Proper portfolio construction is a tricky problem to solve, but fortunately, there are several ways to approach it. I use three methods when constructing my portfolio: 1. Pearson Correlation Matrix - My general rule of thumb is to avoid combining strategies with a correlation above 0.20. However, this isn't a hard rule and can be overridden based on the second method. 2. Strategy Logic and Market Regimes - Correlation alone doesn't tell the full story. Two strategies may have a correlation above 0.20 but still complement each other if they exploit different market conditions. For example, one strategy might perform best in strong, trending markets, while another performs better in weaker market environments. Because of this, around 10–20% of the strategies in my portfolio may exceed my 0.20 correlation threshold, provided their underlying logic is sufficiently different and they offer meaningful diversification across regimes. 3. Downside Diversification - Rather than looking only at overall correlations, I also analyze how strategies behave during the portfolio's worst 10% of trading days. The goal is to understand whether strategies that appear diversified under normal conditions still provide diversification when the portfolio is experiencing losses. After all, diversification matters most when things are going wrong. I combine all three approaches when constructing my portfolio. None is sufficient on its own, but together they give a more complete picture of how strategies interact.