Volatility in Commodity Market


Writer requirement:
Must be familiar with Finance and econometric and Eviews, especially ARCH model and its family models which are GARCH(1,1), EGARCH,TARCH, GARCH M and PGARCH . Then descriptive statistic (correlogram, return), unit root test, augmented dickey fuller test, AIC test, Bai Perron, Dummy variable, forecasting. The writer must be familiar with Eviews and writing in LaTex

The following are the guideline I got from my supervisor and teaching assistant. You will be able to see what I have written so far in the Zip file that i downloaded from LaTex. Please follow the guideline but please feel free to make any improvement. I have a chance that I got something wrong as well, in this case please correct me.

Topic: Volatility in Commodity Market

Commodity include Gold, Silver, Copper, Cotton, Orange juice, Corn , sugar, Cocoa (please pick 6) .All the data will be attached. Data sets are daily price from 2007 to 2017.


Table of content:

Literature review: Please look at at least 15 papers, must read papers are 1) Engle. R.F Autoregressive conditional heteroscedasticity with estimates of variance of United Kingdom inflation. 2)Nelsons, D. B. 1991, Conditional heteroskedasticity in asset returns: A new approach 3)Bollerslev. T 1986 Generalised autoregressive conditional heteroskedasticity. Journal of econometrics.

No equation needed.


3.1 Data description
3.2 Preliminary Tests:
3.2.1 Descriptive statistic such as correlogram, return chart;
3.2.2. Unit Root test, please use Augmented Dickey Fuller test and AIC test together in Eviews to test for unit root and find optimal lag length
3.2.3 Global L break via Bai Perron Test: please use this test to find TWO break days
3.2.4 Dummy variable: Please use dummy variable for the break

4: Methodology

Please refer to stationarity, unit root test, also refer to paper from ADF (1979), Bai Perron (1998, 2003), ARCH and GARCH models. Please add extra model detail as well from textbook and relevant sources

5: Results:

5.1 ARCH , please use Lags and no dummy variable

5.2: ARCH family model testing: please use dummy variable. (not sure if i need to use lag but i dont think so) that include: GARCH(1,1), TARCH, EGARCH, (pick one between GARCH M and PGARCH) and compare which one is the best.

Chapter 6: conclusion 1 page

Appendix: please put relevant eviews graphs, and do forecasting in eviews in the section straight after you do GARCH model tests. Please forecast for the future 20 days and use both Dynamic and static. Please also use GARCH Graph (i havent done this bit yet)

Other information:

Please use 1.5 spacing
• Number the pages
• Use 2.5 or 3cm on all margins
• The word count is 7,000 words (excluding table of contents, tables, graphs, footnotes, references, appendices etc). A discretionary +/-10% is acceptable and won’t be penalised.