If you invest monthly to index based ETF, does it matter at which stage of economical cycle you start?
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.
img Update Youtube link, link to data and development environment. Jan 1, 2019
.gitignore Create mp4 video from charts. Dec 30, 2018
README.md Add possible improvements. Jan 2, 2019
config_example.py Create mp4 video from charts. Dec 30, 2018
env_example.sh First version of analysis. Dec 8, 2018
msci-world-index-historical-data-1969-2018.xlsx Create mp4 video from charts. Dec 30, 2018
requirements.txt First version of analysis. Dec 8, 2018


Economical Cycle Timing

If you invest monthly to index based world wide ETF, does it matter what is the stage of economical cycle when you start?

Check out the economical-cycle-timing.ipynb for the data analysis.

Related blog post

Here is the original blog in Finnish.

Summary of results

Here is a summary of results in video. It shows 10 year cumulative profits by different starting year for a monthly investor.

Click the image to view on Youtube (0 min 26 sec). alt


The msci-world-index-historical-data-1969-2018.xlsx file contains the investment data.

The data is from here. Remember to allow Flash.

Development environment

I used Ubuntu subsystem for Windows for development.

The env_example.sh contains a template script that you can paste to Ubuntu editor when starting a development session.

Python 3.6 has been used. In Ubuntu 18 python 3.6 should be pre-installed. Try it with command python3 -V.

Install virtual environment by pip3 install virtualenv. Create a new virtualenv by virtualenv venv.

Possible improvements

  • Add 2% compound inflation line to charts
  • Add index line to charts
  • Take into account that only full shares can be purchased
  • Make a sliding analysis since the beginning with 15 years investment time
  • Calculate different strategies
    • Only buy after 5 months of down hill
    • Use machine learning to determine best buy and sell strategies
  • Better documentation for the notebook