How do brick-and-mortar businesses leverage financial markets in China?

Abstract

Brick-and-mortar businesses can be categorized into net financing and net investment types based on the diverse approaches they employ in financial markets, with these distinctions rooted in internal factors within the businesses themselves. Research indicates that brick-and-mortar businesses predominantly opt for a net financing approach when engaging with financial markets. Nevertheless, there is a diminishing trend in the proportion and scale of net financing-oriented companies, while those net investment-oriented companies are on the rise. Disparities exist in financial market utilization across industries, regions, and ownership structures. Companies experiencing high-performance volatility and minimal agency issues tend to favor net financing strategies, with performance volatility impacting the method of utilization by influencing free cash flow. Agency issues within businesses contribute to managerial self-interest, consequently influencing their preferred utilization method. This study introduces a novel micro perspective and quantitative approach to investigate the fundamental relationship between brick-and-mortar businesses and financial markets. It aids regulators in assessing the modes and extents of financial services provided to the real economy from a micro-enterprise standpoint.

JEL Code: G1 G2 M2

Keywords

Agency problems financial markets net financing net investment performance volatility

1. Introduction

China’s financial system has embarked on a trajectory of market-oriented development, accelerating the transition of China’s financial market functions from a financing-centric model to a more investment and financing balanced model (Chen et al., 2018; Wu and Fang, 2021). From a micro-enterprise perspective, the number of listed companies engaging in both investment and fundraising in the financial market has surged from 412 in 2013 to 936 in 2020, comprising 29.0% of the total listed companies, up from 18.8% in 2013. Conversely, the number of listed companies prioritizing fundraising over investment has increased from 1776 in 2013 to 2294 in 2020, yet their proportion has declined from 81.2% to 71.0%. This shift in the relationship between physical enterprises and financial market capital fundamentally reflects changes in enterprises’ utilization of financial market mechanisms. Such transformation not only mirrors the significant alterations in China’s financial scale, structure, format, function, and influence but also impacts enterprises’ profitability and risk, the capacity of financial services to support the real economy, and the sound development of the financial market. Excessive investment by numerous physical enterprises in the financial market not only hampers the real economy’s progress but also risks resource misallocation, asset bubble expansion, and potential triggers of financial and economic crises.

To explore the connection between the real economy and financial markets, existing literature has analyzed the financing and investment flowing into the real economy from a macro viewpoint while assessing the efficacy and extent of financial services to the real economy from the standpoint of the financial market (Chaney et al., 2012; Claessens and Laeven, 2003). Although financial development plays a crucial role in fostering real economic growth, its impact is contingent upon various objective factors (Dinç, 2005; Liu et al., 2007; Porta et al., 1998). The financial market operates holistically by catering to individual micro-enterprises, which frequently seek financing and engage in investments within the financial market (Chen et al., 2018; Hachem and Song, 2021). Disparities in financing and investment volumes across different enterprises reflect the varying levels of financial service accessibility to real enterprises. Consequently, solely examining the relationship between financial markets and the real economy from a macro perspective fails to uncover the heterogeneity in how micro-individual enterprises utilize financial markets, nor does it adequately evaluate the efficacy and extent of financial services to the real economy from the perspective of real enterprises. Thus, this article endeavors to scrutinize the connection between individual enterprises and financial markets from a micro perspective, shedding light on the significant differences in how various enterprises utilize financial markets.

In response to the functional offerings of financial markets, micro-enterprises navigate diverse financial decisions, including investing in or securing financing from these markets, with variations in the amounts involved across different enterprises. While external factors like economic cycles, credit policies, and monetary policies exert significant influence on enterprise capital structures, factors such as GDP cycles, equity index growth rates, broad money cycles, and legal reserve ratios notably impact enterprises’ financial investments (Almeida and Campello, 2007; Bernanke, 1990; Cook and Tang, 2010; Ding et al., 2021; Drobetz and Wanzenried, 2006; Korajczyk and Levy, 2003; Lv and Shi, 2014). However, external factors alone struggle to elucidate the marked differences in investment and financing decisions among enterprises within the same industry and even within the same enterprise over time, with the financial market itself representing a pivotal external factor. Hence, this article explores the intrinsic reasons behind the differential utilization of financial markets by micro-enterprises, examining the role of the financial market in delivering anticipated returns and facilitating risk diversification for enterprises. Performance volatility of a company primarily reflects its operational risks and profitability, thereby influencing its propensity to leverage the financial market to balance returns and risks—an internal factor discussed herein. In practice, prevalent occurrences of zero borrowing, irrational investment, and excessive debt highlight instances where managers sporadically utilize financial markets without adequately maximizing enterprise value. Moral hazard and adverse selection among managers compel them to employ financial market functions in ways that diminish enterprise value, thereby spotlighting the principal-agent problem as another internal factor under scrutiny in this article (Bai et al., 2024; Devos et al., 2012; Dong and Guo, 2019; Heaton, 2002; Korajczyk and Levy, 2003; Xiong et al., 2023).

This study reveals that current trends indicate a shift toward net financing among real enterprises, wherein they increasingly rely on financial markets to acquire funds for bolstering their operational endeavors. However, there is a discernible decline in the proportion and scale of net financing-oriented companies, juxtaposed with a rising prevalence of net investment-oriented companies. This suggests that enterprises are progressively viewing the financial market as a resource allocation avenue on par with the product market. Notably, there exists significant heterogeneity in the utilization of financial markets among enterprises across various industries, regions, and ownership structures. Extensive analyses demonstrate that enterprises exhibiting high performance volatility and minimal agency issues tend to favor a net financing approach. Conversely, enterprises characterized by low performance volatility and pronounced agency problems tend to opt for a net investment strategy. Mechanism tests underscore the influence of performance volatility on financial market utilization by shaping free cash flow dynamics, while managers’ self-interest contributes to agency problems, thereby influencing their chosen utilization approach.

This article presents several potential innovations and contributions: (1) The article introduces an innovative research perspective by shifting the focus from the traditional macro-level examination of the relationship between finance and the real economy, typically analyzed from the viewpoint of funding providers (financial markets) to a micro-level exploration from the perspective of diverse demanders—enterprises. This approach broadens the scope of relevant research concerning the real economy and financial markets, offering insights into their relationship through the lens of enterprises. (2) This article offers innovative research content by providing an objective depiction of the varied ways real enterprises engage with financial markets, supported by empirical data. It explores the underlying reasons for this differentiation from within the enterprise, thereby furnishing empirical evidence to analyze the relationship between the real economy and financial markets from a micro-enterprise perspective. (3) The article introduces innovative quantitative indicators by distinguishing financial markets from product markets, a distinction often overlooked in existing literature. Inspired by the Penman-Nissim framework (Nissim and Penman, 2001) and the concept of reconstructing financial statements proposed by Penman (2013), the article employs the concepts of net financing and net investment to quantify the funding relationship between enterprises and financial markets. This quantitative approach, valuable for understanding the fundamental relationship between enterprises and different markets, sheds new light on how real enterprises utilize financial markets.

The rest of the article is structured as follows: Section 2 develops the research hypotheses based on theoretical analysis, followed by Section 3 which outlines the research design, sample selection and the data. Empirical analysis is presented in Section 4, including an exploration of influencing mechanisms. The article concludes with Section 5 dedicated to summarizing findings and providing implications.

2. Theoretical analysis and research hypotheses

2.1. Quantifying how real enterprises utilize financial markets

The utilization of financial markets by real enterprises primarily manifests through their funding relationships. Financial markets play a critical role in risk mitigation (Aikman et al., 2015; Duchin et al., 2017; Koopman et al., 2009) for enterprises by offering funding and optimizing resource allocation (Banerjee and Duflo, 2007; Hsieh and Klenow, 2009; Restuccia and Rogerson, 2017). Existing literature has predominantly focused on shaping specific capital structures for enterprises through financial market financing, which introduces varying financial risks and leverage values (Stulz, 1990). With the ascent of financialization, scholars have delved into explaining enterprise financial asset allocation behavior, considering surplus and crowding-out effects, with distinct financial asset allocations yielding varying effects on enterprises (Orhangazi, 2008; Schumpeter, 1936; Theurillat et al., 2010; Tornell, 1990; Seo et al., 2012). In light of the functional supply of financial markets, micro-enterprises engage with financial markets for both financing and financial asset investment. However, focusing solely on financing or financial asset investment fails to capture the full spectrum of how real enterprises leverage financial markets. Narrowing the description of how real enterprises utilize financial markets to just the amount of financing overlooks the objective reality that enterprises may also invest in financial markets. Similarly, solely focusing on investment amounts neglects the possibility that enterprises may obtain financing from financial markets.

