Atoti Market Risk overview

Atoti Market Risk provides a set of tools designed for market risk and product control management processes across multiple asset classes.

This page gives you an overview of what comes out of the box with the solution.

Key features out of the box

VaR aggregation

Here’s the list of Value-at-Risk (VaR) methodologies available within Atoti Market Risk:

• Historical VaR (Weighted/Unweighted)
• Parametric VaR
• Stressed VaR
• Component VaR
• Incremental VaR
• Harrel Davis VaR
• Expected Shortfall
• Taylor VaR

As PnL vectors are fed into Atoti Market Risk, the aggregation of VaR happens in real time with the transparency to slice data down to the most granular trade level information.

Taylor VaR

Atoti Market Risk offers out-of-the-box Taylor Series approximation in its computation of Taylor VaR. Taylor VaR algorithms rely on the use of sensitivities produced by a firm to estimate the VaR of its portfolios in response to market data movements over a historical look-back period. This technique offers the following benefits to the business:

• Ability to break down the simulation scenarios: Taylor VaR helps you understand risk attribution from different sensitivities.
• Enables VaR Explain exercises: Taylor VaR can be used in conjunction with full revaluation VaR, with Taylor VaR aiding in understanding how VaR is attributed by different sensitivities. In cases of unexplained VaR, you can investigate its root cause.

Sensitivity aggregation

Out of the box, the following sensitivities are included: Delta, Gamma, Cross Gamma, Theta, Vega, Volga, and Vanna. The solution can be extended to add more sensitivities that aid in providing an analytical layer to understand exposures and how sensitive they are to market risk factors.

Profit and Loss Explain for product control processes

PnL Explain

The Atoti Market Risk toolkit comes with an out-of-the box aggregation logic to compute PnL Explain by utilizing sensitivities at the risk-factor level and associated market data moves. This provides full transparency of the PnL attribution to both risk analysts and product controllers.

Key PnL Explain sensitivities available out of the box:

• Delta PnL Explain
• Gamma PnL Explain
• Theta PnL Explain
• Vega PnL Explain
• Cross Gamma PnL Explain
• Volga PnL Explain
• Vanna PnL Explain

Market data

If the exact vertices/tenor points referenced by the sensitivities don’t coincide with the market data received, Atoti Market Risk can apply linear interpolation or cubic spline techniques for estimating missing or intermediate data points within market data. The interpolation logic can be enabled or disabled.

For more information, see Market Data API.

What-if scenario and book hierarchy analysis

Atoti Market Risk comes with the following What-If Scenarios out of the box:

• PnL Scenario Scaling
• PnL Scenario Add-on
• PnL Scenario Rollover
• Book migration checks across desks to analyze new impact on the firm’s overall risk
• File Upload capability

What-if scenarios enable adjustments to pre-loaded PnL scenarios to simulate potential impacts to the firm’s VaR. These scenarios help you answer questions such as:

• How would VaR change if the PnL vectors linked were increased or reduced?
• What would be the effect on VaR if an add-on were introduced?

By scaling risk factor inputs, these scenarios provide a framework for assessing potential impacts on portfolio positions, risk metrics, and PnL, including the recalibration of PnL vectors used in VaR calculations.

Additionally, File Upload lets you upload alternative datasets for any of the components (VaR, PnL, Sensitivities) as a scenario. Examples include:

• New Market Data set for PnL Explain
• New PnL vector set for VaR
• New Sensitivity set for exposure monitoring

User-defined measures

Atoti Market Risk lets you define your own calculated measures within the UI, and see instant results displayed in your dashboard when adding these metrics.

Customizable and modular design

Atoti Market Risk provides flexibility for implementing the full suite or to select specific individual components, such as VaR, PnL, or Sensitivities, based on business needs.