If you look at the eviews output for a G/ARCH equation, the mean equation is the top part of the output (i.e. the standard output whenever you run a linear regression model), and the variance equation is the lower half of the eviews output that adjusts your model for heteroscedasticity concerns (i.e. depending on which model you are running, the variance equation will include several terms including squared lagged residuals, lagged GARCH term, etc)
Hello Sarveshvar. I need some help. I am evaluating a time series data using EGARCH model. In the results, alpha comes negative and gamma comes positive. I read that gamma shall be negative to capture the leverage effect but I get it positive. Please help.
A couple of suggestions: (1) the sum of the EGARCH coefficients is often more important than the sign of an individual coefficient; (2) a lot of the generalizations made about stock data (e.g. leverage, clustering, differencing the levels just once, etc) do not apply every time. Assuming your EGARCH has no other problems with serial correlation, normality, etc, it is possible that EGARCH is not the appropriate model for your particular stock data. That would not be surprising. There is a reason there are so many variations on the ARCH model; (3) without access to your data, one thing to check is that you are using stock returns rather than stock price levels. This conversion is necessary for a whole raft of different statistical reasons but it also has some bearing on the variance equation.
preethi raju the cheapest way is to download the historical price data from a site like Yahoo! Finance and then import that .csv file into eviews (using the “open a foreign file” option)
Jesse C I think it is the daily (or weekly) returns on the Indian nifty 50 index (the US etf equivalent would be INDY and you can download historical price data from sites like Yahoo!, just remember to convert the daily closing prices to daily returns)
sir i am facing some problem in volatility estimation of GARCH model in e views. i enter FTSE 100 data from 1995 to 2007 and estimate it but it shows me please enter continuous data . i want to know what does it mean this statement.
I think it is the daily (or weekly) returns on the Indian nifty 50 index (the US etf equivalent would be INDY and you can download historical price data from sites like Yahoo!, just remember to convert the daily closing prices to daily returns)
Yes, it is very easy. From your GARCH equation output, just select the "forecast" button at the top of the output page and then select the type of forecasting you want to perform. If you are not familiar with dynamic vs. static, how to adjust the forecast sample, etc, there are plenty of videos on YT and 90+% of what you do in an eviews LS forecast will apply to an eviews GARCH forecast.
Sir, i have a question and plz excuse my english.. i'm working on stock index so time series. when i run ARCH-LM test for the residual with EVIEWS i couldn't find any heteroskedasticity !!! what should i do ? thx
@@SarveshwarInani you can use a free program like audacity to take your existing audio recording, boost the sound levels and repost your YT videos in just a couple of minutes. It is common complaint with statistics tutorials on YT, it will make a huge difference to your likes & subs if you make that small change and costs absolutely nothing.
@@tattou3238 Couple of suggestions: (1) you may be using the wrong number of lags for the ARCH-LM test; (2) if you run the ARCH-LM test after running a G/ARCH equation and there is no residual heteroskedasticity that is what you want. Give yourself a pat on the back; or (3) if you are running the ARCH-LM test on a regression model before you incorporate the G/ARCH features (i.e. under equation estimation/estimation settings/method is the ARCH heteroscedasticity test being run on LS or ARCH?) then you may not need to run a G/ARCH because you don't have a heteroscedasticity problem to resolve. You may want to run some alternate heteroscedasticity tests (BPG, White, etc) and also check the residuals visually to confirm that there is not actually any heteroscedasticity. But if the ARCH heteroscedasticity test is run correctly then you may be in the fortunate position of not needing to run a G/ARCH at all (I see that quite often with stock data, despite its reputation for volatility clustering, leverage, etc).
![[Live] : ONE 169 วันนี้!! "อนาโตลี vs อูมาร์"](http://i.ytimg.com/vi/IK3YZVWuFJg/mqdefault.jpg)
hello DR, do you give any teaching session on ARCH and GARCH for day of the week effect ?
Thanks a lot. Impressive and simple presentation 🙏🏻
Thanks sir... it's better to bit more audible
Sir how we will treat with control variables in our mean equation if we are using some control variables also.
Can I use panel data for garch like 6 countries against 8 independent variables?
What is mean equation, variance equation that you mentioned in the video?
