Inverse Volatility with Stock2vec Clustering

Strategy

0. Stock2vec Clustering
   • make dailty log-return series Embedding Vectors by Word2vec-Skipgram
   • pass the embedding vectors to K-Means Clustering Algorithm
     • num_clusters : 5
1. Stock2Vec Clustering + Double Inverse Volatility
   • inverse volatility between assets in one cluster
   • inverse volatility between clusters
2. Stock2Vec Clustering + inverse volatility + min volatility
   • inverse volatility between assets in one cluster
   • minimize portfolio volatility using clusters
3. Risk parity
4. Minimum Volatility
5. Maximum Sharpe Ratio
6. Equal Weight
7. Inverse Volatility

Implemenation

python main.py

Data

• ETF (2008.01 ~ 2022.02)
  • VTI : Vanguard Total Stock Market Index Fund ETF
  • VEA : Vanguard Developed Markets Index Fund ETF
  • VWO : Vanguard Emerging Markets Stock Index Fund ETF
  • IAU : iShares COMEX Gold Trust ETF
  • DBC : Invesco DB Commodity Index Tracking Fund
  • XLB : Materials Select Sector SPDR Fund
  • XLE : Energy Select Sector SPDR Fund
  • XLF : Financial Select Sector SPDR Fund
  • XLI : Industrial Select Sector SPDR Fund
  • XLK : Technology Select Sector SPDR Fund
  • XLP : Consumer Staples Select Sector SPDR Fund
  • XLU : Utilities Select Sector SPDR Fund
  • XLV : Health Care Select Sector SPDR Fund
  • XLY : 미 경기소비재 ETF

Result

Requirements

finance-datareader==0.9.31
gensim==4.1.2
matplotlib
numpy==1.19.3
pandas==1.3.4
scikit-learn==1.0.2
scipy==1.8.0
tqdm==4.62.3