The coexistence of financing and investment among most enterprises arises from factors such as the high liquidity of financial assets, sporadic fundraising, and irrational decision-making within enterprises. This coexistence signifies the ongoing enhancement of China’s financial markets, facilitating the efficient transfer of economic resources across time and space (Aikman et al., 2015; Duchin et al., 2017; Koopman et al., 2009), thereby mitigating risks for enterprises and optimizing resource allocation (Banerjee and Duflo, 2007; Hsieh and Klenow, 2009; Restuccia and Rogerson, 2017). The prevalence of idle funds, detachment from reality, and shadow banking underscores that some enterprises utilize financing obtained from financial markets for financial investments. Consequently, to accurately capture the specific funding relationship between enterprises and financial markets, it is imperative to consider both the net financing and net investment amounts sourced from the financial market. This comprehensive approach enables a thorough depiction of how the real economy leverages financial markets from the micro-enterprise perspective, allowing for an evaluation of the efficacy and extent of financial services to the real economy from the vantage point of micro-enterprise utilization.

As per the Penman-Nissim framework (Nissim and Penman, 2001) and insights from Penman (Penman, 2013), restructuring financial statements involves the equation “operating assets + financial assets = operating liabilities + financial liabilities + owner’s equity.” Further refinement of this equation, drawing from the financing hierarchy theory (Myers and Majluf, 1984), yields “operating assets + financial assets = operating liabilities + retained earnings + prior-year shareholder contributions + financial liabilities + current-year shareholder contributions.” Despite being funds acquired from financial markets, equity capital invested in previous years typically remains within enterprises and seldom exits through delisting or stock repurchases. Consequently, these funds primarily reflect the relationship between the enterprise and its shareholders, exerting minimal influence on the funding supply of financial markets. Conversely, financial liabilities, current-year equity capital invested, and financial assets actively alter the funding relationship between the enterprise and financial markets in the current period. The dynamics of financial liabilities and assets will continue to shape the future funding relationship between the enterprise and financial markets, thereby influencing the funding supply of financial markets over time. Consequently, when excluding the company’s data from the year of listing, the financial asset balance, representing the year-end investment of the enterprise in the financial market, equates to the year-end financing balance, comprising financial liabilities, current-year equity capital invested, and previous-year equity capital invested. If the enterprise’s financing balance exceeds its investment balance in the financial market, it is classified as a net financing type, with its net financing amount calculated as financial liabilities plus current-year equity capital invested minus financial assets. Conversely, if the enterprise’s investment balance in the financial market surpasses its financing balance, it is designated as a net investment type, with its net investment amount determined as financial assets minus financial liabilities plus current-year equity capital invested.

A positive net financing amount signifies that the enterprise primarily secures funds from the financial market to fulfill its funding requirements in the product market. This net financing amount also represents the capital acquired by a specific enterprise from the financial market and subsequently invested in the product market, showcasing the financial market’s contribution of funds to the product market. In this scenario, the financial market operates as a provider of funds to support the product market’s funding needs, with its impact on the enterprise’s value derived from the efficient utilization of the raised funds in the product market.

In contrast, a negative net financing amount (or a positive net investment amount) indicates that the enterprise invests funds in the financial market to allocate financial assets and generate profits. This net investment approach differs from the method where enterprises raise funds from the financial market and invest them directly in the product market. Presently, the financial market serves as an additional avenue for enterprises to deploy funds, potentially leading to crowding-out effects on the enterprise’s funding in the product market. The financial market’s impact on the enterprise’s value stems from the profitability of the financial assets acquired by the enterprise. The pathways and mechanisms through which the financial market’s financial assets and the product market’s operating assets create value for the enterprise are fundamentally distinct. Consequently, the varying funding relationships between enterprises and financial markets reflect the diverse ways in which enterprises utilize financial markets and the differing impacts of the financial market on individual enterprises. Net financing and net investment amounts offer a clear and independent depiction of the relationship between real enterprises and the financial market, aligning with their relationship with the product market. This approach effectively illustrates how the financial market influences real enterprises, a perspective that cannot be adequately captured solely by studying financing or investment activities in isolation.

2.2. Factors underlying the impact of different utilization modes

Real enterprises employ financial markets differently, influenced by external factors like macroeconomic monetary policy, financial market funding availability, product supply, and industry regulations. However, these factors have varying impacts on all enterprises and those within the same industry, making it challenging to explain the diversity in micro-individual enterprise utilization of financial markets. Despite the objective presence of financial markets, enterprises’ internal complexities and diverse factors significantly shape their utilization of these markets. Essentially, the decision to invest or finance in the financial market is a financial choice made by enterprise management teams to enhance returns and mitigate risks based on the financial market’s functional offerings. When making such decisions, management teams consider both their own interests and those of the enterprise. Scholars have also observed a close relationship between enterprises’ financing and investment decisions and their financial characteristics and governance mechanisms (Myers and Majluf, 1984). Therefore, this article explores the various ways micro-enterprises utilize financial markets through the lenses of performance volatility and management’s principal-agent problem.

2.2.1 Impact of performance volatility on financial market utilization in real enterprises

The performance volatility of an enterprise mirrors its profitability, operational risks, and market competitiveness during business operations (Johnson, 2003). Elevated performance volatility often stems from fierce competition within the product market, where enterprises struggle to maintain stable market share and revenue due to insufficient competitive advantages. To mitigate operational risks within the product market, enterprises may pursue breakthroughs in core technology through research and innovation to avoid market elimination resulting from intense competition or inadequate management. Alternatively, they may boost fixed asset investments to facilitate strategic transformations, achieve economies of scale, and upgrade products (Comin and Philippon, 2005). The funds necessary for these tangible investments primarily originate from financing within the financial market.

Conversely, when heightened performance volatility results in unstable profits, enterprises not only face challenges in managing operational risks within the product market, thereby diminishing their capacity to bear risks for investments in the financial market, but also encounter a shortage of funds to acquire financial assets. Consequently, their utilization mode of financial markets is more inclined toward net financing.

In contrast, when an enterprise exhibits low performance volatility, it enjoys a stable income source, ensuring consistent and steady profits. To mitigate the uncertainty associated with new investments, enterprises tend to adopt a more conservative approach in their tangible investments, aiming to reduce operational risks stemming from such investments and maintain their existing advantageous positions (Han and Qiu, 2007). Consequently, this significantly diminishes their demand for financing from the financial market to support tangible investments. Simultaneously, the stable performance and cash flow empower these enterprises to garner higher returns through the financial market, reflecting the surplus effect of financial investment (Orhangazi, 2008). A dynamic trading market for financial assets coupled with swift liquidity conversion facilitates seamless fund transfers between the financial market and the product market, thereby enhancing fund returns and mitigating liquidity risks. Furthermore, lower performance volatility corresponds to reduced operational risks for the enterprise, bolstering its ability to withstand potential losses from financial asset investments that might jeopardize operational sustainability, thereby enhancing its inclination to invest in financial assets. Based on the above analysis, we propose the following hypothesis:

Hypothesis 1 (H1). The greater the performance volatility, the more likely an enterprise is to adopt a net financing mode of financial market utilization. Conversely, the smaller the performance volatility, the more likely an enterprise is to adopt a net investment mode of financial market utilization.

2.2.2 Impact of principal-agent problems on financial market utilization in real enterprises

When making financial decisions regarding investment and financing, enterprise managers consider both their and the enterprise’s interests. However, in the presence of principal-agent problems within the enterprise, managers may prioritize their interests, aiming to minimize risk and maximize short-term profits (Orhangazi, 2019) rather than fostering the enterprise’s long-term competitiveness and development. Severe principal-agent problems necessitate more effective supervision of managers with discretionary power over financial asset allocation and a lack of effective incentive mechanisms to align management incentives. In such cases, management has a strong incentive to allocate funds to financial assets to maximize short-term profits. This is because financial assets offer short cycles, high liquidity, and high returns, incentivizing managers to pursue excess profits from the financial market (Bai and Zhang, 2024; Orhangazi, 2008; Wang et al., 2017).

