Our project Cartorisk consists in synthesizing and classifying the risks that may in a near future impact one of 5000 listed companies in all business sectors, or in a specific geographic region, and establish a “Risk Map” which will be posted on our web site.
It conforms to the current trend called “Big Data”, where data sets that are very complex are collected. The main value of our project is to keep the specialists in the financial world informed of the risks susceptible to affect a company, a sector or a continent. A web site will be created, containing the description of our entire approach as well as various graphics allowing an immediate understanding of the displayed and processed information.
This project is very valuable for risk managers: it will enable them to have a global vision of future risks and it will guide their future risk-aware strategies relative to a business.
First, the collection of reports on the website of the SEC (Securities and Exchange Commission). Secondly, the identification and scoring of risks, i.e. we have to create an algorithm capable of identifying all risks and assign a score according to the impact which it would have on company. Thirdly, the storage of information in the database; and finally, the creation of the website which will show the restitution of the risk factors, trends and correlations.
We have encountered many difficulties in realizing this project, mainly due to the fact that our algorithm should be efficient to process all the reports. Moreover the execution of our program on the website should also be fast. A large part of our time was spent in research.
The finally of Cartorisk being to predict risks, it relies greatly on reports from previous years from companies that compose the database. In this way, the system is given the capacity to predict the risks along many different possible scenarios in the following years.