In this video we did a quick comparison of the portfolio optimization methods. In addition to classic methods such as Mean Variance, HRP and CLA, we also tested two exotic methods: the first is based on the idea of using LSTM model directly to optimize Sharpe value, and the second is a pretrained model that predicts future allocations. Also we created a simple strategy for dynamic rebalancing of the portfolio based on a given model and compared the results.
00:24 Theory and Methods
03:00 Comparison of Allocations
04:30 Testing
06:00 Results
If you want to support this channel:
commerce.coinb...
Code: github.com/Clo...
LSTM optimization methods: paperswithcode...
pyportfolioopt: pypi.org/proje...
14 окт 2024