Conversely, debt financing imposes the obligation of mandatory repayment and interest payments on enterprises, exposing them to the risk of default if their operations are poorly managed. This external supervision pressure on managers arises (Jensen and Meckling, 2019). In instances where managers do not face pressure from ownership, compensation incentives, or active regulation—indicating severe principal-agent problems—they tend to avoid debt financing to mitigate the risks associated with debt obligations while maximizing their interests (Berger et al., 1997). Consequently, in enterprises grappling with severe principal-agent problems, managers prioritize reducing debt financing to safeguard their interests, even if the overall level of enterprise financing remains unchanged. Compared to long-term, low-return tangible investments and high-risk research and development (R&D) endeavors, managers are more inclined to invest enterprise profits and borrowed funds into financial assets. In this scenario, the enterprise’s utilization mode of financial markets is more likely to skew toward net investment.

In contrast, in enterprises with low principal-agent problems, the disparity between the goals, information, risk preferences, and responsibilities of managers and shareholders is minimal. Consequently, managers are more inclined to make financial decisions considering the long-term development of the enterprise. With a focus on maintaining the enterprise’s competitiveness in the market, managers prioritize R&D innovation and operational investments, directing profits and borrowed funds from the financial market toward tangible activities, and even seeking additional financing avenues to support real investments. In this scenario, the enterprise’s utilization mode of financial markets is more likely to lean toward net financing. Based on this, we propose the following hypothesis:

Hypothesis 2 (H2). The smaller the principal-agent problems, the more likely an enterprise is to adopt a net financing mode of financial market utilization. Conversely, the more severe the principal-agent problems, the more likely an enterprise is to adopt a net investment mode of financial market utilization.

3. Research design

3.1. Data sources and sample selection

This study uses the A-share listed companies in the Chinese Stock Market from 2013 to 2020 as the sample. We start our sample period from 2013 for the following reasons: First, 2013 marks the year when the People’s Bank of China (PBOC) implemented a major financial reform, fully liberalizing the regulation of lending rates and capping only deposit rates. This reform provides a more flexible financing environment for enterprises, enabling them to adjust their financing strategies according to their own needs and market conditions, thus making better use of financial market resources. Therefore, choosing 2013 as the starting point of the study captures the impact of this policy change on enterprises’ financial market utilization behavior, providing richer context and depth for the study. Second, the choice of the starting point of the study is also based on considerations of data availability and completeness. Due to advancements in information technology and the improvement of financial market information disclosure systems, data from 2013 and beyond are more comprehensive and accurate, meeting the study’s requirements for data quality. In addition, the selection of relatively recent data reduces data bias and uncertainty that may arise from a long-time span, thereby improving the accuracy and reliability of the study.

Among all the A-shares, we select our sample based on the following criteria: (1) We exclude companies with abnormal operations that are marked as ST or ST* by the exchanges. ST, or special treatment, generally refers to companies that have operated at a loss for two consecutive years, and ST* is a company with losses for three consecutive years that is at risk of delisting and is subject to special treatment. (2) The object of research in this article is the entity enterprise, which refers to enterprises engaged in economic activities such as the production, processing, and sale of actual products or services, with a fixed place of operation and a material production base.¹ Therefore, we exclude financial and insurance listed companies. (3) We remove the firm-year with IPO. And (4) we remove companies with missing data. After applying these filter rules, we retained a total of 21,647 observations. The data primarily originate from the Guotai-An database, and all continuous variables are winsorized at the 1% tails.

3.2. Variable definition

According to the preceding analysis, the net financing amount is calculated as financial liabilities plus shareholder investment minus financial assets. When the net financing amount is positive, enterprises utilize financial markets in a net financing mode. Conversely, when the net financing amount is negative (indicating a positive net investment amount), enterprises adopt a net investment mode in utilizing financial markets. In the baseline regression, we measure a firm’s primary usage of the financial market as a binary variable Fin1, which takes the value of 1 if the net financing amount is positive, and 0 otherwise. In the robustness test, we measure with the continuous variable Fin2, which is calculated as the net financing amount scaled by total assets.

Our two main independent variables are constructed following literature. Based on Cheng (2008) and Comin and Philippon (2005), we employ two methods to measure performance volatility: (1) the standard deviation of return on equity (SdROE); and (2) the standard deviation of return on assets (SdROA). Based on Ang et al. (2000) and Singh and Davidson (2003), we also measure principal-agent problems in two ways: (1) the ratio of management expenses to operating income (Ac1); and (2) the ratio of the sum of management expenses and selling expenses to operating income (Ac2). These indicators primarily reflect the principal-agent problems between managers and shareholders.

Following prior literature, including Zhou et al. (2022), Yang et al. (2024) and Meng et al. (2023), we select a range of firm characteristics, documented to influence the firm’s financing modes, as control variables: company size (Size), proportion of tangible assets (Tangi), years since listing (Age), growth (Growth), registered area (East), property rights (SOE), shareholding proportion of the largest shareholder (Top1), proportion of independent directors (Inde), and CEO duality (Duality). In addition, this study incorporates year-fixed effects and industry-fixed effects to account for macroeconomic and industry policies. Detailed definition and calculation of the variables are summarized in Table 1.

Table 1.

Variable definitions.

Variable definition	Symbol	Variable construction
How real enterprises utilize financial markets	Fin1	Dummy variable, taking the value of 1 if net financing amount is positive, and 0 otherwise.
How real enterprises utilize financial markets	Fin2	Net financing amount, continuous variable, calculated as the sum of financial liabilities and shareholder investment, minus financial assets, scaled by total assets.
Performance volatility 1	SdROE	The standard deviation of return on equity over the past 3 years
Performance volatility 2	SdROA	The standard deviation of return on assets over the past 3 years for the enterprise
Principal-agent problems 1	Ac1	operating expenses divided by operating income
Principal-agent problems 2	Ac2	the sum of operating expenses and selling expenses divided by operating income
Company size	Size	The natural logarithm of total assets
The ratio of tangible assets	Tangi	(Fixed Assets + Inventory)/Average Total Assets
Growth	Growth	(Current period operating revenue—Prior period operating revenue)/Prior period operating revenue
Registration Area	East	Eastern regions take 1, non-eastern region take 0
Age of Listing	Age	ln (current year − year of launch + 1)
Nature of ownership	SOE	State-owned enterprises take 1, non-state-owned enterprises take 0
Percentage of shareholding of the largest shareholder	Top1	Shares held by the largest shareholder/total shares
Percentage of independent directors	Inde	Number of independent directors/Total number of directors
CEO duality	Duality	1 if CEO and the chairman of the board are the same person, and 0 otherwise
Annual dummy variables	Year	Controlling for annual fixed effects
Industry dummy variables	Industry	Controlling industry-fixed effects

This table lists all the variables used in the study, including the dependent variables Fin1 and Fin2; the independent variables SdROE, SdROA, Ac1, and Ac2; and all the control variables.

In Table 2, we present the descriptive statistics of the variables. When measuring the volatility of firm performance, SdROE exhibits higher values and greater volatility compared to SdROA. Geographically, 69.3% of the sample firms are located in the Eastern region of China, a region characterized by relatively higher and faster economic growth. Regarding ownership structure, 35.3% of the firms are state-owned, while the remainder are non-state-owned. On average, 37.7% of the board members are independent directors, the largest shareholders own approximately 34.1% of the shares, and 27.8% of the firms have the CEO and the chairman of the board as the same individual. For our main variables, the net financing amounts of a firm, Fin1 and Fin2, we will examine their values and distributions in detail in Section 3.4.

Table 2.

Descriptive statistics of the main variables.

Variable	N	Mean	Std. Dev	Median	Min.	Max.
SdROE	21,647	0.089	0.263	0.025	0.001	2.129
SdROA	21,647	0.031	0.053	0.014	0.001	0.351
Ac1	21,647	0.095	0.082	0.074	0.009	0.524
Ac2	21,647	0.170	0.138	0.131	0.014	0.765
Size	21,647	22.240	1.286	22.070	19.790	26.190
Tangi	21,647	0.353	0.178	0.340	0.011	0.788
Age	21,647	2.212	0.756	2.303	0.693	3.296
Growth	21,647	0.164	0.452	0.094	−0.628	2.951
East	21,647	0.693	0.461	1.000	0.000	1.000
SOE	21,647	0.353	0.478	0.000	0.000	1.000
Inde	21,647	0.377	0.054	0.364	0.333	0.571
Top1	21,647	0.341	0.146	0.320	0.091	0.737
Duality	21,647	0.278	0.448	0.000	0.000	1.000

This table presents the descriptive statistics of the independent and control variables defined in Table 1. For each variable, we report the number of observation (N), mean (Mean), standard deviation (Std. Dev), median (Median), minimum (Min.), and maximum (Max.) values.

