The Information Content of Green Innovations: U.S. Evidence

Abstract

The study examines whether corporate green innovation enhances the information content of stock prices among U.S. publicly listed firms from 1997 to 2023. Using green patent data from the United States Patent and Trademark Office, we find that firms engaging in green innovation exhibit significantly lower stock price synchronicity, indicating greater firm-specific information reflected in their prices. This finding is robust to alternative innovation measures, the lead dependent variable, the exclusion of the COVID-19 period, and the propensity score matching with Rosenbaum bounds sensitivity analysis. Exploiting the 2018 America’s Water Infrastructure Act as an exogenous regulatory shock, we find that the informativeness effect is substantially expanded following the shock. Citation-based analyses confirm that higher-quality green patents generate a stronger impact after adjusting for truncation bias. We identify analyst coverage and institutional ownership as two economic mechanisms through which green innovation improves the corporate information environment. Our findings suggest that green innovation represents a valuable source of firm-specific information content with meaningful implications for capital market efficiency.

1. Introduction

Stock price informativeness, the degree to which firm-specific value-relevant information is reflected in stock prices, has long been recognized as a critical attribute of well-functioning capital markets [1,2]. When individual stock prices move excessively in tandem with market and industry returns, they convey little firm-specific information. Conversely, when prices reflect firm-specific fundamentals, stock price informativeness is higher [3,4]. Bai et al. [5] point out that while U.S. financial markets have become more efficient over time, the sources of this informativeness continue to evolve. Given the pivotal role that stock prices play in guiding managerial decisions, facilitating efficient capital allocation, and reducing the cost of capital [6,7,8], understanding the determinants of the information content of prices remains a central question in financial economics.

At the same time, the global imperative to address climate change has elevated green innovation to the forefront of corporate strategy. Climate change presents a substantial risk to modern corporations, with firm-level exposure varying significantly across industries [9]. Green innovation, generally defined as the creation of new products, processes, and technologies that mitigate environmental risks and promote sustainability [10], has attracted growing attention from investors, regulators, and the public. The theoretical foundations for how environmental concerns affect asset prices are now well established: Pástor et al. [11] show that green assets command lower expected returns in equilibrium, Bolton and Kacperczyk [12] provide evidence that carbon-transition risk is globally priced, and Ilhan et al. [13] state that significant carbon tail risk in equity options markets.

Prior research has demonstrated that green innovative firms enjoy superior financial performance [14,15], improved access to credit [16,17], enhanced stock liquidity [18], and lower stock price crash risk [19]. However, despite this growing body of literature, an important question remains largely unexplored: does corporate green innovation enhance stock price informativeness?

This question is both timely and economically significant. Institutional investors increasingly incorporate environmental signals into their investment decisions [20], and active ownership engagement on ESG-related issues has been associated with improved firm valuation [21]. Unlike general CSR disclosures, which are susceptible to greenwashing [22], green patents represent costly, externally verified investment that credibly convey information about a firm’s technological capabilities, regulatory preparedness, and long-term strategic direction [18,23]. We argue that such credible signals reduce information asymmetry between corporate insiders and external investors, thereby facilitating the incorporation of firm-specific information into stock prices.

We propose two conceptually distinct but complementary informational channels through which green innovation enhances price informativeness. The first is a signaling channel. Green patents themselves are costly, externally verified signals [24] that directly convey value-relevant information about a firm’s technological capabilities and environmental commitments [25,26]. The second is an information production and aggregation channel. Green innovation attracts greater analyst coverage and attention from institutional investors [19], where analysts serve as critical information intermediaries [25] and institutional investors accelerate the incorporation of private information into prices through monitoring and informed trading [27,28].

To test those propositions, we construct a comprehensive panel dataset of U.S. publicly listed firms from 1997 to 2023, excluding financial and utility firms. We measure stock price informativeness using the inverse of synchronicity, following the standard approach in the literature [1,4], where lower synchronicity values indicate that stock prices reflect more firm-specific information. Our primary independent variable, $Green\_post$, is an indicator that equals one from the year a firm first applies for a green patent onward, capturing the persistent information effect of entering the green innovation space.

Our baseline results demonstrate that green innovation significantly enhances the information content of stock prices. Firms that have initiated green patenting activity show stock prices that reflect a greater proportion of firm-specific information. This finding is robust to an alternative continuous measure of green innovation, is persistent when we employ a lead dependent variable, and survives the exclusion of the COVID-19 period. Citation-based analyses further reveal that higher-quality green patents, as measured by forward citations adjusted for truncation bias [29], generate a substantially stronger effect on price informativeness. The progressive strengthening confirms that the market assigns greater informational weight to green innovations of higher technological significance.

To support causal interpretation, we implement several rigorous identification strategies. We exploit the America’s Water Infrastructure Act (AWIA) of October 2018 as an exogenous regulatory shock in a difference-in-difference framework. The results indicate that the informativeness effect of green innovation is substantially amplified following the shock, consistent with the view that increasing regulatory pressure elevates the informational salience of environmental preparedness. We address selection bias using propensity score matching across four specifications, with Rosenbaum bound confirming robustness to hidden bias up to a critical Gamma of 16.9. The Oster bounds test [30] provides additional assurance against omitted variable bias. We identify analyst coverage and institutional ownership as two economic mechanisms: green innovation firms attract significantly more analyst following and institutional capital, both of which facilitate the production and incorporation of firm-specific information into prices.

Our study makes several contributions. First, we contribute to the growing literature on the capital market consequences of green innovation [18,19,31] by establishing a clear relationship between green patenting activity and stock price informativeness. Second, we enrich the research on the influential factors of price informativeness [4,26,32,33] by identifying green innovation as a new and economically meaningful determinant. Third, we advance the broader literature on how sustainability factors shape financial markets [11,12,20] by demonstrating that green innovation improves not only asset pricing outcomes and trading costs but also the incorporation of firm-specific information into stock prices.

The remainder of this paper is organized as follows. Section 2 reviews the relevant literature and develops the expected channels. Section 3 describes the data, sample, and variable constructions. Section 4 presents empirical results and robustness tests. Section 5 concludes the paper.

2. Literature Review

2.1. Green Innovation and Its Economic Consequences

Green innovation encompasses all forms of creation of new products, processes, and technologies that contribute to reduced environmental risks and enhanced sustainability [10]. In the face of escalating environmental challenges and the rise of socially responsible investing, green innovation has become a core business strategy for firms seeking to address increasingly rigorous sustainability requirements [34,35,36]. The urgency of this strategy has been amplified by the recognition of climate change as a systemic financial risk. Sautner et al. [9] develop a comprehensive measure of firm-level climate change exposure using machine learning applied to earnings call transcripts, documenting substantial heterogeneity in how firms are affected by physical, regulatory, and opportunity dimensions of climate risk. Their findings reveal that higher climate change exposure predicts greater green patent activity, underscoring the strategic importance of green innovation as a corporate response to tightening ESG-related conditions.

