SA
This section describes how requirements in BCBS 457 are implemented in the FRTB (Standardised Approach) Accelerator.
For full details of the specific measures, see the Standardised Approach.
The formulae and definitions in the BCBS document are not repeated here – instead, references are provided to the relevant paragraphs, as appropriate.
Calculation steps
Standardised approach capital requirement for trading desks that are either out-of-scope for model approval or that have been deemed ineligible to use the internal models approach is given as:
$C_U = SBM + RRAO + DRC_{SA}$
Sensitivity Based Method (SBM)
note
BCBS-457 Reference [MAR21] sets out the calculation of the sensitivities-based method under the standardised approach for market risk.
For each of the “high correlations”, “medium correlations” and “low correlations” scenarios as described in [MAR21.6], [MAR21.7](1) states “the bank must simply sum up the separately calculated delta, vega and curvature capital requirements for all risk classes to determine the overall capital requirement for that scenario”.
So the SBM Risk Charge is defined to be to the highest value of three measures: High Risk Charge, Medium Risk Charge, and Low Risk Charge
See SBM Risk Charge.
Calculations
The FRTB Accelerator contains measures for the calculation of risk charges from sensitivities [MAR21.3]. The calculation from sensitivity to risk charge takes place in several steps where the output from one step feeds into the next, creating a chain of calculations. The chain exposes the intermediate measures for:
-
Raw (input) sensitivities on a drill through panel and / or pivot view
-
Conversion of input sensitivities to a reference currency (if the source systems provide native currency sensitivities)
-
Calculation of Weighted Sensitivities
-
Calculation of the Risk Position
-
Calculation of the Risk Charge
Measures
The output of each stage of the calculation chain is available as a measure that can be displayed.
For each risk class, the aggregated risk charge is as follows:
$RiskCharge\_MediumCorr^{Delta}+RiskCharge\_MediumCorr^{Vega}+RiskCharge\_MediumCorr^{Curvature}$
See Aggregated RiskCharge by Class.
note
BCBS-457 Reference
Calculations [MAR21.4] to [MAR21.7] and risk factor definitions [MAR21.8] to [MAR21.16].
Default Risk Charge (DRC)
note
BCBS-457 Reference
[MAR22] sets out the calculation of the default risk capital requirement under the standardised approach for market risk.
For information on configuring the data sets, see DRC Trade Level Files.
DRC Non-Securitisations (Non-Sec)
The approach for the DRC comprises a four-step procedure:
-
The JTD loss amounts for each instrument, subject to default risk, are provided as input to the FRTB Accelerator. The gross JTD can either be supplied to the FRTB Accelerator or it can be calculated by the FRTB Accelerator given the LGD, notional and market value.
-
Net long and net short amounts are produced in distinct obligors, by the offsetting of the JTD amounts of long and short exposures with respect to the same obligor (where permissible).
-
The net short exposures are discounted by a hedge benefit ratio.
-
Default risk weights are applied, to complete the process of computing the capital charge.
Offsetting and Hedging
Offsetting is the netting of exposures to the same obligor (where a short exposure may be subtracted in full from a long exposure), while hedging refers to the application of a partial hedge benefit from the short exposures (where the risk of long and short exposures in distinct obligors do not fully offset due to basis or correlation risks). [MAR21.1]
The procedure is specified in the references cited above.
note
BCBS-457 Reference
The default risk weights as defined in [MAR22.24] and the LGD thresholds in [MAR22.12] are configurable and held as part of data sets.
Scaling Factors and Weightings
The scaling factor for maturities less than one year can be calculated by the FRTB Accelerator (see references below). A scaling indicator in the data instructs the accelerator whether or not to apply the scaling at the granularity of the input data or if the maturity should default to a pre-defined value such as 3 months. The DRC is calculated per bucket and then aggregated. The Hedge Benefit ratio (HBR) and DRC are calculated as described in BCBS 457(see below).
note
BCBS-457 References
Scaling factor for maturities less than one year is defined in [MAR22.20].
Scaling indicator is defined in [MAR22.20]
Hedge Benefit ratio (HBR) is calculated as described in [MAR22.23]
DRC is calculated according to [MAR22.26]
DRC Securitisations Non-Correlation Trading Portfolio (Sec Non-CTP)
note
BCBS-457 References
This section is concerned with [MAR22.31] through to [MAR22.35].
The functionality provided is similar to DRC Non-Sec but with bucketing and risk weights configured appropriately.
The default risk weights for “Sec non-CTP” exposures are the SEC-ERBA risk weights for long-term ratings.
note
BCBS-374 Reference
These risk weights are defined in BCBS 374 - Paragraph 68, and adjusted for non-senior tranches as per BCBS 374 - Paragraph 69.
DRC Securitisations Non-Correlation Trading Portfolio (Sec Non-CTP)
note
BCBS-457 Reference
This section is concerned with [MAR22.40] through to [MAR22.45].
The functionality provided is similar to DRC Non-Sec but with bucketing and risk weights configured appropriately.
The default risk weights for “Sec CTP” exposures are the SEC-ERBA risk weights for long-term ratings.
note
BCBS-374 Reference
These risk weights are defined in BCBS 374 - Paragraph 68, and adjusted for non-senior tranches as per BCBS 374 - Paragraph 69.
Residual Risk Add-On (RRAO)
note
BCBS reference 457
[MAR23] sets out the calculation of residual risk add-on under the standardised approach for market risk.
The risk weight information is available in [MAR23.8].
The FRTB Accelerator requires the input data to provide a notional per trade along with an indication as to whether the trade is exotic or not. See input file format for Trade Attributes.
The risk weight is applied at 1% for exotics and 0.1% for non-exotics. To set different RRAOs under different weightings use the Parameters sets provided in the FRTB Accelerator. See Parameterisation and context values.