SA-CCR Explained: The Standardized Approach for Counterparty Credit Risk

Why SA-CCR Was Needed

The Current Exposure Method (CEM), which most banks used for counterparty credit risk exposure before SA-CCR, had a fundamental problem: it calculated potential future exposure (PFE) as a flat percentage of notional, with fixed add-on factors by asset class and maturity, regardless of whether the trade was collateralized. A $100M 5-year interest rate swap generated the same CEM add-on whether it was traded under a daily-margining two-way CSA or was completely uncollateralized. That design made no economic sense and understated exposure for uncollateralized portfolios while overstating it for tightly margined ones.

CEM also had weak netting recognition. Under CEM, netting set PFE was calculated as NGR × Gross PFE (where NGR was the net-to-gross ratio of current MTM), but the method could not accurately reflect the diversification benefit of a mixed long/short portfolio. A netting set with equal and offsetting pay-fixed and receive-fixed swaps could still show significant PFE under CEM despite having near-zero net exposure.

The Basel Committee published the Standardized Approach for Counterparty Credit Risk (SA-CCR) in March 2014 (BCBS279). It addressed both weaknesses: it explicitly distinguishes margined and unmargined netting sets through different maturity factors and replacement cost formulas, and it uses a rigorous hedging set aggregation framework with prescribed correlations to reflect actual netting. SA-CCR became effective in the US under the final rule (effective January 2022 for large banks), in the EU under CRR II (June 2021), and in the UK post-Brexit (January 2022). It now governs EAD for all OTC derivatives, exchange-traded derivatives, and long settlement transactions for banks subject to Basel III capital requirements.

The SA-CCR EAD Formula

The fundamental SA-CCR equation is:

EAD = α × (RC + PFE)

Where α = 1.4 (the supervisory-set multiplier, representing the uncertainty in the relationship between current exposure and modeled PFE), RC is Replacement Cost, and PFE is Potential Future Exposure. For banks using IMM (Internal Models Method), α may be lower if the supervisor approves an internal estimate, but the floor is 1.2. For the vast majority of banks using SA-CCR as their non-modelling approach, α = 1.4 is fixed.

Breaking down each component:

• "Walk me through the SA-CCR EAD calculation for a simple interest rate derivatives netting set step by step: Single bilateral netting agreement (ISDA Master Agreement), one counterparty, no CSA (unmargined). Three trades: (1) Pay-fixed 5Y USD IRS, $50M notional, current MTM to us = +$1.4M, time to maturity = 4.8Y; (2) Receive-fixed 7Y USD IRS, $30M notional, current MTM to us = -$0.6M, time to maturity = 6.7Y; (3) Long 1Y ATM SOFR cap, $20M notional, current MTM to us = +$0.15M, delta ≈ 0.40. No collateral posted (C = 0). Calculate: (1) Replacement Cost: RC = max(V − C, 0) where V = sum of MTMs; (2) the adjusted notional for each trade using the IR formula (notional × supervisory duration factor); (3) maturity factor for each trade (unmargined); (4) supervisory delta for each trade (linear = 1 for swaps, Black-Scholes delta for the cap); (5) effective notional for each trade; (6) aggregate AddOn for the IR hedging set using the correlation aggregation formula; (7) multiplier; (8) PFE; (9) EAD = 1.4 × (RC + PFE). Show every intermediate step."

Replacement Cost: RC

RC represents the cost of replacing the trades in the netting set if the counterparty defaults today. It is the current net exposure after eligible collateral offset.

For unmargined netting sets:

RC = max(V − C, 0)

Where V = sum of current MTM values across all trades in the netting set (positive = asset to us, negative = liability), and C = value of eligible collateral received net of eligible collateral posted. RC cannot be negative — if the netting set is net negative MTM, RC = 0 (we owe the counterparty money; if they default, we don't lose anything on replacement).

For margined netting sets:

RC = max(V − C, TH + MTA − NICA, 0)

The additional term (TH + MTA − NICA) accounts for the fact that under a margined CSA, there is a buffer below which no margin call is made: TH is the threshold (below which no VM call is triggered), MTA is the minimum transfer amount (granularity floor on each call), and NICA (Net Independent Collateral Amount) is independent collateral already posted. The maximum of (V−C) and (TH+MTA−NICA) ensures that RC reflects the exposure that could build up before the next margin call is made.