A substantial body of research documents the economic benefits of green innovation. From a financial performance perspective, Farza et al. [14] and Tang et al. [15] provide evidence that green innovation improves both financial and operational outcomes. In credit markets, Xing et al. [16] and Liao [17] demonstrate that firms investing in green innovation receive more favorable lending terms. In capital markets, Chen et al. [37] show that green innovation enhances stock liquidity by mitigating information asymmetry, and Qian et al. [18] establish corroborating evidence among U.S. firms. Zaman et al. [19] find that eco-innovative firms attract greater analyst coverage and experience lower stock price crash risk, suggesting that green innovation fundamentally alters the firm’s information environment.

The antecedents of green innovation have also received considerable attention. At the firm level, ownership structure [38], board diversity [39], corporate governance [40], and retail investor attention [41] all influence the propensity of green patenting. At the macro level, environmental regulation [42] and green credit policies [43] stimulate green innovation, while economic policy uncertainty inhibits it [44]. Despite this rich literature documenting the causes and consequences of green innovation, one important question remains largely unexplained: whether and how green innovation affects the informational efficiency of stock prices.

2.2. Stock Price Informativeness

In an efficient market, stock prices aggregate information dispersed among heterogeneous market participants [45,46]. The degree to which prices reflect firm-specific rather than market-wide information varies considerably across firms and is commonly captured by stock price synchronicity, defined as the co-movement of individual stock returns with market and industry returns [1,2]. High synchronicity indicates that prices primarily reflect systematic factors, while low synchronicity suggests that firm-specific information is being actively incorporated into prices through informed trading. Bai et al. [5] provide evidence that U.S. financial markets have become more informative over time, highlighting the evolving nature of price efficiency.

Research on the determinants of stock price informativeness spans several domains. From a corporate governance perspective, Ferreira and Laux [4] show that better governance structure increases informativeness by reducing managerial opacity. Ng and Rezaee [26] document that large foreign ownership improves the information content of stock prices, and Derrien et al. [28] extend this finding to an international context. Gul et al. [47] find that board gender diversity enhances informativeness, while Busch and Obernberger [48] demonstrate that share repurchases improve the information reflected in prices.

From a disclosure perspective, Haggard et al. [45] establish that voluntary disclosure increases informativeness by expanding the set of publicly available firm-specific information. Grewal et al. [25] build on this by showing that material sustainability disclosures reduce stock price co-movement, providing direct evidence that environmental information contains value-relevant content. Edmans et al. [8] examine the origins of information in prices, finding that the source of information has distinct implications for real corporate decisions. Brogaard et al. [32] further decompose price movements and conclude that private, firm-specific information is the primary driver of idiosyncratic variation, reinforcing the view that information production by market participants is central to the informativeness of stock prices.

The consequences of informative prices extend well beyond capital markets. Grounded in theoretical frameworks linking information frictions to market quality [49], empirical research confirms that price informativeness optimizes capital investment efficiency [6], improves labor allocation decisions [7], and catalyzes innovation [50]. These real effects underscore the economic significance of understanding what factors enhance or impair the information content of stock prices.

2.3. Green Innovation and the Corporate Information Environment

A rapid growing literature established that market participants recognize and price climate-related information, creating the informational foundation upon which green innovation can influence stock prices. Investors derive non-pecuniary utility from holding green assets [11], and these assets have been shown to outperform during periods of heightened climate concern [51]. Equity markets actively price carbon-transition risk on a global scale [12] and penalize heavy emitters through the pollution premium [52]. Even amid uncertainty in ESG ratings [53], environmental disclosure reflects fundamentals that are values by market participants [54].

Institutional investors are central to this pricing mechanism. They actively view climate risks as financially material [20], and shareholder engagement on environmental issues has been shown to generate significant gains [55], with successful interventions yielding positive abnormal returns [21]. The pricing of environmental exposure extends into derivative markets as well, where carbon tail risk is distinctly priced in option markets [13]. At the macroeconomic level, Hong et al. [51] emphasize the increasing weight of climate-related data in asset pricing, while Engle et al. [53] develop methodologies for hedging against climate news risk. Fulgence et al. [43] demonstrate that political uncertainty undermines stock price informativeness, suggesting that the information environment surrounding policy and regulation plays a meaningful role in shaping how firm-level signals are processed by markets.

This body of evidence collectively establishes two important premises for our study. First, environmental information is financially material and actively processed by sophisticated market participants. Second, the information environment surrounding climate and sustainability topic has direct implications for stock price efficiency. What remains unexplored is whether green innovation, as a specific and verifiable form of corporate environmental strategy, actively enhances the firm-specific information content of stock prices.

2.4. Expected Channels

We propose that green innovation enhances stock price informativeness through two informational channels. The first channel operates directly through signaling, where the green patent itself transmits firm-specific information to the market. The second channel operates indirectly through information production and aggregation, where green innovation creates new fundamentals about which sophisticated intermediaries produce private information that is subsequently impounded into prices through analyst research and informed trading.

2.4.1. The Signaling Channel

Green patents represent costly, externally verified capital investments that are difficult to replicate [18,23]. Unlike qualitative CSR disclosures, which may be subject to greenwashing and rating inconsistencies, green patents serve as tangible and externally verified signals of a firm’s technological capabilities and environmental commitments [30,31,36]. As financial markets increasingly price climate-related information [9,12,52], the signaling value of green patents is amplified: they provide concrete evidence of environmental preparedness that investors and analysts use to form more precise assessments of firm value. Firms with stronger environmental performance also exhibit a higher propensity for voluntary transparency [33,34], further expanding the set of firm-specific information available to the market and reducing information asymmetry between corporate insiders and outside investors [35,56].

2.4.2. The Information Production and Aggregation Channel

Beyond the direct signaling effect, green innovation attracts heightened attention from institutional investors and financial analysts, thereby stimulating the production and dissemination of firm-specific information [11,19,20]. Through active monitoring and informed trading, institutional investors accelerate the incorporation of private information into market prices [26,27,28]. Concurrently, the expanded analyst coverage drawn to eco-innovative firms generates new research that broadens the set of publicly available firm-specific analysis [24]. Both mechanisms increase the volume and quality of information available to market participants, contributing to lower synchronicity.

2.4.3. Relationship Between the Two Channels

The two channels are conceptually distinct but operationally complementary. The signaling channel describes how a credible information signal is created, while the information production and aggregation channel describes how that signal is processed and transmitted into prices by market participants. Both channels predict a negative association between green innovation and stock price synchronicity, but they have different empirical signatures: the signaling channel implies that informativeness effects scale with patent quality, while the information production channel implies that the effect operates through observable changes in analyst coverage and institutional ownership.