If you look at the eviews output for a G/ARCH equation, the mean equation is the top part of the output (i.e. the standard output whenever you run a linear regression model), and the variance equation is the lower half of the eviews output that adjusts your model for heteroscedasticity concerns (i.e. depending on which model you are running, the variance equation will include several terms including squared lagged residuals, lagged GARCH term, etc)
Hello Sarveshvar. I need some help. I am evaluating a time series data using EGARCH model. In the results, alpha comes negative and gamma comes positive. I read that gamma shall be negative to capture the leverage effect but I get it positive. Please help.
P.S. I am using rugarch package in RStudio.
A couple of suggestions: (1) the sum of the EGARCH coefficients is often more important than the sign of an individual coefficient; (2) a lot of the generalizations made about stock data (e.g. leverage, clustering, differencing the levels just once, etc) do not apply every time. Assuming your EGARCH has no other problems with serial correlation, normality, etc, it is possible that EGARCH is not the appropriate model for your particular stock data. That would not be surprising. There is a reason there are so many variations on the ARCH model; (3) without access to your data, one thing to check is that you are using stock returns rather than stock price levels. This conversion is necessary for a whole raft of different statistical reasons but it also has some bearing on the variance equation.
Hello sir can you please tell me how to import stock data into eviews
preethi raju the cheapest way is to download the historical price data from a site like Yahoo! Finance and then import that .csv file into eviews (using the “open a foreign file” option)
What does the NIFTY_RET variable mean?
Jesse C I think it is the daily (or weekly) returns on the Indian nifty 50 index (the US etf equivalent would be INDY and you can download historical price data from sites like Yahoo!, just remember to convert the daily closing prices to daily returns)
sir i am facing some problem in volatility estimation of GARCH model in e views.
i enter FTSE 100 data from 1995 to 2007 and estimate it but it shows me please enter continuous data . i want to know what does it mean this statement.
What is the frequency of the data?
remove all NA in your data. also check for double decimal points or little overlooked errors in your data
Couldn't understand about NIFTY_RET variable. Kindly give some detailed explanation regarding this variable as i am new learner.
I think it is the daily (or weekly) returns on the Indian nifty 50 index (the US etf equivalent would be INDY and you can download historical price data from sites like Yahoo!, just remember to convert the daily closing prices to daily returns)
@@lawjef Sir how to convert daily closing prices to daily return.
how can we do forecasting of GARCH (1 1) model in eviews?
Yes, it is very easy. From your GARCH equation output, just select the "forecast" button at the top of the output page and then select the type of forecasting you want to perform. If you are not familiar with dynamic vs. static, how to adjust the forecast sample, etc, there are plenty of videos on YT and 90+% of what you do in an eviews LS forecast will apply to an eviews GARCH forecast.
Frequemcy of the data is 3240.
is it daily or monthly?
daily
I have never come across such problem for daily data. Please check if all dates in date column are in proper and consistent format.
can i send u my data sets link? you can it and give me suggestion regarding this problem.
www.google.com/finance/historical?q=INDEXFTSE%3AUKX&ei=rkPCWKAfg9S4BObHuYgB
this is the link of my data sets an di m taking data from1995 to 2007.
we hear nothing :(
Thanks for pointing out. Next time, I will try to improve!
Sir, i have a question and plz excuse my english.. i'm working on stock index so time series. when i run ARCH-LM test for the residual with EVIEWS i couldn't find any heteroskedasticity !!! what should i do ? thx
@@SarveshwarInani you can use a free program like audacity to take your existing audio recording, boost the sound levels and repost your YT videos in just a couple of minutes. It is common complaint with statistics tutorials on YT, it will make a huge difference to your likes & subs if you make that small change and costs absolutely nothing.
@@tattou3238 Couple of suggestions: (1) you may be using the wrong number of lags for the ARCH-LM test; (2) if you run the ARCH-LM test after running a G/ARCH equation and there is no residual heteroskedasticity that is what you want. Give yourself a pat on the back; or (3) if you are running the ARCH-LM test on a regression model before you incorporate the G/ARCH features (i.e. under equation estimation/estimation settings/method is the ARCH heteroscedasticity test being run on LS or ARCH?) then you may not need to run a G/ARCH because you don't have a heteroscedasticity problem to resolve. You may want to run some alternate heteroscedasticity tests (BPG, White, etc) and also check the residuals visually to confirm that there is not actually any heteroscedasticity. But if the ARCH heteroscedasticity test is run correctly then you may be in the fortunate position of not needing to run a G/ARCH at all (I see that quite often with stock data, despite its reputation for volatility clustering, leverage, etc).
@@lawjef thank you
itna jorse kyun bol rahe ho.... aur dhire bolo.....