• "Replacement cost calculation for a margined netting set: Single bilateral netting set with a daily-margining two-way CSA. CSA terms: zero threshold (TH = 0), MTA = $250,000, NICA = $500,000 (we hold $500k of independent collateral from the counterparty). Current portfolio: 6 IRS trades, net MTM = +$3.2M to us. VM posted by counterparty under last margin call: $2.9M (based on yesterday's MTM). Total collateral held by us (C): $2.9M VM + $0.5M NICA = $3.4M. Compute: (1) V − C = $3.2M − $3.4M = −$0.2M (we're holding more collateral than the current MTM); (2) TH + MTA − NICA = 0 + $0.25M − $0.5M = −$0.25M; (3) RC = max(−$0.2M, −$0.25M, 0) = 0. Explain why RC = 0 in this case and what economic situation this represents. Then: recalculate RC if rates move and the portfolio MTM jumps to +$5.1M intraday before the next margin call, showing the gap RC formula in action."

PFE and the AddOn Calculation

PFE is the estimate of how much exposure could grow over the future life of the netting set, beyond the current replacement cost. The SA-CCR PFE formula:

PFE = multiplier × AddOn_aggregate

AddOn_aggregate is computed separately for each asset class (Interest Rates, FX, Credit, Equity, Commodity) and summed, because Basel treats the five asset classes as uncorrelated at the netting set level. Within each asset class, trades are grouped into hedging sets and effective notionals are aggregated with prescribed correlations.

The core building block for each trade's contribution to the AddOn is:

d_i = Adjusted Notional_i × Supervisory Factor_i × Supervisory Delta_i × Maturity Factor_i

Where:

Adjusted Notional for IR trades = Trade notional × Supervisory Duration = notional × (exp(−0.05 × S) − exp(−0.05 × E)) / 0.05, where S and E are the start and end dates of the longest trade in the hedging set (or the trade itself). For FX, adjusted notional = trade notional converted to reporting currency. For Equity and Credit single-name, adjusted notional = number of units × current price per unit.

Maturity Factor: For unmargined trades: MF = √(min(M, 1) / 1), where M is the remaining maturity in years (capped at 1 year, so MF maxes at 1.0). For margined trades: MF = 1.5 × √(MPOR / 1 year), where MPOR (margin period of risk) is typically 10 business days (~0.04 years) for bilateral OTC and 5 business days for CCP-cleared, giving MF ≈ 0.30 for bilateral margined trades versus up to 1.0 for unmargined.

Supervisory Delta adjusts the effective notional for the directionality of the trade: δ = 1 for long linear instruments (a pay-fixed swap receiving floating has positive exposure to rates rising), δ = −1 for short linear instruments, and for options it uses the supervisory Black formula: δ = Φ(λ × (ln(P/K) + 0.5 × σ² × T) / (σ√T)) where Φ is the standard normal CDF, λ = +1 for purchased calls and written puts, λ = −1 for purchased puts and written calls, P is the current price of the underlying, K is the strike, T is time to option expiry, and σ is the supervisory volatility (prescribed by asset class: 50% for equity, 80% for commodity, 15% for IR options at the 1-year tenor point).

• "SA-CCR adjusted notional and supervisory delta for an FX option: We have a long USD call / JPY put option: $25M notional, spot USD/JPY = 155.40, strike = 158.00 (out of the money), 6-month expiry (T = 0.5Y). Supervisory volatility for FX options under SA-CCR: σ = 15%. The trade is unmargined (bilateral, no CSA). Step through the SA-CCR calculations: (1) adjusted notional = $25M converted to JPY at spot (or keep in USD — which currency does SA-CCR require?); (2) supervisory delta using the Black formula: calculate ln(P/K) = ln(155.40/158.00), compute (ln(P/K) + 0.5 × 0.15² × 0.5) / (0.15 × √0.5), apply Φ(·) with λ=+1 for a purchased call; (3) maturity factor for unmargined trade with 0.5Y maturity: √(0.5/1) = 0.707; (4) supervisory factor for FX = 4%; (5) effective notional d = adjusted notional × delta × MF × SF; (6) how does this effective notional compare to the same FX forward (delta = 1) of the same notional?"
• "SA-CCR interest rate AddOn aggregation for a multi-tenor hedging set: IR hedging sets are aggregated by currency. We have four USD IRS trades in one netting set: (1) pay-fixed 2Y IRS: effective notional d₁ = +$8.2M (positive = long rates / exposure to rising rates); (2) receive-fixed 5Y IRS: d₂ = −$14.6M; (3) pay-fixed 10Y IRS: d₃ = +$22.1M; (4) long 5Y/10Y swaption (payer): d₄ = +$6.3M. Under SA-CCR IR aggregation, trades are assigned to tenor buckets (under 1Y, 1-5Y, over 5Y) and aggregated with prescribed correlations: ρ = 0.70 within-bucket, ρ = 0.30 across adjacent buckets (1Y-5Y and 5Y-10Y), ρ = 0.00 between under-1Y and over-5Y buckets. Assign each trade to its tenor bucket, compute the bucket-level effective notionals, aggregate across buckets using the correlation formula: AddOn_IR = √(Σᵢ Σⱼ ρᵢⱼ × Dᵢ × Dⱼ), and compare to the fully netting case (Dnet = Σdᵢ) and fully gross case to illustrate the netting benefit."

