
40,000,000
30,000,000
20,000,000
10,000,000
0
12
20
CSA agreement in place
CSA in place
Threshold: 0
Bilateral
MTA: 0
Frequency: daily
Currency: EUR
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
CSA drastically minimizes risk exposure on cross-currency
EPE
Credit risk
Loan equivalent
4,500,000
4,000,000
3,500,000
Peak exposure:
4 181 994
Market CVA:
227 864
(2.40 bps p.a)
Historical CVA:
22 381
(0.24 bps p.a)
3,000,000
2,500,000
2,000,000
1,500,000
1,000,000
500,000
0
12
20
13
20
14
20
15
20
16
20
17
20
18
20
19
20
20
20
21
20
A new pricing framework
Collecting Information
Market Information
In the new pricing framework, market prices remain of course a central source of information but one has to
interpret them consistently.
A price is at least dependent with the CSA of the product priced and when one observe the price of an OTC
product on a screen the question of the implicit CSA of this price has to be solved.
To fix this problem, a standardization of market practice arise
Standard swap are assumed cleared with a collateral in the currency of the swap with a collateral remuneration at
OIS rate.
Multi-currency products are often assumed collateralized in USD with a collateral remuneration at USDOIS rate.
Physical settlement swaption quotation often assumes the underlying swap to be a standard swap and quotes a
forward premium in order to avoid the LVA impact on the discounted premium.
Once the implicit CSA question is answered, it becomes possible to strip from market prices cleaned
market data (IR curves per tenor, FX forward, default probabilities, inflation forward, liquid volatilities, etc)
Additional Observable Information
Our own funding level will also intervene in the new pricing framework even if its monetisation isn’t
straightforward.
From the CVA point of view, some new deal information are to be taken into account like break clause.
Lastly, the main new information type to input and which will impact all quotations is all information related
to the client (CSA, netting agreement, rating, …)
27
Forecasting Data
Diffusion Data
In addition to market data and client data, we have seen that most of xVA are linked to global portfolio
measures based on Monte Carlo simulations.
Once the diffusion model is defined, its calibration can partly be implicit (calibration on a set of market prices)
but will mainly linked to historical market behaviour.
Like for any diffusion model design for structured derivatives pricing, the calibration process is dependent with
the sophistication of the diffusion process (from a theoretical point of view) and availability of implicit and
historical information.
A key point to focus on is the joint behaviour between all class of market risks (let’s say correlation to
simplify) implicitly defined in the diffusion process. It is indeed well known that this point will drive market risk
netting from the global portfolio point of view.
Missing Credit Data
For a great number of counterparties there is no quoted CDS.
The CVA needs information on default probabilities and LGD for any counterparty. This information is to be
forecasted.
Based on all internal works on counterparty risk management (sector classification, internal ratings,
historical recovery rates, …), it becomes possible to map any counterparty on market information.
28
“Stand alone Price” + “Portfolio-Linked Adjustment”
29
The new pricing workflow can be summarized as follows:
By definition, the “stand alone” price of a product is its value independent with the portfolio of deals
Calculation Engines
Stripping
under CSA
assumptions
Raw
Market Data
Cleaned
Market Data
Internal
Client
Portfolio
Internal
Client
Data
Data Forecasting
Engine
Global Simulation
Engine
Bilateral
CSA Input
Portfolio adjustment
(CVA, DVA, FVA,
non standard LVA)
+
Product Description
(Termsheet)
Client
t0
t1
t2
t3
t
4
t5
t6
Stand-Alone
Pricing Engine
Stand alone Price
About Portfolio-Linked Adjustments
A Non Linear Adjustment
Generally speaking, portfolio-linked adjustments aren’t linear i.e. the value adjustment of a given deal
depends upon the portfolio (and market data) at the moment the valuation.
Same trade with two counterparties, with different credit profiles or portfolio vis-à-vis the bank, will have
different market prices.
Example: Client receives Fixed on 10y Swap
Mid Market 10y Swap Rate = 2.31%
–
–
Client 1: Rated A- receives 2.27%
Client 2: Rated B+ receives 2.23%
4bp CVA charge
8bp CVA charge
In practice, an incremental portfolio value adjustment relating to the new trade is used. It corresponds to the
differential of portfolio value adjustment with and without the deal.
This incremental value can either be positive or negative and a pricing policy is to be defined.
Adjustment Breakdown
The global portfolio value adjustment is made with several sources of risks (credit, liquidity, funding) and with
several market data qualities (pure market data versus forecasted data).
To look into these, it is meaningful to evaluate the global portfolio value adjustment breakdown based on
those different sources of risk and data qualities.
