Presentation is loading. Please wait.

Presentation is loading. Please wait.

Will There Be Jobs For All of Us Financial Econometricians? Ben Kallo Ben Kallo James Katavolos James Katavolos Luke Panzar Luke Panzar Ryan Carl Ryan.

Similar presentations


Presentation on theme: "Will There Be Jobs For All of Us Financial Econometricians? Ben Kallo Ben Kallo James Katavolos James Katavolos Luke Panzar Luke Panzar Ryan Carl Ryan."— Presentation transcript:

1 Will There Be Jobs For All of Us Financial Econometricians? Ben Kallo Ben Kallo James Katavolos James Katavolos Luke Panzar Luke Panzar Ryan Carl Ryan Carl

2 Let’s look at Financial Jobs in the U.S. Economy Data gathered from the St.Louis Federal Reserve Data gathered from the St.Louis Federal Reserve Also includes jobs in the real estate and insurance fields Also includes jobs in the real estate and insurance fields Monthly Data from 1946 to 2003 Monthly Data from 1946 to 2003

3 Objective of Model Using a time-series model, we should be able to forecast the trend of jobs in the finance industry for at least the next six months. Using a time-series model, we should be able to forecast the trend of jobs in the finance industry for at least the next six months.

4 Total Jobs in the Finance Industry

5 Correlogram of FINJOBS Points to an evolutionary series

6 Histogram of FINJOBS

7 Dickey-Fuller Test for FINJOBS ADF stat confirms the presence of an evolutionary times series

8 Mission: Stationarity In order to make the series stationary (not dependent on time), we will perform a logarithmic transformation as well as difference the series. In order to make the series stationary (not dependent on time), we will perform a logarithmic transformation as well as difference the series. The new data set will be called: The new data set will be called:DLNFINJOBS

9 DLNFINJOBS The economic interpretation of DLNFINJOBS is the month over month percentage change in the number of jobs in the finance industry in the United States. The economic interpretation of DLNFINJOBS is the month over month percentage change in the number of jobs in the finance industry in the United States.

10 Line Graph for DLNFINJOBS Looks Stationary

11 Correlogram of DLNFINJOBS The structure in the ACF and the spike at lag 1 of the PACF suggest that we should use an ARONE model

12 Histogram of DLNFINJOBS

13 Dickey-Fuller Test of DLNFINJOBS Stationarity Confirmed

14 Model Output for DLNFINJOBS

15 Correlogram for Model

16 Next Model Output

17 Correlogram This new model only seems to create more structure, with spikes at lags 4,5, 10,11,13

18 Correlogram of Squared Residuals To see whether a ARCH-GARCH model is appropriate After running an ARCH-GARCH, the model did not improve

19 Now try: AR1,MA1,MA13 Again, this new model seems to add structure Again, this new model seems to add structure Final Conclusion: Final Conclusion: Our best model is our original one: AR1,MA1

20 Actual, Fitted, Resid for Model

21 Total Jobs in the Finance Industry 2003:04 0.001130 2003:04 0.001130 2003:05 0.000932 2003:05 0.000932 2003:06 0.000896 2003:06 0.000896 2003:07 0.000817 2003:07 0.000817 2003:08 0.000783 2003:08 0.000783 2003:09 0.000751 2003:09 0.000751 2003:10 0.000720 2003:10 0.000720 2003:11 0.000690 2003:11 0.000690 2003:12 0.000662 2003:12 0.000662 2004:01 0.000634 2004:01 0.000634

22 Line Graph w/forecast to 2004

23 Conclusion Over the next ten months, the number of jobs in the finance industry should be increasing (according the model’s forecast), however at a decreasing rate. Interpretation: Can you say, “Do you want fries with that?”


Download ppt "Will There Be Jobs For All of Us Financial Econometricians? Ben Kallo Ben Kallo James Katavolos James Katavolos Luke Panzar Luke Panzar Ryan Carl Ryan."

Similar presentations


Ads by Google