heteroscedasticity 1 this sequence relates to assumption a.4 of the regression model assumptions and...

22
HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates to the distribution of the disturbance term in a regression model. 1 X Y = 1 + 2 X Y X 3 X 5 X 4 X 1 X 2

Upload: dominick-price

Post on 18-Jan-2018

237 views

Category:

Documents


0 download

DESCRIPTION

3 If there were no disturbance term in the model, the observations would lie on the line as shown. HETEROSCEDASTICITY 11 X Y =  1 +  2 X Y X3X3 X5X5 X4X4 X1X1 X2X2

TRANSCRIPT

Page 1: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

HETEROSCEDASTICITY

1

This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates to the distribution of the disturbance term in a regression model.

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 2: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

1

X

Y = 1 +2X

Y

2

We will discuss it in the context of the regression model Y = 1 + 2X + u. To keep the diagram uncluttered, we will suppose that we have a sample of only five observations, the X values of which are as shown.

X3 X5X4X1 X2

HETEROSCEDASTICITY

Page 3: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

3

If there were no disturbance term in the model, the observations would lie on the line as shown.

HETEROSCEDASTICITY

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 4: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

1

X

Y = 1 +2X

Y

4

Now we take account of the effect of the disturbance term. It will displace each observation in the vertical dimension, since it modifies the value of Y without affecting X.

X3 X5X4X1 X2

HETEROSCEDASTICITY

Page 5: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

5

The disturbance term in each observation is hypothesized to be drawn randomly from a given distribution. In the diagram, three assumptions are being made.

HETEROSCEDASTICITY

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 6: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

6

One is that the expected value of u in each observation is 0 (Assumption A.3). The second is that the distribution in each observation is normal (Assumption A.6). We are not concerned with either of these and we will assume them to be true.

HETEROSCEDASTICITY

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 7: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

7

The third, Assumption A.4, is that the variance of the distribution of the disturbance term is the same for each observation. In the present case, that means that the normal distributions shown all have the same variance.

HETEROSCEDASTICITY

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 8: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

8

If Assumption A.4 is satisfied, the disturbance term is said to be homoscedastic (Greek for same scattering).

HETEROSCEDASTICITY

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 9: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

9

Each observation is then potentially (before the sample is drawn) an equally reliable guide to the location of the line Y = 1 + 2X.

HETEROSCEDASTICITY

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

Page 10: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

10

Once the sample has been drawn, some observations will lie closer to the line than others, but we have no way of anticipating in advance which ones these will be.

1

X

Y = 1 +2X

Y

X3 X5X4X1 X2

HETEROSCEDASTICITY

Page 11: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

11

Now consider the situation illustrated by the diagram above. The distribution of u associated with each observation still has expected value 0 and is normal. However Assumption A.4 is violated and the variance is no longer constant.

X3 X5X4X1 X2

1

X

Y = 1 +2X

Y

HETEROSCEDASTICITY

Page 12: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

12

Obviously, observations where u has low variance, like that for X1, will tend to be better guides to the underlying relationship than those like that for X5, where it has a relatively high variance.

X3 X5X4X1 X2

1

X

Y = 1 +2X

Y

HETEROSCEDASTICITY

Page 13: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

13

When the distribution is not the same for each observation, the disturbance term is said to be subject to heteroscedasticity.

HETEROSCEDASTICITY

X3 X5X4X1 X2

1

X

Y = 1 +2X

Y

Page 14: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

14

There are two major consequences of heteroscedasticity. One is that the standard errors of the regression coefficients are estimated wrongly and the t tests (and F test) are invalid.

HETEROSCEDASTICITY

X3 X5X4X1 X2

1

X

Y = 1 +2X

Y

Page 15: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

15

The other is that OLS is an inefficient estimation technique. An alternative technique which gives relatively high weight to the relatively low-variance observations should tend to yield more accurate estimates.

HETEROSCEDASTICITY

X3 X5X4X1 X2

1

X

Y = 1 +2X

Y

Page 16: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

In the scatter diagram manufacturing output is plotted against GDP, both measured in US$ million, for 30 countries for 1997. The data are from the UNIDO Yearbook. The sample is restricted to countries with GDP at least $10 billion and GDP per capita at least $2000.

16

0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

1800000

0 1000000 2000000 3000000 4000000 5000000 6000000 7000000 8000000

GDP

Man

ufac

turin

g

HETEROSCEDASTICITY

Page 17: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

The scatter diagram is dominated by the observations for Japan and the USA and it is difficult to detect any kind of pattern.

17

Japan

USA

0

200000

400000

600000

800000

1000000

1200000

1400000

1600000

1800000

0 1000000 2000000 3000000 4000000 5000000 6000000 7000000 8000000

GDP

Man

ufac

turin

g

HETEROSCEDASTICITY

Page 18: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

However it those two countries are dropped and the scatter diagram rescaled, a clear picture of heteroscedasticity emerges.

18

0

50000

100000

150000

200000

250000

300000

0 200000 400000 600000 800000 1000000 1200000 1400000

GDP

Man

ufac

turin

g

HETEROSCEDASTICITY

Page 19: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

The reason for the heteroscedasticity is that variations in the size of the manufacturing sector around the trend relationship increase with the size of GDP.

19

0

50000

100000

150000

200000

250000

300000

0 200000 400000 600000 800000 1000000 1200000 1400000

GDP

Man

ufac

turin

g

South Korea

Mexico

HETEROSCEDASTICITY

Page 20: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

South Korea and Mexico are both countries with relatively large GDP. The manufacturing sector is relatively important in South Korea, so its observation is far above the trend line. The opposite was the case for Mexico, at least in 1997.

20

HETEROSCEDASTICITY

0

50000

100000

150000

200000

250000

300000

0 200000 400000 600000 800000 1000000 1200000 1400000

GDP

Man

ufac

turin

g

South Korea

Mexico

Page 21: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

Singapore and Greece are another pair of countries with relatively large and small manufacturing sectors. However, because the GDP of both countries is small, their variations from the trend relationship are also small.

21

0

50000

100000

150000

200000

250000

300000

0 200000 400000 600000 800000 1000000 1200000 1400000

GDP

Man

ufac

turin

g

Singapore

Greece

HETEROSCEDASTICITY

Page 22: HETEROSCEDASTICITY 1 This sequence relates to Assumption A.4 of the regression model assumptions and introduces the topic of heteroscedasticity. This relates

Copyright Christopher Dougherty 2012.

These slideshows may be downloaded by anyone, anywhere for personal use.Subject to respect for copyright and, where appropriate, attribution, they may be used as a resource for teaching an econometrics course. There is no need to refer to the author.

The content of this slideshow comes from Section 7.1 of C. Dougherty, Introduction to Econometrics, fourth edition 2011, Oxford University Press.Additional (free) resources for both students and instructors may be downloaded from the OUP Online Resource Centrehttp://www.oup.com/uk/orc/bin/9780199567089/.

Individuals studying econometrics on their own who feel that they might benefit from participation in a formal course should consider the London School of Economics summer school courseEC212 Introduction to Econometrics http://www2.lse.ac.uk/study/summerSchools/summerSchool/Home.aspxor the University of London International Programmes distance learning courseEC2020 Elements of Econometricswww.londoninternational.ac.uk/lse.

2012.11.10