Forecast volatility with GARCH(1,1) (FRM T2-24)

Поділитися
Вставка
  • Опубліковано 20 січ 2025

КОМЕНТАРІ • 15

  • @noname_whatsoever
    @noname_whatsoever 6 років тому +14

    You have created the most comprehensive and well made content library on financial/statistical topics on UA-cam. With it, you have helped countless students. People in my postgraduate finance class were joking they got their degrees at BionicTurtle university, since the quality of your explanations is so high and you cover just about everything.
    It's just amazing work. Thank you!

  • @andreapatratti9291
    @andreapatratti9291 5 років тому +5

    how do we decide the right value for Beta?

  • @simonejongemans7399
    @simonejongemans7399 6 років тому +3

    Very- very good explination. Great work!

  • @elijahli6546
    @elijahli6546 5 років тому +1

    God bless you sir!

  • @anarkulovaaizhan
    @anarkulovaaizhan 4 роки тому

    Thank you! Your explanations are the best!

  • @aminurrahmanchowdhury1964
    @aminurrahmanchowdhury1964 3 роки тому

    Where is the lagged component?

  • @andyb1336
    @andyb1336 4 роки тому +2

    I am looking everywhere for a clear explanation for where you find the parameter values alpha beta etc, but there is no clear answer anywhere. Why did you not explain why and how these are obtained?

    • @joejohnoptimus
      @joejohnoptimus 4 роки тому

      Apparently the most common approach to obtain those values is to use the "Maximum Likelihood Estimate"
      ua-cam.com/video/Xef4LL2KUy0/v-deo.html

  • @seineyumnam4374
    @seineyumnam4374 6 років тому

    so how do we get the estimated vol sigma-t that is used in forecasting sigma-t+1?

    • @bionicturtle
      @bionicturtle  6 років тому

      right at the beginning I explain that σ(n) is today's estimate as given by a GARCH(1,1) volatility estimate model where σ(n) = ω + α*µ(n-1)^2 + β*σ(n-1)^2. The forecast is derived from this same GARCH(1,1), so it's all GARCH. The previous video is this same place list goes deeper on GARCH(1.1): ua-cam.com/video/7PEaWHIGTFs/v-deo.html

  • @Bmmhable
    @Bmmhable 3 роки тому

    Hi Professor, where does the forecasted volatility/variance equation come from? It seems not to require the previous returns (u_(n+j)) at all, which is neat. I also was under the assumption that the expectation of the conditional variance is the long-range variance V_L, so why does it represent a forecasted variance here?

  • @Jupiter1423
    @Jupiter1423 2 роки тому

    That intro to stats class doesnt even get you close to where you need to be

  • @aminurrahmanchowdhury1964
    @aminurrahmanchowdhury1964 3 роки тому

    It's not useful at all.