碳排放管理對於企業股價之影響
Names _____
5/17/2022
研究方法
利用MSCI ACWI Europe ESG Leaders Index 中的個股與其碳排管理指標，進行單因子迴歸分析，以探討「碳排管理是否能帶來超額報酬」之課題。由於產業間業務性質差異甚大，因此也會將個股按照產業區分，以達到產業中立之目的。
研究動機
分析師報告中證實MSCI ACWI ESG Leaders Index的確在2020年疫情爆發之後較MSCI ACWI Index取得相對好的報酬。近期歐盟祭出2050年碳中和目標與一系列政策，企業因此面臨低碳轉型風險，因此我們希望將樣本池鎖定歐盟企業，且著重於ESG中的碳排放管理指標，期許能夠針對這個課題得出一些新發現。
分析與結果
Exploratory Data Analysis
(說明)
1. Correlation Plots
(說明)
(說明)
2. Avergae Emission Score Across EU
(說明)
Modeling
First, seperate data into carbon heavy industries (those 4), and carbon light industries I used two models here, linear regression and random forest, you guys can compare the performance of both models
(說明) might want to show how you build model, what varibles you included
Model Results
1. Linear Model Result
(說明)
##
## Call:
## lm(formula = TotalReturn_2y ~ IVA_COMPANY_RATING + INDUSTRY_ADJUSTED_SCORE +
## GOVERNANCE_PILLAR_SCORE + CARBON_EMISSIONS_SCORE + Industry,
## data = heavy_train)
##
## Residuals:
## Min 1Q Median 3Q Max
## -209.16 -86.16 -43.93 18.93 2168.31
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 309.497 143.109 2.163 0.0311 *
## IVA_COMPANY_RATINGAA -27.509 42.833 -0.642 0.5211
## IVA_COMPANY_RATINGAAA -8.713 77.247 -0.113 0.9103
## IVA_COMPANY_RATINGB -86.623 101.055 -0.857 0.3918
## IVA_COMPANY_RATINGBB -84.282 71.065 -1.186 0.2363
## IVA_COMPANY_RATINGBBB 3.597 42.413 0.085 0.9325
## IVA_COMPANY_RATINGCCC -164.480 137.015 -1.200 0.2307
## INDUSTRY_ADJUSTED_SCORE -20.674 23.366 -0.885 0.3768
## GOVERNANCE_PILLAR_SCORE 6.762 7.877 0.858 0.3912
## CARBON_EMISSIONS_SCORE -6.389 5.085 -1.256 0.2097
## IndustryMaterials -15.058 24.623 -0.612 0.5412
## IndustryUtilities -38.844 28.921 -1.343 0.1800
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 186.5 on 411 degrees of freedom
## Multiple R-squared: 0.05353, Adjusted R-squared: 0.02819
## F-statistic: 2.113 on 11 and 411 DF, p-value: 0.01852##
## Call:
## lm(formula = TotalReturn_2y ~ IVA_COMPANY_RATING + INDUSTRY_ADJUSTED_SCORE +
## GOVERNANCE_PILLAR_SCORE + CARBON_EMISSIONS_SCORE + Industry,
## data = light_train, na.action = na.omit)
##
## Residuals:
## Min 1Q Median 3Q Max
## -199.37 -60.38 -19.63 23.93 2901.11
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 47.169 59.027 0.799 0.424323
## IVA_COMPANY_RATINGAA -9.005 15.692 -0.574 0.566102
## IVA_COMPANY_RATINGAAA -20.020 30.353 -0.660 0.509613
## IVA_COMPANY_RATINGB 10.086 38.559 0.262 0.793679
## IVA_COMPANY_RATINGBB -4.380 26.954 -0.162 0.870933
## IVA_COMPANY_RATINGBBB 8.100 16.128 0.502 0.615576
## IVA_COMPANY_RATINGCCC 37.840 54.187 0.698 0.485055
## INDUSTRY_ADJUSTED_SCORE 3.022 8.430 0.358 0.720034
## GOVERNANCE_PILLAR_SCORE 1.793 3.014 0.595 0.552030
## CARBON_EMISSIONS_SCORE -2.483 2.372 -1.047 0.295219
## IndustryConsumer Discretionary 33.965 15.819 2.147 0.031897 *
## IndustryConsumer Staples -20.376 16.623 -1.226 0.220408
## IndustryFinancials 34.516 15.003 2.301 0.021511 *
## IndustryHealth Care 15.628 16.297 0.959 0.337691
## IndustryIndustrials 57.072 15.115 3.776 0.000164 ***
## IndustryInformation Technology 44.555 15.424 2.889 0.003911 **
## IndustryReal Estate -5.495 18.395 -0.299 0.765190
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 152.9 on 2008 degrees of freedom
## (7 observations deleted due to missingness)
## Multiple R-squared: 0.02906, Adjusted R-squared: 0.02132
## F-statistic: 3.756 on 16 and 2008 DF, p-value: 6.56e-072. Random Forest Result
(說明)
##
## Call:
## randomForest(formula = TotalReturn_2y ~ IVA_COMPANY_RATING + INDUSTRY_ADJUSTED_SCORE + GOVERNANCE_PILLAR_SCORE + CARBON_EMISSIONS_SCORE + Industry, data = heavy_train, control = rpart.control(cp = 0.002, minsplit = 30), importance = TRUE)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 1
##
## Mean of squared residuals: 35526.41
## % Var explained: 0.51##
## Call:
## randomForest(formula = TotalReturn_2y ~ IVA_COMPANY_RATING + INDUSTRY_ADJUSTED_SCORE + GOVERNANCE_PILLAR_SCORE + CARBON_EMISSIONS_SCORE + Industry, data = light_train, control = rpart.control(cp = 0.002, minsplit = 30), importance = TRUE, na.action = na.omit)
## Type of random forest: regression
## Number of trees: 500
## No. of variables tried at each split: 1
##
## Mean of squared residuals: 23971.04
## % Var explained: -0.433. Compare Out-of-Sample Performance
(說明) you can use rmse_lm_h to show the variable
## [1] 115.2414## [1] 115.2414## [1] 126.2252## [1] 96.92479Conclusion
(說明)