3.3. Model design

Following Zhang and Hao (2022), we use the logit model to investigate the relationship between the way firms use financial markets and their performance volatility and agency problems. In the baseline regression model (1), we use the logit regression model to regress the dummy variable Fin1 on our explanatory variables, to test H1 and H2. H1 is considered to be supported if the coefficient $α_{1}$ is significantly positive, while H2 is considered to be supported if the coefficient $α_{2}$ is significantly negative. In the robustness test, the OLS model is utilized to establish model (2) to test both H1 and H2.

\begin{array}{l} F i n 1_{i, t} = α_{0} + α_{1} S d R O E_{i, t} o r S d R O A_{i, t} + α_{2} A c 1_{i, t} o r A c 2_{i, t} + \sum C o n t r o l s \\ + \sum Y e a r + \sum I n d u s t r y + ε_{i, t} \end{array}

(1)

\begin{array}{l} F i n 2_{i, t} = α_{0} + α_{1} S d R O E_{i, t} o r S d R O A_{i, t} + α_{2} A c 1_{i, t} o r A c 2_{i, t} + \sum C o n t r o l s \\ + \sum Y e a r + \sum I n d u s t r y + ε_{i, t} \end{array}

(2)

Where Controls are the control variables, and Year and the Industry are the year and industry dummies used for the fixed effects.

3.4. Real enterprises’ utilization of the financial market

Table 3 shows the number and percentage of net financing-oriented and net investment-oriented firms within the sample for each year. From the table, we observe that, cross-sectionally, the majority of enterprises annually secure net financing from financial markets to support their operations. However, from a time-series perspective, the proportion of net financing-oriented companies, as well as those maintaining this status for three consecutive years, shows a declining trend from 2013 to 2020. Conversely, the proportion of net investment-oriented companies, along with those persisting for 3 years, exhibits an upward trend.

Table 3.

Number and proportion of the two types of companies.

Year	Total N.	Net financing companies				Net investment companies
Year	Total N.	N.	Prop.	Persist N.	Persist prop.	N.	Prop.	Persist N.	Persist prop.
2013	2188	1776	0.812			412	0.188
2014	2214	1769	0.799			445	0.201
2015	2331	1842	0.790	1467	0.629	489	0.210	212	0.091
2016	2504	1888	0.754	1438	0.574	616	0.246	233	0.093
2017	2772	1997	0.720	1460	0.527	775	0.280	277	0.100
2018	3126	2206	0.706	1543	0.494	920	0.294	347	0.111
2019	3282	2268	0.691	1632	0.497	1014	0.309	482	0.147
2020	3230	2294	0.710	1800	0.557	936	0.290	590	0.183
Total	21,647	16,040	0.741	9340	0.431	5607	0.259	2141	0.099

This table presents the number and percentage of net financing and net investment companies within the sample for each year. Total N. is the total number of companies. N. is the number of companies. Prop. is the proportion of each type of companies as a percentage of the total number of companies. The Persist N. is the number of companies that are persistently classified as net financing or net investment companies for three consecutive years up to and including the given year. Persist prop. is the proportion of such companies as the percentage of the total companies.

Table 4 shows the distribution of the amount of net financing across firms in each sample year, as well as in the whole sample. As shown in the table, for net financing companies, both the mean and quartiles of net financing amounts are decreasing. In contrast, for net investment companies, net investment amounts—including the mean and quartiles—are increasing. This indicates a growing number and scale of net investment companies, reflecting a shift toward enterprises prioritizing net investment in financial markets. This trend highlights a transformation in the role of financial markets, evolving into an increasingly significant resource allocation channel parallel to product markets.

Table 4.

Distribution of the amount of net financing.

Year	Net financing of net financing companies				Net investment in net investment companies
Year	Mean	q1	q2	q3	Mean	q1	q2	q3
2013	0.249	0.096	0.236	0.378	0.097	0.004	0.044	0.140
2014	0.247	0.102	0.229	0.360	0.115	0.013	0.067	0.169
2015	0.256	0.116	0.237	0.370	0.123	0.019	0.086	0.186
2016	0.250	0.112	0.233	0.364	0.131	0.023	0.089	0.205
2017	0.228	0.097	0.216	0.330	0.150	0.040	0.110	0.220
2018	0.217	0.093	0.203	0.312	0.143	0.032	0.097	0.218
2019	0.220	0.097	0.198	0.317	0.140	0.037	0.107	0.206
2020	0.224	0.096	0.201	0.326	0.131	0.030	0.094	0.198
Total	0.235	0.100	0.216	0.341	0.133	0.026	0.091	0.200

This table shows the distribution of the amount of net financing across companies in each sample year, as well as in the whole sample. The amount of net financing is calculated as the sum of financial liabilities and shareholder investment, minus financial assets, scaled by total assets. A positive value represents net financing, while a negative value indicates net investment, expressed as its absolute value.

How enterprises use financial markets often vary significantly across industries due to their unique characteristics and financial needs. We hence separate our sample companies by their industries and look at the distribution of net financing versus net investment firms across industries.

Table 5 and Figure 1 highlight significant industry disparities in financial market utilization. In sectors such as Electricity, Heat, Gas, and Water Production and Supply (D) and Real Estate (K), over 84% of enterprises engage in net financing, with these industries also recording some of the highest mean net financing amounts. Conversely, industries like Education, Culture, and Entertainment (PR) and Information Transmission, Software, and Information Technology Services (I) exhibit lower proportions of net financing-oriented companies and correspondingly lower mean net financing amounts.

Table 5.

The proportion of net financing companies across industries.

Year	2013	2014	2015	2016	2017	2018	2019	2020
Agriculture, forestry, animal husbandry and fishery (A)	0.906	0.909	0.818	0.800	0.667	0.806	0.737	0.757
Mining (B)	0.857	0.875	0.864	0.824	0.800	0.843	0.829	0.845
Manufacturing (C)	0.792	0.787	0.778	0.734	0.699	0.678	0.676	0.698
Electricity, heat, gas, and water production and supply industry (D)	0.963	0.928	0.917	0.906	0.933	0.948	0.942	0.952
Construction (E)	0.948	0.931	0.897	0.831	0.870	0.916	0.816	0.859
Wholesale and retail trade (F)	0.826	0.730	0.712	0.745	0.736	0.732	0.667	0.711
Transportation, storage, and postal industry (G)	0.882	0.920	0.920	0.848	0.807	0.826	0.832	0.848
Social Services (HLMNQ)	0.800	0.772	0.793	0.772	0.767	0.732	0.718	0.732
Information transmission, software, and information technology services (I)	0.755	0.726	0.709	0.721	0.621	0.601	0.574	0.563
Real Estate (K)	0.933	0.924	0.933	0.925	0.925	0.918	0.847	0.848
Education, Culture, and Entertainment (PR)	0.571	0.610	0.698	0.522	0.458	0.448	0.368	0.424
Comprehensive (S)	0.824	0.688	0.722	0.611	0.722	0.778	0.737	0.667

This study follows the industry classification standard set forth by the China Securities Regulatory Commission in 2012. It consolidates certain sectors into broader categories for clarity. For instance, accommodation and catering (H), leasing and business services (L), scientific research and technical services (M), water conservancy, environmental, and public facilities management (N), and health and social work (Q) are merged into the “social services industry.” Similarly, the education industry (P), culture, sports, and entertainment industry (R) are merged into the “education, culture, and entertainment industry.” To streamline presentation, only the proportion of net financing company is listed, with the proportion of net investment companies derived as 1 minus the proportion of net financing companies. The proportion of net financing companies in a specific industry is calculated by dividing the number of net financing companies in that industry by the total number of companies in that industry.

Figure 1.

Average net financing across net financing companies by industry.

These differences suggest the diverse strategies industries adopt in leveraging financial markets, influenced by factors such as operational risks, tangible asset ratios, and financing constraints. Capital-intensive industries, such as Electricity, Heat, Gas, and Water Production and Supply (D) and Real Estate (K), are more likely to be net financing enterprises due to their need for substantial capital investments and extended payback periods. In contrast, asset-light and high-risk industries, such as Education, Culture, and Entertainment (PR) and Information Transmission, Software, and Information Technology Services (I), are more inclined to rely on internal accumulation or venture capital to fund investments and growth. These sectors often have lower capital intensity and may use financial markets primarily to grow their reserves or diversify income streams, rather than to raise external funding.