Similarly, this breakdown can be computed for incremental portfolio value adjustments at the deal level.
30
Centralized Risk Management
Hedging Risks
Stand-alone Value Risk Management
A priori, the stand-alone product value can be hedged trading desk per trading desk without particular need
of centralisation.
Nevertheless, the multiplicity of CSA may induce “not so material” additional sources of risks for trading desks
(like cross-currency basis swap margin risks) which can pollute trading desk risk management and which
could be centralised.
We have to keep in mind that any transaction has, from valuation and risk management point of view, a
stand-alone part and a portfolio part. The implicit CSA of the stand-alone part of the deal isn’t necessarily
the actual CSA of the deal.
Funding Risk
Funding risk could in theory be managed on a stand-alone basis since this risk isn’t explicitly linked to the
portfolio view.
But this risk can’t be directly risk managed in the market.
Hence, this risk has to be risk managed jointly with ALM. Indeed, the stake is the remuneration or the
charge of this adjustment. This has to be made consistently with the way these remuneration/charge are
monetised by ALM.
Global simulation tools can of course be very used to obtain a complete view of the distribution of funding
needs at any future maturity (expectation, standard deviations, … of funding needs including collateral
simulation).
32
Hedging Risks
Portfolio Adjustment Value Risk Management
By definition, this adjustment has to be centrally risk-managed. Indeed:
The value to risk manage make sense at the portfolio level only.
Global metrics to evaluate (EPE) are time consuming and depend upon a wide range of market data then
standard complete local management is just numerically impossible and alternative risk management
strategies has to be developed.
Some adjustments are based on forecasted/mapped data which as to be specifically risk managed
At the centralized level, impacts of global events like defaults, CSA changes, Funding changes can be
anticipated, studied and quantified.
Conversely, a centralized desk can be pro-active on all portfolio-linked risks:
CSA change
Waives/Further discounted counterparties (to take into account liquidity providers or to be aligned with the Banks
business strategy.
Wrong way / Right way risk.
RWA management
A centralised desk ensure a unique and consistent FO view internally and externally.
33
Capital requirements for Counterparty Credit Risk (CCR)
Basel I
Capital CCR only covers default risk: MtM + standardized add-on based on issuer type, underlying and maturity
No capital requirement for MTM loss due to change in counterparty credit spread
Basel II
Capital CCR only covers default risk
Internal Ratings-Based approach: Internal assessment of default risk via internal probability of default and LGD
No capital requirement for MTM loss due to change in counterparty credit spread
EPE: counterparty exposure estimated using non-stressed market data, on a 1Y horizon
Basel III
Basel III Capital CCR = Capital Default (Basel II) + Capital CVA
Capital CVA: additional Capital requirement for MtM loss due to change in counterparty credit spread
Capital CVA: Market Risk capital charge estimated using stressed VaR on credit instruments
EPE: counterparty exposure estimated under stressed parameters on the full maturity
CVA VaR is calculated Net of Eligible Hedges
Other adjustments:
–
–
Assuming a higher correlation between financial institutions in the supervisory formula
Extending the collateral lag period from 10 days to 20 days
34
Regulatory impact on capital requirements
Changing Landscape
CVA VaR now a key component in Return on Capital calculations for New Trades
DVA not an allowable offset under Basel III, so competitors using DVA for pricing must consider capital usage
Active CDS Hedging encouraged  Influences CDS pricing and volatility
Innovation needed to help reduce RWA costs
Cost of Novations / Intermediations now closely scrutinised
Need for CVA Pricing Tools to consider Basel III impacts
CVA Capital Methodology linked more closely to Market Dynamics
Regulatory Methodology is entirely Market Based --> No reliance on Historical Default Measures
Basel III extends maturity of Exposure at Default to the Full EPE Profile (previously 1y EPE under Basel II)
Capital Relief provided by Eligible Hedges
Includes Single name CDS, CDS index, CCDS and other hedges that directly reference the Counterparty
Hedging a CVA portfolio releases Capital
Challenge: focus on regulatory influence vs focus on competitive pricing
35
Central Desk Mandate
Centralise and risk manage FO derivatives counterparty and portfolio-linked risk
Minimise xVA P&L impact for FO through:
Systematic Macro hedging of the overall xVA P&L impact
Single Name default hedging (where practical)
Active participation in RWA management
Target business model is to incentivise Sales Force not only on revenue generation but also on RWA and liquidity consumption
Sales recognition methodology has to be consistently defined
Establish a Pricing Policy aligned with the internal xVA methodology
Educate and train sales and support functions in xVA related issues
Involvement in Regulatory Discussions / Decisions affecting xVA
Involvement in ISDA / CSA discussions affecting xVA.
