- Constructs portfolios by constrained quadratic optimization --- minizing volatility, maxizing returns, and reducing over-fitting by encouraging to equally invest in all positions
- Data is sourced from Yahoo finance.
- Frontend tested in Chrome 44 and Firefox 39
- Some D3.js animation is turned off in Firefox.
- Backend written in Python 3.4, requires the following libraries:
- Cvxopt 1.1.7
- Flask 0.10.1
- Numpy 1.9.2
- Pandas 0.16.2
- Get started
- Run "python3 index.py" in terminal.
- Open 127.0.0.1:5000 in a browser.
- Select start- and end-date; add more tickers.
- Click "Quote" to pull data and fit model.
- Fit the model according to your risk-appetite.
- Click red circles in the panel of "efficient frontier" to get constructed portfolios.
- (Over-)simplification about this constructor
- Reinvest dividends.
- Ignore tax, commission fees, price spreads, and price fluctuation in a day.