Axsym Engineering Publications

Macroeconomic Forecasting

A new approach to macroeconomic forecasting is presented in the Economy Dynamics Series publications below to overcome limitations in the current use of semi-empirical functions correlated to historical datasets. The limitation of existing methods occurs when the input variables of the function being used for forecasting are beyond the range of values held within the correlated dataset. Thus, extrapolating the semi-empirical function introduces numerical uncertainty into the forecast results. This requires intervention by forecasters to process differences between model predictions and actual data from the recent past. Because the intervention process can be specific to the individual forecaster, two forecasters using the same model can end up with very different forecasts. The new approach, called Economy Dynamics, avoids the limitation of extrapolating semi-empirical functions. This will enable forecasts to be validated against metrics at a regional level such as economic output, poverty levels and inflation and against the environmental impact of consumer behaviour, which is not currently possible.

Economy Dynamics Series

Paper 00 Towards a comprehensive macroeconomic simulation capability: Economy Dynamics. An overview of the Economy Dynamics methodology for macroeconomic analysis and forecasting.

Paper 01 Modelling decision-making from population data: Household Income Forcing. Household financial decision-making can be captured from population data by using surrogate modelling methods. The probability of a given financial action can then be modelled under different household financial circumstances by replaying the surrogate model using the revised household financial circumstances.

Paper 02 The economy in one equation: The Macroeconomic Primitive. All financial interactions can be represented by a single equation defining the interdependency between income and expenditure of all consumers.

Software architecture for analytical processes

A new approach to the structure of software architecture to integrate analysis software and data in an adaptable and easy to upgrade framework is presented in the Teclab Series publications below. The new approach enables an analysts to select the simulation type, within a single analytical environment, and have full control over the algorithms and options used, without any duplication of software or methods across a department or company. The framework is truly modular, allowing alternative analysis methods to be interchangeable within the simulation, selected at set-up by the analyst. The architecture structure avoids the common problem of analytical tools becoming fragmented and prevents upgrade requirements encountering a dead-end. Implementing our architecture ensures there is no duplication of effort in an analytical processes between staff and that all analysts are working with the latest datasets, reduces data handling errors and reduces analysis turnaround times.

Teclab Series

Teclab 01 Handling data in automated analysis processes. Presenting a new central data management concept for storing varied data ontologies and ensuring rigorous data access control. Its application within Axsym’s software integration environment Teclab is demonstrated by a case study to automate a road network management analysis process.

Teclab 02 Automating analysis processes for complex modelling ontologies. Presenting a new analysis governance concept for intrinsically determining the model call sequence and managing data dependencies. Its application within Axsym’s software integration environment Teclab is demonstrated by a case study to automate local energy system design feasibility studies.