The SA-CCR Multiplier

The multiplier is the mechanism by which SA-CCR reduces PFE when a netting set is currently out of the money. Its formula:

multiplier = min(1, 0.05 + 0.95 × exp((V − C) / (2 × 0.95 × AddOn_agg)))

The intuition: if V − C is large and positive (deep in-the-money netting set), the exponent is large and exp(·) → ∞, but the min(1, ·) cap binds at 1, so the multiplier = 1 and PFE = AddOn_agg. If V − C is large and negative (deep out-of-the-money), exp(·) → 0 and the multiplier approaches 0.05 (the floor), giving PFE = 0.05 × AddOn_agg. The 0.05 floor prevents PFE from reaching zero — even a deeply out-of-the-money netting set retains 5% of the add-on as minimum PFE.

This has a critical implication for netting set optimization: structuring trades so the netting set stays moderately in-the-money to us keeps the multiplier near 1, while restructuring a large positive-MTM portfolio so that offsetting trades drive V − C negative reduces EAD meaningfully — but only via the PFE multiplier, not by changing the AddOn. This is why trade compression and novation can reduce SA-CCR EAD even without changing the notional or tenor of underlying positions.

• "SA-CCR multiplier sensitivity analysis: A netting set has AddOn_aggregate = $12.0M (fixed, based on trade-level effective notionals). The current net MTM to us (V − C) varies across these scenarios. Calculate the multiplier and PFE for each: (A) V − C = +$8M (deeply in the money to us); (B) V − C = +$2M (moderately in the money); (C) V − C = 0 (at the money); (D) V − C = −$3M (out of the money); (E) V − C = −$10M (deeply out of the money). For each: compute the multiplier using the SA-CCR formula, compute PFE = multiplier × $12M AddOn, compute EAD = 1.4 × (RC + PFE) assuming RC = max(V − C, 0) for an unmargined set. Show the full range of EAD across the five scenarios and comment on the practical implications for XVA desk collateral structuring."

AddOn by Asset Class: Supervisory Factors and Aggregation Rules

Each asset class has its own prescribed supervisory factor (SF), hedging set structure, and aggregation formula. Understanding the differences is critical for practitioners comparing SA-CCR across a mixed derivatives book.

Interest Rates (IR): SF = 0.50% for all maturities (a single supervisory factor regardless of duration — the duration is captured in the adjusted notional via the supervisory duration formula). Trades are aggregated within currency-specific hedging sets. Within a hedging set, trades are bucketed by maturity (under 1Y, 1-5Y, over 5Y) and aggregated with the prescribed correlation matrix (ρ_intra = 0.70 for same bucket, ρ_12 = 0.30 between adjacent maturity buckets, ρ_13 = 0.00 between the 1Y and 10Y+ buckets). IR is unique in using the supervisory duration formula to compute adjusted notionals rather than raw notional.

FX: SF = 4.0% (flat across all currency pairs). All FX trades in a single netting set are aggregated within their currency-pair hedging set using simple netting (no correlation bucketing). The adjusted notional is the trade notional in the reporting currency. FX has no within-currency correlation structure; it's purely additive across currency pairs (different currency pairs are treated as independent hedging sets with no offsets between, say, EUR/USD and EUR/GBP).