The utilization of financial markets by enterprises reveals significant geographic disparities across different regions in China, particularly between the Eastern region and other parts of the country. A defining characteristic of China’s economy is the pronounced imbalance in economic development, with the Eastern region consistently outpacing other regions in terms of growth and economic sophistication over the past several decades. To explore these differences, we classify our sample companies based on geographic location into those located in the Eastern region and those in non-Eastern regions.² Table 6 reports the distribution of net financing and net investment companies across these regions. The results reveal that from 2013 to 2020, both the proportion of net financing companies and the average net financing amount in the Eastern region are lower than those in the non-Eastern regions. Conversely, the Eastern region shows a higher proportion of net investment companies and a greater average net investment amount. These findings underscore the regional disparities in financial market utilization by enterprises, largely driven by differences in the development level of financial markets.

Table 6.

Distribution of the two types of enterprises across regions.

Year	Net financing companies				Net investment companies
	Proportion of companies		Mean net financing amount (Fin2)		Proportion of quantity		Mean net investment amount (-Fin2)
	Eastern Region	Non-eastern region	Eastern Region	Non-eastern region	Eastern Region	Non-eastern region	Eastern Region	Non-eastern region
2013	0.789	0.858	0.236	0.274	0.211	0.142	0.101	0.087
2014	0.779	0.841	0.238	0.267	0.221	0.159	0.118	0.105
2015	0.776	0.820	0.241	0.286	0.224	0.180	0.130	0.102
2016	0.746	0.771	0.242	0.267	0.254	0.229	0.138	0.114
2017	0.699	0.772	0.223	0.239	0.301	0.228	0.154	0.139
2018	0.682	0.768	0.211	0.228	0.318	0.232	0.147	0.130
2019	0.669	0.741	0.205	0.236	0.331	0.259	0.145	0.119
2020	0.692	0.751	0.207	0.240	0.308	0.249	0.137	0.110

This table presents the proportion of net financing and net investment companies and the mean net financing amount across different regions.

The financial markets in the Eastern region are relatively well-developed. This economic advantage has enabled the Eastern region to cultivate a more advanced financial ecosystem, providing enterprises with greater access to diverse financial market functions. Companies in these areas can fully leverage the financing and investment functions of the markets. In contrast, financial markets in non-Eastern regions primarily serve real enterprises by providing funds through traditional financing channels, limiting diversification and market integration. Companies in these less developed regions face significant barriers to financial market utilization, reflecting broader regional disparities in economic infrastructure and resources.

Another characteristic of China’s economy is the imbalance in policy support and resource allocation between state-owned enterprises (SOEs) and non-state-owned enterprises (NSOEs). This raises the question of whether companies with different ownership structures exhibit differences in their utilization of financial markets. Table 7 presents the distribution of net financing and net investment companies across various ownership structures. The results reveal that from 2013 to 2020, SOEs consistently had a higher proportion of net financing companies and greater average net financing amounts compared to NSOEs. Conversely, the proportion of net investment companies among NSOEs surpassed that of SOEs, with the average net investment amounts for NSOE net investment companies significantly exceeding those of SOEs since 2015.

Table 7.

Distribution of the two types of companies across state ownership.

Year	Net financing companies				Net investment companies
	Proportion		Mean net financing amount (Fin2)		Proportion		Mean net investment amount (-Fin2)
	SOE	NSOE	SOE	NSOE	SOE	NSOE	SOE	NSOE
2013	0.879	0.764	0.291	0.214	0.121	0.236	0.113	0.091
2014	0.845	0.767	0.284	0.220	0.155	0.233	0.131	0.107
2015	0.817	0.773	0.291	0.232	0.183	0.227	0.104	0.133
2016	0.790	0.733	0.282	0.229	0.210	0.267	0.108	0.141
2017	0.766	0.697	0.263	0.209	0.234	0.303	0.132	0.157
2018	0.775	0.675	0.248	0.200	0.225	0.325	0.122	0.150
2019	0.755	0.661	0.243	0.201	0.245	0.339	0.117	0.145
2020	0.762	0.685	0.240	0.206	0.238	0.315	0.110	0.137

This table presents the proportion of net financing and net investment companies, and the mean net financing amount across state-owned enterprises (SOE) and non-state-owned enterprises (NSOE).

These differences can be attributed to variations in enterprise value, the degree of financing constraints, and financial objectives tied to ownership structure. SOEs often benefit from easier access to government support and low-cost financing due to their special status. However, they also bear greater social responsibilities and policy burdens, which incline them toward net financing strategies. In contrast, NSOEs, while facing more intense market competition and greater financing constraints, typically enjoy higher operational flexibility and quicker market responsiveness, making them more inclined toward net investment strategies in financial market utilization.

4. Empirical analysis

4.1. Univariate analysis

To investigate potential differences in explanatory and control variables between net financing and net investment companies, we analyze the mean values of each variable for the two groups and conduct t-tests to assess the statistical significance of these differences. Table 8 shows that while there is no significant difference in the mean return on assets (SdROA) between the two groups, net financing companies exhibit a significantly higher mean return on equity (SdROE) compared to net investment companies. Furthermore, agency problems (Ac1, Ac2) are significantly lower for net financing companies than for net investment companies. This analysis suggests that enterprises with higher performance volatility are more likely to adopt net financing methods in financial markets, while those with lower performance volatility tend to prefer net investment methods. Similarly, enterprises with lower agency problems are inclined toward net financing, whereas those with more pronounced agency problems favor net investment. These findings provide preliminary support for H1 and H2.

Table 8.

Univariate analysis.

Variable name	Mean
Variable name	Net investment companies	Net financing companies	T-test
SdROE	0.060	0.099	−0.039***
SdROA	0.032	0.031	0.001
Ac1	0.110	0.089	0.021***
Ac2	0.215	0.155	0.060***
Size	21.720	22.420	−0.701***
Tangi	0.277	0.379	−0.102***
Age	2.045	2.271	−0.226***
Growth	0.115	0.181	−0.067***
East	0.746	0.675	0.071***
SOE	0.278	0.380	−0.103***
Inde	0.377	0.377	0.000
Top1	0.350	0.338	0.012***
Duality	0.329	0.261	0.069***

This table presents the mean values of the main variables we use in the analysis between the net financing and net investment firms, and the t-statistics from the t-test which examines the statistical significance of the differences in their mean values. ***, **, * indicate significance at the 1%, 5%, 10% levels, respectively.

Except for the proportion of independent directors (Inde), the group differences for all other control variables were statistically significant at the 1% confidence level. The mean values for enterprise size (Size), the proportion of tangible assets (Tangi), age of listing (Age), growth (Growth), and state ownership (SOE) were all significantly higher for net financing companies compared to net investment-oriented companies. This suggests that larger enterprises, those with a higher proportion of tangible assets, longer listing durations, higher growth rates, and SOEs are more likely to utilize net financing methods in financial markets. In contrast, the mean values for registered region (East), the proportion of the largest shareholder’s holdings (Top1), and duality (Duality) were significantly lower for net financing companies than for net investment companies. This indicates that enterprises located in the eastern region, those with a higher concentration of ownership by the largest shareholder, and those with CEO-chairperson duality are more inclined to adopt net investment methods in financial markets.

4.2. Regression results and analysis

Table 9 presents the regression results of Model (1), with columns (1)–(4) showcasing different combinations of variables used to measure performance volatility and agency problems. Across all specifications, the regression coefficients for performance volatility are significantly positive at the 1% confidence level, indicating that enterprises with higher performance volatility are more likely to utilize net financing methods in financial markets, while those with lower performance volatility tend to prefer net investment methods. These results provide strong support for H1.

Table 9.

Logit regression results.

Variable	(1) Fin1	(2) Fin1	(3) Fin1	(4) Fin1
SdROE	1.403*** (13.323)	1.472*** (13.939)
Ac1	−0.992*** (−4.168)		−0.647*** (−2.766)
SdROA			2.998*** (8.497)	3.442*** (9.752)
Ac2		−1.635*** (−12.480)		−1.541*** (−11.878)
Size	0.545*** (28.174)	0.525*** (27.692)	0.535*** (27.903)	0.512*** (27.209)
Tangi	3.678*** (29.338)	3.509*** (27.957)	3.669*** (29.525)	3.495*** (28.100)
Age	−0.063** (−2.221)	−0.049* (−1.701)	−0.030(−1.064)	−0.016(−0.570)
Growth	0.443*** (9.382)	0.397*** (8.506)	0.422*** (9.133)	0.369*** (8.088)
East	−0.181*** (−4.557)	−0.214*** (−5.342)	−0.188*** (−4.750)	−0.220*** (−5.536)
SOE	−0.139*** (−2.994)	−0.188*** (−4.001)	−0.142*** (−3.056)	−0.185*** (−3.956)
Inde	0.868*** (2.645)	0.863*** (2.623)	0.893*** (2.736)	0.899*** (2.745)
Top1	−2.072*** (−16.037)	−2.066*** (−15.932)	−2.079*** (−16.150)	−2.075*** (−16.067)
Duality	−0.011(−0.273)	−0.001(−0.020)	−0.014(−0.360)	−0.004(−0.104)
Constant term	−10.988*** (−23.868)	−10.282*** (−22.849)	−10.875*** (−23.800)	−10.083*** (−22.550)
Industry/year	Control	Control	Control	Control
Sample size	21,647	21,647	21,647	21,647
Pseudo $R^{2}$	0.151	0.157	0.143	0.149

This table reports the results from the logit regression of Model (1) in the text, where the dependent variables are the dummy variables Fin1 which takes the value of 1 if the amount of net financing is positive and 0 otherwise. Values in parentheses are z-values; ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

Conversely, the regression coefficients for agency problems are significantly negative at the 1% confidence level. This suggests that enterprises with fewer agency problems are inclined to adopt net financing methods, whereas those experiencing more severe agency problems are more likely to choose net investment. These findings robustly support H2.

