Ziel
Traders and analysts in the capital markets use two different kinds of information sources: real time (financial) quotes and (political and financial) news. The inherent differences (basically numbers on one side and continuous text on the other side) makes it difficult to automatically combine them in decision support systems. The reduction of the complexity and amount of information, related to financial quotes is achieved by using standard statistical, AI methods, or Neural Nets etc.. Treasury departments in banks and other financial institutions will benefit considerably from a system that will help in the analysis and fusion of real time quote data and textual news. In the HANSA project the forecast was on financial quotes only. This project intends to use news filtering techniques to complement developments in time series analysis.
The objective of this project is to develop an integrated environment for analysing and fusing financial quotes and political and financial news, providing active decision support to traders and analysts in the capital markets.
This project has been initiated by traders and analysts in Bankgesellschaft GNI, ABB, and other financial institutions. This initiative has been given technological shape by IT partners, experienced in R&D and commercial exploitation. This environment will enhance the ability of the user to improve the quality and efficiency of decisions in an area, where fast and flexible reaction based on incomplete and / or inconsistent information is needed. The environment will comprise a number of emerging software technologies such as Neural Nets, Advanced News (Text) Filtering Techniques and Knowledge Based Systems in order to provide a homogeneous framework, which can be used in various market sectors. This approach will lead to the reduction of the "information overload" problem for financial analysts, traders, by means of employing the above mentioned emerging technologies.
The project will combine time series and text filtering techniques, following a predefined categorisation and rule-based deduction of events, according to their significance for the financial markets via a module capturing the mutual interdependency of extracted news and technical analysis output, according to historical data, resulting in the production of three decision supporting outputs: one from time series analysis, one from news analysis and one from the output of the combination. This consists a substantial improvement of todays situation, since the trader has not only decision support from technical analysis, but also from fundamental analysis as well as from their mutual interdependence, something that can prove to be extremely important, especially in critical situations.
The consortium is well balanced: led by end users and supported by IT organisations. The main user and the co-ordinator has extensive experience and strong interest in applying innovative technologies for capital management. In addition to the partners there are (non funded) pilot users.
Experience has shown that computerised trading support is either developed in-house by the user or purchased as a service rather than as a product. The reason is that in the case of a product the customer is asked to pay upfront an amount of money, including training and support, and take on trust that the introduction of this new system will result in return of investment and profits. In the case of buying a service risk is minimised, since it is shared by the provider of the service involving no initial investment at all but leads to revenues if the performance of the system is profitable. It is expected, therefore, that for ACE the provision of value added services will lead to a faster return of investment and higher revenues than immediate attempts to sell it as a product. However, at a second stage, as acceptance grows and service users realise the benefits of ACE, the market for a product will become established and users will be willing to buy the system rather than the service.
Wissenschaftliches Gebiet
Programm/Programme
Thema/Themen
Aufforderung zur Vorschlagseinreichung
Data not availableFinanzierungsplan
CSC - Cost-sharing contractsKoordinator
10789 BERLIN
Deutschland