Forecast volatility with GARCH(1,1)

[quicken02]

another useless finance formula that you will never use in real life

[bionicturtledotcom]

@quicken02 thanks for the constructive input, i really owe you one ...

[quicken02]

@bionicturtledotcom no need to take offense, I am just saying it how it is, your videos are great, i watch them all the time, i am subscribed

[bionicturtledotcom]

@quicken02 How do you know what it is ... such that you can say how it is? Rob Engle won the Nobel Prize for the ARCH class and spawned a entire field that continues 30 years later. The current applications ... won't ... fit ... on ... this ... page. And the math is beautiful. I'm not offended, I just don't understand how somebody can be incurious.

[GolbergisKing]

You really should sell this stuff. It is good. Check out VideotutorA2Z.

[EHDuke29]

Very good video, makes it a lot more understandable. My question is: how would I use this to forecast future stock prices ($7, $8, ?forecast).

[bionicturtledotcom]

@EHDuke29 Thank you! You can't use GARCH(p,q) to forecast prices, it makes a limiting assumption that conditional returns are normal. And, in practice, volatility is less difficult to predict.

[bionicturtledotcom]

@EHDuke29 what i meant is: by making the limiting assumptions that returns are normal, the GARCH implicitly assumes the price has a drift (mu), which is a forecast

[smokenfly514]

@bionicturtledotcom May I ask what degree(s) you have? Seems like you have a Ph.D in math, stats or finance.. I'm having trouble following your videos. Is there any other way I could figure out what exactly this model is and how it works?

[EHDuke29]

@bionicturtledotcom ok, is GARCH ever used in combination with ARIMA as a way of deciding the 95% prediction bounds? Or this is not possible because of heteroskedasticity? Hope my question makes sense.

[Kralj2007]

Thanks a lot for your great videos! :) As one of my preposters I would be very interested in seeing a vid about the parameters estimation for GARCH.

[masrurkhan]

Thank you so much :)

[GustaveFlaubert1821]

Hi David,

Thanks for the video. I am wondering what happened to the return variable in the second equation. Do this imply that returns forecasts are not required after the first step in order to forecast volatility? Would you know of any link that would go through the math?

Cheers.

[couples72]

Hi was was looking on how to assess the extent that interest rate changes effect exchange rates over a period of about a year. Can I use a GARCH chart to find the extent that interst rate changes was able to stabilize exchange rates.

[sdoofette]

Hi David - very clear - a rare gift nowadays!

Quick layman's question: is it possible to use the technique to identify stock with mean reversion behaviours or is there better alternatives? 9or is it ill advised to follow such strategy all together?)

Thanks in advance for the feedback

[waldymar100]

thanks for answer. my concern was only the 1st garch variance. since initially we don't have a lagged garch variance, we should calculate it somehow. what you are saying is that we should calculate it as a standard deviation (or you call it also moving average volatility). how long do you go back to calculate it (my boss suggest me to take 150 last returns) to calculate first garch? according to john hull, the 1st garch variance is equal just to lagged squared return. what is your comment?

[waldymar100]

Hi David,

very useful video. but, some moments are not complete. again, I would like to know how you got initial "blue" variance (cell e12)? you says nothing about parameters estimation. are you using maximum likelihood approach?

thanks

[bionicturtledotcom]

Hi waldymar, Thanks. Cell e12 is the prior variance, so here it is just calculated as an historical variance for (t-1). Then it becomes the lagg in the updated variance estimate (that's the "autogresssive" aspect of GARCH)...right, nothing about param estimation, but yes MLE is most common. Thanks, David

[miankhaliq]

Hey david thanks for this very useful vid. Could u make a vid about the parameters estimation of GARCH model that how to calculate them? If possible then kindly must do that. Thank you.

[asilvarocha]

this video is so good as much as didactics!!

best wishes...and keep posting.

[AndriuhaB]

Thank you,very helpfull video.