Can AI-Based Portfolio Recommendations Outperform Traditional Investment Strategies for Beginner Investors?

This research examines whether AI-based portfolio recommendation systems and quantitative investment strategies can outperform a passive buy-and-hold investment approach for beginner investors. Using market data from 2018 to 2024, the study evaluates momentum investing, moving average crossover strategies, and a Random Forest machine learning model against the SPDR S&P 500 ETF (SPY). Performance is measured using returns, volatility, Sharpe ratio, drawdowns, and statistical testing. Results indicate that none of the active strategies significantly outperformed passive investing on a risk-adjusted basis, while the machine learning model achieved below-random predictive accuracy. The findings support the Efficient Market Hypothesis and suggest that low-cost index investing remains the most evidence-based approach for beginner investors.