VaR Spec is a of with functions , keys and example object that represents a collection of specification parameters that define how the Value at Risk pertaining to the price of a given expressed in units of a specified reference Tradable and a certain time horizon is calculated.
Regardless of the exact definition, the Value at Risk depends on the probability distribution of the tradable's price as of the horizon date, which typically is the next business day.
The construction of the probability distribution is based on the assumption of what the random risk factors are and how they evolve.
This information is supplied through the input
As explained in , all stochastic process coefficients are assumed constant for the purpose of calculating the value of each risk factor at the end of each elementary time step.
This "constant coefficient" assumption can be mitigated by increasing the number of time steps in the simulation, but this is hardly needed for the usual one day VaR horizon.
The correlation between the various risk factors is given by a symmetric square matrix of dimensions NxN, where N is the total number of risk factors involved.
The simulation of the correlated risk factors is afforded by random numbers generated according to and can be optionally set as antithetic.
The confidence interval is defined as part of the simulation specification details through a supplied object of type
Finally the degree of detail contained in the VaR output is controlled by the setting