Gives an overview of each project in Econometrics Modelling.
Through this project, I attempted to apply what I had learnt about time series models and forecasting during my undergraduate course using R.
The project uses ARIMA, ARDL and VAR models to forecast Singapore's Consumer Price Index using data from previous periods and with other variables, namely the Domestic Supply Price Index and the Composite Leading Index of Singapore.
I wanted to do this project to find out if R was suitable for time series analysis, as compared to Stata or EViews, which I had used during my undergraduate coursework. It was not as simple and straightforward as these softwares, and the disadvantages have been highlighted in my discussion.
The aim of this project was to study the combined effects of the 10Y/2Y Treasury Yield Spread, the CBOE Volatility Index, and the U.S. Initial Jobless Claims on the S&P 500, measured by the SPDR S&P 500 ETF closing price.
ARIMA and ARDL models were used to estimate SPY returns and forecast 4 periods ahead for June 2022.
In the project Forecasting Singapore CPI Using Different Time Series Models, I had ignored the effects of seasonality on the Singapore Consumer Price Index when estimating the models.
In this project, I attempted to find out if seasonality does affect Singapore CPI using ARIMA, SARIMA, ARIMAX and SARIMAX models. I used either seasonal differencing in the case of SARIMA and SARIMAX or added seasonal dummies as external regressors.
This project sought to understand non-stationary processes in time series analysis and the difference between deterministic and stochastic trend. It also assessed if differencing and de-trending to transform non-stationary variables into stationary processes affects the accuracy of out-of-sample forecast for time series with a deterministic trend.
The project used a simulation of random walk and ARIMA models and added a drift and/or trend term to simulate non-stationary time series data.