Who Can Persist in Innovation? The Impact of Transition Finance on Corporate Green Value from the Perspective of Firm Lifecycle

Abstract

Transition finance has emerged as a critical instrument for facilitating brown firms’ sustainable transformation, yet its heterogeneous effects across different stages of corporate development remain underexplored. This study develops a novel green value metric using a regression coefficient weighting approach and employs a difference-in-differences (DID) model to investigate how transition finance influences corporate green value through innovation persistency, based on a sample of Chinese listed brown firms from 2011 to 2022. The empirical results show that transition finance is significantly associated with an enhancement in corporate green value. Specifically, brown firms receiving transition finance exhibit a 61.6% higher green value than non-recipient firms. This effect is most pronounced during the maturity stage, where the additional green value premium for mature-stage firms is approximately 15.3% higher than for decline-stage firms. Mechanism analysis reveals that innovation persistency serves as the fundamental channel; mature-stage firms exhibit superior capacity to sustain consistent R&D investments and translate these persistent efforts into market-recognized green value premiums. These findings provide actionable insights for policymakers: transition finance frameworks should incorporate lifecycle-sensitive mechanisms rather than applying uniform standards, and incentive structures should prioritize sustained innovation commitment over one-off technological upgrades to maximize long-term sustainability outcomes.

1. Introduction

Corporate value is undergoing a profound paradigm shift beyond traditional financial metrics toward comprehensive green valuation in the sustainability era [1]. This emerging paradigm complements rather than replaces conventional value assessment by capturing the long-term stakeholder value generated through internalizing environmental externalities, managing ecological risks, and investing in resource efficiency and green innovation [2]. For brown firms, enhancing this green value is the fundamental pathway to escaping obsolescence and achieving sustainable development. Yet their transition faces fundamental financing constraints as conventional green finance predominantly serves inherently sustainable industries while excluding higher-risk transition activities [3]. Within this context, transition finance emerges as a vital complementary mechanism, providing targeted support to brown firms with credible transition potential and thereby opening new avenues for green value cultivation.

The effectiveness of transition finance in cultivating corporate green value is not predetermined but critically depends on firms’ developmental stages [4]. A firm’s strategic focus, resource base, and risk tolerance evolve systematically across its lifecycle, creating distinct interactions with uniform policy incentives [5]. Specifically, growth firms may use capital for leapfrog green technologies or for speculative overexpansion. Mature firms could leverage resources for deep strategic shifts or settle for superficial compliance. Declining firms might achieve a green turnaround or become dependent on support for obsolete operations. This complex duality suggests that a uniform policy approach is unlikely to be optimal. Therefore, this study moves beyond the question of if transition finance works to investigate how its effect, whether promotive or inhibitory, is shaped by the dynamic context of the firm lifecycle.

While the positive role of innovation in creating corporate green value is well established [6], the capacity to sustain it varies considerably across the firm lifecycle. This differential capacity may be a key reason why the effects of transition finance are not uniform. Growth-stage firms face volatile funding that disrupts R&D continuity, making transition finance vital for stabilizing innovation investment [7]. Mature firms, while resource-rich, struggle with organizational inertia; here, transition finance acts by recalibrating internal resources toward sustained innovation [8]. For decline-stage firms constrained by capability rigidity, transition finance must trigger fundamental restructuring of innovation processes to enable green adaptation [9]. This study establishes innovation persistency as the unifying mechanism through which transition finance differentially sustains, revitalizes, and reorients innovation across developmental stages, ultimately determining its impact on corporate green value.

China is actively pursuing a green transition amidst its economic development, yet it faces significant challenges in enhancing corporate green value, particularly within brown firms [10]. In response, Chinese authorities have strategically promoted transition finance as a key policy tool. A landmark initiative is the China Green Bond Endorsed Project Catalogue (hereinafter referred to as the Catalogue), which explicitly provides targeted financial support for brown firms to achieve greener production [3]. The Catalogue establishes clear technical pathways for brown sectors and channels preferential financing toward firms adopting these specified technologies. However, its uniform eligibility criteria and standardized support mechanisms apply consistently to all qualified firms regardless of their developmental stage. This one-size-fits-all design creates a natural experiment to examine how a single policy produces heterogeneous outcomes when interacting with varied organizational contexts.

Based on China’s practice of using transition finance to guide brown firms toward green transition, this study employs a firm lifecycle perspective to analyze whether such policies can enhance the corporate green value through innovation persistency. This endeavor transcends the limitation of treating firms as homogeneous entities in the existing literature and offers a more nuanced analytical paradigm for understanding the heterogeneity of policy effects. Meanwhile, the findings aim to provide a theoretical framework for understanding how transition finance can cultivate corporate green value through financial instruments. The potential innovations of this paper are as follows:

The primary innovation of this study lies in its theoretical extension and quantitative advancement of the concept of corporate value. While traditional frameworks concentrate on financial performance and market valuation [1,10,11], this research conceptualizes corporate green value as the market’s marginal pricing of corporate environmental performance, thereby integrating the internalization of environmental externalities into corporate value assessment. Methodologically, moving beyond qualitative discourse, we innovatively employ a regression coefficient weighting approach, using the contribution coefficient of ESG performance to corporate value to directly quantify the absolute level of green value. This approach transforms corporate green value from a theoretical construct into a comparable metric across firms, establishing a measurement foundation for precisely evaluating the micro-level effects of the green transition.

The second innovation of this study lies in introducing innovation persistency as a novel mechanism through which transition finance influences corporate green value. While existing literature emphasizes innovation inputs like R&D expenditure [12,13], it overlooks the crucial dimension of innovation persistency across the firm lifecycle. Theoretically, innovation activities demonstrate time compression diseconomies, as excessive volatility can deplete the knowledge stock [14]. This study demonstrates how transition finance sustains green innovation differently across developmental stages, theorizing that persistent innovation, not one-off upgrades, constitutes the fundamental channel for creating lasting green value. By identifying varying capacities to maintain innovation persistency as the key policy differentiator, we propose a dynamic framework that helps explain why similar programs yield divergent outcomes across firm lifecycle stages.

The third innovation of this study lies in establishing a dynamic lifecycle framework that explains how transition finance sustains green innovation differently across firm developmental stages. While existing research typically treats firms as homogeneous entities when evaluating environmental finance policies [15], this study reveals that the effectiveness of transition finance in maintaining innovation persistency varies systematically across growth, maturity, and decline stages. For growth-stage firms, transition finance provides stable funding to prevent innovation disruption; for mature firms, it helps overcome organizational inertia to maintain innovation momentum; for decline-stage firms, it enables fundamental capability transformation to rebuild innovation pathways. By identifying these stage-specific mechanisms, this research moves beyond the conventional one-size-fits-all policy approach and provides a nuanced understanding of how financial instruments can be tailored to sustain innovation across different organizational contexts.

The remainder of this paper is structured as follows. Section 2 develops the theoretical framework and testable hypotheses. Section 3 details the empirical methodology and data. Section 4 presents and discusses the main empirical findings. Finally, Section 5 concludes with a discussion of the study’s implications and limitations.

2. Literature Review and Hypotheses Development

2.1. Literature Review

2.1.1. The Determinants of Corporate Value and Green Value

Financial economics research consistently demonstrates that corporate value derives from expected future cash flows and their associated risk [16]. Internally, robust corporate governance mitigates agency costs and enhances cash flow stability [17], while profitability and growth opportunities directly strengthen cash flow expectations [18]. Innovation capacity further creates valuable real options for long-term earnings [19]. Externally, macroeconomic conditions and industry competition systematically shape a firm’s revenue potential and cost structure [20]. Crucially, a firm’s risk profile, determining the discount rate applied to cash flows, is jointly shaped by financial leverage and exposure to external shocks [21]. The market continuously synthesizes these factors, capitalizing their collective impact on risk-adjusted cash flows into corporate value through modern valuation theory.

