We have a defined trading logic for AI algorithms developed.
Losing trades are not closed immediately. Another set of AI algorithm is employed to “nurse” the losing trade over a certain period to minimize the lost or recover net position. However, if position is deemed unrecoverable, AI algorithm will take the lost and closed position to prevent further lost. With other winning trades, this lost can be net off.
AI Algorithms do not use historical data and do not predict. They use information available “at the moment” to decide whether to enter a particular trade.
Algorithms have artificial intelligence, which allow them to trade dynamically in the market. They are trained to be self-learning according to the market condition “at the moment”.
Probabilistic-based Approach: AI algorithms use probability based trading to decide execution of trade “at the moment”
Hardware intensive backend operations: We have and need more than a thousand of servers to support the dynamic AI algorithm trading system.