tsuz / backtest-options Goto Github PK

Why

The backtest performance can take minutes if the data goes back 18 years. We should complete this within seconds.

Currently, these make the performance slow:

• Reading the CSV file (this can take a minute or more depending on the size)
• Making an options chain from than CSV file (this can take longer than the data sometimes)

Why

A command-line should output the results of a strategy, and one obvious strategy is covered calls.

Acceptance Criteria

• strategy covered-call should run a covered call strategy
• once the command is run, it should output the results such as returns

Why

EOD bid/ask data seems to be more representative than the 1545 data. OHLCV covers traded prices, so this may not be the best representation of simulated trading if one is trying to cross the spread.

Acceptance Criteria

• Import all columns
• Use EOD col instead of 1545
• Use Mid price of EOD bid/ask to get the fair price of options.

Why

Importing data can take a while since data is zipped into many files, so there should be a normalization process to combine into one file. This would save time to read each file when it's looking back many years of daily data.

Acceptance Criteria

• There should be a data/options/${symbol}.csv folder that can create a combined CSV from imported files.
• There should be an import command to achieve this

Why

By buying the underlying stock, selling a near-term call, and buying a far away put will give us a premium while covering the tail risk. There is no name for this, so it will be called PIP.

Acceptance Criteria

• A hedging strategy, PIP, can be specified via command line

Why

One should understand drawdowns and cumulative returns

What

• Create a cumulative profit column
• Show drawdowns in results