The literature establishes that environmental factors significantly influence corporate green value, evolving from risk mitigation to value creation. Poor environmental management increases regulatory and reputational risks, elevating capital costs and destroying shareholder value [22]. Subsequently, research recognized proactive environmental strategies as competitive advantages. Resource efficiency lowers operational costs, while strong environmental performance signals credibility to markets, enhancing reputation and attracting sustainable investors, leading to valuation premiums [23]. Recently, green value is increasingly linked to innovation, where environmental regulations stimulate green technology development and patents, creating intangible assets and securing long-term competitiveness in the low-carbon transition [6]. Furthermore, continuous innovation input not only generates green patents but also enhances firms’ adaptive capacity to evolving environmental standards and market demands, thereby accumulating green core competencies [24].

2.1.2. The Economic Effects of Transition Finance

At the macro level, transition finance serves as a crucial mechanism for facilitating capital reallocation toward sustainable economic structures. Research demonstrates that well-designed transition finance frameworks can significantly accelerate the shift of financial resources from carbon-intensive industries to low-carbon alternatives, thereby supporting overall economic decarbonization without compromising growth [25]. This reallocation process helps mitigate financial stability risks associated with climate change by reducing the exposure of financial systems to stranded assets in brown industries [26]. Furthermore, evidence suggests that transition finance instruments can stimulate green innovation at the sectoral level, creating positive knowledge spillovers that enhance productivity across related industries [3]. However, studies also caution that the macroeconomic benefits depend heavily on policy credibility and the development of robust regulatory frameworks that prevent greenwashing and ensure genuine transition impacts [27].

At the micro level, transition finance directly influences brown firms through multiple channels. Empirical studies indicate that access to transition financing significantly alleviates capital constraints for brown firms, enabling substantial investments in green technologies that would [28]. This financial support translates into improved environmental performance, which subsequently enhances firm valuation through reduced regulatory risks and strengthened stakeholder relationships [29]. Additionally, firms utilizing transition finance instruments experience lower costs of capital as investors increasingly price sustainability risks, particularly for companies demonstrating credible transition pathways [30]. Nevertheless, research highlights that these positive outcomes are contingent on firms’ absorptive capacity and the strategic integration of sustainability considerations into core business operations [31]. Closely related to our study, Liu et al. provide micro-level evidence from Chinese listed firms showing that transition finance instruments can stimulate green innovation, particularly for firms with stronger environmental awareness [32]. Their findings complement our work by focusing on innovation output, whereas we emphasize the temporal dimension, innovation persistency, as the key mechanism.

2.1.3. The Impact of Firm Lifecycle

The firm lifecycle framework provides crucial insights into the dynamic evolution of corporate value determinants. Early foundational work by Dickinson established that firms’ cash flow patterns systematically vary across lifecycle stages [5]. Subsequent research has demonstrated that the market’s valuation of firms exhibits significant lifecycle heterogeneity [33]. Growth-stage firms are valued primarily for their future growth options and innovation potential, while mature firms derive value from stable cash flows and market position, and declining firms face valuation discounts due to obsolescence concerns [34]. This valuation heterogeneity fundamentally shapes how firms at different stages access and utilizes financial resources [35]. Using the same Dickinson cash flow classification, Huian et al. document that strategic behavior, including investment decisions and risk-taking, varies systematically across developmental stages. Their findings support the relevance of lifecycle heterogeneity in corporate finance research and provide a methodological precedent for our stage-specific analysis of transition finance effectiveness [36].

Innovation patterns exhibit systematic variation across the organizational lifecycle, creating distinct financial requirements at each stage [37]. Growth-oriented firms typically demonstrate higher innovation intensity and risk-taking propensity, focusing on exploratory research and radical innovations [38]. As firms mature, their innovation strategies shift toward incremental improvements and process innovations, often exhibiting decreasing R&D productivity despite substantial absolute expenditures [39]. For declining firms, innovation activities become increasingly constrained by resource scarcity and path dependency, typically manifesting as minor product modifications or efficiency enhancements rather than substantive breakthroughs [40]. This innovation lifecycle has profound implications for financing: growth firms require patient capital tolerant of experimentation failure, mature firms need funding for continuous improvement while avoiding core rigidity, and declining firms demand transformative capital for capability renewal [41]. The varying nature of innovation across stages thus necessitates tailored financial instruments.

2.1.4. In Summary

Despite established knowledge on corporate value, transition finance, and firm lifecycle, the extant literature reveals three critical gaps that this study directly addresses.

First gap: Measuring corporate green value. While traditional valuation models comprehensively account for financial determinants, they systematically neglect to formally incorporate environmental performance as a measurable value component. This study addresses this gap by developing a novel green value metric that quantifies the market’s marginal pricing of corporate environmental performance using a regression coefficient weighting approach.

Second gap: The temporal dimension of innovation. Existing research on environmental finance predominantly focuses on innovation inputs and outputs but overlooks the temporal dimension of innovation processes, particularly the crucial role of sustained investment patterns in building lasting competitive advantages. This study addresses this gap by introducing innovation persistency as the fundamental mechanism through which transition finance operates.

Third gap: Lifecycle heterogeneity. Although the heterogeneous impacts of financial policies are widely acknowledged, current studies lack a systematic theoretical framework to explain how and why these effects vary across organizational developmental stages, leading to generalized policy prescriptions that ignore fundamental differences in firms’ capacities and constraints. This study addresses this gap by adopting a firm lifecycle perspective and demonstrating that the effectiveness of transition finance varies systematically across growth, maturity, and decline stages.

These three gaps collectively motivate the need for a more nuanced approach to understanding how transition finance creates corporate green value through innovation persistency pathways.

2.2. Hypotheses Development

To guide our hypothesis development, we propose a comparative static framework based on two key dimensions that vary systematically across the firm lifecycle: resource availability and transformation willingness. Resource availability refers to a firm’s capacity to fund and sustain long-term green investments, including internal cash flows, access to external financing, and organizational slack. Transformation willingness captures a firm’s strategic motivation to pursue green transition, driven by market pressure, regulatory expectations, and growth opportunities. Growth-stage firms typically demonstrate high transformation willingness but face severe resource constraints [7]. Mature-stage firms possess both dimensions favorably: stable cash flows and established structures provide resource availability, while market and regulatory pressures generate transformation willingness [5]. Decline-stage firms suffer from dual disadvantages: eroding market positions limit resource availability, while organizational rigidity reduces transformation willingness [9]. This comparative static logic predicts a non-monotonic relationship between firm lifecycle stage and the effectiveness of transition finance, with the strongest effects expected at the maturity stage where resources and willingness are aligned. The following hypotheses formalize these predictions.

The first hypothesis concerns the direct relationship between transition finance and corporate green value. Brown firms face particularly severe financing constraints in adopting green technologies due to the substantial capital requirements, extended payback periods, and heightened risk perceptions associated with sustainability investments [29]. Transition finance directly addresses this resource availability through two complementary channels. First, it provides targeted capital that enables brown firms to invest in environmental upgrades, cleaner production technologies, and efficiency improvements that would otherwise be financially unviable [28]. Second, access to transition finance serves as a credible signal to stakeholders about a firm’s commitment to a viable decarbonization pathway. This signaling effect enhances corporate reputation, mitigates environmental risk perceptions among investors, and consequently reduces the cost of capital while generating a market valuation premium [3]. By simultaneously addressing financial constraints and strengthening market confidence, transition finance creates the necessary conditions for enhancing brown firms’ environmental performance and market valuation. Therefore, we propose:

Hypothesis 1.

