Case Study | MF Global Statistical Modelling | Intellisolve

Intellisolve Case Study

MF Global Statistical Modelling

Statistical Modelling Project at MF Global

A statistical model takes the guesswork out of thinking there are patterns in your data. Using a statistical model, you can mathematically prove if there are relationships between elements of a dataset and if so you can assign a number to the degree of the relationship. When you model real world data, relationships are a grey area. There are exceptions to rules due to the complexity of real world interactions. Intellisolve sets up statistical models the same way as financial models, focusing on the business need first and building the model around that.

The Challenge

Take a list of 10,000+ stock price movements and group them into similar transactions. Once the transactions were grouped, run simulations to infer the best trading strategies for specific categories of transactions. Once the best trading strategies were figured out, run optimization algorithms to infer the best parameters for the trading strategies to minimize risk while optimizing profit. All this to be put into a standard reporting format for further analysis.

The Solution

A proper statistical and data analysis solution implemented within an agile framework to give the most relevant solution in the least possible time. Intellisolve has a standard methodology of delivering analysis solutions that has worked well in the past. The share price movements were held in a master data set. The solution was organized in such a way that the parameters could be changed in order to run simulations and scenario analysis and eventually work out the best trading strategies for specific categories of transactions.