Author(s)

Yijiao Chen, Xinyi Sun, Xuyuan Zhu

Corresponding Author

Yijiao Chen

ABSTRACT

Investors can make significant gains in today's booming economy by buying and selling gold and bitcoin, two volatile assets. Given this market demand, this paper develops a model that uses daily price to date to determine whether traders should buy, hold or sell assets in their portfolios daily, based on the daily pricing of gold and Bitcoin. To maximize profits for the $1,000, it is invested over the five-year trading period from September 11, 2016, to September 10, 2021. To counter the problem of investment capital benefit maximization, this article obtains from the correlation analysis the initial data processing after using the curve fitting function. It is considered market VWAP trading strategy, taking the original data accumulation generation sequence, then set up the grey differential equation, the mathematical method of b-b computing, the price of the grey prediction model is established. Under this model, the optimal strategy is obtained in this paper: sell bitcoin shares when the amount of bitcoin shares exceeds 30% of the previous day's amount, and sell gold shares when the amount of gold shares exceeds 15% of the previous day's amount. If the price of another product drops after the transaction, use 65% of the proceeds to buy another product. Follow this strategy, and it will have a maximum investment value of $3340. To verify the optimality of the model after descriptive statistical analysis, the intrinsic consistency reliability of the model was analyzed by calculating the Kronbach coefficient, based on which the rationality of the model and its dependence on related factors were analyzed. Considering the precision of the machine learning model, this paper through the machine learning model for gold stocks and currency respectively to forecast stock trading results, which compares the results with the actual trading results. By observing its error size test of goodness of fitting model, the last of careful verification is proved the optimality of the model. Then, in order to study the strategy of the sensitive degree of transaction costs, the transaction costs on the result of policy and situation, this article calculates the correlation coefficient between stock returns and time mathematical methods, using the Pearson coefficient of deformation after as gold stocks and currency trading price impact on the strategic significance test. Then, the program is written and run through Python, and the influence degree of the rise and fall of the gold stock and bitcoin stock trading price on the strategy, the influence degree of the strategy on the stock return result, and the distribution of the final stock profit and loss are presented in the chart.

KEYWORDS

Curve fitting, Grey prediction, Consistency reliability, Pearson correlation coefficient