Data Intelligence.
Our data intelligence framework transforms diverse raw data into reliable, structured and actionable insights that power our forecasts and strategies — from many sources of input, through a proprietary database, to forecasts, distributions and market maps.
1. Input Data
The data used for forecasting price trends in transportation, commodity and raw-material markets consist of historical time series normalised through technical and fundamental analysis tools, data from derivative markets, production cost indicators, supply and demand factors, as well as other economic, commercial and industry-specific data that directly or indirectly affect market price formation.
Market Data
- Historical price time series
- Derivative market data
- Freight rate indices
Fundamentals
- Production cost indicators
- Supply & demand factors
- Economic & industry data
2. Data processing
Forecasting of future commodity prices and transportation costs across different trade routes is carried out through mathematical models developed by the Company for the analysis of dynamic non-linear systems, incorporating the principles of multi-criteria analysis and a Bayesian approach to probabilistic forecasting.
Multi-Criteria Analysis
- Pareto Optimality Theory
- Multi-Attribute Utility Theory (MAUT)
- Multi-Objective Optimization (MOO)
- Soft Set Theory
Bayesian Approach
- Prior probabilities estimation
- Relationship identification (NN & correlation analysis)
- Information updating and ranking
Forecasting Output
- Price forecasts
- Cost forecasts
- Probabilistic scenarios
- Risk assessment
3. Proprietary framework
All research, forecasts and strategies developed by the Company are based on its proprietary analytical framework, comprising mathematical models, statistical and econometric methods, data-processing algorithms, software applications and other technologies created and continuously refined by the Company.