What's New
This page provides a brief overview of the user-facing new features and improvements in the latest version of Atoti FRTB.
For bug fixes and developer-facing changes, see our Changelog.
5.3.1
Summary
- What-If: Improved support for distributed what-if.
- What-If: Improved performance of what-if.
5.3.0
Summary
- EBA reporting template updates
- New engine
- Databricks
- DirectQuery IMA Support
- Enable Categories hierarchy for internal reporting
- IRT Desk Improvements
EBA Reporting Template Updates
New bookmarks are being added to match the new EBA reporting templates. New fields have also been added to the data model for the SA cube to collect the additional information required for the templates.
For details, see the Atoti FRTB Release Notes.
New Engine
Atoti FRTB is powered by the latest version of Atoti Server (6.0.14-sb3) and Spring Boot (3.2). The result is a faster, safer application, which you can customize using the most modern stack available.
Databricks
DirectQuery support for connecting to a Databricks database has been added. Read more about DirectQuery and how to connect to a Databricks Database.
DirectQuery IMA Support
DirectQuery can now be used with the IMA-ES cube. P&L vectors stored in the database can be used to calculate the CA value.
Enable Categories hierarchy for internal reporting
You can now enable the Categories hierarchy for use in internal reporting.
IRT Desk Improvements
Membership of the Internal Risk Transfer (IRT) Desk is now specified directly at the book level. Additionally, it is possible to dynamically change which books are members of the IRT Desk.
5.2.2
Summary
- Bug fixes : This release fixes some known issues.
5.2.1
Summary
- Atoti Server version: Upgraded to Atoti Server 6.0.13
- Bug fixes : This release fixes some known issues.
5.2.0
Summary
- US FRB Jurisdictions: Added support for NPR from US FRB, including new US-NPR parameter set and reconciliation with a beta version of the ISDA unit tests. See US FRB support
- Stress Calibration P&L Vector : The P&L vector for the stress calibration calculations has been replaced with a PV vector and a base PV value.
- SA Summary Data: SA summary files can now be loaded into the SA cube. These files can be created by exporting data from the SA cube.
- Upgraded CRIF file format: Upgraded CRIF file format to ISDA FRTB-SA version 1.6.
- Added support for DRC Sec non-CTP Variant 2 in CRIF file format
- Database Schema: Removed unnecessary tables from the DirectQuery database schema.
New features
SA Summary Data
New file formats have been added to load SA data at the summary level, instead of just at the trade level. The new formats are the same as the existing ones, except the “trade id” has been replaced with “book” and “legal entity”.
DEE export templates have been added so that data loaded at the trade level can be exported at the summary level, ready for reloading.
5.1.3
Summary
- Reinstating the Categories hierarchy: The Categories hierarchy that was removed in 5.1.0 can be added back with a configuration property.
- Bug fixes This release fixes some known issues.
5.1.2
Summary
- Atoti Server version: Upgraded to 6.0.13
- Bug fixes This release fixes some known issues.
5.1.1
Summary
This release fixes some bugs.
5.1.0
Summary
- Complete DirectQuery implementation for SA
- Support for FX conversions of IMA ES P&L vectors
- The Interpretation and Implementation Guide has been expanded to include details for SA DRC.
- Upgrade to What-If 2.0 with support for distributed deployments
- Definition of regulatory desks is now independent of the organization hierarchy
- Upgrade to ISDA unit tests v3.3.1 with the addition of UK-PRA
- Upgrade to Atoti Server 6.0.9 and Atoti UI 5.1.x
- Memory use improvements over 5.0
- Microsoft SQL Support: DirectQuery support for connecting to a Microsoft SQL Database has been added
New features
Complete DirectQuery implementation for SA
The integration of DirectQuery that was previewed in 5.0 has been completed.
Capital analytics users can now effortlessly access historical data directly from their warehouse within Atoti FRTB. With this technology, market risk users can delve deep into the data and analyze trends without the need for the in-memory loading step. DirectQuery not only enables historical analysis but also optimizes system resources.