Need to remain Competitive in Pricing …
Must remain business-focused and incentivise the right types of trades:
– Waiver / Exemption for Target Clients / Businesses
– Additional Charge for Wrong Way trades
– Benefits for Right Way Trades / Unwinds / Portfolio Diversifications
– Need for feedback from clients relating to current market practices
– Encourage risk mitigation (eg negotiation of Credit Support Annex (CSA) with counterparties)
36
Interaction with other FO teams
Central Desk is there to Support the Business:
In accurately pricing Credit and Liquidity into new trades, and
To manage the credit risk of derivatives portfolio through dynamic CVA hedging
To achieve this Central Desk needs:
Active Dialogue with Trading / Sales / Structuring on new trades, and existing portfolio (risk reduction opportunities)
Involvement in Complex Trades, especially where replacement costs in default are hard to define
Clear Understanding of all key components of the trades
Identification of potential risk mitigants - break clauses, legal details (CSA, netting), etc
Feedback from Clients !
... and on problem names
To ‘wall cross’ the CVA team in case non-public information needs to be shared. To check PV details in front/back office systems.
Involve Legal team to ensure appropriate modus operandi is followed
Goal is to streamline process
Revised guidelines for involvement of Central desk in trade analysis (e.g. where CVA > 50k EUR, ... thresholds to be discussed/revised)
Improvement of CVA / LVA pricing tools (including capital costs/usage)
Clear communication is critical
... and to Adapt to Change
Methodology is dynamic and intended to support business initiatives
37
Disclaimer
38
© 2012, CRÉDIT AGRICOLE CORPORATE AND INVESTMENT BANK All rights reserved.
The information in this document (the “Information”) has been prepared by Crédit Agricole Corporate and Investment Bank or one of its affiliates (“Crédit Agricole CIB”)
for informational purposes only.
Nothing in this document is to be construed as an offer for services or products or as an offer or solicitation for the purchase or sale of securities or any other financial
product. The Information has no regard to the specific investment objectives, financial situations or particular needs of any recipient.
While the Information is based on sources believed to be reliable, no guarantee, representation or warranty, express or implied, is made as to its accuracy, correctness
or completeness. Crédit Agricole CIB is under no obligation to update the Information.
Crédit Agricole CIB does not act as an advisor to any recipient of this document, nor owe any recipient any fiduciary duty and the Information should not be construed
as financial, legal, regulatory, tax or accounting advice. Recipients should make their own independent appraisal of the Information and obtain independent professional
advice from appropriate professional advisers before embarking on any course of action.
In no event shall Crédit Agricole CIB or any of its directors, officers or employees have any liability or responsibility to any person or entity for any direct or
consequential loss, damage, cost, charge, expense or other liability whatsoever, arising out of or in connection with the use of, or reliance upon, the Information.
Furthermore, under no circumstance shall Crédit Agricole CIB have any liability to any person or entity for any loss or damage, in whole or in part, caused by, resulting
from, or relating to, any error (negligent or otherwise), omission, condition or other circumstances within or outside the control of Crédit Agricole CIB or any of its
directors, officers or employees in connection with the procurement, collection, compilation, analysis, interpretation, communication or delivery of the Information.
This document and the Information are confidential and may not be copied, reproduced, redistributed, passed on, published, reproduced, transmitted, communicated or
disclosed, directly or indirectly, in whole or part, to any other person without Crédit Agricole CIB’s prior written consent.
Recipients of this document in jurisdictions outside the United Kingdom should inform themselves about and observe any applicable legal or regulatory requirements in
relation to the distribution or possession of this document to or in that jurisdiction. In this respect, Crédit Agricole CIB does not accept any liability to any person in
relation to the distribution or possession of this document to or in any jurisdiction. This document is not directed at, or intended for distribution or use by, any person or
entity who is a citizen or resident of any jurisdiction where such distribution, publication, availability or use would be contrary to applicable laws or regulations of such
jurisdictions.
United Kingdom: Crédit Agricole Corporate and Investment Bank is authorised by the Comité des Etablissements de Crédit et des Entreprises d’Investissement
(CECEI) and supervised by the Commission Bancaire in France and subject to limited regulation by the Financial Services Authority. Details about the extent of our
regulation by the Financial Services Authority are available from us on request. Crédit Agricole Corporate and Investment Bank is incorporated in France and registered
in England & Wales. Registered number: FC008194. Registered office: Broadwalk House, 5 Appold Street, London, EC2A 2DA.