ark_invest

daily weekday report of @arkinvest ETF activity + data collection

This script was created to:

  1. Extract and save daily csv’s from ARKInvest’s holdings.
  2. Quick command-line report about security additions and removals within each of the 6 ARK ETFs ($ARKK, $ARKW, $ARKG, $ARKQ, $ARKF, $PRNT)
  3. Quick command-line report about significant changes in security positions within each of the 6 ARK ETFs ($ARKK, $ARKW, $ARKG, $ARKQ, $ARKF, $PRNT)
  4. Option to save day-over-day changes in holdings within each of the 6 ARK ETFs ($ARKK, $ARKW, $ARKG, $ARKQ, $ARKF, $PRNT)

Current suggested process flow as follows:

From command-line:

  1. Run ‘import_status()’:
    • Saves current csv from ARKInvest site for each of the 6 ETFs as ‘temp.csv’ in respective data subdirectories.
    • Identifies whether the temp file is new, or a duplicate of your previous day’s run. *ARK appears to upload new csv’s at 7PM EST.
    • If step 2 does not identify a new import, the program will exit().
  2. Run ‘allocation()’:
    • Compares ticker presence within the two most recent confirmed files.
  3. Run ‘changes(significance)’:
    • Identifies significant increases or decreases in shares within the two most recent confirmed files. Significance parameter = float.
    • Example: changes(6.0) -> Within each ETF, return shares increase >= 6.0% AND return shares decrease <= -6.0%
  4. Run ‘all_dfs()’ if you would like to access each of the ETF’s most recent day-over-day report. The function will return a list of pandas dataframes. *See below:
    • DoD_reports = all_dfs()
    • arkk = DoD_reports[0]
    • arkw = DoD_reports[1]
    • arkg = DoD_reports[2]
    • arkq = DoD_reports[3]
    • arkf = DoD_reports[4]
    • prnt = DoD_reports[5]

Still a work in progress but wanted to share with everyone asap. Curious to see if this is helpful!

Some items I am working through now:

  1. What if ARK updates some of the files, but not all?
  2. Is it convenient to exit program after identifying that a new csv is not available?
  3. You will notice a ‘queue’ variable that I am playing around with. This is to potentially only pass eligible import items to different functions.
  4. Considering using a SQLite db file instead of saving files to subdirectories (obviously not ideal)
  5. Considering dashboard or API endpoint on simplimoku.com (a simple site for ichimoku charting and ichimoku signal screener)

GitHub

https://github.com/import-pandas/ark_invest