counterparty-risk
Skill16 starsChanged 8 months agoMITView on GitHub
Margin, collateral, CVA/DVA/FVA, central clearing, netting, wrong-way risk, and credit support framework. Credit Exposure - The loss if a counterparty defaults today = max(0, mark-to-market of all trades with that counterparty) - Only positive MTM creates exposure (you lose if they owe you and default) - Negative MTM: no credit risk to you (but you are their credit risk — DVA perspective) Expected Exposure (EE) - EE(t) = E[max(0, V(t))] — expected positive exposure at future time t - Computed…
1 other project uses a version of this.
Adopted by
Most recently taken up by
- IngAs8/ClaudeSkilltheir own version, first seen Sep 28, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- IngAs8/ClaudeSkill★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository83 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- live-trading-monitoringSkill
- monte-carlo-simulationSkill
- overfitting-preventionSkill
- research-notebooksSkill
- strategy-evaluationSkill
- walk-forward-optimizationSkill
- cross-chain-arbitrageSkill
- crypto-derivativesSkill
- defi-protocolsSkill
- mev-strategiesSkill
- on-chain-analyticsSkill
- stablecoin-analysisSkill
- tokenomicsSkill
- yield-farmingSkill
- alternative-dataSkill
- data-qualitySkill
- data-vendorsSkill
- feature-engineeringSkill
- sentiment-analysisSkill
- web-scraping-financeSkill
- binomial-treesSkill
- black-scholesSkill
- credit-derivativesSkill
- exotic-optionsSkill
- hedging-strategiesSkill
- interest-rate-derivativesSkill
- monte-carlo-pricingSkill
- structured-productsSkill
- vol-surfacesSkill
- vol-tradingSkill
- execution-algosSkill
- latency-optimizationSkill
- market-microstructure-theorySkill
- order-flow-analysisSkill
- performance-metricsSkill
- smart-order-routingSkill
- tick-data-processingSkill
- transaction-cost-analysisSkill
- central-bank-analysisSkill
- commodities-fundamentalsSkill
- economic-indicatorsSkill
- fx-fundamentalsSkill
- geopolitical-riskSkill
- yield-curve-analysisSkill
- alternative-allocationsSkill
- benchmark-trackingSkill
- black-littermanSkill
- factor-investingSkill
- markowitz-optimizationSkill
- portfolio-attributionSkill
- rebalancing-strategiesSkill
- risk-paritySkill
- alpha-researchSkill
- bayesian-methodsSkill
- deep-learning-tradingSkill
- factor-modelsSkill
- kalman-filtersSkill
- ml-for-financeSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
5 rules. About 0 are common, 5 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Only positive MTM creates exposure (you lose if they owe you and default)
- Distinctive: line 122
Must be identified trade by trade.
- Distinctive: line 135
Default management: CCP has established waterfall for member default
- Distinctive: line 149
Default fund mutualization: surviving members bear losses of defaulting member
- Distinctive: line 203
All OTC counterparties must have approved credit limits with PFE monitoring
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Counterparty Credit Risk
Margin, collateral, CVA/DVA/FVA, central clearing, netting, wrong-way risk, and credit support framework.
