PortfolioScience RiskAPI

Value at Risk (VaR) Calculation


Value At Risk (VaR) is a calculation used to estimate the magnitude of a portfolio's extreme or unlikely future gain or loss. Rather than looking to predict how much a portfolio could make or lose on a typical day, VaR's goal is to calculate, with a certain degree of certainty, large, out of the ordinary profit & loss events that a portfolio might experience. Knowing a portfolio's VaR can help aid in understanding if an investor is taking too much risk.

Who Uses VaR

Value at Risk software is used by banks, hedge funds, mutual funds, brokers, and many other financial service firms. Many of these firms employ VaR to predict the size of future outlying losses or gains that their (or their clients') portfolios might experience. Many firms use VaR to determine the amount of collateral needed from an execution client for a margin loan used to trade financial instruments, for example. Buy-side entities, such as hedge funds, use VaR to determine if a portfolio's allocation exceeds a current risk tolerance or investment mandate.

Value At Risk Calculation Methods

Value at risk is calculated using several different methods. There is no single standard method as each method has advantages and disadvantages that relate to complexity, calculation speed, applicability to certain financial instruments, and many other factors. In general, a VaR method must be appropriate in the way it models the behavior of the components contained in the portfolio being analyzed. For example, if a technique used to calculate VaR is incapable of taking non-linearity into account and is used on a portfolio with financial instruments that behave in a non-linear manner (such as options on equities), the Value at Risk estimate will be incorrect (i.e. over-estimating or under-estimating the future gain/loss of a portfolio). Value at Risk calculation techniques fall into the following categories:

  1. Observation-based
  2. Parametric
  3. Historical Simulation
  4. Monte Carlo

The RiskAPI System

PortfolioScience offers RiskAPI - a totally encapsulated, on-demand service that generates multi-model VaR on portfolios, individual positions, and hypothetical baskets. RiskAPI incorporates historical data, computing power, and analytics all combined into one simple interface to allow individual portfolio managers, hedge funds and service providers to rapidly connect to and employ a VaR engine via a customizable Application Programming Interface (A.P.I). Users can access the service via Excel, using the RiskAPI Add-In, or any of several programming environments using the powerful RiskAPI Enterprise client. Included in the RiskAPI System are:

  • Back-end market data
  • Robust, dynamic analytics
  • Global computing power with dynamic fail-over infrastructure
  • Sophisticated client software