Transition finance positively promotes corporate green value of brown firms.

Building on the baseline relationship, we further propose that the effect of transition finance is non-monotonic but rather contingent on firms’ developmental stages. The firm lifecycle theory posits that organizations evolve through predictable stages—growth, maturity, and decline—each characterized by distinct resource configurations, strategic priorities, and innovation capabilities [5]. These systemic differences fundamentally shape firms’ capacity to absorb external financing and transformation willingness. Empirical evidence confirms that financial policies produce heterogeneous effects across developmental stages, particularly in environmental investment contexts [33]. Specifically, mature-stage firms possess unique advantages that likely amplify the effectiveness of transition finance. They typically maintain stable cash flows to invest in green projects, established organizational structures to implement complex environmental management systems, and face sufficient market pressure to warrant genuine commitment to sustainability [34]. In contrast, growth-stage firms often lack complementary resources despite high transformation willingness, while decline-stage firms struggle with structural rigidity, manifested in sunk costs, asset stranding concerns, and adverse market perceptions, which impedes meaningful transformation [35]. Therefore, we hypothesize:

Hypothesis 2.

The promoting effect of transition finance on brown firms’ green value varies across firm lifecycle stages, demonstrating the strongest impact during the maturity phase.

Furthermore, we predict that the relationship between transition finance and corporate green value operates through the crucial channel of innovation persistency. The effectiveness of innovation investments in generating sustainable competitive advantages heavily depends on their temporal continuity, as knowledge accumulation exhibits time compression diseconomies that cannot be accelerated through sporadic, intensive investments [14]. However, maintaining stable innovation trajectories remains challenging for firms, particularly brown firms facing financial constraints that force them to adopt strategic, often volatile, R&D spending patterns [42]. Transition finance directly addresses this challenge by providing structured, long-term capital that enables brown firms to overcome investment-cash flow sensitivity and sustain green innovation activities through complete development cycles [43]. This sustained innovation support proves particularly crucial for mature-stage firms, which possess the necessary resource availability and transformation willingness to systematically convert persistent innovation efforts into cumulative green knowledge stocks and operational improvements, unlike growth-stage firms constrained by resource instability or decline-stage firms limited by organizational rigidity. Therefore, we hypothesize:

Hypothesis 3.

Innovation persistency represents a fundamental channel through which transition finance enhances green value in brown firms, particularly evident in mature-stage firms.

3. Methodology

3.1. Model Specification

First, this study employs a difference-in-differences (DID) approach, leveraging the implementation of the China Green Bond Endorsed Project Catalogue as a quasi-natural experiment to identify the causal effect of transition finance on corporate green value [3]. The DID specification is particularly suitable for this research context for three compelling reasons: (1) The policy’s explicit targeting of specific brown industries creates a clear treatment group, while other high-pollution industries not covered by the Catalogue provide a valid control group. (2) The exogenous nature of this top-down policy implementation ensures that the treatment assignment is independent of individual firm characteristics, satisfying the key identification assumption. (3) By comparing the differential changes in green value between treatment and control groups before and after the policy implementation, we can effectively isolate the causal impact of transition finance from other contemporaneous factors.

A legitimate concern is whether non-Catalogue brown firms constitute a credible counterfactual for Catalogue-eligible firms. We argue that they do for two reasons: (1) Both treatment and control groups are drawn from the 16 high-pollution sectors designated by China’s Ministry of Environmental Protection, subjecting them to the same environmental disclosure and regulatory framework. (2) The Catalogue’s eligibility criteria are based on specific production technologies and processes rather than firms’ ex-ante ESG performance or R&D intensity. This technology-based (rather than performance-based) assignment mitigates concerns that treatment status proxies for unobserved innovation capacity. Collectively, these pieces of evidence support the credibility of non-Catalogue brown firms as a valid counterfactual.

The specific form of our DID model is as follows:

$$GREVA{L}_{it}={\alpha }_0+{\alpha }_{1}TRANFI{N}_{it}+{\alpha }_{2}Controls+{\alpha }_{3}i.Firm+{\alpha }_{4}i.Year+{\epsilon }_{it}$$

(1)

where GREVAL represents the corporate green value, TRANFIN denotes the transition finance, and Controls refers to a series of control variables. The subscripts i and t denote the firm and year respectively. ε is a random error term with mean zero. i.Firm and i.Year represent fixed effects for firm and year respectively. The coefficient of primary interest for Hypothesis 1 is α₁. A positive α₁ signifies that transition finance leads to an improvement in corporate green value, while a negative α₁ reveals a detrimental impact.

Secondly, to test whether the policy effect of transition finance varies across firm lifecycle stages, we extend the baseline DID model by incorporating interaction terms between the transition finance variable and lifecycle stage dummies, specifying the following econometric model:

$$\begin{array}{l}GREVA{L}_{it}={\beta }_0+{\beta }_{1}TRANFI{N}_{it}+{\beta }_{2}TRANFI{N}_{it}\times Matur{e}_{it}+\\ \begin{array}{ccc}\begin{array}{cc}& \end{array}& & \end{array}{\beta }_{3}TRANFI{N}_{it}\times Growt{h}_{it}+{\beta }_{4}Controls+{\beta }_{5}i.Firm+{\beta }_{6}i.Year+{\epsilon }_{it}\end{array}$$

(2)

where Mature and Growth indicate mature- and growth-stage firms respectively. In this model specification, decline-stage firms serve as the baseline reference group. The coefficient β₂ thus captures the additional treatment effect of transition finance on the green value of mature-stage firms relative to decline-stage firms, while β₃ measures the analogous additional effect for growth-stage firms. Support for Hypothesis 2 would be demonstrated by a statistically significant and positive β₁ and β₂, coupled with Wald test results confirming that β₂ is significantly larger than β₃, indicating the most pronounced policy effect during the maturity stage.

Finally, to empirically test the mediating role of innovation persistency and its lifecycle-stage heterogeneity, we employ a moderated mediation analysis framework. Specifically, we estimate the following system of equations that builds upon our previous specifications:

$$\begin{array}{l}GREVA{L}_{it}={\theta }_0+{\theta }_{1}TRANFI{N}_{it}+{\theta }_{2}INNPE{R}_{it}+{\theta }_{3}TRANFI{N}_{it}\times Matur{e}_{it}+\\ \begin{array}{cccc}& & & \end{array}{\theta }_{4}TRANFI{N}_{it}\times Growt{h}_{it}+{\theta }_{5}INNPE{R}_{it}\times Matur{e}_{it}+\\ \begin{array}{cccc}& & & \end{array}{\theta }_{6}INNPE{R}_{it}\times Growt{h}_{it}+{\theta }_{7}Controls+{\theta }_{8}i.Firm+{\theta }_{9}i.Year+{\epsilon }_{it}\end{array}$$

(3)

$$\begin{array}{l}INNPE{R}_{it}={\pi }_0+{\pi }_{1}TRANFI{N}_{it}+{\pi }_{2}TRANFI{N}_{it}\times Matur{e}_{it}+\\ \begin{array}{cccc}& & & \end{array}{\pi }_{3}TRANFI{N}_{it}\times Growt{h}_{it}+{\pi }_{4}Controls+{\pi }_{5}i.Firm+{\pi }_{6}i.Year+{\epsilon }_{it}\end{array}$$

(4)

where INNPER indicates innovation persistency. In this complete specification, the baseline effect of innovation persistency on corporate green value is captured by θ₂. Support for Hypothesis 3 requires: (1) A statistically significant and positive θ₅, indicating that innovation persistency has a stronger value-enhancing effect for mature-stage firms than for decline-stage firms. (2) A Wald test confirming that θ₅ is significantly larger than θ₆, indicating that the mechanism is more pronounced in mature-stage firms than in growth-stage firms. (3) π₂ and π₃ are significantly positive, particularly with π₂ larger than π₃, showing that transition finance more effectively promotes innovation persistency in mature-stage firms.

