THE POTENTIAL FOR REAL-TIME TESTING OF HIGH-FREQUENCY TRADING STRATEGIES THROUGH A DEVELOPED TOOL DURING VOLATILE MARKET CONDITIONS
This study presents a method for testing high-frequency trading (HFT) for algorithms on GPUs using kernel parallelization, code vectorization, and multidimensional matrices. The research evaluates HFT strategies within algorithmic cryptocurrency trading in volatile market conditions, particularly du...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Polish Association for Knowledge Promotion
2023-06-01
|
| Series: | Applied Computer Science |
| Subjects: | |
| Online Access: | http://www.acs.pollub.pl/index.php?option=com_content&view=article&id=566:the-potential-for-real-time-testing-of-high-frequency-trading-strategies-through-a-developed-tool-during-volatile-market-conditions&catid=97:vol-19-no-22023&Itemid=171 |
| Summary: | This study presents a method for testing high-frequency trading (HFT) for algorithms on GPUs using kernel parallelization, code vectorization, and multidimensional matrices. The research evaluates HFT strategies within algorithmic cryptocurrency trading in volatile market conditions, particularly during the COVID-19 pandemic.
The study's objective is to provide an efficient and comprehensive approach to assessing the efficiency and profitability of HFT strategies. The results show that the method effectively evaluates the efficiency and profitability of HFT strategies, as demonstrated by the Sharp ratio of 2.29 and the Sortino ratio of 2.88. The authors suggest that further study on HFT testing methods could be conducted using a tool that directly connects to electronic marketplaces, enabling real-time receipt of high-frequency trading data and simulation of trade decisions. Finally, the study introduces a novel method for testing HFT algorithms on GPUs, offering promising results in assessing the efficiency and profitability of HFT strategies during volatile market conditions.
|
|---|---|
| ISSN: | 1895-3735 2353-6977 |