FX Conversions of IMA ES P&L vectors
IMA ES P&L vectors can now be converted to your regulatory currency.
We introduce a solution for how to aggregate IMA for local teams in their risk-free currency while enabling global teams to report in the currency of the group level. Previously a single FX rate was used for the P&L vectors. However, for the FX risk-class, the scenarios involved shocking the FX rates.
Now, you have the ability to provide simulated PV data instead of PL, and dynamically apply simulated FX rates to ensure accurate handling of the risk-free currency. The shocked PV can be converted using the shocked FX rate, while the base PV is converted using the base FX rate. This enhancement guarantees seamless IMA aggregation across multiple balance sheet currencies.
Separation of Regulatory Desks and Organization Hierarchy
We have introduced the concept of regulatory desks as distinct entities from the organizational hierarchies utilized for internal risk management. Recognizing that organizations often have multiple organizational hierarchies, our aim is to help simplify the management of these hierarchies.
With this update, every trade is now linked not only to one or more organizational structures via books but also to a regulatory desk. This linkage has a direct impact on the capital treatment of each trade, whether it falls under IMA or SA. This refined approach ensures greater flexibility and accuracy in managing trade-specific regulatory requirements while maintaining the integrity of organizational hierarchies.
What-if simulations on the organization hierarchy will not impact the desk-level calculations, and these simulations are no longer restricted by regulatory desks. Additionally, new what-if simulations can move books between desks without impacting the organization hierarchy.
Overrides With DirectQuery
Overrides can now be specified in remote Database tables to define overrides. The Override tables are outlined within the Database section of the docs. For a step-by-step walk-through guide, see Overrides With DirectQuery.
Microsoft SQL Database Support
Atoti FRTB now supports connecting to a Microsoft SQL database as a data source through the use of DirectQuery. Please refer to the Connecting to a Database section for a guide on how to connect to a Microsoft SQL database.
Improvements
- ISDA unit tests v3.3.1: The distribution now contains updated ISDA unit tests, this includes the UK-PRA parameter set.
- Reduced Memory Use: The memory use for the in-memory database is now closer to what it was in 4.1
- Improved ETL performance: The ETL is faster and uses less memory.
- What-If 2.0: The new version of the what-if simulations now support distributed deployments.
5.0.1+1
Known issues
- LIM-581 - FRTB 5.0.1 is not compatible with the Limits Module.
Summary
- Fixed known issue in 5.0.1 where What-Ifs created through the Parameter Sets widget were not visible in the What-If Manager widget. You can now view /delete them through the What-If Manager widget.
5.0.1
Known issues
- Updated 2023-03-20: The What-If Manager does not show Parameter Sets widget changes.
- LIM-581 - FRTB 5.0.1 is not compatible with the Limits Module.
Summary
- Switch Desk Model improvements : Redundant fields have been removed and minor defects preventing the form from being populated correctly have been fixed.
- Database Schema Update: Legal Entity Attributes database definition now includes Not Null fields.
5.0.0
Known issues
- Updated 2023-03-20: The What-If Manager does not show Parameter Sets widget changes.
- Updated 2023-02-27: FRTB-2682 - The documented mechanism for customizing the ETL is not working.
- Updated 2023-02-14: FRTB-2701 - Distribution is not working for IMA. It is only working for SA.
- LIM-581 - FRTB 5.0.0 is not compatible with the Limits Module.
Summary
- Added preview support for the Standardised Approach with DirectQuery
- Added support for overlapping tranches in CSR Sec non-CTP. New columns added to Delta, Vega, Curvature, and Underlying Description Overrides files
- Merged summary and trade-level cubes. The IMADRCSummaryCube and the PLSummaryCube have been removed. Data are now in the IMADRCCube and the PLCube, respectively.
New features
DirectQuery Preview (FRTB-SA)
This version of Atoti FRTB includes a preview of DirectQuery support for the Standardised Approach.
