One of the considerations when setting up your organisation will be the source of rates. Each rate in the system belongs to a rate source which may be "public" or "private".

Public Rates

Vector Risk is able to source and maintain vendor rates and make them available to customers. The process we follow is that the customer must licence the rates and facilitate an introduction between the rate vendor and Vector Risk to document the specific access rights of the customer and to agree a rate capture mechanism. 

Rates are collected on a daily basis from these vendors, and history sets are gradually built up. Where history is lacking for an organisation's requirements Vector Risk can work with the rate vendor to source the history and make it available. Currently ICE data services provide a large current and historic data set with three targeted time zones: Asia, Europe and USA. 

Vector Risk is in discussions with major vendors to add wide ranging FRTB data sets. These data sets will include long historic series suitable for FRTB internal models and must only include rates were committed or traded to pass modelability tests.

Private Rates

Rates can also be loaded by your organisation. These rates are not visible to other Vector Risk customers.
Private rates can be loaded at two levels; organisation wide (all environments can use them) or environment specific. The naming convention is as follows:

• Organisation wide rates use a rate source that matches the Org name.
• Environment specific rates use a rate source that matches the Environment name (or more specifically "org.environment").

All rates are loaded against curve definitions, where a curve is defined by four keys (Market, CurveType, Asset and Currency). An organisation can load its own curve definitions but where possible it is recommended that it load rates against standard curve definitions found under the "Default" rate source. 

Note: Pricing functions use built curves rather than raw rates. Vector Risk is able to build curves (eg zero curves built from deposit and swap rates). The curve building can be tailored to a customer's requirements and can be stored under the customer's private rate sources. It is possible to include public rates that the customer has a subscription to (see public rates above) in this curve building process, and still have the built curves stored under a customer's private rate source.

Each curve has assumptions associated with it. Chief among these are the models used for evolution in simulation processes, the interpolation model used to estimate rates in between terms, and the length of time a rate is considered valid (expiry). For example some rates (eg proxy credit premiums) might only be collected monthly so they need a long expiry, whereas FX spot prices are collected daily and should have an expiry of 1 (1 day). These are the sort of considerations needed if the organisation decides to load its own curve definitions.

Customers using the system to run the FRTB standardised approach will be aware of detailed risk factor classifications in that model. These classifications are captured in the Vector Risk's default curve definitions.

Note: Test and UAT environments can be configured to access production curves and rates. Alternatively, rates can be loaded directly to those environments and not accessible to production. Production level rates can only be loaded into production environments.

Process Parameters and Correlations

If you are running Monte Carlo simulations and require process parameters and correlations. you can either load these or we can easily switch on the relevant tasks in your daily (or monthly) process to generate them from historic rate series. See the section Parameter Estimation for more information.