The following is a project on a relatively new method of portfolio optimisation - "Hierarchichal Risk Parity". This technique has three sub-parts namely - "Tree Clustering", "Quazi-Diaganolization" and "Recursive Bisection". The last part of this project is to compare the results of this technique to the results of other methods.
This technique being a completlely different one than the comtemporary methods of portfolio optimisation, allowed me to learn a lot of new intricacies of this subject and helped me to get a better look on the problems that a quantitative researcher has to face in real life.
By using various paramenters like 'Maximum Drawdown' and 'Sharpe Ratio', I have compared this method with the other comtemporary methods like 'Method of Uniform Weights' and 'Minimum Variance method' to find that this technique of portfolio optimisation worked much better than other on most parameters.