3. Data, Sample, and Measurement

3.1. Sample Construction

Our sample comprised the universe of U.S. publicly listed firms from 1997 to 2023. We established 1997 as the start of our sample period to align with the adoption of the Kyoto Protocol, a catalytic event in global environmental policy that spurred corporate green innovation worldwide. We obtained daily stock return data from the Center for Research in Security Prices (CRSP), restricting the sample to ordinary common shares (share codes 10 and 11) traded on the NYSE, AMEX, and NASDAQ. Firm-level financial data were sourced from Compustat North America and merged with the CRSP dataset using the standard CRSP-Compustat linking table. Consistent with prior literature [1,30], we excluded financial institutions (SIC codes 6000-6799) and utility firms (SIC codes 4900-4999) due to their unique regulatory and accounting environments.

Green patent data were sourced from the United States Patent and Trademark Office (USPTO). To identify relevant innovation, we utilized section Y of the Cooperative Patent Classification (CPC) scheme, which categorizes technologies pertinent to climate change mitigation. We linked these patents to our firm universe via PERMCO identifiers utilizing established matching procedures [42]. The dataset was further augmented with patent citation metrics [57].

3.2. Variable Construction

3.2.1. Measure of Stock Price Synchronicity

We measured stock price informativeness using the inverse of stock price synchronicity, adopting a standard methodological framework widely utilized in the literature [1,4]. We first aggregated daily CRSP returns to a weekly frequency by compounding log returns, requiring a minimum of 30 valid trading weeks per firm-year to ensure robust estimation [35]. For each firm $i$ in year $t$ by week $w$, we estimated the following market-model regression:

$${\mathrm{r}}_{\mathrm{i},\mathrm{w}}={\mathsf{\alpha }}_{\mathrm{i}}+{\mathsf{\beta }}_1·\mathrm{M}\mathrm{K}{\mathrm{T}}_{\mathrm{i},\mathrm{w}}+{\mathsf{\beta }}_2·\mathrm{M}\mathrm{K}{\mathrm{T}}_{\mathrm{i},\mathrm{w}-1}+{\mathsf{\gamma }}_1·\mathrm{I}\mathrm{N}{\mathrm{D}}_{\mathrm{i},\mathrm{w}}+{\mathsf{\gamma }}_2·\mathrm{I}\mathrm{N}{\mathrm{D}}_{\mathrm{i},\mathrm{w}-1}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{w}}$$

where ${\mathrm{r}}_{\mathrm{i},\mathrm{w}}$ represents the weekly excess return for firm $\mathrm{i}$ in week $\mathrm{w}$. $\mathrm{M}\mathrm{K}{\mathrm{T}}_{\mathrm{i},\mathrm{w}}$ denotes the firm-excluded value-weighted market return, and $\mathrm{I}\mathrm{N}{\mathrm{D}}_{\mathrm{i},\mathrm{w}}$ represents the firm-excluded value-weighted industry return based on two-digit SIC codes. The lagged market and industry returns were included to capture delayed price adjustments caused by trading frictions.

Stock price synchronicity (SYNCH) mathematically derived from the ${\mathrm{R}}^2$, the coefficient of determination, of the regression above. Because ${\mathrm{R}}^2$ is bounded between 0 and 1, we applied a logistic transformation to yield a continuous measure: $\mathrm{S}\mathrm{Y}\mathrm{N}\mathrm{C}{\mathrm{H}}_{\mathrm{i},\mathrm{t}}=\mathrm{l}\mathrm{n}({\displaystyle \frac{{\mathrm{R}}_{\mathrm{i}.\mathrm{t}}^2}{1-{\mathrm{R}}_{\mathrm{i},\mathrm{t}}^2}})$. A lower value of $\mathrm{S}\mathrm{Y}\mathrm{N}\mathrm{C}\mathrm{H}$ indicates higher price informativeness, signifying that a greater proportion of the stock’s return variation is driven by firm-specific information, captured by the residual ${\mathsf{\epsilon }}_{\mathrm{i},\mathrm{w}}$, rather than systematic market or industry shocks.

3.2.2. Green Innovation

Our primary independent variable, $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$, is an indicator designed to capture the persistent information effect of entering the green innovation space. Specifically, this dummy variable takes the value of one for the year of a firm’s first green patent application and all subsequent years. For all prior firm-year observations, as well as for all observations of firms that never produced a green patent during our sample period, the variable is strictly coded as zero.

To capture the intensive margin of green innovation, we employed two alternative continuous measures following established practices [58]: $\mathrm{L}\mathrm{n}\_\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}$, defined as the natural logarithm of one plus the annual green patent count, and $\mathrm{L}\mathrm{n}\_\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{C}\mathrm{i}\mathrm{t}\mathrm{e}$, defined as the natural logarithm of one plus the forward citation count. Finally, to conduct falsification testing, we constructed analogous placebo measure utilizing non-green patents, calculated as the firm’s total patent output minus its green patents.

3.2.3. Control Variables

To mitigate potential omitted variable bias, our empirical models incorporated a standard vector of firm-level control variables known to influence the corporate information environment. Firm size was measured as the natural logarithm of CPI-adjusted total assets. Book leverage was defined as the ratio of total debt to total assets. We measured R&D intensity as research and development expenditures scaled by sales; to retain observations, missing R&D values were set to zero. Finally, asset tangibility was calculated as a weighted average of receivables, inventory, net property, plant and equipment, and cash, scaled by total assets. Cash flow was constructed as income before extraordinary items plus depreciation, minus preferred and common dividends, scales by lagged total assets [59].

All continuous variables were winsorized at the 1st and 99th percentiles before conducting our empirical analyses to avoid the influence of extreme outliers.

3.3. Empirical Model

To estimate the impact of green innovation on stock price informativeness, we employed the following baseline panel regression model:

$$\mathrm{S}\mathrm{Y}\mathrm{N}\mathrm{C}{\mathrm{H}}_{\mathrm{i},\mathrm{t}}=\mathsf{\alpha }+\mathsf{\beta }·\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}{\mathrm{t}}_{\mathrm{i},\mathrm{t}}+\mathsf{\gamma }·{\mathrm{X}}_{\mathrm{i},\mathrm{t}}+{\mathsf{\delta }}_{\mathrm{i}}+{\mathsf{\delta }}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{i},\mathrm{t}}$$

where $\mathrm{i}$ and $\mathrm{t}$ index firm and year, respectively. SYNCH represents stock price synchronicity. $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ is the green innovation indicator. $\mathrm{X}$ is a vector of firm-level control variables. ${\mathsf{\delta }}_{\mathrm{i}}$ and ${\mathsf{\delta }}_{\mathrm{t}}$ denote firm and year fixed effects, controlling for time-invariant firm heterogeneity and common time trends. Standard errors were clustered at the firm level to account for serial correlation within firms.