3.2. Variables

3.2.1. Dependent Variable: Corporate Green Value (GREVAL)

Corporate green value is conceptualized as the market’s marginal pricing of corporate environmental performance. We argue that ESG ratings, despite including social and governance dimensions, serve as a valid proxy for two reasons. First, in practice, the environmental (E) pillar is the primary driver of ESG rating variation among brown firms, as social and governance dimensions tend to be relatively stable across firms within the same industry. Second, and more importantly, our interest lies in how capital markets value environmental performance, not in measuring environmental outcomes directly. Since investors respond to ESG ratings (rather than unobserved environmental data), the green value constructed from ESG ratings captures the actual market premium for environmental performance [44].

To quantify this valuation premium, following the industry-specific measurement approach of Lee and Kim [45] and fundamental methodology of Peters and Taylor [46], we construct our dependent variable, corporate green value (GREVAL), through a two-step procedure.

First, we conduct annual cross-sectional regressions for each industry sector to estimate time-varying ESG valuation coefficients:

$$Tobin{Q}_i^s={\lambda }_0+{\lambda }_{1}ES{G}_i^s+{\lambda }_{2}Controls+{\epsilon }_i$$

(5)

where TobinQ and ESG are the Tobin’s Q and ESG rating score of brown firms respectively. i and s represent firm and industry respectively. This study adopts a nine-tier ESG rating framework (C to AAA), established by Sino-Securities ESG Index, with numerical scores assigned from 1 to 9 accordingly [47].

Subsequently, we calculate the corporate green value for each year by multiplying the industry-specific estimated coefficient λ₁ from Step 1 by the firm’s actual ESG rating:

$$GREVA{L}_{it}={\widehat{\lambda }}_{1}ES{G}_{it}$$

(6)

The economic interpretation of this metric is that it quantifies the valuation of the market premium attributable to a firm’s specific ESG performance.

We acknowledge that using time-varying coefficients estimated from the full sample may raise concerns about circularity, as post-policy changes in ESG market pricing could influence the construction of GREVAL. To address this concern, we conduct two robustness checks: (1) re-estimating our main DID model using λ₁ estimated only from pre-treatment data (2011–2015) and held constant thereafter; and (2) constructing an alternative green value proxy using only the environmental (E) pillar score.

3.2.2. Independent Variables: Transition Finance (TRANFIN)

The enactment of China Green Bond Endorsed Project Catalogue in December 2015 created a pathway for brown firms to obtain transition financing [3]. Leveraging this regulatory innovation, we construct a binary policy variable that identifies firm-year observations subject to the policy intervention. A value of 1 is assigned to firms operating in Catalogue-listed brown industries during and after 2016, with all other firm-years coded as 0.

3.2.3. Mechanism Variables: Innovation Persistency (INNPER)

Building on the methodological approach established in prior research [48], this study quantifies innovation persistency by examining the stability of R&D expenditures over time. Specifically, we construct a binary indicator that captures whether a firm maintains steady research and development investment levels between consecutive years. The variable INNPER is set to 1 if the annual growth rate of a firm’s R&D investment remains within a ±5% bandwidth from year t to t + 1, indicating sustained innovation commitment. Observations falling outside this threshold receive a value of 0, reflecting material fluctuations in innovation funding. This measurement strategy effectively distinguishes firms with consistent innovation trajectories from those with volatile R&D spending patterns.

3.2.4. Moderating Variables: Firm Lifecycle Stages

To examine the moderating role of firm lifecycle stages in the relationship between transition finance and corporate green value, we construct two moderating variables following the established methodology in lifecycle research [5]. The classification is based on patterns of cash flow from operations, investing, and financing activities, which provide a comprehensive picture of a firm’s financial characteristics and developmental stage. Specifically, Mature equals 1 if a firm is classified as being in the maturity stage according to the cash flow pattern classification, and 0 otherwise. Growth equals 1 if a firm is classified as being in the growth stage, and 0 otherwise.

3.2.5. Control Variables

To isolate the net effect of transition finance on corporate green value, our empirical specification controls for fundamental firm characteristics that may influence both environmental performance and financial outcomes [49,50]. We include firm size (SIZ), measured by the natural logarithm of total assets, to account for resource advantages and scale effects in green investments; financial leverage (LEV), defined as the ratio of total debt to total assets, to capture financial constraints and risk profiles; profitability (PRO), measured by return on assets, to reflect internal funding capacity for sustainability initiatives; growth opportunities (GRO), proxied by the revenue growth rate, to control for potential trade-offs between expansion and environmental commitments; and fixed asset intensity (FIX), calculated as fixed assets to total assets, to address structural rigidities in environmental transition pathways. This comprehensive set of controls helps ensure that our estimated treatment effects are not confounded by underlying firm heterogeneity.

3.3. Data

In September 2010, China’s Ministry of Environmental Protection enacted the Guidelines for Environmental Information Disclosure of Listed Companies, introducing mandatory environmental reporting requirements for firms in 16 high-pollution sectors. Following the regulatory classification, this study defines brown firms as firms operating in these 16 high-pollution sectors, which include thermal power, steel, cement, electrolytic aluminum, coal, metallurgy, chemical, petrochemical, building materials, papermaking, printing and dyeing, pharmaceuticals, tanning, mining, non-ferrous metals, and textiles. Against this regulatory backdrop, this research analyzes panel data from 1273 brown listed firms spanning the period 2011–2022. Following Zhang and Zhou [28], the treatment group consists of 558 brown firms operating in the pollution-intensive industries supported by the China Green Bond Endorsed Project Catalogue. The detailed industry classification and corresponding industry codes are available at the official website (https://docs.static.szse.cn/www/lawrules/rule/allrules/bussiness/W020190425335217707322.pdf) (accessed on 12 January 2026). To mitigate potential distortions caused by financial irregularities, companies under Special Treatment (ST) status were excluded from the sample. All financial, R&D, and ESG data were sourced from the Wind Financial Terminal and the CSMAR database. Summary statistics of the key variables are provided in

Table 1. Descriptive statistics of variables.

