Question 1.2 (49)

In this question you need to analyse the trend and drivers of food price inflation in South Africa. According to theory, the drivers of food inflation can include domestic supply and demand dynamics, production prices as well as changes in exchange rates and commodity prices.

You are given the following data:

Where: CPI_FOOD = CPI: South Africa, All urban areas - Food (Index)

CPI_ALL = CPI: South Africa, All urban areas - Headline: All Items (Index)

PPI = PPI for final manufactured goods: Food products (Index)

RETAIL = Retail trade sales: Retailers of food, beverages and tobacco in specialised stores. Current prices (R million)

ZAR_DOL = Exchange rates: SA cent per: USA - Dollar (monthly averages)

ZAR_EUR = Exchange rates: SA cent per: EU - Euro (monthly averages)

OIL = Brent crude oil - US Dollar (USD/barrel)

The data is available in sheet DATASET _ Q1 of the MS EXCEL file ECS4863 OctNov 2022 data.xls.

DATASETS: For questions 1.2 and 4, you must use the correct dataset, depending on the first letter of your surname.

DATASET 1: Students whose surname start with A-H;

DATASET 2: Students whose surname start with I-Q;

DATASET 3: Students whose surname start with R-Z.

Please make sure to download the correct datafile (Excel), depending on your surname. The Excel files are clearly marked DATASET 1, DATASET 2 and DATASET 3. A mark of zero will be allocated for students not adhering to thus rule.

Please complete the following before answering this question: My Surname is: DATASET used: (indicate number)

(a) Write a paragraph in which you discuss the developments in food inflation versus headline inflation during the period. You can use your own graphs, tables, calculations,etc. to support your arguments.

(b) Explain the theoretic justification of the potential inclusion of the retail trade sales variable when modelling food price inflation. Will you make any adjustments to the variable before including it in the model? Explain.

(c) Test the variables CPI_FOOD, PPI and OIL for stationarity (transform all applicable variables). Also comment on their respective orders of integration.

(d) In this question you need to build a long run equation and test it for possible cointegration. (i) Estimate the following long-run equation: (2) LCPI_FOODt = + LPPIt + LOILt+ Copy/paste your EViews Estimation output window (include all information).

(ii) Generate the residual series: [use the first three letters of your surname as suffix, e.g. res_JOU]. Demonstrate, how you can use the residual series to test if cointegration exists between the variables in the long run model. Provide all details of the test(s) you perform and their criteria. Copy/paste your EViews Estimation output window (include all information).

(e) In this question you need to build a short run (Error Correction Model) (i) Estimate the ECM using the equation below. D(LCPI_FOOD)t =+ D(LZAR_DOL)t +D(LRETAIL)t + D(LPPI)t + D(LZAR_EUR)t+ (RESIDUAL)t1

(e) In this question you need to build a short run (Error Correction Model) (i) Estimate the ECM using the equation below. (3) (ii) Evaluate the ECM statistically.

(iii) Comment on the inclusion of both exchange rates in the ECM. Does this make sense according to economic theory? Which other variable(s) can be included to capture the effect of exchange rate volatility?

(f) Perform diagnostic checks on the ECM. Test only for the potential problems as indicated below. In each instance, provide the name and a brief description of the test you perform. Also, clearly indicate how you have drawn your conclusions based on the results. (4) (i) Normality (ii) Heteroscedasticity

(g) Create a model statement in EViews to combine your long-run and ECM. Provide your model statement (3 marks), and a copy of the graph depicting the actual and modelled values (2 marks), in the space provided.

(h) Critically evaluate the overall model and its performance. (4) (i) Choose any one of the explanatory variables and apply a permanent shock to it. Use the first three letters of your name to label your shocked variable, e.g. retail_FAN. Post a graph that include both the original and shocked variable. (You may choose the starting date, size and direction of the shock). Comment on how you expect the shocked variable to influence the results of your model (you do not have to re-run the model).