3.4. Descriptive Statistics

Table 1. Summary statistics.

Variable	N	Mean	Std. Dev.	P25	Median	P75
Synchronicity	79,704	−1.0673	1.1397	−2.0598	−1.2767	−0.4904
Green × Post	79,704	0.1977	0.3983	0	0	0
Firm Size	79,704	18.8482	2.0949	17.2959	18.7557	20.2973
Book Leverage	79,369	0.2252	0.2326	0.0149	0.1722	0.3532
Tangibility	78,410	0.4807	0.2002	0.3452	0.4667	0.5852
R&D Intensity	79,704	0.1886	0.4686	0	0.0067	0.1211
Cash Flow	77,088	−0.0447	0.3267	−0.0759	0.0562	0.1146

present the summary statistics for our variables used in regression analysis. The mean value of SYNCH was $-1.0673$ with a standard deviation of $1.1397$. Our main independent variable, $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$, had a mean of $0.1977$, indicating that approximately $19.8\%$ of the firm-year observations corresponded to firms that had entered the green innovation space. Regarding the firm-level controls, the mean of firm size was $18.8482$, book leverage averaged $0.2252$, tangibility had a mean of $0.4807$, R&D intensity averaged $0.1886$, and cash flow had a mean of $-0.0447$.

In

Table 2. Univariate test: comparison between GI and Non-GI firms.

Variable	GI Firm	Non-GI Firm	Difference	t-Statistics
Synchronicity	−1.0912	−0.8498	−0.243 ***	9.91
Firm Size	19.781	18.5058	1.2752 ***	4.33
Book Leverage	0.21	0.2307	−0.0207 ***	−13.82
Tangibility	0.4918	0.4766	0.0153 ***	9.36
R&D Intensity	0.2973	0.1487	0.1486 ***	4.82
Cash Flow	−0.0404	0.0064	−0.0468 **	−3.31

** indicates p < 0.05, two-tailed; *** indicates p < 0.01, two-tailed.

, we reported our univariate test results for our sample firms with green innovations and without green innovations. We found that firms engaging in green innovations had lower stock price synchronicity, larger size, lower book leverage, higher assets tangibility, higher R&D intensity, and lower cash flow.

4. Empirical Results and Robustness Tests

4.1. Baseline Results

Table 3. Baseline regression relating green innovation to stock price synchronicity.

Independent Variables	Dependent Variable: Synchronicity
	Baseline	Alternative Measure	Lead Synch	Excluded COVID	Industry × Year Fixed Effect	Included ESG Rating
	1	2	3	4	5	6
Green_post	−0.0659 ***		−0.0644 **	−0.0624 **	−0.0405 **	−0.0658 **
	(0.0242)		(0.026)	(0.0249)	(0.0206)	(0.0259)
Ln(Green)		−0.0450 ***
		(0.0137)
Firm Size	0.2822 ***	0.2831 ***	0.2244 ***	0.2873 ***	0.2766 ***	0.2824 ***
	(0.0078)	(0.0078)	(0.0087)	(0.008)	(0.0075)	(0.0084)
Book Leverage	−0.2444 ***	−0.2457 ***	−0.2833 ***	−0.2833 ***	−0.2133 ***	−0.2452 ***
	(0.0285)	(0.0285)	(0.032)	(0.0295)	(0.028)	(0.0305)
Tangibility	0.3206 ***	0.3211 ***	0.3204 ***	0.3045 ***	0.2616 ***	0.3198 ***
	(0.0357)	(0.0357)	(0.0386)	(0.0372)	(0.0351)	(0.0382)
R&D Intensity	0.0790 ***	0.0799 ***	0.0391 **	0.0849 ***	0.0643 ***	0.0578 ***
	(0.0151)	(0.0151)	(0.0165)	(0.0158)	(0.0142)	(0.0136)
Cash Flow	0.0772 ***	0.0760 ***	0.0692 ***	0.0643 ***	0.1007 ***	0.0770 ***
	(0.0188)	(0.0188)	(0.0208)	(0.0198)	(0.0183)	(0.0201)
						−0.0065 **
						(0.0017)
Constant	−6.6966 ***	−6.7206 ***	−5.5444 ***	−6.7917 ***	−6.5703 ***	−5.7944 ***
	(0.1509)	(0.151)	(0.1685)	(0.155)	(0.1452)	(0.2117)
Observations	74,103	74,103	63,625	69,579	74,096	51,250
Year FE	Yes	Yes	Yes	Yes	Yes	No
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes
Industry × Year FE	No	No	No	No	No	Yes
Adjusted R-squared	0.5456	0.5456	0.5503	0.5513	0.5654	0.5725

** indicates $p<0.05$, two-tailed; *** indicates $p<0.01$, two-tailed. “Yes” indicates that the specific fixed effects were included, and “No” indicates that the specific fixed effects were no included.

presents our baseline regression results examining the relationship between green innovation and stock price informativeness. All specifications include firm and year fixed effects with standard errors clustered at the firm level.

In Column 1, we regress our synchronicity measure on the green innovation indicator ($\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$) incorporating the full set of firm-level controls. The coefficient on $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ is $-0.066$ and statistically significant at the 1% level ($\mathrm{t}=-2.73$). This result supports our central hypothesis: firms that engage in green patenting exhibit significantly lower stock price synchronicity, reflecting a meaningful increase in the proportion of firm-specific information incorporated into their stock prices. The inclusion of firm fixed effects ensures that this estimate captures within-firm variation, isolating the structural shift in a firm’s informational environment that occurs when it initiates green patenting rather than reflecting persistent cross-sectional differences between firms.

Column 2 extends the analysis to the intensive margin by employing a continuous measure of innovation quantity, $\mathrm{l}\mathrm{n}(1+\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n} \mathrm{P}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t})$. The estimated coefficient is significantly negative at $-0.045$, with $\mathrm{t}=-3.28$. This confirms that beyond the initial onset, the accumulated volume of green patenting activity is also associated with enhanced price informativeness. The information content generated by green innovation scales with the firm’s total output of environmentally oriented patents.

Column 3 examines whether the informativeness effect persists beyond the contemporaneous period by utilizing lead synchronicity as the dependent variable. The coefficient on $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ remains negative and significant with $\mathsf{\beta }=-0.064$ and $\mathrm{t}=-2.48$. This persistence indicates that green innovation facilitates a sustained improvement in the corporate information environment that extends into the following year rather than producing a transient market reaction.

Column 4 addresses potential confounding effects arising from the extreme market volatility during the COVID-19 pandemic. Upon excluding observations from 2020 and 2021, the coefficient on $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ remains negative and significant with $\mathsf{\beta }=-0.062$ and $\mathrm{t}=-2.51$. This validates that our findings reflect the informational dynamics of green innovation and are not driven by pandemic-induced market distortions.

