(EViews10): How to Perform GARCH Diagnostics #garch #diagnostics #garchdiagnostics #archdiagnostics

This video explains how to perform GARCH diagnostics using an approach that beginners can grasp. The GARCH Modeling series has 9 collections on the following topics: (1) ARCH versus GARCH (Background), (2) Basics of GARCH Modeling, (3) how to estimate a simple GARCH model, (4) ARCH versus GARCH (Estimations), (5) how to estimate GARCH-in-Mean models, (6) how to estimate Threshold GARCH (GJR GARCH) models, (7) how to estimate Exponential GARCH models, (8) GARCH models and diagnostics and (9) how to forecast GARCH volatility. So, what is GARCH? Generalised autoregressive indicates that heteroscedasticity observed over different time periods may be autocorrelated; conditional informs that variance is based on past errors; heteroscedasticity implies the series displays unequal variance. Popularised by Tim Bollerslev in 1986.

Why use GARCH: Models the attitude of investors not only towards expected returns but also towards risk (uncertainty); Relates to economic forecasting and measuring volatility; Techniques  GARCH, GARCH-M, TGARCH, EGARCH, PGARCH, CGARCH, IGARCH and several other extensions; Concerned with modeling the volatility of the variance; Conditional and time-varying variance; Deals with stationary (time-invariant mean) and nonstationary (time-varying mean) variables; Nonstationary  varying mean; Heteroscedastic  varying variance; Concerns financial and macroeconomic time series; Duration  daily, weekly, monthly, quarterly (high frequency data); Financial/economic series  stock prices, oil prices, bond prices, inflation rates, exchange rates, interest rates, GDP, unemployment rates etc. What is conditional variance? The assumption of homoscedasticity (constant variance) is very limiting, hence preferable to examine patterns that allow the variance to depend (conditional) on its history. Volatility Clustering: Periods when large changes are followed by further large changes and periods when small changes are followed by further small changes. Shows wild and calm periods.

Some Lessons Learnt: The time-varying variance is modeled by the procedure called autoregressive conditional heteroscedasticity (ARCH); GARCH simply conveys that the series in question has a time-varying variance (heteroscedasticity) that depends on (conditional on) lagged effects (autocorrelation); GARCH model is intuitively appealing because it explains volatility as a function of the errors. These errors are called “shocks” or “news” by financial analysts. They represent the unexpected!; The larger the shocks, the greater the volatility in the series; Since variance is often used to measure volatility, and volatility is a key element in asset pricing theories, GARCH models have become important in empirical finance; Most financial time series like stock prices, exchange rates, oil prices etc. exhibit random walks in their level form, that is, nonstationary (time-varying means)

Need the data used in the video? Click on these links:

References and Readings: Asteriou and Hall (2016) Applied Econometrics, 3ed; Hill, Griffiths and Lim (2008) Principles of Econometrics, 3ed; Roman Kozhan (2010) Financial Econometrics with EViews; Gujarati and Porter (2009) Basic Econometrics, International Edition; R. Engle, “Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation,” Econometrica, vol. 50. No. 1, 1982, pp. 987–1007; A. Bera and M. Higgins, “ARCH Models: Properties, Estimation and Testing,” Journal of Economic Surveys, vol. 7, 1993, pp. 305–366; Bollerslev (1986); Amadeus Wennström (2014) Volatility Forecasting Performance: Evaluation of GARCH type volatility models on Nordic equity indices; Bollerslev, T (1986)“Generalised Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation,” Journal of Econometrica, vol. 31, pp. 307–327; Tsay, R.S. (2002) Analysis of Financial Time Series, John Wiley & Sons, Inc., New York.