Credit (single-name): SF ranges from 0.38% (investment grade, IG) to 6.0% (HY and unrated). Credit CDS are aggregated by reference entity within a single hedging set. Same-entity same-currency positions net fully (ρ = 1.0); different entities within the same credit bucket have correlation ρ = 0.35 (systemic factor). The adjusted notional = notional × supervisory duration for CDS (using the CDS maturity for the duration formula).

Equity: SF = 32% for single-name equity, 20% for equity indices. Single-name equity is aggregated by issuer. The adjusted notional = number of shares × current share price. Within a hedging set, different-issuer equity positions are aggregated with ρ = 0.50 (correlation between individual stocks within the same index cluster).

Commodity: SF ranges from 18% (precious metals) to 40% (electricity). Commodities are grouped into hedging sets by commodity type (energy, metals, agriculture, other). Different commodities within the same type aggregate with ρ = 0.99 for the same grade and delivery location, and ρ = 0.40 for different grades/locations of the same commodity. Energy and other commodity hedging sets aggregate with zero correlation between types.

• "SA-CCR AddOn for a mixed-asset portfolio: Single netting set, unmargined, containing: (1) 5Y pay-fixed USD IRS, $50M notional, effective notional d = +$6.8M (after supervisory duration and SF); (2) EUR/USD FX forward, €20M notional equivalent $21.8M, long EUR, delta = 1, MF = √(0.25) = 0.5, SF = 4%; (3) 3Y CDS selling protection on an IG corporate (BBB), $15M notional, long credit risk (beneficiary if spreads tighten, lose if widen), delta = +1 (sold protection = long credit), MF = √(1) = 1, SF = 0.38%; (4) Long equity call option on S&P 500 ETF, $10M notional, delta = 0.45 (supervisory Black formula), MF = √(0.5) = 0.707, SF = 20%. Calculate: effective notional for each trade, asset-class-level AddOn (each asset class is a separate hedging set, summed without correlation), aggregate AddOn = sum across all asset classes, then apply multiplier (assume V−C = +$1.2M, AddOn_agg will come from your calculation), compute PFE and EAD = 1.4 × (RC + PFE). Show how EAD changes if you add an offsetting EUR/USD FX forward of the same size."
• "SA-CCR vs CEM comparison for a vanilla IRS: I have a pay-fixed 10Y USD IRS, $100M notional, ATM at 4.5%, current MTM = +$2.1M to me, unmargined. CEM calculation (legacy): Current Exposure = max(MTM, 0) = $2.1M. Add-on under CEM = 1.5% × $100M = $1.5M (CEM IR add-on for over-5Y maturity). CEM EAD = $2.1M + $1.5M = $3.6M. SA-CCR calculation: (1) RC = max($2.1M − 0, 0) = $2.1M; (2) adjusted notional = $100M × (exp(−0.05×0) − exp(−0.05×10)) / 0.05 = $100M × (1 − 0.6065) / 0.05 (use S=0, E=10); (3) delta = +1 (pay-fixed, long rates); (4) MF = √(1) = 1.0 (unmargined, M = 10Y capped at 1 in the maturity factor formula — confirm: for unmargined, MF = √(min(M,1)/1), so for M=10Y, MF = √(1/1) = 1.0); (5) SF = 0.5%; (6) AddOn_IR = adjusted notional × delta × MF × SF; (7) multiplier = min(1, 0.05 + 0.95×exp($2.1M / (2×0.95×AddOn))); (8) PFE = multiplier × AddOn; (9) EAD = 1.4 × ($2.1M + PFE). Compare EAD under CEM vs SA-CCR and comment on which is more conservative for this specific trade."

Margined Netting Sets: MPOR and the Maturity Factor

The SA-CCR treatment of margined netting sets is one of its most important improvements over CEM. The key insight: a daily-margined netting set cannot accumulate exposure beyond what can build up during the margin period of risk (MPOR) — the time between when a margin call is missed and when the surviving party can close out and replace positions. Only exposure that grows during the MPOR is relevant for PFE.

The SA-CCR maturity factor for margined netting sets:

MF_margined = 1.5 × √(MPOR / 1 year)

Prescribed MPOR values under Basel SA-CCR: 10 business days (~0.04 years) for bilateral OTC derivatives subject to daily margining. This gives MF = 1.5 × √(0.04) ≈ 0.30. For CCP-cleared trades: 5 business days for client clearing (MF ≈ 0.21), 10 business days for house accounts. For netting sets with 5,000 or more trades, the MPOR doubles (20 business days for bilateral, 10 for CCP), reflecting operational complexity risk. For re-margining periods longer than 1 day: MPOR = (re-margining period in days + 9) business days.

