We are a small team of academics and professionals, some focused on economics, others in systems engineering. We decided to combine strategies in the framework of classical financial data technical analysis along with ´ad hoc´ artificial neural network algorithms to produce a software aiming at financial forecasting. To be specific, we are developing and testing neural algorithms providing simple and accurate stock quotes trend indicators. We decided, in this preliminary stage, the forecasting signals to be available ´free´ to the public. This key choice is motivated to have a feedback from the vast community of on-line home traders interested in very simple and yet accurate trading signals. To be specific, we focus on about 500 live stock quote prices from 7 markets: the US New York Stock Exchange, the US NASDAQ Stock market, the London Stock Exchange, the Deutsche BÃ¶rse Frankfurt Stock Exchange, the Euronext Paris, the Borsa Italiana Milan Stock Exchange and the SIX Swiss Exchange in Zurich. Based on public data available on internet, for each stock quote we provide every five minutes:
- two sets of trend indicators, respectively short and medium term.
- each set consists of 5 qualitative trend indicators: strong uptrend,weak uptrend,lateral movement, weak downtrend, strong downtrend.
- data are represented in a row of a table, with a marker whenever there is a trend change.
- based on trend changes, different trading strategies can be performed. Some of them are reported for each quote and recorded, to show the effectiveness of the trend signals.
Notice the team at this project stage does not guarantee neither the correctness of stock quote prices, nor the accuracy of forecasts, so that the data must not be considered investment advices. However, by recording the trend indicators, it can be verified ´a posteriori´ the accuracy of the neural trend forecast.