the impact of air pollution emissions and related human health risks on the cross-section of stock...

Post on 27-Dec-2015

214 Views

Category:

Documents

1 Downloads

Preview:

Click to see full reader

TRANSCRIPT

THE IMPACT OF AIR POLLUTION THE IMPACT OF AIR POLLUTION EMISSIONS AND RELATEDEMISSIONS AND RELATED

HUMAN HEALTH RISKS ON THEHUMAN HEALTH RISKS ON THECROSS-SECTION OF STOCK RETURNSCROSS-SECTION OF STOCK RETURNS

Dinah A. KoehlerDinah A. KoehlerEPAEPA

Bernell K. StoneBernell K. Stone

BYUBYU

CONCERNCONCERN

For a cross section of companies/industries, does air pollution,

especially health damaging air pollution,have a systematic effect on valuation?

HYPOTHESESHYPOTHESES

Null: no effect

Alternative l: Positive effect, i.e.: More pollution greater value

Alternative 2: Negative effect, i.e.:

More pollution lower value

Comments:

1. In equilibrium, value and return realization are inversely related.

2. Revaluation from new information and/or from changes in how stocks are valued can be source of noise, possibly distortion.

Pollution

Ret

urn

Fair

Pollution

Ret

urn

Fair

Pollution

Ret

urn

Fair

Null

HYPOTHESESHYPOTHESES

Positive

Negative

ASSESSMENT ISSUESASSESSMENT ISSUES

• Fair return for time and risk

• Other cross-sectional return dependency variables

• Interpretative complexities

FAIR RETURN MODELSFAIR RETURN MODELS

CAPM (Capital Asset Pricing Model)2 return factors:

- time (short-term riskless interest rate)- systematic market risk (beta)

Rs = Rf + (RM - Rf)βs

Multifactor- Endogenous Factors (APT)- Exogenous Factors (specified variables) Example: “Fama-French 3 Factor”

OTHER CROSS SECTIONAL OTHER CROSS SECTIONAL RETURN DEPENDENCY RETURN DEPENDENCY

VARIABLESVARIABLES

• Other so-called anomaly variables

• Taxes and financial structure

• Growth, profitability, and other performance attributes

• Industry

Summary of Return Impacting Variables

VARIABLE NAME SYMBOL VARIABLE DEFINITION

Beta = Cov(Rs-Ro, RM – Ro) / Var(RM – Ro) measured over 3 years of past monthly returns, where Ro is the riskless

rate and RM is the market index.

Book-to-Market ratio B2M Ratio of BV/MV where BV is accounting book value (total common equity) and MV is market value of common stock

Market Cap (Market Value)

MV The market value of common stock at a point in time

Earnings Yield EY The ratio of Net Income to market value, the reciprocal of the price-earnings ratio

Dividend Yield DY The ratio of Annual Dividends to Market Value

Financial Structure FS The fraction of Total Investment provided by debt and preferred stock

Effective Tax Rate ET The ratio of Tax Payments to Net Income

Return on Investment ROI The ratio of Operating Income (before extraordinary income and expenses) to Total Investment

Return on Equity ROE The ratio of Net Income to Book Value

Sales Intensity SI The ratio of Sales to Total Investment

Sales Growth SAG Five-year average sales growth

Average Margin AM The Ratio of Operating Income to Sales

Sustainable Growth SUG The growth of common equity from retained earnings

INTERPRETATIVE INTERPRETATIVE COMPLEXITIESCOMPLEXITIES

The usual specification problems– Measurement error– Omitted variables– Improperly specified functional dependencies– Correlation distortion

State of the economyState of the market

– Changing level of interest rates – Changing term structure– Changing currency values– Hot and cold styles (e.g., growth vs value)

Industry-specific effectsArrival of new value changing information

STUDY LOGICSTUDY LOGIC

DATA

1. Rank Order into Industry Portfolios2. Step-Wise use of MAP

Scatter PlotsRegression Fits

Hypothesis Tests

STUDY LOGICSTUDY LOGIC

DATA

1. Rank Order into Industry Portfolios2. Step-Wise use of MAP

Model 1: No control restrictionsModel 2: CAPM betaModel 3: Fama-French 3-FactorModel 4: 7-Factors: EY, DY, FS, ETModel 5: Add Growth and Profitability

