<img height="1" width="1" style="display:none" src="https://www.facebook.com/tr?id=974250883405448&amp;ev=PageView&amp;noscript=1">

Risk adjusted modeling and simulation powered by Azure

Read more! Read more!

PUBLISHED: 21 June 2021


About the Customer

One of the largest holding companies in the middle-east region with several billion in assets under management, its investment portfolio covers a diversified range of industries, ranging from financial services to hospitality.


The challenge

Given the nature of customer’s business, they needed a strong probabilistic solution to manage risk-adjusted modeling and planning to consolidate their large portfolio of assets and investments.

The solution had to be:

  • Easily adaptable to each business characteristic and peculiarities
  • Compatible with a hierarchical company structure
  • A full probabilistic solution easily accessible as a web application
  • Powerful in terms of graphical reporting for financial communication
  • Best-in-class in quantitative risk analysis
  • Field-proven for ERM and Financial modeling
  • With a flexible licensing plan to fit the large number of users with different profiles


The solution

Techedge has in its portfolio Riskturn, a solution that allows users to define their financial models and input data and perform simulations based on Montecarlo methods, a broad class of computational algorithms that rely on a repeated random sampling of input data to obtain a distribution of numerical outcomes.

Riskturn was fitting with the majority of customer’s requirements but the original on-premise architecture was not the best one to provide the scalability and cost-effectiveness for this specific use case, with the key need to perform simultaneous simulations on hundreds of different financial cases, reflecting the multi-industry business of our customer.

We decided that the solution needed to be redesigned to leverage the full extent of a cloud-native architecture based on Microsoft Azure with the introduction of the following logical components:

  • Web Application: Asp .Net Core web application for data configuration and simulation management. Hosted on an Azure App Service.

  • Simulation Engine: Azure Batch application for simulation execution. 

  • Database: Sql Server database containing the application data and simulation outputs.

  • Cache: to boost simulation performance storing intermediate results of each run we selected Azure Radis Cache.  

This technological upgrade allowed to provide the customer with a fully satisfactory solution, because:

  • The introduction of Azure Batch for high-performance computing brings in several advantages in probabilistic simulation performance, thanks to pools of dedicated resources for calculation, automatic horizontal scaling and out-of-the queue system for concurrent executions.

  • With the help of Azure Redis Cache, each simulation output is stored in memory, quick access and temporary storage support the execution of full-scale Montecarlo methods. 

  • Azure App service is cheaper than a virtual machine and Azure Batch can provide better performance at a fraction of the cost, with an extremely positive impact on ROI.

How the solution helps

Moving to Azure avoided the management and maintenance of on-premise servers, saving time and furtherly reducing costs by delegating these activities directly to Microsoft services.

In this specific use case, the running costs for the infrastructure were decreased by 1/3, thanks to the pay-as-you-go pricing model of Microsoft Azure and a proper choice of the involved services.

Moreover, the use of cloud-native services highly increased the application performance and stability: the time to perform tens of thousands of cycles of Montecarlo simulation decreased from 10 to 2 minutes for each run, enabling the customer to easily manage all the company assets and perform quantitative analysis even for the whole portfolio in a reasonable amount of time, increasing analytical capabilities and providing a better decision making support to business users.