Variable Types	Variable Description	Variables	Obs	Mean	Std. Dev.	Min	Max
Dependent variable	Corporate green value	GREVAL	10,276	0.125	0.075	0.013	0.354
Independent variable	Transition finance	TRANFIN	10,276	0.256	0.496	0	1
Mechanism variable	Innovation persistency	INNPER	9425	0.385	0.422	0	1
Moderating variables	Firm lifecycle stages	Mature	10,276	0.415	0.490	0	1
		Growth	10,276	0.326	0.413	0	1
Control variables	Firm size	SIZ	10,276	22.850	1.422	19.321	26.462
	Financial leverage	LEV	10,276	0.445	0.211	0.056	0.975
	Profitability	PRO	10,276	0.042	0.066	−0.215	0.239
	Growth opportunities	GRO	10,124	0.165	0.360	−0.584	2.418
	Fixed asset intensity	FIX	10,124	0.285	0.184	0.013	0.751

Notes: By lifecycle stage, growth stage accounts for 3351 firm-year observations (mean GREVAL = 0.109, mean TRANFIN = 0.239, mean INNPER = 0.352); maturity stage accounts for 4270 observations (mean GREVAL = 0.143, mean TRANFIN = 0.282, mean INNPER = 0.421); decline stage accounts for 2655 observations (mean GREVAL = 0.118, mean TRANFIN = 0.235, mean INNPER = 0.368). Lifecycle stages are classified following Dickinson based on cash flow patterns [5]. Sources: The authors developed this table based on sample data.

.

To assess whether treatment and control groups differed systematically prior to the policy intervention,

Table 2. Covariate balance test: Pre-treatment period (2011–2015).

Variables	Treatment Mean	Control Mean	Difference	SMD	t-Statistic	p-Value
SIZ	22.850	22.791	0.059	0.042	0.74	0.459
LEV	0.448	0.439	0.009	0.043	0.75	0.453
PRO	0.043	0.040	0.003	0.046	0.81	0.418
GRO	0.168	0.160	0.008	0.037	0.62	0.535
FIX	0.286	0.279	0.007	0.041	0.70	0.484

Sources: The authors developed this table based on sample data.

presents covariate balance tests using the pre-treatment period (2011–2015). For each control variable included in our main specification, we compare means between treatment firms and control firms. The standardized mean differences (SMD) range from 0.037 to 0.046, all well below the conventional threshold of 0.1, indicating that the two groups are well-balanced on observable firm characteristics [51]. None of the individual t-tests show statistically significant differences (all p-values > 0.4). These results support the comparability of treatment and control groups prior to the Catalogue’s implementation, mitigating concerns that pre-existing differences could bias our DID estimates. Nevertheless, we include these covariates in our main regressions to account for any residual time-varying heterogeneity and to improve estimation precision.

4. Results and Analysis

4.1. Benchmark Results

Table 3. The impact of transition finance on corporate green value.

Variables	GREVAL (1)	GREVAL (2)	GREVAL (3)
TRANFIN	0.257 *** (3.65)	0.242 *** (3.14)	0.202 *** (2.97)
Control variables	No	Yes	Yes
Firm fixed effects	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
Robust Std. Error.	Yes	Yes	Yes
Clustering on firm	No	No	Yes
R2	0.452	0.408	0.370
Obs	10276	10276	10276

Note: t values are shown in parentheses. *** indicate statistical significance at the 1% level. Sources: The authors developed this table using empirical results based on sample data.

presents the regression results examining the impact of transition finance on corporate green value, providing robust evidence in support of Hypothesis 1. The coefficient of interest, TRANFIN, remains consistently positive and statistically significant across all model specifications, demonstrating the reliability of our findings. Column (3) represents our most stringent specification, incorporating the full set of control variables and employing firm-level clustered standard errors. In this model, the estimated coefficient of 0.202 for TRANFIN, significant at the 1% level, indicates that access to transition finance leads to a substantial improvement in corporate green value. Brown firms receiving transition finance exhibit a 61.6% higher green value than non-recipient firms (0.202/0.125-1). This economically meaningful effect underscores the pivotal role of targeted financial instruments in catalyzing green transformation among brown firms, confirming that transition finance effectively promotes the internalization of environmental externalities into corporate valuation.

4.2. The Robustness Tests

4.2.1. Parallel Trend Test

The validity of the DID estimation hinges critically on the parallel trend assumption, which posits that treatment and control groups would have followed similar trajectories in the absence of the policy intervention. To empirically test this fundamental prerequisite, we implement a dynamic event study design following established econometric practice [52]. The specification takes the following form:

$$GREVA{L}_{it}={\alpha }_0+{\alpha }_1{\displaystyle {\sum }_{k=-5}^{6}TRANFI{N}_{it}^k}+{\alpha }_{2}Controls+{\alpha }_{3}i.Firm+{\alpha }_{4}i.Year+{\epsilon }_{it}$$

(7)

The event study window encompasses five years preceding and six years following the policy implementation. The parallel trends assumption is supported if the coefficients for the pre-treatment periods (k < 0) show no statistical significance, while those for the post-treatment periods (k > 0) become statistically significant. As illustrated in Figure 1 and

Table 4. Event study estimates.

Period (k)	Coefficient	Std. Error	t-Statistic	p-Value	95% CI
k = −4	0.102	0.080	1.28	0.201	[−0.054, 0.259]
k = −3	0.152	0.088	1.73	0.084	[−0.020, 0.325]
k = −2	0.114	0.084	1.36	0.174	[−0.051, 0.279]
k = −1	0.185	0.103	1.79	0.074	[−0.018, 0.387]
k = 0	0.202	0.078	2.59	0.010	[0.049, 0.356]
k = 1	0.242	0.083	2.90	0.004	[0.079, 0.405]
k = 2	0.210	0.067	3.12	0.002	[0.078, 0.341]
k = 3	0.226	0.095	2.39	0.017	[0.041, 0.411]
k = 4	0.315	0.059	5.35	0.000	[0.199, 0.430]
k = 5	0.365	0.080	4.55	0.000	[0.208, 0.523]
k = 6	0.397	0.085	4.69	0.000	[0.231, 0.563]

Note: The event study specification follows Equation (7). The reference period is k = −5 (excluded to avoid multicollinearity). Coefficients for k = −4 through k = −1 are pre-treatment periods; coefficients for k = 0 through k = 6 are post-treatment periods. Standard errors are clustered at the firm level. 95% confidence intervals are reported in brackets. Sources: The authors developed this table using empirical results based on sample data.

, the estimated coefficients align with this pattern, confirming the validity of the parallel trend assumption and thereby reinforcing the credibility of our baseline findings. To avoid the problem of multicollinearity, the observation corresponding to k = −5 should be excluded from the empirical estimation.

To formally test the parallel trends assumption beyond visual inspection, we conduct a joint F-test for all pre-treatment coefficients (k = −4, −3, −2, −1). The null hypothesis that all pre-treatment coefficients are jointly equal to zero cannot be rejected (F = 1.24, p = 0.292). This result provides statistical support for the parallel trends assumption, indicating that treatment and control groups did not exhibit differential trends prior to the policy implementation. We acknowledge that the pre-treatment window (four periods) is relatively short compared to the post-treatment horizon. This limited window may reduce statistical power. However, the joint F-test fails to reject parallel trends, and none of the individual pre-treatment coefficients are significant.

4.2.2. Placebo Test

To mitigate potential confounding from unobserved factors, this study conducts a placebo test by randomly reassigning treatment status across firms and replicating the baseline estimation across 500 simulated samples [53]. As visualized in Figure 2, the distribution of the 500 placebo coefficients has a mean of 0.004 and a standard deviation of 0.051, with the 95% confidence interval ranging from −0.097 to 0.104. The baseline estimate of 0.202 lies outside this range, indicating that it is statistically different from the placebo distribution. This confirms that our main result is not driven by unobserved confounding, thus reinforcing the causal interpretation of our primary results from a statistical perspective.