STUDY LOGICSTUDY LOGIC

DATA

1. Rank Order into Industry Portfolios2. Step-Wise use of MAP

Scatter PlotsRegression Fits

Hypothesis Tests

Measured Cross Section of Returns Measured Cross Section of Returns on Matched Portfolioson Matched Portfolios

y = -0.0244x + 0.2034

R2 = 0.1621

-0.2

-0.1

0

0.1

0.2

0.3

0.4

-3.5 -1.5 0.5 2.5 4.5 6.5log kg

1998

-200

2 Por

tfolio

Retu

rn

Figure 1 y = 0.0049x + 0.1497R2 = 0.0135

-0.1

0

0.1

0.2

0.3

-6 -4 -2 0 2 4log kg/va

1998

-200

2 Po

rtfoli

o Re

turn

Figure 2

Measured Cross Section of Returns Measured Cross Section of Returns on Matched Portfolioson Matched Portfolios

y = -0.3059x - 0.4848R2 = 0.5567

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

-3.5 -3 -2.5 -2 -1.5 -1log total pm2.5 risk/VA

1998

-200

2 Po

rtfoli

o Re

turn

Figure 4

Measured Cross Section of Returns Measured Cross Section of Returns on Matched Portfolioson Matched Portfolios

y = -0.314x - 0.5059R2 = 0.591

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

-3.5 -3 -2.5 -2 -1.5 -1log total TRI + pm2.5 risk/VA

1998

-200

2 P

ortfo

lio R

etur

n

Figure 5

Linear Regression Results: 1998-2002Linear Regression Results: 1998-2002LOG [KILOGRAMS DIRECT MASS EMISSIONS]

Model Coefficient T-value F-value R2

1 -0.02 -1.28 1.63 0.14

2 -0.03 -1.63 2.65 0.21

3 -0.03 -1.74 3.03 0.23

4 -0.01 -0.32 0.10 0.01

5 -0.02 -1.39 1.93 0.16

LOG [DIRECT KG/VA]

Model Coefficient T-value F-value R2

1 -0.0272 -1.8042 3.2552 0.2656

2 -0.0226 -1.2258 1.5026 0.1431

3 -0.0179 -1.0864 1.1803 0.1159

4 0.0070 0.4533 0.2055 0.0223

5 0.0049 0.3508 0.1231 0.0135

Linear Regression Results: 1998-2002Linear Regression Results: 1998-2002LOG [TOTAL TRI RISK/VA]

Model Coefficient T-value F-value R2

1 -0.23 -1.98 3.92 0.23

2 -0.15 -1.48 2.20 0.14

3 -0.17 -1.67 2.80 0.18

4 -0.30 -5.01** 25.10** 0.66

5 -0.31 -4.08* 16.65* 0.56

LOG [TOTAL PM2.5 RISK/VA]

Model Coefficient T-value F-value R2

1 -0.21 -2.42 5.84 0.29

2 -0.16 -1.84 3.39 0.19

3 -0.15 -1.88 3.55 0.20

4 -0.27 -3.21* 10.32* 0.42

5 -0.31 -4.19** 17.58** 0.56

* p < .01 ** p < .001

Linear Regression Results: 1998-2002Linear Regression Results: 1998-2002

LOG [(TOTAL TRI + PM2.5 RISK) / VA]

Model Coefficient T-value F-value R2

1 -0.23 -2.54 6.44 0.32

2 -0.18 -2.12 4.50 0.24

3 -0.14 -1.78 3.18 0.19

4 -0.25 -3.07* 9.40* 0.40

5 -0.31 -4.50** 20.23** 0.59

* p < .01 ** p < .001

MAJOR CONCLUSIONSMAJOR CONCLUSIONS

• Tonnage of air pollution is not significant

• Cancer health risk is significant

• Lung health risk is significant

INTERPRETATIVE COMPLEXITY:INTERPRETATIVE COMPLEXITY:EQUILIBRIUM VS REVALUATIONEQUILIBRIUM VS REVALUATION

“Equilibrium” (no pollution-related revaluation)-market failure in that companies in industry groups producing more health damaging air pollution are more highly valued than otherwise identical companies-there is a need for pertinent regulation

Cross-time revaluation-the market believes that there is the possibility that health-related air pollution will be more regulated (with costs to the polluting companies), and/or -the market becomes concerned with health-related legal liability (as occurred for asbestos and tobacco)

METHODOLGY CONCLUSIONSMETHODOLGY CONCLUSIONS

Many variable impacts

For pollution, it is crucial to control for– Fair return for time and risk– Taxes and financial structure– Growth and profitability– Industry attributes

ON-GOING AND FUTURE ON-GOING AND FUTURE RESEARCHRESEARCH

Size scaling alternatives

Outlier control

Industry impacts/distortion

More portfolios per cross-section

More sensitivity analysis

Additional years?

top related