Column 5 includes industry-year fixed effects to absorb all time-varying shocks common to firms within the same industry-year cell, such as industry-wide regulatory changes, sector-specific innovation waves, and shifts in industry-level investor preferences. The coefficient on Green_post remains negative and significant. In Column 6, we use the ASSETS4 database to construct our measure of overall ESG rating and enter ESG rating as an additional control for firm information environment [29]. Our result reveals that ESG rating as a proxy of firm information environment has a negative effect on stock price synchronicity, and the significance of Green_post still holds.

The estimated coefficients on control variables align with established theoretical expectations. Firm size is positively associated with synchronicity, consistent with the well-documented tendency of larger firms to co-move with the broader market [1,4]. Book leverage exhibits a negative coefficient, suggesting that higher debt levels generate firm-specific default risk information that the market prices accordingly. Tangibility enters positively, while R&D intensity is also positive and significant, consistent with the interpretation that broad R&D signals are often processed as industry-level rather than firm-level news. Cash flow is positively associated with synchronicity, indicating that firms generating stronger internal cash flows tend to show greater co-movement with market returns.

4.2. Difference-in-Difference: The 2018 AWIA Shock

To strengthen causal identification, we employ a difference-in-difference design exploiting the America’s Water Infrastructure Act (AWIA) of October 2018 as an exogenous regulatory shock. The AWIA mandated comprehensive water quality standards and risk resilience assessments for public water systems, intensifying environmental regulatory pressure and increasing the informational salience of corporate green innovation [18]. Following the $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ construction defined in Section 3.2.2, the treatment group consists of firm-year observations for which $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}=1$, and the control group consists of firm-year observations for which $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}=0$. We also define $\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}\_2019$ as an indicator equal to one for years 2019 onward and construct the interaction $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}\_\mathsf{{\rm X}}\_\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}\_2019$ to capture the incremental effect of the regulatory shock on green innovators. Because year fixed effects absorb all common time variation, the $\mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}\_2019$ indicator does not appear separately in the specification.

Table 4. Difference-in-differences: the 2018 AWIA shock.

Independent Variables	Dependent Variable: Synchronicity
	AWIA Shock	Event Study
	1	2
Green_post	−0.0482 *	0.023
	(0.025)	(0.0414)
Green_Post × Post2018	−0.0769 ***
	(0.0257)
Green_Post × Year 2014		−0.0682 *
		(0.0376)
Green_Post × Year 2015		−0.0358
		(0.0466)
Green_Post × Year 2016		−0.075
		(0.0472)
Green_Post × Year 2017		−0.1923 ***
		(0.0481)
Green_Post × Year 2019		−0.0136 *
		(0.0071)
Green_Post × Year 2020		−0.1980 ***
		(0.0484)
Green_Post × Year 2021		−0.2796 ***
		(0.0515)
Green_Post × Year 2022		−0.1063 **
		(0.0452)
Green_Post × Year 2023		−0.1800 ***
		(0.0489)
Firm Size	0.2821 ***	0.2819 ***
	(0.0078)	(0.0078)
Book Leverage	−0.2462 ***	−0.2477 ***
	(0.0284)	(0.0284)
Tangibility	0.3164 ***	0.0781 ***
	(0.0357)	(0.015)
R&D Intensity	0.0784 ***	0.3161 ***
	(0.0151)	(0.0357)
Cash Flow	0.0775 ***	0.0775 ***
	(0.0188)	(0.0188)
Constant	−6.6914 ***	−6.6870 ***
	(0.1509)	(0.1511)
Observations	74,103	74,103
Firm FE	Yes	Yes
Year FE	Yes	Yes
Adjusted R-squared	0.5457	0.546

* indicates $p<0.10$, two-tailed; ** indicates $p<0.05$, two-tailed; *** indicates $p<0.01$, two-tailed.

reports the results. The coefficient on $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ is $-0.048$, marginally significant at the 10% level, confirming that a baseline negative association between green innovation and synchronicity exists prior to the regulatory shock. The interaction term yields a coefficient of $-0.077$, statistically significant at the 1% level. This indicates that the AWIA substantially amplifies the informativeness effect of green innovation. Combining both coefficients, the total effect from the shock for green innovators is $-0.125$, representing a pronounced strengthening relative to the pre-shock period.

Before the AWIA, green innovation already contributed to incorporating firm-specific information into stock prices. Following the regulatory event, the market placed greater weight on green innovation signals, as the new legislation elevated the materiality of environmental preparedness for corporate valuation. This pattern is consistent with evidence that regulatory events amplify the informational relevance of sustainability activities by raising both the expected costs of noncompliance and the expected benefits of proactive environmental strategy [52,53].

A natural concern for any difference-in-difference design is whether contemporaneous policy events, rather than the regulatory shock of interest, drive the estimated treatment effect. We therefore consider the major policy and economic events that overlap with our sample window. First, the Tax Cuts and Jobs Act (TCJA), enacted in December 2017, modified corporate tax rates and capital structure incentives uniformly across firms. It is not differentially targeted at green innovators, and its enactment precedes our AWIA shock window. The progressive timing of our DID effect, concentrated in fiscal year 2019 and later, is therefore inconsistent with TCJA being the active driver. Second, the COVID-19 fiscal and monetary stimulus measures enacted in 2020 and 2021 fall within a period we explicitly exclude as a robustness check. Our baseline result is unchanged and slightly strengthens when the pandemic window is removed ($\mathsf{\beta }=-0.075$, $\mathrm{t}=-2.86$). The slightly larger magnitude in the COVID-19-excluded specification is consistent with the pandemic period introducing macro-level noise that attenuates rather than generates the green-innovation effect. Third, the Securities and Exchanges Commission’s climate-related disclosure guidance, proposed in March 2022 and finalized in March 2024, postdates our post-treatment window and therefore cannot drive the 2019-onward effect we document. Taken together, the timing of major contemporaneous policy events, combined with our pre/post-structure and explicit exclusion of the pandemic window, supports the interpretation that our DID estimates capture the AWIA-driven effect rather than confounding policy shocks.

To formally test the assumption of parallel trend underlying our difference-in-difference design, Column 2 of Table 4 reports the dynamic event-study specification. We replace the static $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}\times \mathrm{P}\mathrm{o}\mathrm{s}\mathrm{t}2018$ interaction term with a series of year-specific interaction terms spanning 2014 to 2023. The year 2018 is the omitted reference category, so each coefficient measures the difference relative to that year. The pre-shock coefficients for 2014, 2015, and 2016 are statistically insignificant. This supports the parallel trend assumption, as treated and control firms followed similar trajectories before the regulatory shock. The coefficient for 2017 is negative and significant. One plausible explanation is that AWIA was introduced to Congress in 2017 as bill S.3021, well before its enactment in October 2018. Sophisticated managers and investors typically begin updating their expectations once a bill is formally proposed. The effect materializes gradually from 2018 onward, reflecting the time required for the regulation’s implications to be fully incorporated into firm behavior and market expectations. The post-shock coefficients for 2019 through 2023 are all negative and significant. This confirms that the informativeness effect of green innovation strengthened following the AWIA shock.

