automobile sales and the general economy econ240a
DESCRIPTION
Automobile Sales and the General Economy ECON240A. Group #1 Deepti Goyal Rory Tyler Hofstatter Hairuo Hu Joel Benjamin Lindenberg Sooyeon Angela Shin Michael John Stromberg Kathy Zha Ling Zhu. Introduction. Dependent Variable Amount of Auto Sales Independent Variables Unemployment - PowerPoint PPT PresentationTRANSCRIPT
Automobile Sales and the Automobile Sales and the General EconomyGeneral Economy
ECON240AECON240AGroup #1Deepti Goyal
Rory Tyler HofstatterHairuo Hu
Joel Benjamin LindenbergSooyeon Angela Shin
Michael John StrombergKathy Zha
Ling Zhu
IntroductionIntroduction
Dependent Variable◦Amount of Auto Sales
Independent Variables◦Unemployment◦Price of Oil◦Average Mileage per Gallon◦Income per Capita
Why Study Such Variables?Why Study Such Variables?
Trend towards vehicles with better fuel efficiency
Automobile sales have been decreasing, particularly for bigger vehicles notably in the past couple of years
Impact of current recession on the auto
sales industry
Auto Sales by MakeAuto Sales by Make
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
20080.00
20.00
40.00
60.00
80.00
100.00
120.00
others
Hyundai
Honda
Nissan
Toyota
Volksw agen
GM
Ford
Daimler
Chrysler
American Motors
Trucks vs. CarsTrucks vs. Cars
year
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000 Trucks
Cars
How the Study is ConductedHow the Study is Conducted
Exploratory Data Analysis◦Histograms◦Box Plots◦Scatter Diagram◦Time Series Trend
Regression Analysis◦Correlation Diagram◦Bi-variate Regression using OLS method◦Normality test using Jarque-Bera Statistics◦Heteroskedasticity
Data GatheringData Gathering
Auto Sales: Ward’s Automotive GroupUnemployment Rate:
US Bureau of Labor StatisticsAnnual Crude Oil Prices:
US Bureau of Labor StatisticsIncome per Capita:
US Department of Commerce, Bureau of Economic Analysis
35 Years (1974-2008)
Variable HistogramsVariable Histograms
Histogram of AUTO
AUTO
Frequency
10000 12000 14000 16000 18000
02
46
8
Histogram of AVGMPG
AVGMPG
Frequency
18 20 22 24 26 28 30
02
46
810
12
14
Histogram of INCOME
INCOME
Frequency
10000 20000 30000 40000
01
23
45
6
Histogram of OIL
OIL
Frequency
0 20 40 60 80 100
05
10
15
Histogram of UNEMP
UNEMP
Frequency
4 5 6 7 8 9 10
02
46
810
12
Variable BoxplotsVariable Boxplots
20
22
24
26
28
30
Boxplot of AVGMPG
AVGMPG
5000
150002
50003
5000
Boxplot of INCOME
INCOME
20
40
60
80
Boxplot of OIL
OIL
45
67
89
Boxplot of UNEMP
UNEMP
Auto Sales vs. TimeAuto Sales vs. Time
Unemployment Rate vs. TimeUnemployment Rate vs. Time
Oil Price vs. TimeOil Price vs. Time
Income per Capita vs. TimeIncome per Capita vs. Time
Average Mileage vs. TimeAverage Mileage vs. Time
Correlation between VariablesCorrelation between Variables
AUTO
20 24 28 20 40 60 80
12000
16000
20
24
28
AVGMPG
INCOME
5000200003
5000
20
40
60
80
OIL
12000 16000 5000 20000 35000 4 5 6 7 8 9
45
67
89
UNEMP
Correlation between VariablesCorrelation between Variables
Correlation – Auto Sales and Other Correlation – Auto Sales and Other VariablesVariables
Negative Slope!
20 22 24 26 28 30
12000
14000
16000
18000
AVGMPG
AU
TO
5000 15000 25000 35000
12000
14000
16000
18000
INCOME
AU
TO
20 40 60 80
12000
14000
16000
18000
OIL
AU
TO
4 5 6 7 8 9
12000
14000
16000
18000
UNEMP
AU
TO
Regression EquationRegression Equation
Auto Sales = c1*Avgmpg + c2*Income+c3*oilprice + c4 * Unemployment + constant
Regression IRegression I
Highly Significant F-statistic
Barely Significant at 5% level
All other variables are significant at 5% level
Dependent Variable: TOTALAUTOS
Method: Least Squares
Date: 11/28/09 Time: 14:30
Sample(adjusted): 1 35
Included observations: 35 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
AVGMPG 226.2330 91.73632 2.466122 0.0196INCOME 0.084573 0.039173 2.158980 0.0390
OILPRICE -34.56953 16.32058 -2.118156 0.0426
UNEMP -714.5807 211.1398 -3.384395 0.0020
C 12504.28 2335.651 5.353661 0.0000
R-squared 0.741293 Mean dependent var 14794.74Adjusted R-squared 0.706799 S.D. dependent var 2107.357
S.E. of regression 1141.092 Akaike info criterion 17.04892
Sum squared resid 39062750 Schwarz criterion 17.27112
Log likelihood -293.3562 F-statistic 21.49035
Durbin-Watson stat 0.572137 Prob(F-statistic) 0.000000
Diagnostic of Regression IDiagnostic of Regression I
Residual vs. Fitted Values
Slightly skewed to the left
But still normally distributed
Heteroskedasticity?Heteroskedasticity?