This preview includes all SA calculations, but does not have full multi-jurisdictional support and has limited customizability.
Currently, DirectQuery does not support what-ifs, sign-off, or data updates. These features are only available using the in-memory datastore.
Summary cubes
The IMADRCSummaryCube and the PLSummaryCube have been removed. Data are now in the IMADRCCube and the PLCube, respectively. For details, see Migration Notes
CSR Sec non-CTP tranche overlaps
To determine when two tranches overlap, the Accelerator now collects additional details for each tranche: Pool, Attachment, and Detachment.
These fields can be provided in the Delta, Vega, Curvature, and Underlying Description Overrides.
These fields are optional in all four files. In the Delta, Vega, and Curvature files, if the pool is not provided then the Tranche
field is copied to the Pool
field.
4.1.0+1
2022-12-02
Known issues
- Updated 2022-11-24: FRTB-2624 - The release of Spring boot 3.0.0 compiled for Java 17 is preventing the accelerator from compiling with Java 11. The workaround is to set the version of
spring-boot-maven-plugin
to2.4.2
in the top levelpom.xml
. - Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
Summary
This is a UI-only release that has updated Accelerator-sdk and ActiveUI dependencies.
For details, see the Release notes
4.1.0
Known issues
- Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
Summary
- Added Stress Calibration Cube for identifying period with the highest capital requirements for IMCC and SES.
- Weekly omega values for IMCC
- Merged IMA ES trade-level and summary cubes. For details, see Migration Notes
- SA DRC support for equity positions in high yield and distressed funds.
- Support for Limits Module integration
New features
CSR Bucket override for CRR2/CRR3
The integration of CSR buckets for CRR2 and CRR3 has been simplified. If a CSR sensitivity input file ("**/SBM_*Sensitivities(SECCTP|SECNONCTP|CSRNS).csv") has no
selected bucket (the Bucket
field), the quality and sector (CSRQuality
and CSRSector
fields) are used to select the bucket.
For BCBS, the bucket is computed and set as usual on the UnderlyingDescription store. For other jurisdictions (CRR2 and CRR3), if the bucket is different from the BCBS one, an override entry is created automatically on the UnderlyingDescriptionOverrides store with the right bucket. As a result, registering the CRR2 or CRR3 buckets on the “**/Underlying_Desc_Overrides*.csv” file is no longer needed if the bucket is not set on the sensitivity file.
The files “**/CSR_Bucket_NONSEC*.csv”, “**/CSR_Bucket_SECCTP*.csv”, and “**/CSR_Bucket_SECNONCTP*.csv” have been modified to handle the jurisdiction:
- An extra
ParameterSet
column has been added, allowing you to override the bucket selection for CRR2/CRR3 jurisdiction. - The
asOfDate
parameter now represents the start date that the bucket definition takes effect, rather than the current asOfDate. - As the files are not daily files anymore, they are now located in the /resources/data/configuration folder.
- The files are now loaded at the very first stage of the initial load, under the ConfigurationData topic, which is not dependent on the asOfDate.
Handling of multiple hierarchies and level depths higher than 1 for sign-off scopes
Until version 3.1.0, the scope of a Sign-off task could only be defined for the first level of one hierarchy. Now the scope definition can contain multiple hierarchies, and within each hierarchy, a level at a depth that can be higher than 1. See Multiple/Parent-Child scopes in the Sign-Off documentation.
note
The hierarchies and levels that can be used in the scopes of the Sign-off task must be part of the ones defined in configuration properties. See details in signoff.properties.
4.0.0
Known issues
- Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
- Update 2022-06-07: We’ve identified a bug in the Frequent Value Compression ActivePivot feature (new since AP 5.10). FRTB 4.0.* depends on ActivePivot 5.10.6 and is therefore impacted. We strongly recommend disabling Frequent Value Compression by setting the following property when launching ActivePivot:
-Dactiveviam.chunk.compression.enabledFrequentValue=0.