4.3. Robustness tests

4.3.1 Lagged variables

In the previous analysis, H1 and H2 are both supported. However, it is possible that firms that find it easier to access external funding from financial markets may engage in riskier investments and operations, leading to more volatile performance. Conversely, firms that face greater challenges in raising funds from financial markets may adopt more cautious investment and operational strategies due to their reliance on limited internal resources, resulting in lower performance volatility. This suggests the potential for reverse causality. To address this concern, we follow Zhou et al. (2021) and rerun the regression in Model (1), using lagged variables for our main explanatory factors, performance volatility, and agency problems. Specifically, performance volatility (L. SdROE, L. SdROA) and agency problem variables (L. Ac1, L. Ac2) are measured with a one-period lag. As shown in Table 10, the regression coefficients for lagged performance volatility variables are significantly positive at the 1% confidence level, while the coefficients for lagged agency problem variables are significantly negative at the same confidence level. These findings confirm that firms with higher performance volatility are more likely to adopt a net financing mode of financial market utilization, while those with lower volatility lean toward a net investment mode. Similarly, firms with fewer agency problems are more inclined toward net financing, whereas those with more severe agency problems favor net investment. Thus, even after accounting for the potential issue of reverse causality, the results remain robust, reinforcing the validity of H1 and H2.

Table 10.

Regression using the lagged independent variables.

Variable	(1) Fin1	(2) Fin1	(3) Fin1	(4) Fin1
L. SdROE	1.287*** (10.586)	1.311*** (10.784)
L. Ac1	−1.720*** (−6.522)		−1.440*** (−5.537)
L. Ac2		−1.928*** (−13.256)		−1.860*** (−12.879)
L. SdROA			2.380*** (5.669)	2.682*** (6.381)
Size	0.556*** (26.331)	0.543*** (26.273)	0.543*** (25.891)	0.527*** (25.659)
Tangi	3.558*** (26.115)	3.407*** (24.951)	3.552*** (26.256)	3.396*** (25.047)
Age	−0.101*** (−2.917)	−0.089** (−2.564)	−0.062* (−1.798)	−0.051(−1.483)
Growth	0.473*** (9.542)	0.485*** (9.825)	0.454*** (9.354)	0.473*** (9.784)
East	−0.186*** (−4.297)	−0.220*** (−5.054)	−0.194*** (−4.516)	−0.229*** (−5.268)
SOE	−0.157*** (−3.130)	−0.206*** (−4.079)	−0.159*** (−3.172)	−0.204*** (−4.044)
Inde	0.979*** (2.740)	0.958*** (2.670)	1.008*** (2.833)	0.996*** (2.786)
Top1	−2.276*** (−16.057)	−2.249*** (−15.796)	−2.282*** (−16.151)	−2.260*** (−15.923)
Duality	0.001(0.026)	0.011(0.249)	−0.001(−0.015)	0.009(0.217)
Constant term	−11.814*** (−23.302)	−11.289*** (−22.703)	−11.605*** (−22.992)	−11.017*** (−22.225)
Industry/year	Control	Control	Control	Control
Sample size	18,075	18,075	18,075	18,075
Pseudo $R^{2}$	0.148	0.154	0.142	0.148

This table reports the results from the logit regression of Model (1) in the text, but all the main explanatory variables are measured with one year’s lag. Values in parentheses are z-values; ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

4.3.2 Alternative model

The previous regression results demonstrate how performance volatility and agency problems influence the likelihood of a company deciding how to utilize financial markets, highlighting a qualitative relationship. In this section, we further explore the quantitative relationship. To this end, we estimate Model (2), where the dependent variable is a continuous variable representing the net financing amount.

As shown in Table 11, the regression coefficients for performance volatility are consistently positive and statistically significant at the 1% confidence level. Likewise, the regression coefficients for agency problems are consistently negative and also significant at the 1% confidence level. These findings are consistent with the earlier analysis and further reinforce H1 and H2. Apart from the statistical significance of the effects of performance volatility and agency problems on the use of financial markets, our regression also indicates economic significance. For example, the coefficient for SdROE is 0.164, suggesting that when a firm’s performance, measured by ROE, changes by one standard deviation, the firm’s net financing increases, on average, by 16.4% of its total assets. This analysis not only corroborates the qualitative insights but also highlights the robustness of the results, offering additional evidence for the relationship between performance volatility, agency problems, and financial market utilization.

Table 11.

OLS regression on continuous net financing measurement.

Variable	(1) Fin2	(2) Fin2	(3) Fin2	(4) Fin2
SdROE	0.164*** (32.508)	0.168*** (33.864)
Ac1	−0.089*** (−4.845)		−0.050*** (−2.659)
Ac2		−0.165*** (−15.874)		−0.159*** (−14.962)
SdROA			0.486*** (18.315)	0.528*** (20.272)
Size	0.059*** (47.511)	0.057*** (46.566)	0.059*** (46.394)	0.056*** (45.054)
Tangi	0.339*** (40.808)	0.322*** (38.848)	0.348*** (41.274)	0.330*** (39.200)
Age	−0.004* (−1.829)	−0.003(−1.233)	−0.001(−0.235)	0.001(0.375)
Growth	0.022*** (7.766)	0.018*** (6.403)	0.022*** (7.388)	0.017*** (5.801)
East	−0.022*** (−7.610)	−0.024*** (−8.506)	−0.024*** (−8.177)	−0.026*** (−9.098)
SOE	−0.013*** (−3.757)	−0.016*** (−4.854)	−0.011*** (−3.280)	−0.015*** (−4.274)
Inde	0.049** (2.030)	0.050** (2.119)	0.050** (2.063)	0.053** (2.190)
Top1	−0.182*** (−19.152)	−0.180*** (−19.097)	−0.183*** (−18.947)	−0.182*** (−18.920)
Duality	0.005(1.531)	0.006* (1.896)	0.005(1.528)	0.006* (1.907)
Constant term	−1.236*** (−43.687)	−1.155*** (−41.771)	−1.245*** (−43.146)	−1.153*** (−40.803)
Industry/year	Control	Control	Control	Control
Sample size	21,647	21,647	21,647	21,647
$R^{2}$	0.301	0.308	0.278	0.285

This table reports the results from the logit regression of Model (2) in the text, where the dependent variables are the continuous measurement of the amount of net financing Fin2. Values in parentheses are z-values; ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

4.4. Analysis of the influence mechanism

The earlier analysis demonstrates that companies with high-performance volatility and low agency problems are more likely to adopt a net financing approach, while those with low volatility and significant agency problems tend to favor a net investment approach. The next step is to explore the underlying mechanisms driving this phenomenon. Accordingly, this article advances to a theoretical exploration and empirical investigation of the transmission mechanisms through which performance volatility and agency problems influence the formation of net financing utilization.

4.4.1 Free cash flow

High-performance volatility reflects heightened competition and business risks in the enterprise’s product market. When faced with intense competition and challenges in maintaining market share and revenue, a firm’s operating income becomes unstable, resulting in erratic free cash flow. In such scenarios, the firm may lack sufficient internal funds to sustain its business activities. According to the pecking order theory (Myers and Majluf, 1984), firms prioritize raising external funds from financial markets to meet their capital needs in the product market. Furthermore, the scarcity of surplus funds limits the firm’s ability to invest in financial markets. As a result, the firm is more likely to adopt a net financing approach in its financial market utilization.

To investigate the mediating role of free cash flow, this article constructs mediating effect models (3) and (4), following the methodology outlined by Baron and Kenny (1986). These models incorporate control variables, including agency problems and other relevant factors, as specified in prior research.