When to Activate
- User evaluating counterparty exposure in OTC derivative portfolios
- Computing CVA, DVA, or FVA for derivative pricing
- Designing collateral agreements (CSA) and margin frameworks
- Assessing the impact of central clearing mandates
- Analyzing wrong-way risk in derivative portfolios
- Managing bilateral vs. cleared counterparty risk
Core Concepts
Counterparty Exposure Fundamentals
Credit Exposure
- The loss if a counterparty defaults today = max(0, mark-to-market of all trades with that counterparty)
- Only positive MTM creates exposure (you lose if they owe you and default)
- Negative MTM: no credit risk to you (but you are their credit risk — DVA perspective)
Expected Exposure (EE)
- EE(t) = E[max(0, V(t))] — expected positive exposure at future time t
- Computed via Monte Carlo simulation of risk factor paths
- EE profile: typically hump-shaped for swaps (increases then decreases as payments are exchanged)
Potential Future Exposure (PFE)
- PFE(t) = quantile of the positive exposure distribution at time t (e.g., 97.5%)
- Represents a worst-case exposure measure analogous to VaR
- Used for credit limit monitoring: ensures exposure stays within approved counterparty limits
Expected Positive Exposure (EPE)
- EPE = (1/T) * integral of EE(t) dt from 0 to T — time-averaged expected exposure
- Used in regulatory capital calculations (SA-CCR, IMM)
- Effective EPE: non-decreasing version of EPE (captures roll-off risk)
Netting and Close-Out
Close-Out Netting
- ISDA Master Agreement allows netting of all trades with a counterparty upon default
- Net exposure = max(0, sum of MTM across all trades) — dramatically reduces exposure
- Without netting: exposure = sum of max(0, MTM_i) — each positive trade is at risk
- Netting benefit is greatest when portfolio has offsetting trades (some positive, some negative MTM)
- Legal enforceability: netting must be legally valid in the counterparty's jurisdiction
Payment Netting
- Net cashflows on the same date in the same currency
- Reduces settlement risk (Herstatt risk) but smaller impact than close-out netting
Margin Framework
Initial Margin (IM)
- Posted at trade inception to cover potential future exposure during close-out period
- Covers the expected change in portfolio value during the margin period of risk (MPOR)
- Bilateral: ISDA SIMM (Standard Initial Margin Model) based on sensitivities
- Cleared: CCP-determined, typically based on historical simulation VaR at 99%+ confidence
- Two-way exchange: both parties post IM (segregated at a third-party custodian)
Variation Margin (VM)
- Daily (or intraday) mark-to-market settlement
- Winning party receives VM; losing party posts VM
- Reduces credit exposure to approximately one day's potential move
- Minimum transfer amount (MTA): typically $500K-$1M to avoid operational burden
- Threshold: exposure level below which no VM is exchanged (reduces to zero under new bilateral rules)
Margin Period of Risk (MPOR)
- Time from last margin exchange to close-out of defaulted portfolio
- Bilateral: typically 10 business days
- Cleared: typically 5 business days (more liquid, standardized products)
- Includes: margin call, dispute resolution, close-out/replacement of trades
Methodology
Credit Valuation Adjustment (CVA)
CVA is the market price of counterparty credit risk — the expected loss due to counterparty default:
Unilateral CVA
- CVA = (1 - R) * integral from 0 to T of EE(t) * dPD(t)
- R = recovery rate (typically 40% for senior unsecured)
- PD(t) = cumulative default probability (from CDS spreads or internal ratings)
- EE(t) = expected exposure at time t (from Monte Carlo)
- Discrete approximation: CVA = (1-R) * sum over time buckets of EE_i * (PD_{i} - PD_{i-1})
Bilateral CVA (with DVA)
- DVA = debit valuation adjustment = value of own default to the counterparty
- DVA = (1 - R_own) * integral of ENE(t) * dPD_own(t)
- ENE = expected negative exposure (exposure from counterparty's perspective)
- Net adjustment: CVA - DVA
- DVA is controversial: own credit deterioration appears as a profit
CVA Sensitivities
- CVA spread sensitivity: how CVA changes with counterparty CDS spread
- CVA exposure sensitivity: how CVA changes with underlying risk factors
- CVA must be hedged: CDS on counterparty (for spread risk), underlying hedges (for exposure risk)
Funding Valuation Adjustment (FVA)
- FVA captures the cost of funding uncollateralized derivative positions
- FVA = integral of (funding_spread) * (expected_funding_needs(t)) dt
- Uncollateralized trades require funding; collateralized trades are self-funding
- FVA is debated: some argue it is a real cost, others argue it double-counts with DVA
- In practice: most dealers include FVA in pricing, adding 1-10bp to uncollateralized derivative prices