The next maintenance release of FRTB will include an upgrade to ActivePivot, which resolves this issue. - Regulatory calculations for DRC Sec CTP have been added, but the “Beyond Basel” calculations are missing for incremental and marginal measures, and multi-jurisdictional overrides.
- Lookback Rank measures do not work with Desk filters. As a result, PLA bookmark KS Test tab Desk filter will break charts.
- The custom Book Hierarchy widget has not been ported to ActiveUI5 yet. For now, this functionality can be achieved by using the File Upload widget. See UI Upgrades.
- Deleting a what-if simulation in the What-If Manager deletes the branch.
- You cannot switch a desk model back to what it was on the master branch.
- When switching a desk’s model, the new model does not default to the opposite of the current model.
Summary
- Upgraded to ActivePivot 5.10.5
- The Branch Manager and WhatIf Manager screens have been combined into a single WhatIf Manager Widget. This widget shows all created branches with the audit details and the ability to show differences against the master branch. Users will also be able to delete branches from this screen. Please note: the merge functionality has been removed.
- Added support to specify a staging directory through a new “scenarios” scope key for DLC requests. The value of this scope key is the directory name where the data is located.
- The enable-signoff Spring Profile has been added to the FRTB Accelerator, grouping all Sign-off functionality in a single toggled configuration.
- Performance benchmark testing results and methodology have been included.
- Limits Module for limit monitoring
- The integration of CSR buckets for CRR2 and CRR3 has been simplified.
New features
PLAT and Backtesting measures have been moved to Combined Cube
See the migration guide for details
Added data extraction template orders to DEETemplateOrders directory
The following extraction orders can be executed via DEE REST endpoint:
- For IMA, IMADRC, and PL trade level data for load into respective summary cubes
- For exporting sensitivity data in CRIF file format
For details, see Using DEE Template Orders.
Improvements to PL Cubes (for backtesting and PLAT)
- Loading APL, HPL, and RTPL values into trade-level cube
- Exporting aggregate data from trade-level cube to summary cube
- Loading P&L vectors into summary cube (including support for desk-switching what-if in IMA backtesting)
See the migration guide for details
Datastore Viewer Changes
- As part of the upgrade to ActiveUI5, the Datastore Viewer has been removed. In its place we now have the brand new Admin-UI, a separate web application that is part of the core ActiveUI product. The Admin-UI is also replacing the Content Server UI.
- The Admin UI allows authorized users to log in to a separate web application where they will be able to see both the familiar Content Server UI and also the new Datastore Viewer. The new Datastore Viewer is strictly a viewer, it does not allow modification to datastore values or what-if simulations. What-If simulations can be done through the File Upload widget.
- It can be accessed through the URL of the Content Server, followed by “/content”. This is the same URL that the content server UI was always accessible through.
Performance benchmark Results
- Details of the testing results and methodology we use are now provided. See the Performance Benchmarks for details.
Limits Module
The FRTB Accelerator uses the separate Limits Module to handle the limit monitoring workflow. Limits is a comprehensive workflow solution that lets you create, update, and approve limits easily. For more information, see Limit monitoring or the dedicated Limits Module documentation.
CSR Bucket override for CRR2/CRR3
The integration of CSR buckets for CRR2 and CRR3 has been simplified. If a CSR sensitivity input file ("**/SBM_*Sensitivities(SECCTP|SECNONCTP|CSRNS).csv") has no selected bucket (the “Bucket” field), the quality and sector (“CSRQuality” and “CSRSector” fields) are used to select the bucket. For BCBS, the bucket is computed and set as usual on the UnderlyingDescription store. For other jurisdictions (CRR2 and CRR3), if the bucket is different from the BCBS one, an override entry is created automatically on the UnderlyingDescriptionOverrides store with the right bucket. As a result, registering the CRR2 or CRR3 buckets on the “**/Underlying_Desc_Overrides*.csv” file is no longer needed if the bucket is not set on the sensitivity file. The files “**/CSR_Bucket_NONSEC*.csv”, “**/CSR_Bucket_SECCTP*.csv”, “**/CSR_Bucket_SECNONCTP*.csv” have been modified to handle the jurisdiction:
- An extra
ParameterSet
column has been added, allowing you to override the bucket selection for CRR2/CRR3 jurisdiction. - The
asOfDate
parameter now represents the start date that the bucket definition takes effect, rather than the current asOfDate. - As the files are not daily files anymore, they are now located in the /resources/data/configuration folder.