F i n 2_{i, t} = β_{0} + β_{1} S d R O E_{i, t} + \sum C o n t r o l s + \sum Y e a r + \sum I n d u s t r y + ε_{i, t}

(3)

F C F_{i, t} = β_{0} + β_{1} S d R O E_{i, t} + \sum C o n t r o l s + \sum Y e a r + \sum I n d u s t r y + ε_{i, t}

(4)

F i n 2_{i, t} = β_{0} + β_{1} S d R O E_{i, t} + β_{2} F C F_{i, t} + \sum C o n t r o l s + \sum Y e a r + \sum I n d u s t r y + ε_{i, t}

(5)

The intermediate variable in this analysis is free cash flow (FCF), calculated as: (net cash flow from operating activities in the cash flow statement + total interest expense—cash from fixed, intangible, and long-term asset acquisitions)/total assets at the end of the period. A mediating effect is confirmed if the coefficient on FCF is significant in both Model (3) and Model (4), and if it remains significant in Model (5) with a reduced absolute value compared to Model (3). The mediation test results are summarized in Table 12. Columns (1)–(3) presents the regression results for Models (3)–(5), respectively, demonstrating that free cash flow serves as a mediator in the relationship between performance volatility and companies’ financial market utilization. We also perform the same test using SdROA and Ac2 as alternative measurements. The results are consistent.

Table 12.

Mechanism analysis on free cash flow.

Variable	(1) Fin2	(2) FCF	(3) Fin2
SdROE	0.164*** (32.508)	−0.017*** (−8.278)	0.153*** (31.387)
FCF			−0.656*** (−39.604)
Ac1	−0.089*** (−4.845)	−0.106*** (−14.500)	−0.158*** (−8.880)
Size	0.059*** (47.511)	−0.002*** (−3.298)	0.058*** (48.303)
Tangi	0.339*** (40.808)	−0.012*** (−3.655)	0.331*** (41.265)
Age	−0.004* (−1.829)	0.013*** (15.391)	0.005** (2.240)
Growth	0.022*** (7.766)	−0.004*** (−3.788)	0.020*** (7.020)
East	−0.022*** (−7.610)	0.001(1.131)	−0.021*** (−7.576)
SOE	−0.013*** (−3.757)	−0.004*** (−2.647)	−0.015*** (−4.603)
Inde	0.049** (2.030)	−0.018* (−1.940)	0.037(1.579)
Top1	−0.182*** (−19.152)	0.046*** (12.263)	−0.152*** (−16.474)
Duality	0.005(1.531)	−0.005*** (−4.226)	0.001(0.446)
Constant term	−1.236*** (−43.687)	0.017(1.548)	−1.224*** (−44.824)
Industry/year	Control	Control	Control
Sample size	21,647	21,647	21,647
$R^{2}$	0.301	0.059	0.348

This session tests the mediating role of free cash flow. The columns (1)–(3) report the regression results from the models (3)–(5) in the text. t-values are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

4.4.2 Minimizing the agency problem

In firms burdened by severe agency problems, managers exert significant influence over financial decisions, often driven by personal interests arising from incomplete contracts with shareholders (Hooghiemstra, 2008; Jensen, 1986; Shleifer and Vishny, 1997; Xiao et al., 2021). This influence manifests in two primary ways: First, managers constrained by financing restrictions imposed by financial markets often seek to limit external financing. Second, managers, aiming for short-term success, adopt low-risk, low-cost financial strategies that prioritize assets yielding quick returns. As a result, their self-serving approach fosters a preference for conservative financial policies with ample liquidity, ultimately leading to a net investment-oriented financial market utilization pattern.

On the other hand, in companies with minimal agency problems, the objectives of managers and shareholders are more closely aligned. Managers are motivated to prioritize long-term corporate growth over self-interest. They allocate operating surpluses into core operations while actively seeking external financial market funding for investments, such as R&D, and are more willing to assume the associated risks. This approach creates a higher demand for external financing. However, these managers operate under stricter constraints, curbing their discretion in financial decisions, especially concerning large-scale investments in financial markets. Consequently, in companies with fewer agency problems, managers’ reduced self-interest fosters a net financing approach in financial market utilization.

To investigate the mediating role of managerial self-interest, this study employs Models (6)–(8). The “Controls” include the set of control variables, such as performance volatility, consistent with prior research definitions.

F i n 2_{i, t} = β_{0} + β_{1} A c 1_{i, t} + \sum C o n t r o l s + \sum Y e a r + \sum I n d u s t r y + ε_{i, t}

(6)

A b m a n a c o m p_{i, t} = β_{0} + β_{1} A c 1_{i, t} + \sum C o n t r o l s + \sum Y e a r + \sum I n d u s t r y + ε_{i, t}

(7)

F i n_{2 i, t} = β_{0} + β_{1} A c 1_{i, t} + β_{2} A b m a n a c o m p_{i, t} + \sum C o n t r o l s + \sum Y e a r + \sum I n d u s t r y + ε_{i, t}

(8)

The mediating variable, managerial self-interest (Abmanacomp), is proxied by managerial overcompensation. Following the methodology of Wang and Wang (2018), this is calculated as the difference between managerial compensation and entitlement compensation, derived using established research methodologies. Managerial compensation represents the average total compensation of the top three executives, logarithmically transformed for analysis, while entitlement compensation (Manacomp) is estimated through regression model (9).

\begin{array}{l} M a n a c o m p_{i, t} = γ_{0} + γ_{1} R O A_{i, t} + γ_{2} R O A_{i, t - 1} + γ_{3} S O E_{i, t} + γ_{4} D u a l_{i, t} + γ_{5} M a n a h l d_{i, t} \\ + γ_{6} L e v_{i, t} + γ_{7} B d s_{i, t} + γ_{8} S i z e_{i, t} + \sum Y e a r + \sum I n d u s t r y + ε_{i, t} \end{array}

(9)

The variables are defined as follows: Manahld represents the percentage of management ownership, Bds denotes the asset-liability ratio, and other variables retain consistent definitions from prior studies. Specifically, Manahld signifies management’s shareholding ratio, Bds indicates the balance sheet ratio, and Bds also represents board size (with all other variables retaining their previously established definitions). A higher value of managerial self-interest (Abmanacomp) reflects more pronounced self-serving behavior. Due to missing values for some variables, the sample size for this mechanism test is 20,436. The analysis results, presented in Table 13, columns (1)–(3), demonstrate that managerial self-interest acts as a mediating factor in the relationship between agency problems and firms’ utilization of financial markets. We also perform the same test using SdROA and Ac2 as alternative measurements. The results are consistent.

Table 13.

Mechanism analysis on managerial self-interest.

Variable	(1) Fin2	(2) Abmanacomp	(3) Fin2
Ac1	−0.097*** (−5.124)	0.525*** (10.188)	−0.091*** (−4.766)
Abmanacomp			−0.013*** (−4.879)
sdROE	0.161*** (30.485)	0.026* (1.842)	0.161*** (30.563)
Size	0.060*** (46.707)	0.015*** (4.240)	0.060*** (46.858)
Tangi	0.339*** (39.742)	−0.047** (−2.041)	0.339*** (39.691)
Age	−0.001(−0.225)	−0.015** (−2.519)	−0.001(−0.311)
Growth	0.023*** (7.729)	−0.005(−0.602)	0.023*** (7.713)
East	−0.023*** (−7.778)	0.183*** (22.875)	−0.021*** (−6.913)
SOE	−0.016*** (−4.563)	0.063*** (6.737)	−0.015*** (−4.331)
Inde	0.062** (2.530)	−0.068(−1.016)	0.061** (2.497)
Top1	−0.180*** (−18.482)	−0.218*** (−8.255)	−0.183*** (−18.743)
Duality	0.006* (1.952)	−0.009(−1.033)	0.006* (1.918)
Constant term	−1.267*** (−43.411)	−0.378*** (−4.772)	−1.272*** (−43.574)
Industry/year	Control	Control	Control
Sample size	20,436	20,436	20,436
	0.304	0.036	0.305

This session tests the mediating role of managerial self-interest. The columns (1) to (3) report the regression results from the models (6) to (8) in the text. t-values are reported in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.

