Setup > Tags 

Tags

Risksvr™ filters, accumulates and displays results according to:

  1. Dimension(s).
  2. Tag hierarchy.
  3. Tag.
  4. Filters.
  5. Formulas (Beta).

Tags are a powerful mechanism to slice and dice results .Tags are designed to associate meaningful user-defined or engine generated tokens with intermediate values and results.

Tag Specifications
Every tag or slash separated tags belong to a Dimension.

A Tag is either generated by the engine or defined by the user.

You can mix and invoke generated and user tags in your reports.

 Types of User Defined Tags

 


How do you use Tags ?

 

Slice Data You must define your Tag(s): In the Trade Header or in the Trade legs. In the Credit Accounts or In the Engine Setup (Analysis).

 

Dice the Data in your Reports You can then invoke specifically each tag value by it's name  to build a custom report.

Dimensions and Tags "filter" or namespace. You first scope the Dimension then the tag(s). i.e. PORTFOLIO/TAG1/TAG2. If you don't invoke a Tag you get all the tags in the Dimension namespace. If you don't call a specific dimension, you get all the data produced for that report. If you define output data or reports.

 

 

 

Slice DataThis depends on Tag Type


  1. Market Risk Tags are attached  to Trades:
    You create a TAG token that is associated to a Dimension and one or more strings separated by forward slashes "/".

    For example: in your trade TAG  PORTFOLIO/customer1/Alpha

    If no Dimension is defined, the Dimension is assumed the Default dimension with Id 0.
    If you are using one single dimension,  no specific dimension name or identifier is required.


  2. Credit Risk:
    3. Credit Tags are attached to Trades and Party Accounts.

    Credit-Tags are attached to the Trade or the Party Account.
    Credit Tags are called with the CREDITTAG token instead of TAGtoken used for market risk accumulation.
    Credit Tags have a netting status that can be disabled, enabled and forced on or off !

    If no Dimension is defined, the Default Dimension is assumed, with Id 0.

    As with Market Tags you don't need to call the dimension name or id if only one single dimension is called.

  3. Yield Curve Buckets are used to aggregate results according to User specified Maturity Bands or Time-Buckets.

    If you define 3 Month Time-bands over ten years, every flow will be aggregated according to respective Risk Factor Maturity.

    Time Buckets are defined in the Analysis section or at the Trade and Leg level.

    Time Buckets can be defined as a time period and a number of bands, a Frequency and Maturity or a series of variable Time-Bands.

    Time Buckets are invoked with MATURITYTAGS and or TAGPERIOD  or TAGMATURITY tokens.


    For Example to Define a series of Quarterly Buckets spanning Ten Years you can define:

    User Defined Variable Bands:

    MATURITYTAGS  40
    TAGMATURITY 91 182 273 365 .....

    Fixed Period Time-Bands:

    MATURITYTAGS  40
    TAGPERIOD  0.25

    or

    TAGMATURITY  10Y
    TAGPERIOD  0.25

    or

    TAGMATURITY  120M
    TAGPERIOD  3M

    or

    MATURITYTAGS 40
    TAGPERIOD  91D

    etc


    Note: Results will be very close but NOT identical as the engine is converted to Days per Annum.

    If you are computing with 6 digits precision you should use Time-Bands defined in Days.

Why are tags SO Important !

Tags are very important for advanced Risk analysis. Tags are used to accumulate or aggregate risk factor values at each and every simulation horizon.

Possibilities are endless as results can be re-combined, renamed and invoked in other reports !

 How do you Define Tags ?

  1. Market Risk Tags are defined in the Position. 
  2. Credit  Risk Tags are defined in the Position and Account DataSources.

What is a Tag Hierarchy ?

 

A Tag hierarchy is a sequence of Tags. 
Each Tag is separated by a forward slash "/". 
A Tag hierarchy looks very much like a folder hierarchy  or web address URI, but with or without the first slash (ie root slash).

For example, you might associate a tag hierarchy to your positions as:

strategy_1/ bonds/ high-yield for, say, a corporate bond.
Strategy_1/ bonds/riskless-asset for a government bond,

As the name implies the  strategy_1/ bonds hierarchy will contain the total (fair-value or combined risk values) for both the government and corporate bonds included above.

etc.

