Predicting trends in financial time series is one of the crucial aspects for traders and investors. Moving Averages (MAs) are widely used in technical analysis to identify trends and generate trading signals, However there is vast variability in terms of using Moving AverageCross Over Strategy, and subjectivity in MACS strategy arises from different factors. In this paper, we propose a novel approach called Moving Average Randomized Trees (MARTs) that combines Moving Averages with tree-based approach to enhance the performance of MACS. MARTs leverage the power of decision trees to capture complex patterns in financial data and generate accurate trading signals. By integrating Moving Averages with Random Forests, MARTs provide a more robust and adaptive framework for trend prediction in financial markets. We conduct extensive experiments on historical stock price data to evaluate the effectiveness of the proposed approach. Our results demonstrate that MARTs outperform traditional MACS and other benchmark models in terms of accuracy and profitability. The proposed model offers a promising solution for traders and investors seeking to improve their trading strategies and achieve better returns in financial markets.