The stock market is a critical element of the financial system, where individual and institutional investors engage in the pursuit of optimized returns. However, systemic and non-systemic risks are inherent in stock markets, making it essential to identify, address and reduce these risks to maintain market stability and ensure investors’ safety. A stock market is a complex system with properties like non-linearity, emergence and phase transition. Investors’ comparative analysis of stock returns generates underlying stock return network in a stock market, which introduces systemic risks including random crashes and heightened, uncontrolled structural entropy. In this study, we propose an algorithm to reduce specific systemic risks of a stock market by controlling the topology of the underlying stock return network. The topology of the generated network can then be used to construct stock portfolio with reduced volatility-return ratio. Using Dhaka Stock Exchange trading data from 2015 to 2016, we demonstrate the effectiveness of our method in reducing risks and providing a more stable investment environment.