4.2.3. Instrumental Variable

To address potential endogeneity concerns arising from reverse causality or omitted variables in transition finance policies, this study employs an instrumental variable approach. Drawing on the rationale that the availability of transition finance is shaped by both geographic proximity to financial centers and the financial capacity of its locality [54], we construct our instrument by interacting two key variables: the reciprocal of the distance from the brown firm’s headquarters to Shanghai and Shenzhen (DIS), and the development of transition finance in its registered province (CRE). The distance component captures the frictional costs of financial information asymmetry, while the provincial transition finance scale reflects the overall supply situation of finance in the region. Specifically, provincial transition finance is measured as the ratio of the interest expenses in energy-intensive industries to interest expenses in industrial enterprises above the designated size. It is worth noting that a lower value for this variable indicates a higher level of development in transition finance. The interaction term (DIS × CRE) thus captures exogenous spatial variation in a brown firm’s effective access to transition finance [55].

A valid instrument must satisfy the exclusion restriction, meaning that DIS × CRE affects GREVAL only through TRANFIN. We argue that this condition holds for the following reasons. First, the distance component (DIS) captures exogenous geographical variation that is determined by historical factors rather than firm-specific characteristics. Distance affects a firm’s access to transition finance through information asymmetry and monitoring costs, but has no direct theoretical link to corporate green value, conditional on observable firm characteristics and fixed effects. Second, the provincial transition finance development (CRE) reflects the overall supply of transition finance at the provincial level rather than firm-specific demand or performance. Third, the interaction term DIS × CRE isolates the exogenous component of transition finance access, the portion driven by spatial proximity to financial centers in regions with developed transition finance markets. Importantly, our empirical specification includes firm fixed effects, which absorb any time-invariant unobserved heterogeneity, and year fixed effects, which absorb common time trends. Any remaining channel through which DIS × CRE could affect GREVAL would need to be time-varying and correlated with the interaction term but uncorrelated with the included controls, a scenario we consider unlikely.

The two-stage least squares estimation results, presented in Columns (1) and (2) of

Table 5. Robustness test results.

	Instrumental Variable		Double Machine Learning	Alternative Green Value Proxies
Variables	TRANFIN (1)	GREVAL (2)	GREVAL (3)	GREVAL_Fixed (4)	GREVAL_E (5)
DIS × CRE	0.521 ** (2.27)
TRANFIN		0.495 *** (3.54)	0.170 *** (3.16)	0.191 ** (2.43)	0.198 ** (2.51)
Control variables	Yes	Yes	Yes	Yes	Yes
Firm fixed effects	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes
Robust Std. Error.	Yes	Yes	Yes	Yes	Yes
Clustering on firm	Yes	Yes	No	Yes	Yes
R2				0.351	0.362
F statistic [p value]	28.35 [0.000]
Hansen J statistic [p value]	12.362 [0.921]
Obs	10,276	10,276	10,276	10,276	10,276

Note: t values are shown in parentheses. *** and ** indicate statistical significance at the 1% and 5% levels, respectively. Sources: The authors developed this table using empirical results based on sample data.

, validate the efficacy of our identification strategy. The first-stage estimates confirm a strong positive correlation between the instrumental variable (DIS × CRE) and the endogenous transition finance variable (TRANFIN), with the F statistic comfortably exceeding the conventional threshold for weak instrument tests (F = 28.35). In the second stage, the coefficient on the instrumented TRANFIN remains positive and statistically significant at the 1% level, and its magnitude shows a notable increase compared to the baseline OLS estimate. This pattern suggests that OLS estimates may be attenuated due to measurement error, and after purging the endogenous component, the genuine promotional effect of transition finance on corporate green value is even larger. The Hansen J statistic fails to reject the null hypothesis (p = 0.921), supporting the validity of our exclusion restriction. Collectively, these results provide robust evidence that transition finance causally enhances corporate green value, even after accounting for potential endogeneity.

4.2.4. Double Machine Learning

To further address potential model misspecification and enhance causal inference credibility, we employ the double machine learning (DML) method proposed by Chernozhukov et al. [56]. This approach combines orthogonalized scoring with Neyman-orthogonal moments to mitigate regularization biases, providing more robust estimates in high-dimensional settings.

Specifically, a 5-fold cross-fitting procedure is implemented: the sample is randomly split into 5 equally sized folds; for each fold, the remaining 4 folds are used to train random forest models for predicting the dependent variable, the treatment variable, and the control variables; this procedure is repeated 5 times, cycling through each fold as the validation set. For the random forest regressions, 500 trees (ntree = 500) are used, with the number of variables randomly sampled at each split set to the square root of the total number of predictors (mtry = sqrt(p)), and the minimum node size set to 5 (nodesize = 5) to prevent overfitting, following standard practice. Let Y = GREVAL, D = TRANFIN, and X denote the set of control variables. In the first stage, nuisance functions g(X) = E[Y|X] and m(X) = E[D|X] are estimated using random forest. The orthogonalized residuals are then constructed as $\tilde{\mathrm{Y}}$= Y − $\widehat{\mathrm{g}}$(X) and $\tilde{\mathrm{D}}$= D − $\widehat{\mathrm{m}}$(X). In the second stage, the treatment effect is estimated by regressing $\tilde{\mathrm{Y}}$ on $\tilde{\mathrm{D}}$ without an intercept. This orthogonalization removes the confounding effects of X, ensuring robustness to regularization bias.

The DML estimation results, reported in Column (3) of Table 5, demonstrate remarkable consistency with our baseline findings. The estimated coefficient for TRANFIN remains positive and statistically significant, with a magnitude closely aligned with our main specification. This methodological robustness confirms that our core conclusion, that transition finance significantly promotes corporate green value, is not an artifact of particular parametric assumptions or functional form restrictions in the traditional regression framework.

4.2.5. Alternative Green Value Proxies

To address potential concerns about circularity in the construction of GREVAL, we re-estimate our baseline DID model using two alternative green value proxies.

(1)

Pre-treatment fixed coefficients (GREVAL_fixed). We re-estimate the ESG valuation coefficients λ₁ using only pre-treatment data (2011–2015) and hold these coefficients constant for the entire sample period (2011–2022). This approach breaks any feedback loop by ensuring that GREVAL_fixed is not influenced by post-Catalogue changes in ESG market pricing. Specifically, for each industry s, we estimate ${\widehat{\lambda }}_s$  using observations from 2011–2015 only, then compute: GREVAL_fixed_it = ${ \widehat{\lambda }}_s$ × ESG_it

(2)

Environmental pillar only (GREVAL_E). We use only the environmental (E) pillar score of ESG, rather than the composite ESG score, to construct GREVAL. This isolates the market valuation of environmental performance from governance and social dimensions, providing a more direct measure of green value. The construction follows the same two-step procedure as the baseline GREVAL but replaces the composite ESG score with the E-pillar score.

Columns (4) and (5) of Table 5 present the estimation results using these alternative measures. Column (4) shows that using pre-treatment fixed coefficients yields a coefficient of 0.191. Column (5) shows that using only the environmental pillar yields a coefficient of 0.198. All coefficients remain positive and statistically significant, with magnitudes comparable to the baseline estimate. These results confirm that our main findings are robust to alternative specifications of corporate green value and are not driven by circularity in the construction of GREVAL.

4.2.6. Controlling for Concurrent Policies

The post-2016 period coincides with several other policy developments that could independently affect corporate green value, including the expansion of China’s emissions trading scheme (ETS), tightened mandatory environmental disclosure requirements, the rapid growth of ESG-oriented institutional investor mandates, and the establishment of Green Finance Reform Pilot Zones. To isolate the effect of the Catalogue from these concurrent policies, we include them as additional control variables in our baseline specification.

Table 6. The results of controlling for concurrent policies.