The 1.5 multiplier in the margined MF formula is a supervisory add-on (not present in the unmargined formula) that acts as a buffer for the uncertainty in close-out timing.

Comparing a $50M 5Y IRS under different margining regimes: unmargined MF = 1.0, bilateral margined MF ≈ 0.30, CCP-cleared MF ≈ 0.21. The PFE add-on for the same trade drops to roughly 30% of the unmargined level under bilateral daily margining, and 21% under CCP clearing — showing why central clearing materially reduces SA-CCR EAD and therefore regulatory capital for OTC derivatives.

• "SA-CCR comparison across margining regimes for a 5Y EUR IRS: Trade: receive-fixed 5Y EUR IRS, €60M notional, ATM at 3.20%, current MTM = −€0.8M to us (we're paying floating). Calculate SA-CCR under three regimes: (A) Unmargined (no CSA): RC = max(−€0.8M, 0) = 0; MF = √(min(5Y, 1Y) / 1Y) = 1.0; compute AddOn and EAD. (B) Bilateral daily-margined CSA (TH = 0, MTA = €500k, NICA = 0): RC = max(V−C, TH+MTA−NICA, 0) = max(−€0.8M − (−€0.8M), €0+€0.5M−0, 0) (assume VM = €0.8M held against us, C = €0.8M, so V−C = −€0.8M − (−€0.8M) = 0); MPOR = 10BD; MF = 1.5 × √(10BD / 250BD) = 1.5 × √(0.04) ≈ 0.30; compute AddOn and EAD. (C) LCH SwapClear cleared: same MTM, MPOR = 5BD for client clearing; MF = 1.5 × √(5BD / 250BD) ≈ 0.21; compute AddOn and EAD. Show the full EAD calculation for each and express the capital saving from clearing vs unmargined."

Portfolio-Level SA-CCR: Netting Set Optimization

For risk and capital management teams, SA-CCR creates optimization opportunities that CEM never provided. Because SA-CCR uses proper hedging-set aggregation with prescribed correlations rather than simply adding up notionals, restructuring a portfolio — without changing its economic profile — can materially reduce EAD. Common optimization levers:

Netting set structuring: All trades under a single ISDA Master Agreement with a single counterparty form one netting set. Adding a new offsetting trade to an existing positive-MTM netting set reduces V (the net MTM), which depresses the multiplier and reduces PFE. If the new trade has opposite directionality (opposite delta in the same hedging set), it also reduces the AddOn directly.

Compression: Trade compression (through TriOptima, LCH Compression, or bilateral compression) reduces the number of trades and the outstanding notional while preserving the net risk position. For SA-CCR, compression reduces the adjusted notionals going into the AddOn calculation, directly reducing EAD. This is most effective in IR portfolios where large gross notionals with partial offsetting generate a large AddOn that overstates net risk.

Clearing: Moving bilateral OTC positions to CCP clearing reduces the maturity factor from ≈0.30 to ≈0.21 (for 5BD vs 10BD MPOR), reducing PFE by about 30% on a like-for-like basis. There are also netting benefits at the CCP level if the clearing portfolio has offsetting positions across counterparties that were formerly in separate bilateral netting sets.

• "SA-CCR portfolio optimization analysis for a rates desk: Our desk has $15B notional of bilateral unmargined USD IRS across 85 trades with 12 counterparties. Current aggregate SA-CCR EAD: $420M (across all netting sets). Three optimization actions under consideration: (A) Bilateral compression with counterparties A, B, C: expected to reduce gross notional by $4B while preserving net DV01. How does a $4B notional reduction in the IR hedging set translate to SA-CCR EAD reduction? Approximately model the AddOn reduction. (B) Migrate $3B notional with counterparty D to LCH SwapClear bilateral margining under a two-way CSA (TH = 0): this changes MF from 1.0 (unmargined) to 0.30. Estimate EAD reduction for that netting set. (C) For counterparty E (currently $800M notional, MTM = +$12M, EAD = $48M), add a $50M offsetting receive-fixed 5Y IRS. How does this affect the multiplier and EAD? Estimate the aggregate EAD reduction from all three actions and comment on implementation cost vs. capital benefit."
• "SA-CCR for a cross-asset netting set — business line review memo: Our investment bank is reviewing SA-CCR EAD for a large asset management client. Netting set: 40 trades across IR, FX, and equity across a single ISDA/CSA (zero threshold, daily margining, NICA = $5M). Current aggregate SA-CCR EAD: $95M. Current RWA: $95M × 20% risk weight (financial institution counterparty, investment grade) = $19M. Required capital: $19M × 8% = $1.52M. The client wants to add a new $200M equity total return swap (equity, 2Y). Estimated SA-CCR EAD increase: $8.5M (AddOn from the new equity trade, after multiplier adjustment for the margined netting set). New capital requirement: ($95M + $8.5M) × 20% × 8% = compute. Write a one-page business line memo explaining: the SA-CCR EAD impact of the new trade, the incremental capital cost, the relationship between SA-CCR EAD and the CVA capital charge (brief), and whether the return on capital for this trade is acceptable at our current 12% RoE hurdle."