-3000
-2000
-1000
0
1000
2000
3000
10000 12000 14000 16000 18000
FITTED
RE
SID
UA
LS
White Heteroskedasticity test:
F-statistics 1.409723 Probability 0.239025
Obs*R-squared 10.58868
Probability 0.226112
Regression IIRegression II
Highly Significant F-statistic
All variables are significant at 5% level with income as highly significant
Dependent Variable: TOTALAUTOS
Method: Least Squares
Date: 11/28/09 Time: 14:59
Sample(adjusted): 1 35
Included observations: 35 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
INCOME 0.134491 0.036183 3.717019 0.0008OILPRICE -46.02644 16.87912 -2.726828 0.0104
UNEMP -549.5911 216.0515 -2.543796 0.0162
C 16619.34 1763.169 9.425836 0.0000
R-squared 0.688847 Mean dependent var 14794.74Adjusted R-squared 0.658735 S.D. dependent var 2107.357
S.E. of regression 1231.073 Akaike info criterion 17.17637
Sum squared resid 46981757 Schwarz criterion 17.35412
Log likelihood -296.5865 F-statistic 22.87646
Durbin-Watson stat 0.541308 Prob(F-statistic) 0.000000
Diagnostic of Regression IIDiagnostic of Regression II
0
2
4
6
8
10
12
-2000 -1000 0 1000 2000
Series: ResidualsSample 1 35Observations 35
Mean 3.13E-12Median 55.60103Maximum 2333.201Minimum -2337.957Std. Dev. 1175.507Skewness 0.077039Kurtosis 2.302377
Jarque-Bera 0.744359Probability 0.689230
Correcting the autocorrelation Correcting the autocorrelation functionfunction
Dependent Variable: TOTALAUTOSMethod: Least SquaresSample(adjusted): 2 35Included observations: 34 after adjusting endpointsConvergence achieved after 22 iterations
Variable Coefficient Std. Error t-Statistic Prob. UNEMP -618.2731 165.5458 -3.734755 0.0009
OILPRICE -89.27321 27.99955 -3.188380 0.0035INCOME 0.094867 0.093254 1.017291 0.3177AVGMPG 288.8147 102.7569 2.810661 0.0089
C 10891.13 2666.290 4.084750 0.0003AR(1) 0.809449 0.110204 7.345013 0.0000
R-squared 0.887058 Mean dependent var 14833.03Adjusted R-squared 0.866889 S.D. dependent var 2126.656S.E. of regression 775.8966 Akaike info criterion 16.30470Sum squared resid 16856434 Schwarz criterion 16.57406Log likelihood -271.1799 F-statistic 43.98277Durbin-Watson stat 1.290312 Prob(F-statistic) 0.000000Inverted AR Roots .81
Error Term RegressionError Term Regression
Dependent Variable: ERRORMethod: Least SquaresDate: 12/02/09 Time: 11:22Sample(adjusted): 2 35Included observations: 34 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob. C -6.888265 133.5331 -0.051585 0.9592
ERROR(-1) 0.717434 0.127652 5.620227 0.0000R-squared 0.496752 Mean dependent var 21.87280Adjusted R-squared 0.481026 S.D. dependent var 1080.031S.E. of regression 778.0532 Akaike info criterion 16.20849Sum squared resid 19371737 Schwarz criterion 16.29828Log likelihood -273.5443 F-statistic 31.58696Durbin-Watson stat 0.981066 Prob(F-statistic) 0.000003
Durbin Watson CorrectionDurbin Watson Correction
Dependent Variable: YMethod: Least SquaresDate: 12/02/09 Time: 11:30Sample(adjusted): 2 35Included observations: 34 after adjusting endpointsVariable Coefficient Std. Error t-Statistic Prob.
G -615.9022 159.3205 -3.865807 0.0006F 0.110870 0.059957 1.849160 0.0747E -82.52244 28.51312 -2.894192 0.0071B 283.0815 102.1097 2.772328 0.0096C 3066.207 702.8768 4.362368 0.0001
R-squared 0.639690 Mean dependent var 4156.583Adjusted R-squared 0.589992 S.D. dependent var 1206.765S.E. of regression 772.7143 Akaike info criterion 16.27275Sum squared resid 17315534 Schwarz criterion 16.49721Log likelihood -271.6367 F-statistic 12.87155Durbin-Watson stat 1.133002 Prob(F-statistic) 0.000004
Y = totalautos-.717*totalautos(-1)A = (1-.717)*a
B = avgmpg-.717*avgmpg(-1) E = oilprice-.717*oilprice(-1)F = income-.717*income(-1)G = unemp-.717*unemp(-1)
Y = A + B + E + F + G
ConclusionConclusion
Significant Factors Affecting Automobile Sales: Unemployment Rate Income per Capita Fuel Economy (Avg. Mileage per Gallon) Avg. Price of Crude Oil
Forecasting◦Automobile Sales , when unemployment rate and
income per capita .Room for Future Studies:
◦For stronger R2 (0.74 for Reg. #1 and 0.69 for Reg. #2), additional variables should be studied
QuestionsQuestions??