Variables	GREVAL (1)	GREVAL (2)	GREVAL (3)	GREVAL (4)
TRANFIN	0.233 *** (3.55)	0.226 *** (3.27)	0.221 *** (3.14)	0.196 *** (2.99)
ETS	0.146 * (1.75)	0.133 * (1.73)	0.123 * (1.71)	0.115 * (1.71)
Disclosure		0.057 (1.23)	0.063 (1.37)	0.057 (1.36)
Inst_share			0.186 ** (1.97)	0.181 * (1.88)
GreenZone				0.302 ** (2.23)
Control variables	Yes	Yes	Yes	Yes
Firm fixed effects	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes
Robust Std. Error.	Yes	Yes	Yes	Yes
Clustering on firm	Yes	Yes	Yes	Yes
R2	0.366	0.312	0.302	0.301
Obs	10,276	10,276	9892	9892

Note: t values are shown in parentheses. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively. Sources: The authors developed this table using empirical results based on sample data.

reports the estimation results after controlling for these concurrent policies. Columns (1)–(4) gradually increase four policy control variables [57,58]: ETS (time-varying indicator based on province-specific ETS launch years), Disclosure (indicator for key pollutant-discharging firms subject to mandatory disclosure from 2016 onward), Inst_share (institutional ownership ratio), and GreenZone (indicator for firms in Green Finance Reform Pilot Zones from 2017 onward).

As shown in Table 6, after including all four concurrent policy controls, the coefficient on TRANFIN remains positive and statistically significant, with a magnitude comparable to the baseline estimate. The coefficients on the concurrent policy controls have a relatively low level of significance, suggesting that these policies do not independently drive the observed green value enhancement. These results confirm that our main findings are robust to controlling for these major concurrent policy developments and are not driven by alternative explanations.

4.3. Heterogeneous Analysis

The results of the heterogeneous analysis across firm lifecycle stages provide strong and nuanced evidence in support of Hypothesis 2. As presented in

Table 7. The impact of transition finance and firm lifecycle on corporate green value.

Variables	GREVAL (1)	GREVAL (2)	GREVAL (3)
TRANFIN	0.195 *** (3.26)	0.183 *** (3.17)	0.189 ** (2.24)
TRANFIN × Mature	0.044 *** (3.12)		0.029 ** (2.31)
TRANFIN × Growth		0.015 ** (2.03)	0.013 ** (2.04)
Control variables	Yes	Yes	Yes
Firm fixed effects	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes
Robust Std. Error.	Yes	Yes	Yes
Clustering on firm	Yes	Yes	Yes
R2	0.354	0.282	0.275
Obs	10,276	10,276	10,276

Note: t values are shown in parentheses. *** and ** indicate statistical significance at the 1% and 5% levels, respectively. Sources: The authors developed this table using empirical results based on sample data.

, the interaction term between transition finance and mature-stage firms (TRANFIN × Mature) in column (1) is positive and statistically significant, indicating that the policy has a pronounced promoting effect on green value for mature firms. Similarly, column (2) shows a significant positive coefficient for the interaction with growth-stage firms (TRANFIN × Growth), confirming that the policy is also effective for firms in this developmental stage.

Critically, when both interaction terms are included simultaneously in column (3), the coefficients reveal a clear hierarchy in policy effectiveness. The coefficient for TRANFIN × Mature (0.029) is substantially larger than that for TRANFIN × Growth (0.013), with both significant at the 5% level. This differential is formally confirmed by a Wald test of coefficient equality, which rejects the null hypothesis that the two coefficients are equal at conventional significance levels. Therefore, the effect for growth-stage firms is approximately 6.9% higher than for decline-stage firms (0.013)/0.189). The effect for mature-stage firms is approximately 15.3% higher than for decline-stage firms (0.029/0.189). The results suggest that mature firms, with their stable cash flows, established organizational structures, and greater resource slack, are uniquely positioned to leverage transition finance most effectively for substantial green value creation. While growth-stage firms also benefit from the policy, their relative resource constraints and strategic focus on expansion appear to limit the magnitude of this effect compared to their mature counterparts.

4.4. Mechanism Analysis

The empirical results validate the moderated mediation mechanism proposed in Hypothesis 3, revealing not only that innovation persistency serves as a fundamental channel through which transition finance enhances corporate green value, but also that this pathway operates with significantly greater effectiveness for mature-stage firms. First, the baseline mediation pathway is firmly established, with Column (1) of

Table 8. The role of innovative persistency in promoting corporate green value.

Variables	GREVAL (1)	INNPER (2)
TRANFIN	0.186 ** (2.07)	0.367 ** (2.29)
INNPER	0.475 *** (3.33)
TRANFIN × Mature	0.030 ** (2.31)	0.046 *** (2.77)
TRANFIN × Growth	0.015 ** (1.99)	0.020 ** (2.23)
INNPER × Mature	0.057 *** (2.83)
INNPER × Growth	0.022 ** (2.30)
Control variables	Yes	Yes
Firm fixed effects	Yes	Yes
Year fixed effects	Yes	Yes
Robust Std. Error.	Yes	Yes
Clustering on firm	Yes	Yes
R2	0.217	0.248
Obs	10,276	10,276

Note: t values are shown in parentheses. *** and ** indicate statistical significance at the 1% and 5% levels, respectively. Sources: The authors developed this table using empirical results based on sample data.

demonstrating a statistically significant positive effect of innovation persistency (INNPER) on corporate green value (GREVAL), while Column (2) confirms that transition finance (TRANFIN) significantly promotes innovation persistency, thereby fulfilling the core requirement for mediation.

Second, and central to Hypothesis 3, the analysis reveals a pronounced hierarchical pattern across the firm lifecycle: in the outcome model, the value-enhancing effect of innovation persistency is substantially stronger for mature-stage firms (INNPER × Mature = 0.057) than for growth-stage firms (INNPER × Growth = 0.022), a difference confirmed by a Wald test; simultaneously, in the mediation model, transition finance is shown to be significantly more effective at fostering innovation persistency in mature-stage firms (TRANFIN × Mature = 0.046) compared to their growth-stage counterparts (TRANFIN × Growth = 0.020). This collective evidence demonstrates a complete and differentially effective moderated mediation pathway. Specifically, transition finance disproportionately enhances innovation persistency in mature-stage firms, which in turn generates significantly stronger green value returns for precisely these firms. This pattern underscores the unique capacity of mature firms to both sustain innovation activities and translate them into superior environmental value creation, revealing why the policy produces the most pronounced effects during this particular developmental stage.

To address potential concerns regarding the measurement of innovation persistency, we conduct two sets of robustness checks. First, we re-define INNPER using alternative thresholds of R&D growth rate (±3% and ±10%) instead of the baseline ±5% bandwidth [59]. Second, we construct a continuous measure of innovation persistency, the inverse of the coefficient of variation of R&D expenditure over the previous three years, to avoid potential information loss from binary discretization. As reported in

Table 9. Robustness of innovation persistency.

Specification	INNPER Definition	TRANFIN → INNPER	INNPER → GREVAL
Panel A: Alternative thresholds (binary measure)
(1)	±3% bandwidth	0.428 ** (2.51)	0.448 *** (3.21)
(2)	±10% bandwidth	0.312 ** (2.08)	0.491 *** (3.45)
Panel B: Continuous measure
(3)	Inverse of CV of R&D	0.283 ** (2.24)	0.412 *** (3.18)

Note: Column TRANFIN → INNPER reports the coefficient of transition finance on innovation persistency. Column INNPER → GREVAL reports the coefficient of innovation persistency on corporate green value. All specifications include the same control variables, firm fixed effects, year fixed effects, and firm-clustered standard errors as the baseline model. t values are shown in parentheses. *** and ** indicate statistical significance at the 1% and 5% levels, respectively. Sources: The authors developed this table using empirical results based on sample data.

