Executive Summary Summarise the key results of your report in a few sentences or dot points. Contents 1.Introduction. 3 2.Data. 3 2.1.Data Description. 3 2.2.Data Summary. 3 3.Methodology. 4 3.1.Model4 3.2.Testing Hypotheses. 4 3.3.Model Limitations. 5 4.Results. 5 5.Conclusion. 6 6.Appendix. 6 1.Introduction Use the research question you have chosen, to write an introduction to the report. Your research topic should be broadly related to the contents of the unit. You can modify a question that you found interesting in video lectures or research papers or explore a topic that was not discussed in detail during the lectures but was briefly mentioned. In the introduction, you should: I1.Clearly state the aim or objective of your report in 1 – 3 sentences.This is also known as the research question. I2.Explain why the research question is of interest and why the answer to the research question is useful, in 1-3 sentences. I3.Provide the details for who/what/when your analysis of this question will cover in a single sentence. I4.Provide the source, in brief of the data you will use in a single sentence. I5.Link your results to the literature including the papers that we discussed during the workshops. 2.Data 2.1.Data Description Provide the original source of your data (see the list below).

State the data’s population (what does it cover).For time series data the time-period and frequency of the data. For cross sectional data provide the time-period over which the data set was collected and the unit of analysis or what an observation in the data represents (e.g. household, state, business)

Provide the number of observations.

Provide a table and/or explain your variables:name

description (levels, change or growth, over what period?)

units of measurement (e.g. $, $m, %, )

original, seasonally adjusted, trend/smoothed

nominal or real.

For all the variables in your models for the whole sample, provide summary statistics: min, max, median, mean, and sample standard deviation.Convert categorical variables into dummy variables prior to calculating the sample statistics.

Provide scatter plots of your dependent variable against two of your continuous independent variables.

Provide the correlation matrix of all your variables that shows the simple pair-wise correlation between the dependent variable and each of the independent variables from all models.

For time-series data, provide time-series charts of your y and x variables over the sample range.For cross-sectional data provide charts or tables that show the value of your y and x variables by groups based upon one for your qualitative x variables (or values of a quantitative variable if you have no qualitative variables)

Interpret and comment on D5, D6, D7 and D8 highlighting any signs that there is a relationship between your Y and X variables and the possible nature of that relationship (linear or otherwise).

Calculate the mean and the standard error of the mean of all your variables for model A for:

The following data sources can be used: www.abs.gov.au, www.rba.gov.au, au.finance.yahoo.com, www.quandl.com , www.imf.org, data.worldbank.org, www.economagic.com/aus.htm. If your research project is mainly reliant on historical data, check there are at least 50 observations with significant variation in data. 2.2.Data Summary You can use any software that you familiar with (e.g. Eviews, STATA, R), but this is not compulsory. Excel is also a suitable option. Q1 Sub-periods if it is relevant (for example, the sample statistics a. before and b. during the COVID period). Q2 a. big institutions and b. small institutions. D11.Provide D10 in a table.Identify the independent variables that are most different between a. and b. that may help explain the difference between a. and b. for the dependent variable. Discuss the results in reference to your question. 3.Methodology 3.1.Model M1.Write out the models you will estimate. Be sure to define all terms in the equation (or refer to where they are defined). For parameters use lower case Greek letters such as β with subscripts. Do not use the same lower-case Greek letters and subscript twice. For variables, use the names provided in the question in italics. To estimate your model, you can use any software you like (e.g. Eviews, STATA, R). Excel is powerful enough to handle basic models. A linear regression tool in Excel is available in MyLo. This tool will help you to estimate and analyse the standard linear regression models. M2If you rely on a Monte Carlo simulation in your study, clearly explain the setup of your simulation/experiment. Which distribution you use to generate random numbers (e.g. Normal distribution)? Carefully justify the inputs for the experiment (how you obtained the correlation coefficient, covariances etc.)? 3.2.Testing Hypotheses M2.Formulate Hypotheses that you want to test. Check and explain why the model is appropriate for your data type. Link your hypotheses to the main research question formulated earlier Do not forget about some statistical issues that might be present in your model (the following list is for regression models): Q1 For time-series data: Check for Multicollinearity, Autocorrelation, Stationarity and the normality of residuals. Q2 For cross-sectional data: Check for Multicollinearity. You should clearly state what tests/evidence you are using for each possible violation: its null hypothesis, and whether it is retained or rejected. You can place other details of the test in the appendix but make sure you refer to where and what it is. 3.3.Model Limitations M3.If you believe that your model has an important drawback you need clearly state this issue and clearly explain the consequence of that. For example, you might have a model which helps you to measure and explain consequences of credit risks, but it ignores foreign exchange risks. Alternatively, your model might be suitable for big financial institutions that normally do not experience problems with liquidity, but it might not be applicable to small illiquid FIs. 4.Results R1.Table of Results. Provide a concise table or tables of the results, which shows for the model: estimated coefficients b

either: se(b) or their t-values or their p-values

indicate which is significant with *s.* indicates significant at 10%, ** significant at 5% and *** significant at 1%.

R2, adjusted R2, standard error of regression.

Try to ensure your table is easy and pleasant to read. You may want to change the number of decimal places or rescale your variables (easiest in Excel). Try to make it fit on one page (you can shrink the font) or make sure you repeat the headings on the next page.

You may wish to include the F Stat and its p-value, AIC and SIC.

Compare and interpret the SE of the regression of the model.

Interpret the R2 of the model. Is it satisfactory?

R2.Interpret Explanatory and Predictive Power R3.Interpret the coefficients of the following variables from Model A and state whether they are statistically significant (and if so at what level). Explain why you think the variable has a positive or negative effect. What do these results mean for your research question. Q1: Q2: and …(together) Point out any other interesting features of the results. R4.If your original model can be improved revisit this model and work with a new model B. From Model B, interpret the coefficients of the variables listed above and the additional variables or interaction terms in B.State whether the coefficients are statistically significant (and if so at what level) and interpret the result with regards to your question.Carefully explain why the effects of the variables from R3 varies from model A and model B. R5.Forecast (You can obtain predictions from your model) Q1 – Use the sample means of the independent variables from 1990q1 to 2013q4 and 2014q1 to 2017q4 and your estimated coefficients from model A to provide a forecast of average Variable 1 growth for the two periods. Q2 – Use the sample means of the independent variables for big and small banks and your estimated coefficients from model A to provide a forecast of loses for big and small banks. R6.Use your calculations from R5 to determine which independent variables are most responsible for the difference between… Q1 – losses for big and small banks Q2 – capital growth from 1990q1 to 2013q4 and 2014q1 to 2017q4 …and discuss the result. R7.If you rely on a Monte Carlo simulation, it is a good idea to plot a frequency distribution. For example, simulated distributions of bond prices or simulated distributions for exchange rates. 5.Conclusion C1.State the main finding of your report in regards to your question. C2.What are the implications of your findings? C3.Provide any suggestions for improving similar studies in the future, for example additional variables that could be included. C4.Can your findings be useful for regulators? If yes, explain how. 6.Appendix Attach an appendix containing labelled tables and charts from EViews or MS Excel work refer to them at the appropriate times in your answers to the sections above.

© 2018 |** Intelli Essays Homework Service®**