4.3. Green Innovation Quality: Citation-Based Analysis

If green innovation genuinely improves price informativeness, higher-quality patents should generate stronger effects. We test this premise in

Table 5. Citation-based analysis.

Independent Variables	Dependent Variable: Synchronicity
	Include 2021–2023		Exclude 2021–2023		Exclude Self-Citation
	Raw Citation	Citation Adjusted for Truncation	Raw Citation	Citation Adjusted for Truncation	Raw Citation
	1	2	3	4	5
Citation	−0.0182 ***		−0.0286 ***		−0.0187 ***
	(0.0051)		(0.0054)		(0.0053)
Citation_Adjusted		−0.0361 *		−0.0615 ***
		(0.0194)		(0.0213)
Firm Size	0.2821 ***	0.2818 ***	0.2841 ***	0.2836 ***	0.2821 ***
	(0.0078)	(0.0078)	(0.0083)	(0.0083)	(0.0078)
Book Leverage	−0.2458 ***	−0.2456 ***	−0.2761 ***	−0.2753 ***	−0.2458 ***
	(0.0285)	(0.0286)	(0.0309)	(0.0309)	(0.0285)
Tangibility	0.3231 ***	0.3226 ***	0.3029 ***	0.3028 ***	0.3231 ***
	(0.0357)	(0.0357)	(0.0386)	(0.0386)	(0.0357)
R&D Intensity	0.0797 ***	0.0796 ***	0.0884 ***	0.0886 ***	0.0797 ***
	(0.0151)	(0.0151)	(0.0169)	(0.0169)	(0.0151)
Cash Flow	0.0766 ***	0.0766 ***	0.0746 ***	0.0749 ***	0.0766 ***
	(0.0188)	(0.0188)	(0.0205)	(0.0205)	(0.0188)
Constant	−6.7056 ***	−6.7016 ***	−6.7416 ***	−6.7355 ***	−6.7056 ***
	(0.1514)	(0.1514)	(0.1602)	(0.1603)	(0.1514)
Observations	74,103	74,103	66,703	66,703	74,103
Firm FE	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes
Adjusted R-squared	0.5456	0.5455	0.5498	0.5496	0.5456

* indicates $p<0.10$, two-tailed; *** indicates $p<0.01$, two-tailed.

using forward patent citations as our primary proxy for innovation quality [60]. Raw citation counts, however, suffer from truncation bias: recent patents have less time to accumulate citations than older ones [61]. A patent filed in 2020 with three citations may be equally impactful as a 2005 patent with fifteen citations. To address this temporal bias, we apply a standard truncation adjustment [61] and concurrently test specifications that exclude the most recent three years of our sample (2021–2023).

Table 5 reports four progressively rigorous specifications. Column 1 uses raw citations across the full sample, yielding a coefficient of $-0.018$ ($p<0.10$). Column 2 applies the truncation adjustment, and the effect nearly doubles to $-0.036$ ($p<0.01$). Column 3 drops observations from 2021 to 2023, producing a coefficient of $-0.029$ ($p<0.01$). Column 4 combines both adjustments, yielding the largest magnitude at $-0.062$ ($p<0.01$).

This progressive strengthening demonstrates that unadjusted citation counts understate the true economic impact of patent quality on price informativeness. Once properly corrected for temporal bias, high-quality green innovation exhibits a considerably stronger capacity to enhance the incorporation of firm-specific information into stock prices.

The economic intuition behind citation-based quality measures is straightforward but worth making explicit. Not all patents convey equivalent information to the market. Forward citations capture this distinction by aggregating the external technical community’s revealed assessment of which innovations matter. From the perspective of information transmission to capital markets, this is meaningful. Highly cited green patents signal not only that the firm is investing in environmental innovation but also that the innovation is technologically substantive, externally recognized, and likely to anchor follow-on research and development. The strengthening of our results across citation-weighted specifications therefore reflects this informational content rather than mechanical scaling: investors and analysts assign greater informational weight to innovation outputs that the broader inventor community has validated through citation.

To verify that our citation-quality result reflects external technological recognition rather than within-firm self-referencing, Column 5 re-estimates the baseline citation regression after excluding self-citations from the forward-citation count. Self-citations refer to citations in which the citing and cited patents belong to the same firm. The coefficient remains significant and negative. This confirms that our findings reflect the broader inventor community’s revealed assessment of patent quality rather than artificial inflation from internal self-citation practices that are particularly common in the green-patent literature.

4.4. Propensity Score Matching and Sensitivity Analysis

To address potential selection bias, we use propensity score matching (PSM). Treated observations are matched to control firms using nearest-neighbor matching with replacement. The matching criteria incorporate firm size, book leverage, R&D intensity, tangibility, cash flow, and year.

Table 6. Propensity score matching results.

Independent Variables	Dependent Variable: Synchronicity
	1–3 Matching	1–5 Matching	1–10 Matching	1–20 Matching
	(1)	(2)	(3)	(4)
Green_post	−0.0719 ***	−0.0580 **	−0.0639 **	−0.0647 ***
	(0.0278)	(0.0263)	(0.025)	(0.0243)
Firm Size	0.2565 ***	0.2704 ***	0.2759 ***	0.2840 ***
	(0.0113)	(0.0101)	(0.009)	(0.0083)
Book Leverage	−0.2246 ***	−0.2249 ***	−0.2419 ***	−0.2551 ***
	(0.0406)	(0.0362)	(0.0326)	(0.0306)
Tangibility	0.3609 ***	0.3680 ***	0.3568 ***	0.3369 ***
	(0.0507)	(0.0456)	(0.0406)	(0.0376)
R&D Intensity	0.1065 ***	0.1029 ***	0.1025 ***	0.0893 ***
	(0.0203)	(0.0179)	(0.0163)	(0.0157)
Cash Flow	0.1043 ***	0.0889 ***	0.1018 ***	0.0913 ***
	(0.0287)	(0.0253)	(0.0224)	(0.0203)
Constant	−6.1148 ***	−6.4212 ***	−6.5498 ***	−6.7158 ***
	(0.2291)	(0.2038)	(0.1783)	(0.1622)
Observations	38,544	46,950	58,118	66,907
Year FE	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes
Adjusted R-squared	0.5531	0.5497	0.5451	0.5426

** indicates $p<0.05$, two-tailed; *** indicates $p<0.01$, two-tailed.

reports the PSM results across four specifications: 1-to-3, 1-to-5, 1-to-10, 1-to-20 nearest neighbors. The estimated coefficients on $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ are stable across all specifications, ranging from $-0.058$ to $-0.072$ and consistently significant at the 5% level or better. This stability across different matching structures confirms that our baseline findings are not driven by observable selection into green innovation.

