我目前在学习MGARCH, 特别是Dynamic Conditional Correlation. 但输出结果中 quasi correlation 不知道怎么解释。VCC和DCC的公式也研究了。但是由于数学基础一般。对公式没有一个很深刻的理解。
先拿英文搜了搜发现有这个问题的人还真不少。而问题也都没有得到解决。比如下面这个
Hey guys,
I have a small problem with interpreting the results of Stata MGARCH DCC/VCC models and I hope you are able to help me.
I understand how the models with dynamic correlations (DCC and VCC) work and I also found in the Stata manual that you have to use "predict" in order to calculate the time-varying correlations for these models. But there is one question that I can't answer with reading the manual or searching in this forum/google:
What does the (quasi-)correlation coefficient in the model output stand for?
- Is that the average estimated correlation over time?
- Is that the estimate for the dynamic correlation at the end of the time series?
- Or the estimate for the dynamic correlation at the beginning of the time series?
- Or something completely different?