SA-CCR (Standardized Approach for Counterparty Credit Risk)
Basel III replacement for CEM (Current Exposure Method):
- Replacement cost (RC): max(V - C, 0) where V = portfolio MTM, C = collateral held
- Potential future exposure (PFE): aggregated add-on based on trade notionals, maturity, and asset class
- Exposure at default: EAD = alpha * (RC + PFE), where alpha = 1.4
- Add-on calculation: notional * supervisory factor * maturity factor * delta adjustment
- Supervisory factors by asset class: IR (0.5%), FX (4%), equity (32%), credit (0.54-6%), commodity (18-40%)
- Netting benefit captured through hedging set aggregation
Wrong-Way Risk
Specific Wrong-Way Risk (SWWR)
- Exposure increases when counterparty credit quality deteriorates due to a direct link
- Example: sold a put option to a bank on its own stock — if stock crashes, counterparty is more likely to default AND the option is deep ITM
- Must be identified trade by trade; cannot be modeled statistically
General Wrong-Way Risk (GWWR)
- Positive correlation between counterparty default probability and market exposure
- Example: interest rate swap with a corporate — if rates rise, corporate may struggle AND swap MTM increases
- Modeled by correlating the exposure simulation with the default process
- Increases CVA significantly: can be 50-200% higher than independent case
Central Counterparty (CCP) Clearing
Advantages
- Multilateral netting: reduces total system exposure
- Standardized margining: IM and VM rules, daily settlement
- Default management: CCP has established waterfall for member default
- Transparency: standardized products, centralized reporting
CCP Default Waterfall
- Defaulting member's initial margin
- Defaulting member's default fund contribution
- CCP's own capital (skin in the game)
- Non-defaulting members' default fund contributions
- CCP's remaining capital
- Assessment powers (additional calls on members)
Risks of CCP Clearing
- CCP concentration risk: "too big to fail" CCPs create systemic risk
- Procyclical margin calls: IM increases in stress, forcing liquidation
- Default fund mutualization: surviving members bear losses of defaulting member
- CCP interoperability: multiple CCPs can fragment netting
Examples
CVA Calculation
Interest rate swap with Corporate X:
Notional: $100M, 5-year maturity
Expected exposure profile: EE = [$0, $2M, $3.5M, $4M, $3M] (annual)
CDS spread of Corporate X: 200bp (implies PD approximately 3.3%/year)
Recovery rate: 40%
CVA = (1-0.40) * sum of EE_i * delta_PD_i
= 0.60 * ($0*0.033 + $2M*0.033 + $3.5M*0.032 + $4M*0.031 + $3M*0.030)
= 0.60 * ($0 + $66K + $112K + $124K + $90K)
= 0.60 * $392K = $235K
CVA = $235K or 2.35bp of notional.
This cost is embedded in the swap pricing as a spread adjustment.ISDA SIMM Initial Margin
Portfolio with counterparty Y:
IR delta: $5M DV01 in USD
FX delta: $2M sensitivity to EUR/USD
Equity vega: $500K sensitivity to S&P vol
SIMM calculation (simplified):
IR risk charge: $5M * risk_weight(USD_2Y) = $5M * 61bp = $305K
FX risk charge: $2M * risk_weight(EUR/USD) = $2M * 7.5% = $150K
Equity risk charge: $500K * risk_weight(large_cap) = $500K * 25% = $125K
Aggregated (with diversification): sqrt($305K^2 + $150K^2 + $125K^2) = $370K
Initial margin requirement: $370K posted by each party (two-way).Wrong-Way Risk Impact
CDS sold to Bank Z on a correlated reference entity:
Standard CVA (independent default): $180K
Correlation between Bank Z default and reference entity spread: 0.60
Wrong-way adjusted CVA:
Increase EE by factor reflecting correlation: EE_wwr = EE * (1 + rho * stress_factor)
With rho=0.6 and stress_factor=1.5: EE_wwr = EE * 1.9
Wrong-way CVA = $180K * 1.9 = $342K (90% increase)
This demonstrates why wrong-way risk identification is critical for pricing.Quality Gate
- All OTC counterparties must have approved credit limits with PFE monitoring
- CVA must be computed for all uncollateralized or partially collateralized counterparties
- CVA calculations must use market-implied default probabilities (CDS spreads) where available
- Netting must be validated: ISDA Master Agreement and netting opinion must be in place
- Margin agreements (CSA terms) must be accurately reflected in exposure calculations
- Wrong-way risk must be assessed for every new trade: screen for specific and general WWR
- IM calculations under SIMM must be validated against the counterparty's independent calculation
- CCP exposure must include default fund contribution in total counterparty risk assessment
- Exposure simulations must use at least 1,000 Monte Carlo paths with appropriate time steps
- CVA hedging effectiveness must be monitored: track CVA P&L vs. hedge P&L
- Collateral eligibility, haircuts, and rehypothecation rights must be documented per CSA
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

