looking to implement a realized kernel model to forecast realized variance of around ~140 equities and indices in Python given order book data.

I have read "Realised Kernels in Practice: Trades and Quotes" and the associated papers by Nielsen et al and before reinventing the wheel by implementing from scratch I was hoping someone had a link to a public implementation that I could base my work off of to save time/effort and likely find best practices!

If not, any tips on how to most efficiently implement manually would be much appreciated as there are run time constraints on my model!



  • 2
    $\begingroup$ I don't know any Python package that have implemented the Realized kernel method. However, in R there is the highfrequency package developed and maintained by a couple of professors. You can see the source code on Github and try to reimplement the Realized kernel in Python. $\endgroup$
    – Pleb
    Sep 1, 2021 at 5:38
  • $\begingroup$ Thanks you! this is great and exactly what I was looking for $\endgroup$ Sep 1, 2021 at 5:48


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.