Dear Strangers,
For my master thesis I have to research market timing in mutual funds. I use both the Treynor-Mazuy model and the Henriksson-Merton model.
I run an mvreg command as I have a lot of individual funds(180, as dependable variables) and only 2 independable variables. My command is -mvreg depVar1 DepVar2 DepVarX = IndepVar1 IndepVar2-.
So far everything goes well. But as I have seen in my statistical classes I need to make sure the OLS assumptions of my residuals are met like(homoscedasticity, normallity distribution, no autocorrelation etc.) Or do they not have to be met in my specific case? I can't find anything online about it.
While I can use the normality test just find by -predict residuals, res) -sktest res- or -swilk res-, I cant use -estimate hettest-(for homoscedasticity) or the White test. Autocorrelation doesn't work either -estat dwatson-.
Does anyone have any support for me?
I can provide further information if needed.
Best wishes
For my master thesis I have to research market timing in mutual funds. I use both the Treynor-Mazuy model and the Henriksson-Merton model.
I run an mvreg command as I have a lot of individual funds(180, as dependable variables) and only 2 independable variables. My command is -mvreg depVar1 DepVar2 DepVarX = IndepVar1 IndepVar2-.
So far everything goes well. But as I have seen in my statistical classes I need to make sure the OLS assumptions of my residuals are met like(homoscedasticity, normallity distribution, no autocorrelation etc.) Or do they not have to be met in my specific case? I can't find anything online about it.
While I can use the normality test just find by -predict residuals, res) -sktest res- or -swilk res-, I cant use -estimate hettest-(for homoscedasticity) or the White test. Autocorrelation doesn't work either -estat dwatson-.
Does anyone have any support for me?
I can provide further information if needed.
Best wishes