- The files are now loaded at the very first stage of the initial load, under the ConfigurationData topic, which is not dependent of the asOfDate.
3.1.1
Known issues
-
Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
-
Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
-
Update 2021-09-02: The CSR non-Sec Curvature Risk Position calculations may return incorrect results for CRR2 if the “Parameter Set” hierarchy is included on the rows or columns of the query. In this case, the accelerator will mistakenly treat bucket 16 as the “other” bucket, instead of bucket 18. Similarly for CSR Sec CTP.
-
Real-time queries are not updated with changes to the static data overrides.
Summary
CRR2 support has the following improvements:
- Three different interpretations of the zero risk-weight exposures for DRC non-Sec (see CRR2 DRC non-Sec Risk Weights)
- The ability to turn off the 3-month floor for DRC non-Sec matuirty scaling when offsetting (see CRR2 DRC non-Sec Maturity Scaling 3-Month Floor)
- The “risk-class Delta Sensitivities” measures have been added to the imported measure variations (as the sum of the Long and Short imported measures).
- Performance improvements for DRC non-Sec queries.
See the migration guide for details.
New features
Zero Risk-Weight Exposures for DRC non-Sec (CRR2)
The risk-factor description now includes a flag to indicate if an exposure should have zero risk-weight. For DRC non-Sec, the exposure (risk-factor) description can now be overridden per jurisdiction. In particular, the zero risk-weight flag can be specified per exposure and per jurisdiction.
See the migration guide for details.
3-Month Floor for DRC non-Sec Maturity Scaling (CRR2)
You can now turn off the 3-month floor for maturity scaling when offsetting is possible, which results in the maturity scaling step being combined with the offsetting step in the calculations.
See the migration guide for details.
3.1.0+1
Known issues
-
Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
-
Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
-
Update 2021-09-02: The CSR non-Sec Curvature Risk Position calculations may return incorrect results for CRR2 if the “Parameter Set” hierarchy is included on the rows or columns of the query. In this case, the accelerator will mistakenly treat bucket 16 as the “other” bucket, instead of bucket 18. Similarly for CSR Sec CTP.
-
Real-time queries are not updated with changes to the static data overrides.
-
DLC Rest endpoints are not exposed by default. FRTBConfig must include FRTBRestServicesConfig.class in order to expose these REST endpoints.
Summary
- FRTB 3.1 introduced a bug so that it was not possible to save dashboards. This has been fixed with this release and you can now save dashboards.
3.1
Known issues
-
Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
-
Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
-
Update 2021-09-02: The CSR non-Sec Curvature Risk Position calculations may return incorrect results for CRR2 if the “Parameter Set” hierarchy is included on the rows or columns of the query. In this case, the accelerator will mistakenly treat bucket 16 as the “other” bucket, instead of bucket 18. Similarly for CSR Sec CTP.
-
Real-time queries are not updated with changes to the static data overrides.
Summary
- Added support for subsidiaries (netting sets and imported capital charges) to CRR SA Reporting
- Added multi-jurisdictional support that includes static data overrides (ability to change DRC obligor rating, CSR bucketing, or Equity market cap per jurisdiction)
- Updated the SA Data-model to include Direct Bucket Mappings (SBM) and improvements for DRC and RRAO
- Upgrade to ActivePivot 5.9.6 and ActiveUI 4.3.17
See the migration guide for details.
New features
Support for subsidiaries
The new, optional “netted” and “imported” measure variations added in this release provide the ability to meet the CRR2 required reporting where capital charges are independently calculated and then summed when the netting of positions between different legal entities does not apply. An additional “reported” measure variation has also been added, that includes the impact of the netting sets and imported values. The imported values do not appear when you drill in or drill down on calculations as they are designed for reporting purposes only.
