Garch(1,1) is extensively applicable. Use it as an initial model.

exponential GARCH model is considered as a superior model in analyzing the volatlity.

there is a large repository of literature on volatility transmission between two different segments of the market.

one popular method is to derive the residuals from ARIMA and surrogate the residuals as the proxy for spillover of volatility.

in your case, apply ARIMA on exchange rate and derive the residuals and treate the residuals as volatiity spillover variable from forex market to oil market and you can apply the same procedure for oil market.

further discussion in this regard is most welcome

GJR-GARCH model

https://www.dohainstitute.edu.qa/MEEA2016/Downloads/Thouraya%20Amor_Final.pdf

GARCH and APARCH models

file:///C:/Users/CC/Downloads/imfi_en_2016_04cont_Kutu.pdf

VAR-GARCH model

Dear Ette,

Thank you for your contribution and as you mentioned GARCH (1,1) is applicable, however; my question about the conditional covariance model that could be used.

Sorry maybe my question wasn't clear enough but I appreciate your contribution and helping.

Dear Srikanth,

Thank you for your detailed and helpful answer.

I am planning to use the EGARCH as more Asymmetry to variance model. Also, I am thinking if I can use the BEKK-GARCH and DCC- GARCH especially the DCC which can be used in the correlation analysis.

would you have any suggestion according to your experience?

Regards,

Dear Ashwini,

Thank you for the models and files links. the first link very interested. And would please double check the link for imfi_en_2016_04cont_Kutu.pdf because it seems to be a local PC link?

UPDATE: for reader who would be benefited from Ashwini answer, I have googled the file and found it on this link

https://businessperspectives.org/images/pdf/free/8098/imfi_en_2016_04cont_Kutu.pdf

Regards,

Dear Mohanned,

This topic is extensively studied in the literature using diversity of models ranging from simple VAR to multivariate SV models within the framework of both Bayesian and frequentist approaches. It is extremely difficult to publish this topic in high-ranked journals unless you develop completely different approach and use unique data series. You may consider volatility impulse response functions which is relatively less common in this context.

Regards

Dear Akram,

Thank you for your contribution, in this stage I will use the GARCH for academic thesis, and I have been advised to work in GARCH to be easy to find an examiner

I will take your point of view for the future papers

Thank you again

Dear Tony,

I can’t thank you enough for your contribution!

Eve though I have read about DCC and cDCC, you added more knowledge to what I came up with. I’ll read the attached article.

Kind regards,

Mohanned

Dear you may also try the MSGARCH model the Markov switching generalized conditional heteroskedasticity model . You may try ox metric 7 software the application is just straight.

Dear David,

thank you for your contribution!

I haven’t read about this model so I appreciate if you know any helpf source that could help me to read about.

Would you please provide some articles names if it possible?

kind regard,

Applied Time Series Modelling and Forecasting. by Richard Harris and Robert Sollis To familiarize your self with the ox metric output and interpretation for basic garch models .

for the msgarh i will get back to you

Thank you David for the model and the supported file

Hi,

Modeling Oil Shocks and Exchange Rates is not a simple task. Indeed, one can't choose the model without having a look on the data series that you dispose. However, here are some tips that might help you to identify the adequate model:

1 - It is clear that these time series present changes in volatility, so, you do better go for volatility models.

2 - Please check if there is a periodicity in your data. If yes, try the Periodic GARCH

3 - If the histogram shows an asymetry, refer to the Asymetric GARCH and Exponential GARCH for the leverage effect

4 - In some cases, the histogram shows a multimodality, so the Mixture GARCH models should be more appropriate

5 - To allow the randomness in the volatility, the Stochastic Volatility models are welcome.

6 - if there is some sudden changes in the series behavior, you should investigate in the Markov Switching GARCH models.

7 - The long memory feature(observed in the correlogram of the yield-squares as a persistence of the autocorrelations even for very higher lags) can be captured by Long memory models.

Finally, here are some references (that I already read them) in which, the authors treated oil prices and exchange rates:

BOUSSAHA, Nadia, HAMDI, Faycal, SOUAM, Saïd, et al.Multivariate Periodic Stochastic Volatility Models: Applications to Algerian dinar exchange rates and oil prices modeling. University of Paris Nanterre, EconomiX, 2018.

Article Volatility Modelling in Crude Oil and Natural Gas Prices

Chapter GARCH Models in Forecasting the Volatility of the World’s Oil Prices

Article Use of the GARCH Models to Energy Markets: Oil Price Volatility

Article Mixture Periodic GARCH Models: Theory and Applications

Best regards,

Dear Ouzzani,

Thank you for your remarkable contribution!

I have already read one of them and will go through the others.

Regards,