نوع مقاله : مقاله پژوهشی
نویسندگان
1 دکترا، گروه اقتصاد، دانشکده اقتصاد، دانشگاه مازندران، مازندران، ایران.
2 استاد، گروه اقتصاد، دانشکده اقتصاد، دانشگاه تهران، تهران، ایران.
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
This article explores meta-heuristic methods for optimizing asset portfolios. It calculates the investment efficiency frontier, identifies the minimum risk portfolio, and computes the Sharpe ratio. Optimal currency baskets serve as signals for algorithmic trading, enhancing investment efficiency, particularly in volatile markets like currencies. The proposed algorithmic trading system is based on optimal currency basket selection.The method involves successive optimization of portfolios, using data mining concepts to determine the Value at Risk (VaR) for short-term portfolios. Key topics include non-linear exchange rate forecasting, VaR calculation via the EGARCH model, meta-heuristic optimization of portfolios, and algorithmic trading system design.
To address Markowitz model limitations, future exchange rate predictions using the RNN model are employed. Asset covariance considers exchange rate correlations, scaled by VaR. Random optimization calculates minimum values and asset weights for buying, holding, and selling signals. Selecting 9 out of 28 main currency rates minimizes systematic risk in day trading. Testing the system on 123 daily data points yielded a 27% total return (approximately 4.5% monthly), using only 10% of initial capital and considering transaction costs. The system’s maximum loss was 6%, and the maximum profit was 5.7%.
کلیدواژهها [English]
)