The engine will then generate (seven) risk statistics accumulators for each individual tag and aggregated dimension   as well as the total value for each Dimension.
These values will be stored FOR EACH simulation horizon defined in the Analysis.

Each tag creates a new report section , however each additional tag hierarchy in this dimension will grow the report's size exponentially O(N).

Summary / Examples

 

User Defined Tags


  • Types of Tags Description
    User Defined Tags  
    Market Tags Market Tags are a sequence of User defined tags.
    They can be automated in the Analysis Screen.
    Every tag is separated by a / slash.
    Tags are associated to the Trade with the TAG token
    Credit Tags Credit Tags are a sequence of User defined tags
    They can be automated in the Analysis Screen.
    Every tag is separated by a / slash.
    Credit Tags are associated to Accounts and Trades with the CREDITTAG token.
    User and Engine Defined Tags  
    Time Bucket Tags Time-Buckets are generated by the engine from default Time-Bands or User supplied Time-Bands.
    Engine Defined Tags  
    Risk Factor Tags Risk Factor tags are accumulated according to Risk Factor type. 
    Credit Factor Tags Credit Risk Factor tags are filtered according to their Credit Risk Nature : Value of Recovered/Non-Recovered Portion, Discounted from Forward Value, Hazard, Survival, Default, Forward No-Default.

    •  

    Filters

    To define a filter you invoke the DRILLDOWN, SLICE and or DICE Tokens in the ANALYSIS DataSource

    For Example The Drilldown token will usually include something like this:

    Plain dimension
    DRILLDOWN DIMENSION
    Produces all available Tags, including Totals for the dimension.

    or     Tag Dimension drilldown
    DRILLDOWN DIMENSION/TAG 
    Produces all available Tag hierarchies that begin with the requested "TAG"

    or     Tag Hierarchy dimension
    DRILLDOWN DIMENSION/TAG1/TAG2/TAG3
    Produces data including and below TAG1/TAG2/TAG3

    Available Dimensions


    User Defined and Engine Generated (Automatic) Dimensions:

    TAG. Which is equivalent to MARKETTAG.
    CREDIT
    BENCHMARK
    PORTFOLIO
    BUCKETS

    Engine Defined Dimensions:
    RISKFACTORS.

    Then any Tag if you want to drill any further down.

    Example:

    TAG or MARKETTAG  OFFICEXYZ/DESK456/Trader123/
    CREDITTAG              FINANCE/BANKING/US/EASTCOAST/NY/LEADERS
    BENCHMARK             S&P/DYNAMIC/AGGRESSIVE/XYZ 
    PORTFOLIO             PORTFOLIO1
    PORTFOLIO             PORTFOLIO2
    PORTFOLIO             PORTFOLIO7
    BUCKETS                1M-3M
    RISKFACTORS          EQUITY/INDEX
    RISKFACTORS          EQUITY
    RISKFACTORS          COMMODITY/SPOT
    RISKFACTORS          COMMODITY/FORWARD 
    CREDITRISKFACTORS   DEFAULT (Portfolio Values upon Default event )

    User Defined Advanced Reports.

    In the Advanced Reports section you can choose your results as combinations of different Tags.

    For Example:

    SLICE  TAG/Portfolio1 - CREDITTAG/Portfolio1.

     

    Scatter-Gram Graph Reports:

    To define a scatter-gram report, you can define:

    The X axis as one Dimension /Tag(s).
    The Y axis as one Dimension /Tag(s).
    Values can then be sorted Ascending or Descending with either the Dimension /Tag(s) used for y or x axis.

     

     

    Formulas

    Customers have access to the Formula parser (Beta) A whole series of Names and Values are then accessible: Market and Credit Data you have supplied to the engine. Computed Risk Values: VaR(0.95), DEaR(0.99), Mean or Mean(t), SUM, SUM2,SUM3,SUM4, etc for each and every value or intermediate value, market data or credit data at each and every time-step used in the analysis.