I want to know analytical machine learning methods (the more innovative, the better!) to calculate contributions of sectors (ex. financials, consumer staples, industirals indices) to the market (ex. s&p index).

In other words, I want to answer this question: which sector had the biggest return and risk contribution to the market from t = t1 to t= t2? What percentage of total return and risk is coming from the TMT sector from t = t1 to t = t2?

Performing simple multiple regression is one way, but what are other advanced/innovative ways which make use of machine learnings?

  • $\begingroup$ Not really relevant, but nevertheless: Francis X. Diebold on supervised machine learning being nothing more than regression estimation in "Machines Learning Finance" (regarding a conference/meeting of the same title). $\endgroup$ – Richard Hardy Jun 12 '18 at 19:10
  • $\begingroup$ You need a lot of features to get anything from ML. OLS and mixed effects is the way to go $\endgroup$ – Aksakal Jun 12 '18 at 19:10
  • $\begingroup$ @Aksakal Thank you very much. I have a few questions. 1) What is mixed effects? 2) Ultimately, the number of independent variables in my data set will go up to like 100. In that case, what kind of ML should I use to select a few features that describe the market during a specific time period? Subset selection/Ridge/Lasso? $\endgroup$ – Jun Jang Jun 12 '18 at 19:12
  • $\begingroup$ @Aksakal, why is there a need for a lot of features? What would that yield? $\endgroup$ – Richard Hardy Jun 12 '18 at 19:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.