Back to Community
Identifying statistical mis-pricing

Here is my attempt:

  1. Identify factors that explain weekly returns and compute a hedge portfolio for each asset.
  2. See how the hedge portfolio performs vs. the asset returns

Idea is to enter into a trade if the difference between hedge portfolio and asset returns diverges and close the position when it converges.

Loading notebook preview...
Notebook previews are currently unavailable.
3 responses

Here is one with cumulative returns. The hedge portfolio is co-integrated with the asset.

Loading notebook preview...
Notebook previews are currently unavailable.

Again, this is fit over all the data, then regressed over all the data, right? There's no out-of-sample or cross-validation test at the end, to see if the hedges cointegrate beyond the training period?

Hi Simon,

Here's 25 days out of sample. However this is just for one sample. Any idea how I can check this across all samples and come up with a reliability score?

Loading notebook preview...
Notebook previews are currently unavailable.