How much will a one standard deviation increase

Question

Econometrics Assignment

Length : 4000 words +/- 10%.

Question 1

The file "REER.xlsx" contains panel data on the real effective exchange rate of Singapore, which includes the following variables:

REER is the real effective exchange rate index of Singapore
SIBOR is the 3-month Singapore Interbank Operating Rate
CA is the current account (in billion SGD) of Singapore
CLI is the OECD Composite Leading Index
CPI is the Consumer Price Index of Singapore
STI is the Singapore's stock market Straits Time Index
FED is the US Fed Fund Rate

Suppose you are an economist with the Monetary Authority of Singapore (MAS). To investigate the determinants of Singapore's real effective exchange rate (REER), you build a linear regression model using monthly data from the 2010-2017 period, as follow:

REER_t = β₀ + β₁SIBOR_t+β₂CA_t + β₃CLI_t + β₄CPI_t + β₅STI_t + β₆FED_t + μ_t

1. Use Stata to derive the OLS estimates of β₀, β₁, β₂, β₃, β₄, β₅ and β₆.

2. Comment on significance of the coefficient estimates of your fitted model. Provide economic intuition on the relationships you uncover.

3. How much will a one standard deviation increase in the Singapore's Straits Time Index (STI) affect the country's real effective exchange rate?

4. What is the real effective exchange rate predicted if all the explanatory variables are set at their sample mean values?

5. For each of the following, formulate a null hypothesis and test it using the regression equation.

i. A $1 billion increase in the current account will cause the real effective exchange rate index to increase by 3 points.

ii. The SIBOR and FED will have the same effect on the real effective exchange rate.

iii. The cumulated effect of a point increase in both CPI and CLI will increase the real effective exchange rate by 1.5 point.

Question 2

The file "WAGE.xlsx" contains the cross-sectional data on the national labour statistics of South Korea. It includes the following variables:

WAGE is the monthly wage of employee
EDU measures the years of education of employee
EXP measures the years of working experience of employee
AGE is the age of employee
GEN is a dummy variable which is equal to 1 if the employee is male, 0 otherwise
UNI is a dummy variable which is equal to 1 if the employee is a graduate, 0 otherwise
NAT is a dummy variable which is equal to 1 if the employee is a Korean, 0 otherwise
M is a dummy variable which is equal to 1 if the employee is married, 0 otherwise

To investigate whether there were signs of wage discrimination in the South Korea labour market, you run the following linear regression:

WAGE_i= β₀ + β₁EUU_i+β₂EXP_i + β₃AGE_i + β₄GEN_i + β₅UNI_i + β₆NAT_i + β₇M_i + μ_t

1. Use State to derive the OLS estimates for of β₀, β₁, β₂, β₃, β₄, β₅, β₆ and β₇. (8 marks)

2. Using your fitted model, what would be the increment in monthly wage if the worker obtained a University degree?

3. Are there signs of wage discrimination being practiced in the Singapore labour market? State the null hypotheses for each case and explain your results.

4. What is the expected wage for a 32-year-old Korean female employee who is married and has invested 14 years in education to earn a University degree, but has no prior working experience.

5. Rewrite the equation and test the hypothesis that a non-Korean employee will see a bigger wage increment from each additional year of working experience.

Question 3

The file "GDP.xlsx" contains panel data on the economic condition of Singapore, which includes the following variables:

GDP is the natural logarithm of Singapore real GDP
M2 is the money supply of Singapore (in SGD billion)
FDI is the Singapore's share of global FDI
REER is the Singapore's real effective exchange rate
GOV is the Singapore government budget balance
CLI is the OCED Composite Lead Index

1. Use Stata to build a linear regression model to investigate the determinants of Singapore's real GDP numbers. Comment on the results and provide intuition on your findings.

2. Test for heteroscedasticity in your regression. Suggest how you can correct for the problem of heteroscedasticity in linear regressions.

Attachment:- DATA-Assignment.rar