The calculation of VaR and similar measures (Marginal VaR, Expected Shortfall) form the backbone of Atoti Market Risk. Input data consists of trade-level/position-level vectors of PnL simulations. The reference data model proposes a breakdown by risk factor - which may or may not be used - and a single set of PnLs per trade.
This Trade PnL file type is identified using the pattern: **TradePnLs*.csv (as specified by mr.var.file-patterns.trade-pnl).
This file is loaded using the TradePnLs topic. See the Topic Aliases table for an understanding of the topic aliases associated with each topic.
For information on the glob patterns used and how to customize them, see note on File name patterns
If TradeId comes from multiple systems you may need to prepend source system to the ID for uniqueness. Note that in certain cases, the TradeId could be for adjustment purposes. In such cases we might only have one PnL vector per Book or desk. The TradeId should contain this information clearly (ADDON or ADJ).
“IR_IRSWAP_LIBOR3M”, “EQ_12345677”, etc.
ScenarioSet
Y
N
String
Name of the scenario set for the PnL vector.
“Historical”, “Stress”
CalculationId
Y
N
String
Name of the PnL vector calculation run. There may be several runs per AsOfDate.
1
RiskFactor
Y
N
String
Underlying risk factor (may be more than one) of the risk class. It is expected that the risk factor name encompasses the definition of the risk factor per the FRTB specification (paragraphs 59-66) or remains as close as possible to this regulation. This field is mandatory.
BHP Billiton_Credit spread
RiskClass
N
N
String
Defines the risk class that the PnL vector is computed for.
CSR non-Sec
SensitivityName
N
Y
String
Name of the sensitivity that the PnL is attributed to.
“Delta”
LiquidityHorizon
N
Y
Integer
The Liquidity Horizon in days. This field is optional.