I am creating a tool to analyze asset correlations and exploring performance of portfolios with minimized total absolute correlation between assets. Right now I am using the last 365 days of historical data (daily adjusted close) normalized to % change, but I have seen some suggestions to use monthly asset price movements instead. I also want to filter out assets with "flat" performance even though they may have a low correlation to the portfolio (such as bonds or other fixed income). The objective is to use market timing along with periodic re-balancing to keep the risk exposure low, including international markets.
Is anyone familiar with using low correlation portfolio methods ? There's a lot of theory there, but in practice it's more difficult to diversify properly. You get different results depending on time horizon etc.
Is anyone familiar with using low correlation portfolio methods ? There's a lot of theory there, but in practice it's more difficult to diversify properly. You get different results depending on time horizon etc.