See the migration guide and Netted and Imported for details.
3.0.1
Known issues
-
Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
-
Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
-
In order to ensure proper connection of the UI to the AP Server, the ActivePivotServicesConfig.class must be replaced with APMActivePivotServicesConfig.class in the configuration imports of FRTBConfig.java You can find an example FRTBConfig file on our artifactory page.
Summary
- Data-Connectors upgraded to 1.2.0 to resolve concurrency issues with DLC.
3.0
Known issues
-
Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
-
Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
-
FRTB-2148 Concurrent DLC operations result in NPEs.
Summary
- IMA Performance improvements
- Upgrade to ActivePivot 5.9 and accelerator-sdk 4.1.0
- Replaced XML Cube Configuration
- Integrated Data Connectors component (which includes DLC and DEE)
- Additional CRR2 support
- Documentation on extending the Accelerator
-
- Starting with 3.0.0, the FX Complex Delta flag defaults to “Y” (use filtering) if not provided. Previously (2.4 and earlier), the default was N (use translations).
New features
Replaced XML Cube Configuration
The XML configuration of the ActivePivot Cube schemas and hierarchies has been replaced by ActivePivot’s fluent API java configuration. As part of this, some properties file configuration has been added to more easily modify the configuration for different environments.
See the migration guide for details.
Negative Values in Square Root in Risk Charge Formula
In some cases it is possible to have a negative value inside the square root of the SBM Delta and Vega Risk Charge formula MAR 21.4 (5).
And, this can occur even with the alternative Sb and Sc values from MAR 21.4 (5) (b).
For example, this can happen for Equity Delta in the high correlation scenario
, see test class EquityDeltaNegEVIT.java
for example sensitivities.
In this release, the value inside this square root is now floored at zero (after trying the MAR 21.4 (5) (b) alternative). This is a similar treatment to the Risk Position formula in MAR 21.4 (4).
CRR2 Support
For details on CRR2 support see the CRR2 page. The following items have been added for supporting CRR2.
- Support for DRC non-Sec adjustments
- Added BGN and HRK to CRR2 ERM II currencies (FX risk-weight overrides)
- Added flag to exclude GIRR inflation and cross-currency basis curves when dividing major currency risk weights by sqrt 2.
- Added support for scaling SBM sensitivities to support CRR2 underwriting
- Replaced Adjunct Currencies with FX risk-weight overrides and GIRR correlation overrides
Previous Releases
2.4.1
Known issues
- Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
- Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
Summary
- Support for https protocol when sharing bookmark URL’s through bookmark tree.
- Support for multiple content servers when sharing bookmark URL’s through bookmark tree.
- Support for multiple content servers when sharing bookmark URL’s through File submenu.
2.4.0
Known issues
- Updated 2022-09-12: PIVOT-5759 - Attempting to create partitions for reference stores causes a deadlock.
- Update 2021-10-08: The FX Delta Sensitivities Long/Short measures (used for CRR2 reporting), do not work properly if you are using FX sensitivity translations.
New features
ISDA tests and CRIF format
- Added support for loading the ISDA FRTB-SA CRIF file format (v4 r9)
- Added ISDA Unit tests (version 2.5.1), these can be run when compiling the accelerator
- Added support for starting up the accelerator with a copy of the ISDA test data, loaded from a CRIF file
For details, see CRIF Files
CRR2 Support
- Added Long and Short Sensitivity measures for the EBA reporting template
- Added CRR2 ISDA unit tests (version 2.5.1)
- Added support for Yellow and Orange PLAT zones
For details, see CRR2 Support
Improvements
- SBM performance improvements, by converting the correlation matrix into a block diagonal form (with a block per underlying) and using multi-threaded matrix multiplication.
