This demo utilizes the Coinbase Pro public end points and the Coinbase Pro client for Python
The data files used are provided for convenience:
An SMA Strategy is Backtested using Bitcoin-US Dollar (BTC-USD) data from Coinbase Pro public data-endpoint.
Cumulative:
Returns 1.199487
Strategy 1.027011
Volatility:
Returns 0.140443
Strategy 0.140796
The log returns for BTC-USD are shown in a histogram, illustrating 'fat-tails'.
Linear Ordinary Least Squares (OLS) Regression is backtested and compared to daily returns
Cumulative:
returns 1.254766
strat_ols_1 1.322613
strat_ols_2 1.229719
k-means clustering is demonstrated and compared to OLS results
Cumulative:
returns 1.254766
strat_clus 1.396107
The notebook contains all the source code to reproduce the results summarized above. The data files need to be saved in the same directory as the notebook. The data files can be refreshed using the Coinbase Pro public data end points, which are referenced in comments through-out the notebook. Notebook