Enhancing Trend-Following Strategies Using Machine Learning and Time Series Models
This paper integrates machine learning and time series models with trend-following strategies to enhance financial trading performance and reduce risk.
What it examines
The paper explores improving trend-following trading strategies by integrating machine learning and time series models, particularly using Ichimoku Cloud as a base indicator. It combines classical techniques with advanced algorithms such as XGBoost, Naive Bayes, TCN, and Kalman Filter for enhanced decision-making and risk management in volatile markets.
What it concludes
The study shows that merging machine learning and time series forecasting significantly improves trading strategy performance by increasing returns and reducing risk. Applications include automated trading, financial market prediction, and resource optimization in various assets. Future research should explore deeper models and broader asset classes for more generalized trading solutions.
Evidence objects
A comprehensive study integrates machine learning techniques, including Extreme Gradient Boosting and Naive Bayes, with the classic Ichimoku Cloud strategy, enhancing trading performance and reducing risk through lower maximum drawdowns.
key_findings bullet 1 · key_findings · validation V0
Researchers combine ML classification with series forecasting, employing ensemble methods including Temporal Convolutional Networks and Kalman filters, generating signals, reducing false positives, and optimizing trades in volatile markets.
key_findings bullet 2 · key_findings · validation V0
Extensive back-testing on currency pairs from $$2010$$ to $$2023$$ using total return, mean return, standard deviation, and trade count confirms superior profitability, though inconsistencies and scalability issues persist in tests.
key_findings bullet 3 · key_findings · validation V0
Integrating $machine learning$ with time series analysis, this paper refines FX trend-following strategies using established techniques like the Ichimoku Cloud and ML models such as Naive Bayes and XGB. Although it merges known methods instead of groundbreaking innovations, its balanced approach offers compelling insights and practical enhancements to quantitative trading.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
- … Financial industry primarily in trading and investment is undergoing an infrastructural change due to machine learning (… The ever-increasing volatility of financial markets …
Source row: 715 · abstract type: snippet