We conduct Rosenbaum bounds sensitivity analysis to quantify vulnerability to hidden bias [58]. The critical Gamma value is 16.9 across all four matching specifications, indicating that a hidden bias factor of 16.9 is required before the confidence interval begins to contain positive values. To make this result more interpretable, we translate it into equivalent changes using firm size as an illustration. An unobserved variable would need to produce an effect equivalent to approximately 4.75 to 5.26 standard deviations from the mean of the firm size to overturn our findings. Since a disturbance of such extraordinary magnitude is virtually implausible, we conclude that hidden bias does not pose a threat to our causal inference. The negative effect of green innovation on stock price synchronicity therefore remains robust.

We further evaluate omitted variable bias using the Oster bounds test [30]. The procedure yields a negative proportional selection coefficient, which arises because adding observable firm controls strengthens the baseline coefficient. Under the Oster framework, which assumes that unobservable variables share similar properties with observable ones, controlling unobservable variables would likely reinforce our result. To completely nullify the findings, a hidden confounder would need to influence synchronicity in the opposite direction of all standard firm control, a scenario that is economically implausible.

Together, our Rosenbaum bounds ($\mathsf{\Gamma }=16.9$) and Oster bounds analyses are explicitly designed to bound the influence of unobserved confounders that persist after our matching and difference-in-difference procedures, including time-varying confounders such as changes in firm-level governance practices, shifts in industry-wide investor preferences, or evolving managerial incentives. The magnitude of any such unobserved confounder required to overturn our findings is implausibly large, providing quantitative reassurance that our causal interpretation is robust to the residual endogeneity concerns that conventional matching and difference-in-difference procedures cannot fully address.

4.5. Exploring Underlying Mechanisms

Having established a robust causal relationship between green innovation and stock price informativeness, we investigate the economic channels through which this effect operates. We examine two categories of information intermediaries: financial analysts and institutional investors, both of which play important roles in collecting, processing, and incorporating firm-specific information into stock prices [4,26,28].

4.5.1. Analyst Coverage

Financial analysts function as pivotal information intermediaries who produce and disseminate firm-specific research to market participants [25,62]. Prior studies demonstrate that analyst coverage reduces information asymmetry and improves the efficiency of price discovery [63,64]. If green innovation enriches the corporate information environment, it should attract greater analyst attention, providing a direct channel through which firm-specific information reaches the market [65,66].

Column 1 of

Table 7. Economic mechanisms.

Independent Variables	Dependent Variable
	Analyst Coverage	Synchronicity	Institutional Ownership	Synchronicity	Synchronicity
	1	2	3	4	5
Green_post	0.5743 ***	−0.3779 ***	2.3114 **	−0.3756 ***	−0.3760 ***
	(0.2134)	(0.0225)	(1.1408)	0.0224)	(0.0244)
Analyst Coverage		−0.0046 ***			−0.0062 **
		(0.0133)			(0.0025)
Institutional Ownership				−0.0235 **	−0.0012 *
				(0.0113)	(0.0007)
Analyst Coverage × Institutional Ownership					0.0275 ***
					(0.0043)
Firm Size	1.6796 ***	0.2066 ***	1.3716 ***	0.2158 ***	0.2614 ***
	(0.0671)	(0.079)	(0.183)	(0.0077)	(0.0082)
Book Leverage	−0.0814	−0.1633 ***	−0.4319	−0.1636 ***	−0.2410 ***
	(0.195)	(0.02909)	(0.5469)	(0.02912)	(0.0284)
Tangibility	0.3001	0.2480 ***	−1.7295	0.2493 ***	0.3195 ***
	(0.2495)	(0.0357)	(1.219)	(0.0368)	(0.0357)
R&D Intensity	−0.0195	0.0237 **	0.1362	0.0028 **	0.0814 ***
	(0.0706)	(0.0011)	(0.1484)	(0.0012)	(0.0151)
Cash Flow	0.0816	0.0487 **	−0.2604	0.0491 **	0.0745 ***
	(0.0876)	(0.0197)	(0.2264)	(0.0197)	(0.0188)
Constant	−26.1842 ***	−5.2457 ***	−19.5766 ***	−5.3971 ***	−6.3896 ***
	(1.2737)	(0.1517)	(3.4395)	(0.1499)	(0.1553)
Observations	74,103	74,103	74,103	74,103	74,103
Year FE	Yes	Yes	Yes	Yes	Yes
Firm FE	Yes	Yes	Yes	Yes	Yes
Sobel test		p < 0.01		p < 0.01
Adjusted R-squared	0.8395	0.4836	0.7913	0.5963	0.5466

* indicates $p<0.10$, two-tailed; ** indicates $p<0.05$, two-tailed; *** indicates $p<0.01$, two-tailed.

regresses analyst coverage, measured as the number of analysts following a firm from I/B/E/S, on our $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ indicator. The coefficient is $0.574$, statistically significant at the 1% level ($\mathrm{t}=2.69$). Once a firm begins green patenting, it attracts approximately 0.57 additional analysts. This confirms that green innovation draws analyst attention, consistent with the proposition that environmentally innovation firms generate novel, value-relevant information that warrants dedicated analytical coverage [19,24].

4.5.2. Institutional Ownership

Institutional investors enhance price informativeness through two complementary roles: active monitoring of managerial behavior and informed trading based on superior information processing capabilities [26,27]. The literature documents that institutional presence is associated with greater corporate transparency and more efficient price discovery [28]. We measure institutional ownership using the average quarterly shares held by institutional investors (in tens of millions), obtained from the Thomson Reuters 13F database.

Column 3 of Table 7 reports the results. The coefficient on $\mathrm{G}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\_\mathrm{p}\mathrm{o}\mathrm{s}\mathrm{t}$ is $2.311$, statistically significant at the 5% level ($\mathrm{t}=2.03$). Firms that enter the green innovation space attract substantially more institutional capital. This finding aligns with the growing body of evidence that institutional investors increasingly integrate environmental performance into their portfolio decisions [20,21]. As these sophisticated market participants increase their positions in green innovative firms, they contribute both monitoring intensity and informed trading activity that accelerate the reflection of firm-specific information into stock prices.

Together, the mechanism test reveals that green innovation enhances stock price informativeness through two complementary channels: attracting analyst coverage that produces new firm-specific information and increasing institutional investment whose monitoring and trading activities facilitate the incorporation of that information into prices.