Follow up with soft-notes and updates from CrunchEconometrix:

[CrunchEconometrix]

I want to appreciate all my subscribers from across the globe (Africa, Asia, Europe, the Middle East, The Americas, and The Pacific). Thank you all for your support. I am encouraged by your comments, questions, likes and critiques. They keep me focussed and poised to do better. I will continue to contribute my little quota such that every student and researcher will independently analyse his/her data. My teaching approach is very practical. I adopt a do-as-I-do style. Many thanks to those who have supported me by telling others. Once again, CrunchEconometrix loves to teach, support my Channel with your subscription, likes, feedbacks and sharing my videos with your cohorts. Please do not keep me to yourself (lol) inform your
friends, students and academic networks about my Channel. Tell them CrunchEconometrix breaks down the econometric jargons and teaches with simplicity. Follow me on Facebook, Twitter and Reddit. Love you all, greatly!!!
J

[CrunchEconometrix]

I want to appreciate all my subscribers from across the globe (Africa, Asia, Europe, the Americas, and The Pacific). Thank you all for your support. I am encouraged by your comments, questions, likes and critiques. They keep me focussed and poised to do better. I will continue to contribute my little quota such that every student and researcher will independently analyse his/her data. My teaching approach is very practical. I adopt a do-as-I-do style. Many thanks to those who have supported me by telling others. Once again, CrunchEconometrix loves to teach, support my Channel with your subscription, likes, feedbacks and sharing my videos with your cohorts. Please do not keep me to yourself (lol) inform your friends, students and academic networks about my Channel. Tell them CrunchEconometrix breaks down the econometric jargons and teaches with simplicity. Follow me on Facebook, Twitter and Reddit. Love you all, greatly!!! 

[rachelcxuan]

Thank you so much for the detailed video presented

[johnotalor]

Dr you are simply amazing. Thanks for the video

[mohamedkkamara]

Your videos are extremely helping me to do my dissertation work

[rahulbhirani]

You are a star! Thank you for the amazing econometrics content.

[AlagieBSowe]

Keep up the good job Ma, its been helpful. Thanks a million

[drsunilbhardwaj]

Thanks madam for such meaningful videos. I am an Indian

[michaelbaah]

mommy thank you very much for your videos God bless you from Michael (ghana)

[Amal-galos]

I am Egyptian, I like your videos very much

[anwarqahtan]

Thank you so much, it's very helpful.
Please, we want another video that explains and describes step by step how to get rid and correct Heteroscadisity and serial correlation in different ways not only convert data to (log) but other methods in the panel data.
Thank you.

[estevaomcs]

Great job, CrunchEconometrix. I ask you if in eviews there Is a test like Durbin Watson test for testing the residuals? And how to execute it?

[tobibakare]

Thank you very much for your video's, I really appreciate your help! If we just want to model volatility using GARCH and we have no intention to forecast, does our GARCH model need to pass the no heteroscedasticity & auto-correlation condition, or are the 2 criteria only relevant for forecasting? I am asking as my mean models show ARCH effect, however I get serial correlation and heteroscedasticity for my GARCH equation. Thank you from the UK!

[vitorbuarque]

Hi! Can you tell me exactly which paper/textbook i can find the explanation to choose between models? You wrote: "From the literature, the preferred model(s) should have:" but i couldn't find it anywhere. help me please! greetings from brazil

[windijordan]

HI! could please tell me how to fix if in the Ljung-BOX result for serial correlation within GARCH models is written that: "Probabilities may not be valid for this equation specification"?? I cant rely on probabilities of Q-stat to recognize existance of serial correlation for GARCH models. Thank you

[kamariannisgr]

How do we react on autocorrelation existence on garch models? I mean that on the Correlogram -Q statistics there are 4 observations from the 4th to 8th that their Pvalue is lower than 5%. How can we explain that? Thank you..

[radhikajain]

I have a question regarding diagnostics.
In my data series, all three tests are showing positive results in all three kinds of the model but I am getting R square, adj R square, and the log-likelihood as negative. What should I do?

[ahah]

The question regarding lags. In case of daily date is it fixed that i try first 1 then 36 ? Or i can choose any number ? I got tired trying to get this information

[danielstalmeisters]

Hi Ngozi, I hope that you are well. Just (hopefully) a quick question, why does all empirical research use the ARCH LM and the Q-statistic test for OLS testing but not for testing GARCH-type models?

[ahah]

Good morning mam can you explain how we can select the optimal number of lags? In case i have daily data should i just use 36 ? It can be different than garch lag ? I mean garch (1, 1) does not mean the number of lag should be 1 ? Thanks for your effort