SA-CCR in the Leverage Ratio

SA-CCR is not only used for CCR capital (RWA calculations). Under Basel III, it also replaces CEM as the method for calculating derivatives exposure in the leverage ratio (LR) denominator — the total exposure measure in the denominator of the CET1/Total Exposure ratio. The leverage ratio requires all derivative exposures to be included, without netting against collateral received. This means leverage ratio SA-CCR EAD is computed on a gross basis: EAD_LR = replacement cost (gross positive fair values, no netting across counterparties for LR purposes) + PFE (on a netting set basis).

This creates a tension for banks with large derivatives books: SA-CCR for CCR capital benefits from netting (a natural hedge within a netting set reduces EAD), but the leverage ratio limits how much netting can be applied. A bank optimizing for CCR capital might increase netting set complexity; but the same optimization may do less to improve the leverage ratio, where gross positive fair values drive the denominator.

• "SA-CCR leverage ratio treatment for a netting set: Netting set with counterparty X: 10 IRS trades. Gross positive fair values (sum of positive-MTM trades): $18.4M. Net MTM across all 10 trades: +$7.2M. SA-CCR RC for CCR capital: max($7.2M − collateral, 0) = $7.2M (no collateral). SA-CCR RC for leverage ratio: under BCBS leverage ratio framework, RC in the exposure measure = max(net MTM, 0) across netting sets, PLUS variation margin received can offset positive fair values to the extent it meets specified conditions. Assume daily VM of $6.8M received (cash, same currency). Is VM eligible to offset in the LR calculation under Basel III LR standards? If eligible, LR RC = max($7.2M − $6.8M, 0) = $0.4M. If not eligible, LR RC = $7.2M. Explain the conditions for VM to reduce the LR exposure measure, calculate LR EAD = RC_LR + PFE (using same PFE as CCR capital), and compare to CCR capital EAD = 1.4 × ($7.2M − $6.8M + PFE). Show the divergence between CCR EAD and LR EAD."

Implementing SA-CCR with Claude

SA-CCR is computation-intensive enough that step-by-step walkthroughs are valuable for risk teams validating model implementations, auditors checking capital calculations, and business line heads trying to understand why a specific trade generates a particular capital charge. Claude can work through any netting set at trade-level detail with explicit formulas and intermediate calculations.

The most productive workflows are: (1) validating a specific netting set EAD against your risk system output — provide the trade-level inputs and ask Claude to replicate the calculation, (2) stress testing — varying the MTM, collateral, or supervisory factors to understand EAD sensitivity, (3) new trade impact analysis — calculating the incremental EAD of a proposed trade before booking it, and (4) explaining SA-CCR mechanics to business line heads who need to understand why their derivatives book generates the capital charges it does.

For the related XVA charges that flow from SA-CCR EAD — CVA, KVA, and the Basel IV CVA capital calculation — see XVA Explained. For FRTB market risk capital, which interacts with SA-CCR for the total RWA picture, see FRTB and Claude AI. For the broader derivatives workflow including structuring and client documentation, see Derivatives AI. All Compliance & Risk templates include an SA-CCR Calculator for netting set EAD computation.

