| Abstract [eng] |
The aim of this thesis is by using AI to create a prototype of a trading strategy in the cryptocurrency market that would yield better returns than the buy-and-hold strategy. The objectives of the study are to analyse the characteristics of the cryptocurrency market, to select a technical analysis strategy for trading based on these characteristics, to enhance the cryptocurrency trading prototype using artificial intelligence, and to conduct a test launch of the developed trading strategy prototype on a simulated cryptocurrency exchange. It was found that the cryptocurrency market is characterized by very high volatility, therefore it is advisable to have a clear trading strategy with a risk management system. A fractal and alligator strategy with a trailing stop-loss was tailored for trading cryptocurrencies. This strategy was backtested by using the “TradingView” platform and yielded greater returns than the buy-and-hold strategy. The fractal and alligator strategy with a trailing stop-loss was further developed by using ChatGPT-4. After these improvements with AI, the strategy showed considerably better results and generated an average return of 1083% on the 4-hour chart, which demonstrated the effectiveness of AI in improving cryptocurrency trading strategies. After programming the strategy and alerts in the “Pine Editor” based on the “AutoView” syntax, the prototype was launched on the Binance testnet using API integration. |
