•
Sen et al. (2023) bake off mean–variance optimisation, hierarchical risk parity, and reinforcement-learning allocators on the same playground. Plain-language intro. Not investment advice.
Takeaway. Mean–variance (MVO), hierarchical risk parity (HRP), and reinforcement-learning (RL) allocators are three different answers to “how should weights move?” MVO inverts an estimated Σ and μ; HRP clusters the correlation matrix and allocates risk down a tree (à la López de Prado); RL treats allocation as a sequential decision. Sen and co-authors put them on one empirical bake-off so you can see fragility vs complexity, not pick a mascot.
A Comparative Study of Portfolio Optimization Methods for the Indian Stock Market (Jaydip Sen et al., 2023; arXiv:2305.17523) is a methods comparison—read it as a lab protocol, not as “RL wins.”
Fit for learners: Open-PDF classic/working paper on portfolio — useful for after-work Python quant learners building intuition before coding.
This Hub note is a plain-language intro. Not investment advice.
Tutorials sell one hero method. MVO looks optimal and then implodes on estimation error. HRP is more stable but is still a heuristic with clustering choices. RL can overfit the training window and hide turnover and costs. Without a shared universe, costs, and walk-forward, “which is best?” is slogans.
You get a single weekend scaffold: pandas returns → three weight engines → one tear sheet. That is how you learn portfolio APIs without marrying a theory.
Tiny evening exercise: on 8 ETF-like return columns, compute equal weight vs inverse-vol weights; print turnover if you rebalance monthly. Then write why HRP’s tree is trying to avoid MVO’s inversion. Learning only.
Landing: arXiv abs
Not investment advice. This Hub note is educational only. The MVO/HRP/RL comparison is academic / working-paper research—not a recommendation to allocate capital with any of the three methods.
Single course CTA: SoloQuant Academy’s Introduction to Python Programming—lists and functions first; then three functions that return a weight vector is a weekend notebook.