• "SA-CCR documentation for a model validation team: Our internal model validation group needs to validate the SA-CCR implementation in our risk system. Provide a structured test case with: (A) a 5-trade netting set (2 IRS, 1 FX forward, 1 equity option, 1 credit CDS) with explicit parameters for each trade (notional, maturity, current MTM, CSA status, delta where applicable); (B) worked SA-CCR calculation for each component — adjusted notional, supervisory delta, maturity factor, effective notional, asset-class AddOn, multiplier, PFE, EAD — with every intermediate step shown and the formula source referenced to BCBS279 paragraph numbers where applicable; (C) three sensitivity tests: MTM shift of +$5M across all trades, MPOR change from 10BD to 20BD (netting set exceeds 5,000 trade threshold), and addition of a new offsetting 5Y IRS; (D) expected EAD results for the model validator to benchmark against."
• "SA-CCR briefing for a business line head: Our credit structuring desk runs a $2B notional portfolio of USD and EUR IRS with corporate clients (60% collateralized, 40% uncollateralized). Current SA-CCR EAD: $185M. Current RWA from CCR: $1.48B. Basel IV CVA capital: $32M. Total capital allocated: $122M at 8% × RWA + CVA capital. Cost of capital: 12% RoE → annualized capital cost = $14.6M/year. Write a plain-English briefing for the desk head covering: (1) why we use SA-CCR instead of notional to measure exposure, (2) why uncollateralized trades generate 3-4× the EAD of collateralized equivalents, (3) the three actions that would most reduce our capital consumption — compression, CSA adoption for non-CSA counterparties, and clearing — with estimated capital benefit of each, (4) how the desk's capital cost per dollar of NII (net interest income) compares to the bank's target, and (5) one practical recommendation for the next quarter."

Frequently Asked Questions

What is α (alpha) in SA-CCR and why is it 1.4?

Alpha (α = 1.4) is the supervisory multiplier applied to the sum of RC and PFE to arrive at EAD: EAD = 1.4 × (RC + PFE). It was calibrated by the Basel Committee to account for the fact that the PFE AddOn calculation, while designed to measure potential future exposure at the 95th–99th percentile, does not perfectly capture correlation effects across asset classes, WWR, or the granularity of actual trade portfolios. The 1.4 buffer ensures that SA-CCR EAD is conservative relative to IMM-modeled exposure. Banks using the IMM (Internal Models Method) may apply for a lower alpha if their internal model evidence supports it, subject to a regulatory floor of 1.2. For standard SA-CCR users, α = 1.4 is fixed and non-negotiable.

How does SA-CCR handle equity derivatives versus equity positions in the banking book?

SA-CCR only applies to OTC derivative contracts and exchange-traded derivative contracts in the trading book — it does not apply to cash equity positions, bond holdings, or loans. An equity total return swap is captured under SA-CCR (equity asset class, SF = 32% for single-name, 20% for index). A physical equity investment (e.g., a stake in a corporate held on the balance sheet) is treated under the simple risk-weight approach or IRB approach for credit risk, not SA-CCR. The boundary is the derivative contract: if the instrument is a derivative with an equity underlying, SA-CCR applies; if it is a direct equity holding or debt instrument, different capital frameworks apply.

Does SA-CCR apply to repos and securities financing transactions?

SA-CCR applies to long settlement transactions and can be applied to SFTs (securities financing transactions — repos, reverse repos, securities lending/borrowing) as an alternative to the comprehensive approach or simple approach for collateral. The Basel III final rules give banks the option to use SA-CCR for SFTs where the transaction involves daily re-margining. In practice, most banks use the master netting agreement comprehensive approach for repos rather than SA-CCR, because the repo-specific rules recognize overcollateralization through haircuts more efficiently. SA-CCR is primarily used for OTC derivatives; for repos the applicability varies by jurisdiction and the bank's modeling choices.

How does the SA-CCR supervisory factor for credit derivatives compare to the old CEM factor?

Under CEM, the add-on factor for credit derivatives was 5% for qualifying reference entities (IG) and 10% for non-qualifying (HY/unrated), applied to the full notional regardless of maturity. Under SA-CCR, the supervisory factor is 0.38% (IG) to 6.0% (HY/unrated) — lower numerically but applied to an adjusted notional that includes the supervisory duration factor, which magnifies the base notional for longer-dated CDS. For a 5Y IG CDS, the supervisory duration ≈ 4.35, so the effective add-on factor relative to raw notional is approximately 0.38% × 4.35 ≈ 1.65% — still below the 5% CEM factor, reflecting SA-CCR's intent to reduce capital for IG credit derivatives portfolios with proper netting recognition.