A New Portfolio Optimization Model Considering Hybrid Trading Strategies

Computational intelligence (CI) has been extensively used in financial technology areas to help make smart decisions within enormous solution spaces. This study proposes a new portfolio optimization model using hybrid strategies to concurrently consider long and short investments and compose effective portfolios assessed by an emerging indicator to reach an excellent balance between return and risk. It exploits the trend ratio investment strategy to a detailed extent to evaluate the steady uptrend and downtrend portfolios and provides an in-depth discussion of the U.S. stock market. The proposed method combines long and short selling through a single fund to enhance investment efficiency. Quantum-inspired CI is adopted to construct a near-optimal portfolio efficiently and effectively, and sliding windows are applied to rebalance the portfolio and dynamically discover appropriate periods. The experiment conducts comprehensive statistical tests to show that hybrid strategies significantly improve traditional approaches. The comparison with other traditional methods shows outstanding performances in MDD, PF, RoMaD, return, and trend ratio. The proposed hybrid portfolio optimization application can be complementary, offset the volatility and increase the return, and show great potential in the financial industry.