, the first-stage coefficients (TRANFIN → INNPER) remain positive and statistically significant across all alternative bandwidths, and the outcome model coefficients (INNPER → GREVAL) also remain positive and significant. These results confirm that our mediation findings are not sensitive to the specific choice of bandwidth or the binary definition of innovation persistency.

To assess the sensitivity of our mediation results to potential unobserved confounding, we conduct a formal sensitivity analysis following Imai et al. [60]. As reported in

Table 10. Sensitivity analysis for mediation effect.

	Estimate
ACME (average causal mediation effect)	0.084 **
ADE (average direct effect)	0.118 **
Total effect	0.202 ***
Proportion mediated	0.416
Sensitivity parameter ρ (ACME = 0 threshold)	0.51

Note: ACME is the average causal mediation effect of transition finance on green value through innovation persistency. ADE is the average direct effect. ρ is the correlation between the error terms in the mediator and outcome equations. *** and ** indicate statistical significance at the 1% and 5% levels, respectively. Sources: The authors developed this table using empirical results based on sample data.

, the average causal mediation effect (ACME) remains statistically significant. The sensitivity parameter ρ, the correlation between the error terms in the mediator and outcome equations, is estimated at 0.51, meaning that an unobserved confounder would need to correlate at least 0.51 with both innovation persistency and corporate green value to completely nullify the mediation effect. Given the rich set of observed covariates, we have already controlled for (including firm size, leverage, profitability, growth opportunities, and fixed asset intensity), such a strong correlation from an omitted variable is highly unlikely. This sensitivity analysis thus supports the credibility of innovation persistency as a mediating channel.

4.5. Discussion

The confirmation of our first hypothesis establishes that transition finance does more than merely facilitate compliance or fund isolated green projects; it enables firms to internalize environmental performance into their core market valuation. This finding extends current understanding of sustainable finance mechanisms, which have often emphasized risk mitigation or regulatory compliance [16,29]. Our results suggest that capital markets actively reward brown firms that demonstrate credible transition pathways, assigning tangible value to their environmental improvements. This market recognition creates a complementary mechanism to regulatory pressures, aligning financial incentives with environmental objectives. Consequently, transition finance emerges not merely as a risk management tool, but as a strategic instrument that functions by transforming sustainability from a cost center into a value driver.

The validation of our second hypothesis reveals why similar policy support produces divergent outcomes across firms. Rather than attributing differential effectiveness to sectoral characteristics or managerial attitudes alone, our lifecycle perspective highlights the fundamental role of organizational readiness. For growth-stage firms, high transformation willingness is constrained by resource limitations, limiting the translation of transition finance into green value. For mature-stage firms, the alignment of resource availability and transformation willingness creates an ideal condition for absorbing financial support. For decline-stage firms, structural rigidity, manifested in sunk costs, asset stranding concerns, and adverse market perceptions, impedes meaningful transformation. Having moved beyond the resource constraints of growth-stage firms while avoiding the structural rigidities of decline-stage firms, mature-stage firms are uniquely positioned to benefit from transition finance [34,35]. This finding helps reconcile contradictory evidence in the literature regarding environmental policy effectiveness, suggesting that organizational maturity provides the necessary absorptive capacity to translate financial support into sustainable value creation.

The confirmation of our third hypothesis uncovers why the relationship between transition finance and corporate green value operates with particular strength in mature firms. It is not merely that these firms innovate more, but that they maintain innovation consistently over time. This ability stems from their stable organizational structures, experienced management teams, and established resource allocation processes. While growth-stage firms may pursue innovation opportunistically and decline-stage firms reactively, mature firms demonstrate the strategic discipline to sustain environmental investments through complete innovation cycles [42]. Therefore, innovation persistency emerges as the critical micro-foundation that explains how financial support translates into lasting environmental value. The insight shifts focus from motivating initial environmental investments toward creating conditions that sustain them, particularly important for brown industries where meaningful transformation requires continuous improvement rather than one-time solutions.

Our finding that mature-stage firms exhibit the strongest response to transition finance might seem to suggest prioritizing them for transition funding. However, we caution that such prioritization involves important trade-offs that warrant careful welfare consideration. First, allocating a larger share of transition finance to mature firms may divert resources from growth- and decline-stage firms. Growth-stage firms, despite their weaker average response, may generate higher long-term returns if their innovation projects succeed, suggesting a potential dynamic trade-off between short-term efficiency and long-term transformation. Second, mature firms are generally more resource-rich, and directing additional funding to them could exacerbate existing inequalities in access to green finance. A comprehensive welfare analysis accounting for these distributional consequences, dynamic efficiency gains, and the social cost of delayed transition would be necessary before implementing such targeted allocation policies. Absent such analysis, we recommend a more balanced approach: while mature firms could be prioritized for scalable transition projects, dedicated mechanisms for growth-stage firms (e.g., risk-sharing co-investment funds) and decline-stage firms (e.g., debt restructuring programs focused on capability renewal) should be maintained to ensure equitable and efficient transition support across all lifecycle stages.

5. Conclusions

Amidst the global low-carbon transition, leveraging financial instruments to steer brown firms toward sustainability has become a critical challenge. Grounded in China’s transition finance practice, this study develops a green value measurement system and employs a DID model to examine how transition finance influences corporate green value through the channel of innovation persistency across different firm lifecycle stages. The findings are consistent with our three core hypotheses: First, transition finance significantly promotes the green value of brown firms. Second, this effect exhibits notable heterogeneity across the life cycle, being most pronounced during the maturity stage. Third, innovation persistency serves as a fundamental mechanism, with mature firms demonstrating superior capacity in both sustaining innovation and converting it into green value.

Our findings offer nuanced policy insights for designing targeted transition finance systems. First, regulators may consider incorporating green value metrics into corporate sustainability evaluations. This would help investors distinguish firms with genuine transition potential from those engaging in greenwashing, thereby improving capital allocation efficiency in green financial markets. A standardized green value disclosure framework could also reduce information asymmetry between brown firms and financial institutions.

Second, policy instruments should be differentiated according to firms’ developmental stages. For mature-stage firms, scalable transition funding can be prioritized for long-term green infrastructure projects. For growth-stage firms, policymakers may design risk-sharing mechanisms such as co-investment funds, first-loss guarantees, or innovation vouchers. For decline-stage firms, policy support could focus on debt restructuring programs including debt-for-equity swaps tied to green transition milestones or subsidized loans for decommissioning polluting assets.

Third, policymakers may establish mechanisms that incentivize continuous innovation across all stages. Phased funding arrangements, where subsequent tranches are conditional on sustained R&D commitment, can encourage firms to maintain stable innovation trajectories rather than pursuing one-off upgrades. Additionally, R&D persistence certifications could provide preferential financing rates to firms demonstrating stable innovation patterns, reinforcing the behavioral shift from episodic to persistent green innovation.

This study has several limitations that warrant further investigation. First, although our green value metric may be subject to measurement limitations, our conclusions remain robust across alternative specifications, and future research could refine this measure using physical environmental data. Second, the exclusive focus on listed companies may overlook the unique transition challenges faced by SMEs, suggesting the need for broader sample coverage. Finally, as transition finance evolves with new instruments like sustainability-linked bonds, future research should examine how these tools interact with corporate governance to shape transition pathways.