4.5.3. Mediation Analysis

In Column 2 of Table 7, we formally test the mediation effect of analyst coverage by including Green_post and analyst coverage as explanatory variables to predict stock price synchronicity. With the inclusion of analyst coverage, the magnitude of Green_post on stock price synchronicity decreases. Our Sobel test suggests a significant partial mediation effect of analyst coverage. In Column 4 of Table 7, we perform similar analysis to test the mediation effect of institutional ownership, and our analysis confirms a significant partial mediation effect of institutional ownership.

4.5.4. Interaction Between the Two Mechanisms

In Column 5 of Table 7, we enter an interaction term between analyst coverage and institutional ownership to test the joint effect of these two underlying mechanisms. We report a positive and significant coefficient of the interaction terms. Given that the first order effects of analyst coverage and institutional ownership are negative and significant, the positive coefficient of interaction term indicates a substitution relationship between the two mechanisms. In other words, the effect of analyst coverage (institutional ownership) on stock price synchronicity is weaker in the presence of higher stock ownership (analyst coverage). This pattern is intuitive in the sense that green innovation improves the information environment through both channels, but the two channels substitute for each other rather than operating in full addition.

5. Conclusions

This study investigates whether corporate green innovation enhances the information content of stock prices using a comprehensive sample of U.S. publicly listed firms from 1997 to 2023. We show that firms engaging in green innovation exhibit significantly lower stock price synchronicity, indicating that a greater proportion of firm-specific information is reflected in their prices. This finding is robust to alternative measures of green innovation, lead dependent variables, and the exclusion of the COVID-19 pandemic period. Propensity score matching with Rosenbaum bounds sensitivity analysis confirms that an unobserved confounder would need to increase the odds of treatment by nearly 17 times to invalidate our results, and the Oster bounds test provides further assurance against omitted variable bias. Exploiting the 2018 America’s Water Infrastructure Act as an exogenous regulatory shock in a difference-in-difference framework, we find that the informativeness effect green innovation is substantially amplified following the shock, strengthening the causal interpretation.

Citation-based analyses confirm that the quality of green innovation matters. Higher-quality green patent, properly adjusted for truncation bias, generates a substantially stronger effect on the information reflected in stock prices. This demonstrates that the market does not treat all green patents equally but rather assigns greater informational weight to innovations that receive wider recognition from subsequent inventors. The progressive strengthening of the effect across increasingly rigorous specifications reinforces the economic significance of green patenting for price informativeness.

We identify analyst coverage and institutional ownership as two economic mechanisms underlying this relationship. Green innovation firms attract significantly more analyst following and institutional capital, both of which facilitate the production and incorporation of firm-specific information into stock prices [67]. These findings are consistent with the view that green innovation enriches the corporate information environment by attracting sophisticated intermediaries who actively collect, interpret, and trade on firm-specific news.

Several boundary conditions warrant explicit discussion. Our findings come from the U.S. context, which features a transparent disclosure regime, deep analyst coverage, and a large institutional investor base. These features shape the informational environment in which we identify our effects. The magnitude of our findings may therefore not transfer directly to settings with thinner analyst coverage, weaker institutional ownership, or less stringent disclosure requirements. Three dimensions of cross-country variation are particularly relevant. First, patent-quality standards differ across jurisdictions. The USPTO is among the most rigorous patent-offices globally. Where patent grant standards are more permissive, the signaling channel developed in Section 2.4 may be weaker. Second, our identification strategy exploits a U.S.-specific regulatory shock. Analogous events elsewhere, such as the EU Taxonomy Regulation, Japan’s Green Growth Strategy, or China’s Carbon Emissions Trading Schemes, would provide complementary evidence on whether our findings generalize. Third, the relative roles of mandatory and voluntary disclosure vary across jurisdictions, which may shape how green innovation interacts with the broader information environment. We identify cross-country comparison, especially studies that exploit institutional variation across developed and emerging markets, as a high-priority direction for future research.

A second scope-related consideration concerns our use of green patents as the primary proxy for green innovation. Patents capture innovation outputs that are formally registered with a recognized authority, but they do not capture the full landscape of corporate environmental activity. Process improvements, voluntary emission reductions, internal sustainability initiatives, environmental certifications, and unpatented green R&D investment also represent meaningful forms of green innovation but fall outside our measurement window. We deliberately adopt the patent-based measure because it offers three advantages essential to our research question. First, patent applications are externally verified by the USPTO, providing a standardized and credibility-anchored signal that contrasts with self-reported environmental disclosures susceptible to greenwashing. Second, patents are publicly disclosed in a uniform format with consistent timing, enabling clean panel construction across firms and years. Third, patent-based measures align with the established green-innovation literature in finance and economics, facilitating comparison with prior studies. The complementary measurement strategies that combine patent data with non-patented innovation indicators would yield a more complete picture of corporate environmental activity. This is also a direction for future research.

An important conceptual point concerns the relationship between green patents and the broader ESG and CSR disclosure ecosystem. Green patent position serves as a complementary, not substitute, source of corporate environmental information. ESG ratings and CSR reports remain the predominant lens through which investors evaluate firm-level sustainability, but the literature documents two important limitations of these disclosures. First, ESG ratings exhibit substantial divergence across providers, reflecting differences in methodology and weighting rather than underlying firm fundamentals [59]. Second, voluntary CSR disclosures are susceptible to greenwashing and selective reporting [22]. Green patents address neither limitation directly but provide an independent informational signal anchored in external verification by the USPTO and in the underlying technological substance of the invention. They therefore complement, rather than replace, conventional ESG-based assessments by offering a hard, standardized measure of substantive environmental activity. We view the integration of patent-based measures with ESG ratings, CSR disclosures, and direct environmental performance indicators as an important agenda for future work seeking a fuller account of how the market processes firm-level environmental information [68,69].

Our findings carry several practical implications, each anchored to specific empirical evidence from our analysis. For investors and analysts, green innovation is a valuable source of firm-specific information. The baseline result shows that green patenting reduces stock price synchronicity by $0.066$, with the effect concentrated in higher-quality patents. This implies that valuations are sharper when patents weighted by citation impact rather than raw count. For corporate managers, green innovation can be used to enrich the firm’s information environment. The mechanism evidence shows that initiating green patenting attracts roughly $0.57$ additional analysts and significantly greater institutional ownership. Both channels are well-known drivers of lower information asymmetry and reduced cost of capital, making green innovation a credible strategy for improving access to external financing. For policy makers, well-designed environmental regulation can generate positive externalities for capital market efficiency [70]. The difference-in-difference result indicates that the 2018 AWIA amplified the informativeness effect of green innovation, doubling the baseline magnitude. Through our analysis of underlying mechanisms, we show that environmental regulation strengthens the role of analysts and institutional investors in disseminating firm-specific information, improving capital allocation.