- Improved performance of default cube configuration
- Added DRC Sec non-CTP Region and Asset Class fields to SA cube
2.3.0
Known issues
Menu item “Display as Columns”
In the UI, the “Display as Columns” menu item is showing up as “invalid”, moreover once selected there is no way to return to the “Tree” layout.
Internal reference: FRTB-1914
Status: Open
Last updated: 2020-09-20
Documentation landing page
On the documentation landing page, the “Latest updates - activeviam.com” link points to the 2.2.0 documentation instead of 2.3.0. The correct link is https://docs.activeviam.com/products/accelerators/frtb/2.3/online-help/documentation-measures/ Internal reference: FRTB-1917 Status: Fixed Last updated:2020-09-20
New features
CRR2 Support
A few items have been added to support CRR2. See CRR2 Changes for a detailed description of the difference between the BCBS and CRR2 specifications.
-
A CRR2 parameter set has been added to the configuration files shipped with the accelerator. This parameter set contains the risk weights and correlations used in CRR2.
-
The concept of "adjunct currencies" has been added to the accelerator. These are currencies that receive special treatment relative to the reporting currency. This is aimed at the ERM II mechanism (i.e. the DKK currency).
For details, see CRR2 Support
CRIF file format support
In addition to the existing file formats, SA sensitivities may also be loaded using the CRIF file format which is described in ISDA's Risk Data Standards document.
Note, some functionality in the accelerator is not available when using the CRIF file format. This includes:
-
Base currency and multi-jurisdiction support for the FX risk class
-
Control over which Delta sensitivities are used when Delta stripping in the Curvature calculations.
For details, see CRIF files
New measures
-
IMA ES ratio to track if the reduced set of risk-factors explains at least 75% of the variation of the full set). For details, see ES (Model Variation)
-
SBM Scenario: to track which SBM scenario (low, medium, or high) is used. For details, see SBM Correlation Scenario
Improvements
- Fixed Curvature calculations when no Delta sensitivities are loaded.
- SA DRC performance improvements.
2.2.1
Known Issues
ActivePivot incompatible with latest OpenJDK
WARNING!
- ActivePivot 5.8.7 (and all earlier
versions) are incompatible with the latest version of OpenJDK 11
(Version 11.0.6, released 2020-01-15).
This is due to a breaking change in OpenJDK (JDK-8211919)
New features
None
Improvements
None
2.2.0
warning! Version 2.2.1 of the FRTB Accelerator has been released. This is a maintenance release and fixes known issues with version 2.2.0. We recommend upgrading to 2.2.1.
Known issues
ActivePivot incompatible with latest OpenJDK
WARNING!
- ActivePivot 5.8.7 (and all earlier
versions) are incompatible with the latest version of OpenJDK 11
(Version 11.0.6, released 2020-01-15).
This is due to a breaking change in OpenJDK (JDK-8211919)
Fixed Issues
The following issues have all been fixed in 2.2.1:
-
There is a calculation error for Commodity Curvature when the "other commodity" bucket contains more than one risk-factor (FRTB-1822)
-
The BeyondBasel/What-If bookmark is using a wrong container key for one of its widgets. To fix this issue, change
"containerKey" : "accelerator_ParentChild"
to"containerKey" : "ParentChild"
-
The sample data set still has the incorrect risk weight for SEC-ERBA (long term, senior) BB-. (FRTB-1830)
-
Due to a bug in the react-script module, frtb-activeui hangs after the tests run instead of carrying on with the maven build. As a workaround, update the "test-build" script in frtb-activeui/package.json file. The correct script is
"react-scripts test --env=jsdom --no-watch"
. -
Due to a bug in the react-script module, the frtb-activeui development server will not start (e.g.
$ yarn start
). The workaround is to upgrade the version ofreact-scripts
inpackage.json
from 3.3.0 to 3.4.0
New features
Offline documentation
The accelerator documentation is now bundled with the accelerator so that it is accessible without needing Internet access. A copy of the documentation still remains on activeviam.com.
Data Load Controller
The Data Load Controller (DLC) introduced in 2.1.0 has been improved to support ETL auditing.
