Top Banner
Time Series Techniques and an Example from Roger Simon’s The City Building Process
17

Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Dec 18, 2015

Download

Documents

Derick Lamb
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Time Series

Techniques and an Example from Roger Simon’s The City Building

Process

Page 2: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

What is a Time Series?

• Numerical data ordered by intervals of time in chronological order

• The analysis of a time series involves decomposing the series into its components, e.g., – Trends– Cyclical and Seasonal

Fluctuations– Irregular Fluctuations

Page 3: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Example of a Time Series

1888 2

1889 29

1890 33

1891 40

1892 56

1893 43

1894 39

1895 21

1896 38

Etc… 31

Page 4: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

To Analyze a Time Series…

• Identify the possible processes affecting the series:– Inflation or Deflation

(for series of prices or expenditures)

– Long term trend– Cyclical fluctuations– Seasonal fluctuations– Irregular fluctuations

Page 5: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Steps to Analyze Changing Construction Costs

• Identify a measure:– Construction Cost? Or…. – Construction Cost Per Square Foot

• Convert to Constant Dollars

• Measure the Trend

• Identify cyclical or seasonal factors

• Identify irregular fluctuations

Page 6: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Raw Construction Cost over Time

Page 7: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Properties Built over Time

Page 8: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Size of Buildings Built over Time

Page 9: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Adjusting for Inflation• if yrbuilt >887 and yrbuilt < 912 then let concost2=concost*20• if yrbuilt >911 and yrbuilt < 916 then let concost2=concost/.06• if yrbuilt = 916 then let concost2=concost/.065• if yrbuilt = 917 then let concost2=concost/.077• if yrbuilt = 918 then let concost2=concost/.090• if yrbuilt = 919 then let concost2=concost/.104• if yrbuilt = 920 then let concost2=concost/.120• if yrbuilt = 921 then let concost2=concost/.107• if yrbuilt = 922 then let concost2=concost/.103• if yrbuilt = 923 then let concost2=concost/.102• if yrbuilt = 924 then let concost2=concost/.102• if yrbuilt = 925 then let concost2=concost/.105• if yrbuilt = 926 then let concost2=concost/.106• if yrbuilt = 927 then let concost2=concost/.104• if yrbuilt = 928 then let concost2=concost/.102• if yrbuilt = 929 then let concost2=concost/.102• if yrbuilt=. then let concost2= concost *20• if concost=. Then let concost2 =.

Page 10: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Construction Cost Adjusted for Inflation

Page 11: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Comparing Raw and Inflation Adjusted Construction Costs

Page 12: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Construction Cost per Square Foot (Inflation Adjusted)

Page 13: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Construction Cost per Square Foot (Inflation Adjusted)

Page 14: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

The Model: Cost per square foot = constant(a) + b1*year +

b2*(last year’s cost per square foot)

Dep Var: COSTPER N: 39 Multiple R: 0.885 Squared multiple R: 0.784 Adjusted squared multiple R: 0.772 Standard error of estimate: 4.143 Effect Coefficient Std Error Std Coef Tolerance t P(2 Tail)

CONSTANT 7.330 3.100 0.000 . 2.364 0.024LAGCOSTP 0.589 0.148 0.540 0.325 3.974 0.000YEAR2 0.293 0.103 0.387 0.325 2.851 0.007

Analysis of VarianceSource Sum-of-Squares df Mean-Square F-ratio P Regression 2244.126 2 1122.063 65.375 0.000Residual 617.887 36 17.164

Dep Var: COSTPER N: 39 Multiple R: 0.885 Squared multiple R: 0.784 Adjusted squared multiple R: 0.772 Standard error of estimate: 4.143 Effect Coefficient Std Error Std Coef Tolerance t P(2 Tail)

CONSTANT 7.330 3.100 0.000 . 2.364 0.024LAGCOSTP 0.589 0.148 0.540 0.325 3.974 0.000YEAR2 0.293 0.103 0.387 0.325 2.851 0.007

Analysis of VarianceSource Sum-of-Squares df Mean-Square F-ratio P Regression 2244.126 2 1122.063 65.375 0.000Residual 617.887 36 17.164

Page 15: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

The Model: Cost per square foot = constant(a) + b1*year +

b2*last year’s cost per square foot

• Cost per square foot = $7.33 + .59 (last year’s cost) + .29 (year)

• Number of cases: 39 (1889-1928). 1889=1

• Adjusted R Square: 77.2

Page 16: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

Trend and Residuals

Page 17: Time Series Techniques and an Example from Roger Simon’s The City Building Process.

YearCost Per

Square FootLast Year's Cost

Per Sq. Ft.PredictedCost PerSq. Ft.

Residual

889 $21.91 $22.67 $20.99 0.92890 $20.42 $21.91 $20.83 -0.41891 $17.27 $20.42 $20.25 -2.98892 $17.40 $17.27 $18.68 -1.29893 $19.21 $17.40 $19.05 0.16894 $19.76 $19.21 $20.41 -0.65895 $20.85 $19.76 $21.03 -0.19896 $19.36 $20.84 $21.96 -2.60897 $20.73 $19.36 $21.38 -0.65898 $20.54 $20.73 $22.48 -1.94899 $25.83 $20.54 $22.66 3.17900 $28.09 $25.83 $26.08 2.01901 $28.80 $28.09 $27.70 1.10902 $28.63 $28.80 $28.41 0.22903 $33.00 $28.63 $28.61 4.40904 $31.60 $33.00 $31.48 0.12905 $34.47 $31.60 $30.94 3.53906 $32.93 $34.47 $32.93 0.00907 $35.72 $32.93 $32.32 3.40908 $37.37 $35.72 $34.25 3.12909 $37.97 $37.37 $35.52 2.45910 $34.66 $37.97 $36.17 -1.51911 $35.13 $34.66 $34.51 0.63912 $30.27 $35.13 $35.08 -4.81913 $37.89 $30.27 $32.51 5.38914 $28.78 $37.89 $37.29 -8.51915 $38.23 $28.78 $32.22 6.01916 $37.99 $38.23 $38.08 -0.09917 $40.33 $37.99 $38.23 2.10918 $23.85 $40.33 $39.91 -16.06919 $30.15 $23.85 $30.49 -0.34920 $33.45 $30.15 $34.49 -1.04921 $34.19 $33.45 $36.73 -2.54922 $34.37 $34.19 $37.46 -3.09923 $36.37 $34.37 $37.86 -1.49924 $36.81 $36.37 $39.33 -2.53925 $42.96 $36.81 $39.88 3.08926 $49.20 $42.96 $43.80 5.40928 $53.59 $49.20 $48.07 5.52

. $53.59 . .

The Data:1.Year 2. Cost Per Square Foot3. Last Year’s Cost Per Square Foot4. Predicted Cost5. Residual