# Message Boards

Posted 2 years ago
7254 Views
|
6 Replies
|
12 Total Likes
|
6 Replies
Sort By:
Posted 2 years ago
 Great post. It is possible to define copulas with higher degrees of asymmetry than the Clayton, Frank or Gumbel-Howard varieties by taking [Theta] in the normal copula to be a function of space rather than a constant: Attachments:
Posted 2 years ago
 Excellent post Jonathan. Looking forward to similar ones in the near future. I have a question regarding the calculation of the returns. For example, the S&P 500 on February 1, 2010 was 1089.19 and on the next day it increased to 1103.32 but when applying the original formula the log return is -0.0128895. Shouldn't the formula be reversed? Is my reasoning correct or am I missing something? Thanks in advance for the clarification.
Posted 2 years ago
 - Congratulations! This post is now a Staff Pick! Thank you for your wonderful contributions. Please, keep them coming!
Posted 2 years ago
 Here is a notebook that expalins how to code asymmetric Gaussian copulas in explicit detail:Note: Updated the notebook to refer to c as a copula density, rather than a copula, the latter not being precisely correct. Attachments:
 http://jonathankinlay.com/2017/01/copulas-risk-management/Copulas in Risk Management on Page 4 has: SP500returns = Log[Drop[SP500prices[[All, 2]], 1]] - Log[Drop[SP500prices[[All, 2]], -1]]; NASDAQreturns = Log[Drop[NASDAQprices[[All, 2]], 1]] - Log[Drop[NASDAQprices[[All, 2]], -1]]; Instead of (Drop -1) Minus (Drop 1)