Linear Regression Model Research Paper
Develop a simple linear regression model using one independent variable to explain the short-term, “risk-free” rate or yield (i.e., 90-day U.S. Treasury bill) dependent variable. Use monthly data over a recent U.S. business cycle (e.g., 2002 to 2019). Provide an overview of the paper. State and justify the null and directional (when warranted) alternative hypothesis for the model. Consider selecting one independent variable from among monetary, fiscal, economic, financial, political, demographic and/or another factor you believe relevant. Do not use another interest rate as an independent variable for the first regression model. If your variable increases with time, such as the Consumer Price Index or the Money Supply, change the metric to an annual percent change from the prior year. Is the variable tested significant and consistent with the alternative hypothesis or merely “noise” in the market?
Compare (e.g., T-statistic, R-square, SER) the first simple regression model with a second model evaluating the same period of time (except for three-month lag) and same dependent variable using the implied three-month forward rate of interest computed by pure expectations (lagged three months) as the independent variable. Evaluate the term structure hypothesis best supported by the results of the second model to include the intercept and slope. Do your results support liquidity preference or pure expectations? Why? Ensure that you provide and assess simple or descriptive data for all dependent and independent variables used in the two regressions to include the three-month T-bill and independent variable selected for the first model and the three-month, six-month and forward T-bill rates derived for the second model. Present your findings in a report of approximately five pages (excluding tables, statistical output and/or graphs). Briefly provide a context of the economic and financial environment for the variables tested.
Linear Regression Model Research Paper |
|