Event monitoring has been added to the Data Load Controller DLC.
-
Results returned to the data orchestrator through the REST service are now JSON formatted and suitable for parsing to extract statistics about the data loading.
-
Additional events are generated by the tuple publishers and include information on how many tuples have been inserted into each store.
-
In additional to returning this information in the REST service, an internal component can also receive these events, for example LoggingDlcEventHandler.java
IRT Desk
Support has been added for calculating the capital charge separately for an IRT desk.
Support for Java 11
A jdk11 version of the accelerator is available with this release.
IMA DRC Summary Cube
The IMADRCSummaryCube has been added to calculate the 12 week IMA DRC average as specified in [MAR33.22].
-
This new cube is the same as the existing IMADRCCube, except that the trade level is absent. Instead, the Book and Legal Entity levels are lowest levels available in the booking and organisation dimensions.
-
A REST service has been added to export daily data from the trade-level, ready for import into the summary cube. For details, see Using DEE Template Orders
Active UI 4.3
The new accelerator UI application is based on ActiveUI 4.3.4.
Implementation and Interpretation Documentation
New documentation has been added for each SBM risk class describing the accelerators implementation and interpretation of BCBS 457 for the risk class, this includes data model (and how it maps to the input files, datastore, and cube levels) and calculations (and how they map to measures in the cube).
For details, see FRTB Accelerator Interpretation and Implementation of BCBS 457
Improvements
Common Libraries
The FRTB accelerator now shares common accelerator code (both back-end and front-end) with other ActiveViam accelerators.
2.1.0
Known issues
None
New features
Measures Documentation
New documentation of all the measures has been added. You can browse this documentation here.
Data Load Controller
The Data Load Controller has now been integrated into the accelerator. It provides the ability to better manage loading (and unloading) data. A REST service is provided, so that your data orchestration service (the process that is responsible for delivering data to the accelerator), can trigger the loading of data into ActivePivot.
The REST service will allow loading (and unloading) of topics (or groups of topics, called aliases) by scope. Currently the scope allows the specification of an as-of date, however it can easily be customized to support, for example, desks or legal entities. The scope helps the data load controller find the data when loading, and becomes the criteria for removing data when unloading. For example, it is possible to load (or unload) all SBM sensitivities (the topic alias) for the as-of date 2019-08-09 (the scope).
Spring Boot
Starting from version 2.1.0, the frtb-starter and frtb-activemonitor projects have been migrated to Spring Boot. As an immediate consequence, building these projects will produce executable JAR files instead of WAR files, which are deployed as standalone applications.
Improvements
None
2.0.1
Known issues
None
New features
None
Improvements
None
2.0.0
Known issues
None
New features
BCBS 457 implementation
This release adds support for the IMA changes in BCBS 457, including:
-
New PLA test metrics (Spearman correlation metric and Kolmogorov-Smirnov test metric), and the RAG zones associated with the results of these tests.
-
The Capital Surcharge measure is introduced based on the PLA zones.
-
The SES calculations have been updated with the new formulas and additional Equity risk class for idiosyncratic non-modellable risk-factors.
-
The ACC measure has been updated to the ACR (Aggregated Capital Requirement) and associated RWA (risk-weighted assets).
Note: for purposes of the Capital Surcharge and ACR calculations, the PLA zone of a desk is determined quarterly, and also we expect that the PLA zone will be monitored daily (for internal use), Hence the ACR calculations use the PLA zone provided in the input files instead of using the results directly from the PLA tests.
Note: We have changed the default interpretation of [MAR21.71] (see FRTBParameters.csv changes in the migration guide) to match the BCBS 352 calculations. Previously, we identified three different plausible interpretations of [MAR21.71], and used a "hybrid" approach as the default. However, it has since been discovered that the adjustments in [MAR21.4](5)(b) to avoid taking the square root of a negative number do not always work for this hybrid approach (the value can remain negative). All three configuration options are still available.
Improvements
None