代写Data Sources代写Web开发

1 Data Sources

The data used in this study includes corporate financial data, green bond issuance data, credit risk data, and macroeconomic data. The specific data sources are as follows:

Corporate Financial Data:

Capital Expenditures (CAPEX): obtained to track business investment spending using databases such Bloomberg.

Cash Flow: taken from company cash flow records to assess internal financing capability.

Total Assets, Total Liabilities, Net Profit: gathered from databases including Bloomberg to compute profitability, leverage, and company size as control factors.

Green Activity Data:

Green Innovation: Calculated based on the quantity of green patents registered by companies from databases including Derwent Innovation.

Green Governance: Calculated using companies' ESG ratings derived from databases including Refinitiv ESG.

Green Bond Issuance: Drawn from Bloomberg and Thomson Reuters, a dummy variable showing whether a company issued green bonds in a certain period.

Credit Risk Data:

To gauge company credit risk, credit rating data are gathered from rating companies as Moody's and S&P.

Credit risk is represented by proxies from credit ratings or credit spreads.

Macroeconomic Data:

Institutions like the World Bank and Eurostat provide macroeconomic statistics including GDP growth rate, interest rate, inflation rate, etc., which helps one to control the influence of the macroeconomic environment on company investment behavior.

2 Variable Construction

Dependent Variable:

Investment Expenditure (Investment): Calculated by the capital expenditures (CAPEX) to total assets or new fixed asset ratios..

Core Independent Variables:

Cash Flow (Cashflow): Measured by the ratio of operating cash flow to total assets to represent internal financing capacity.

Green Activities (GreenActivity): Measured by green innovation (number of green patents), green governance (ESG scores), or green bond issuance (dummy variable).

Credit Risk (CreditRisk): Measured by credit ratings or credit spreads to represent corporate credit risk.

Control Variables:

Firm Size (Size): Measured by the natural logarithm of total assets.

Leverage (Leverage): Measured by the ratio of total liabilities to total assets.

Profitability (Profitability): Measured by net profit or return on assets (ROA).

Market Valuation (Tobin's Q): Measured by the ratio of market value to book value.

3 Econometric Methods

This study employs a panel data regression model. The core regression equation is as follows:

Investment_i,t=α+β₁Cashflow_i,t+β₂GreenActivity_i,t+β₃CreditRisk_i,t

+β₄Controls_i,t+ϵ_i,t

Where Investment_i,t represents the investment expenditure of firm i in period t, Cashflow_i,t represents cash flow, GreenActivity_i,t represents green bond issuance, CreditRisk_i,t represents credit risk, and Controls_i,t represents control variables (e.g., firm size, leverage).

To further analyze the moderating effect of credit risk on the impact of green bonds on investment, an interaction term is introduced:

Investment_i,t=α+β₁Cashflow_i,t+β₂GreenActivity_i,t+β₃CreditRisk_i,t

+β₄(GreenActivity_i,t×CashFlow_i,t)+β₅Controls_i,t+ϵ_i,t

To further analyze the interaction effect between cash flow and green activities, an interaction term (Cashflow_i,t×GreenActivity_i,t) is introduced. If β₄ is significantly negative, it suggests that green activities can enhance the impact of cash flow on reducing investment-cash flow sensitivity.

4 Data Processing

Data Cleaning: The data are cleaned to handle missing values and outliers, ensuring completeness and consistency.

Panel Data Construction: Corporate financial data, green bond issuance data, credit risk data, and macroeconomic data are integrated into a panel dataset.

Fixed Effects Model: A fixed effects model is used to control for firm-specific and time-specific effects, ensuring the robustness of the regression results.

Instrumental Variable (IV) Regression: If endogeneity issues are present, instrumental variable (IV) regression is employed to address them.

5 Model Validation

To guarantee the model's precision and dependability, this document utilizes multiple methods of verification. Initially, the paper employs the Hausman test to select between a fixed and a random effect model, aiming to identify the most appropriate model. Next, this paper carries out multicollinearity test. This is done to confirm that there is no high correlation between explanatory variables. In this paper, robust standard error is also used to explain heteroscedasticity. In order to solve the possible endogeneity problem, regression of lagging variable and instrumental variable (IV) is implemented in this paper. Subsample analysis and quantile regression are used to verify the robustness of the model under different conditions.