5. Conclusion and implications

Based on the functional supply of financial markets, real enterprises exhibit significant diversity in their utilization patterns. This study categorizes enterprises into net financing and net investment types, aiming to identify the drivers behind these distinct approaches. The findings reveal a prevalent trend toward net financing utilization, where enterprises frequently leverage financial markets to support real operations. However, a declining trend is observed in the proportion and scale of net financing companies, while net investment companies are on the rise. This shift suggests that financial markets increasingly act as resource allocation channels similar to product markets. This conclusion aligns with regional heterogeneity tests, demonstrating that as China’s financial markets develop, financial progress in non-eastern regions gradually converges with that in the eastern region. A robust financial market provides enterprises with diversified financing channels and investment opportunities, enabling more flexible market utilization. This flexibility fosters the growth of net investment-oriented enterprises.

Furthermore, heterogeneity in financial market utilization is evident across industries, regions, and ownership structures. Enterprises with high-performance volatility and low agency problems are more inclined toward net financing strategies. Further analysis indicates that performance volatility impacts utilization patterns by affecting free cash flow, while agency problems influence these patterns through managerial self-interest. In light of these findings, this study proposes the following recommendations:

First, regulators should strengthen the relevance, effectiveness, and resilience of financial regulations. Given the diverse utilization patterns among enterprises, targeted measures aligned with national strategies are crucial. For net investment-oriented enterprises, stricter oversight and enhanced disclosure of fundraising activities are essential to curb capital idling and shadow banking. Encouraging these enterprises to refocus on core business activities and scale up real investments can enhance their industrial competitiveness while balancing risks and returns. For net financing companies, governments should prioritize the development of diverse financing channels and foster financial innovation. This approach ensures a mutually reinforcing relationship between financial development and the real economy, promoting sustainable economic growth and resilience.

Second, creating a conducive institutional environment is key to increasing real investments and innovation, fostering economic growth. Enterprises with less volatile performance tend to adopt net investment strategies, reflecting a gap between the virtual and real economies and a shortage of viable investment opportunities in the latter. Addressing this requires deeper financial sector reforms to dismantle excessive financial monopoly profits and redirect funds toward real economic activities. Efforts should focus on expanding opportunities for real investment and improving capital allocation channels within real enterprises. Achieving these demands risk-taking, innovation, and comprehensive national system construction. For instance, initiatives like the “new development paradigm,” emphasizing domestic circulation alongside international integration, and the high-quality development goals outlined in the 19th National Congress report, provide new perspectives and opportunities for China’s real economy.

Third, financial market reforms must align with overarching national development goals, ensuring that financial markets consistently serve the real economy. The evolving utilization patterns of China’s enterprises reflect fundamental changes in the financial market’s functions, scale, structure, and industry composition. These shifts, driven by the rapid advancement and structural upgrading of the real economy, impose new demands on financial regulation. Regulators must accurately guide these financial market changes to align with real economic developments, preventing divergence. Ultimately, the financial market should evolve to serve the real economy effectively, ensuring that financial market advancements contribute meaningfully to sustained economic growth.

Key theoretical and practical implications

This study examines how real businesses in China utilize financial markets. It identifies a declining trend in the proportion and scale of net financing companies, while net investment companies are on the rise.

Enterprises exhibit significant variations in financial market utilization across industries, regions, and ownership structures.

Enterprises with high-performance volatility, which impacts their free cash flows, are more likely to adopt net financing strategies.

Enterprises with low agency problems tend to favor net financing strategies, with low managerial self-interest acting as an intermediate factor.

Footnotes

Acknowledgements

We would like to sincerely thank the editor and reviewers for accepting the manuscript and for their valuable comments and suggestions, which greatly helped improve the quality of the work.

Final transcript accepted 12 March 2025 by Philip Gharghori (AE Finance).

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Min Bai

Notes

References

Aikman

Haldane

Nelson

(2015) Curbing the credit cycle. The Economic Journal 125: 1072–1109.

Almeida

Campello

(2007) Financial constraints, asset tangibility, and corporate investment. The Review of Financial Studies 20: 1429–1460.

Ang

Cole

Lin

(2000) Agency costs and ownership structure. The Journal of Finance 55: 81–106.

Bai

Zhang

(2024) The virtual degree of leveraged capital within Chinese enterprises. International Journal of Business & Economics (IJBE) 9: 67–88.

Bai

Zhang

Zhao

(2024) Unveiling the driving forces behind high non-performing loans in zombie enterprises: An executive-centric study. International Journal of Managerial Finance 20: 1320–1341.

Banerjee

Duflo

(2007) The economic lives of the poor. Journal of Economic Perspectives 21: 141–167.

Baron

Kenny

(1986) The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology 51: 1173–1182.

Berger

Ofek

Yermack

(1997) Managerial entrenchment and capital structure decisions. The Journal of Finance 52: 1411–1438.

Bernanke

(1990) The Federal Funds Rate and the Channels of Monetary Transmission. Cambridge, MA: National Bureau of Economic Research.

10.

Chaney

Sraer

Thesmar

(2012) The collateral channel: How real estate shocks affect corporate investment. American Economic Review 102: 2381–2409.

11.

Chen

Ren

Zha

(2018) The nexus of monetary policy and shadow banking in China. American Economic Review 108: 3891–3936.

12.

Cheng

(2008) Board size and the variability of corporate performance. Journal of Financial Economics 87: 157–176.

13.

Claessens

Laeven

(2003) Financial development, property rights, and growth. The Journal of Finance 58: 2401–2436.

14.

Comin

Philippon

(2005) The rise in firm-level volatility: Causes and consequences. NBER Macroeconomics Annual 20: 167–201.

15.

Cook

Tang

(2010) Macroeconomic conditions and capital structure adjustment speed. Journal of Corporate Finance 16: 73–87.

16.

Devos

Dhillon

Jagannathan

, et al. (2012) Why are firms unlevered? Journal of Corporate Finance 18: 664–682.

17.

Dinç

(2005) Politicians and banks: Political influences on government-owned banks in emerging markets. Journal of Financial Economics 77: 453–479.

18.

Ding

Guariglia

Knight

, et al. (2021) Negative investment in China: Financing constraints and restructuring versus growth. Economic Development and Cultural Change 69: 1411–1449.

19.

Dong

Guo

(2019) Management equity incentives, two jobs and corporate over-indebtedness: An analysis based on two agency theories. Contemporary Finance and Economics 1: 119–130. [In Chinese]

20.

Drobetz

Wanzenried

(2006) What determines the speed of adjustment to the target capital structure? Applied Financial Economics 16: 941–958.

21.

Duchin

Gilbert

Harford

, et al. (2017) Precautionary savings with risky assets: When cash is not cash. The Journal of Finance 72: 793–852.

22.

Hachem

Song

(2021) Liquidity rules and credit booms. Journal of Political Economy 129: 2721–2765.

23.

Han

Qiu

(2007) Corporate precautionary cash holdings. Journal of Corporate Finance 13: 43–57.

24.

Heaton

(2002) Managerial optimism and corporate finance. Financial Management 31: 33–45.

25.

Hooghiemstra

(2008) East—west differences in attributions for company performance: A content analysis of Japanese and US corporate annual reports. Journal of Cross-Cultural Psychology 39: 618–629.

26.

Hsieh

Klenow

(2009) Misallocation and manufacturing TFP in China and India. The Quarterly Journal of Economics 124: 1403–1448.

27.

Jensen

(1986) Agency costs of free cash flow, corporate finance, and takeovers. The American Economic Review 76: 323–329.

28.

Jensen

Meckling

(2019) Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure. Gower, MO: Corporate Governance, pp. 77–132.

29.

Johnson

(2003) Debt maturity and the effects of growth opportunities and liquidity risk on leverage. The Review of Financial Studies 16: 209–236.

30.

Koopman

Kräussl

Lucas

, et al. (2009) Credit cycles and macro fundamentals. Journal of Empirical Finance 16: 42–54.

31.

Korajczyk

Levy

(2003) Capital structure choice: Macroeconomic conditions and financial constraints. Journal of Financial Economics 68: 75–109.

32.

Liu

Huang

Cao

(2007) Research on the mechanism of financial agglomeration affecting the real economy. Management World 4: 152–153. [In Chinese]

33.

Shi

(2014) Research on the impact of macroeconomic factors on corporate capital structure—Another discussion on the relationship between three capital structure theories. Contemporary Economic Science 366: 95–105+125–126. [In Chinese]

34.

Meng

Xiong

Xiao

, et al. (2023) The effect of overseas investors on local market efficiency: Evidence from the Shanghai/Shenzhen–Hong Kong Stock Connect. Financial Innovation 9: 42.

35.

Myers

Majluf

(1984) Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics 13: 187–221.

36.

Nissim

Penman

(2001) Ratio analysis and equity valuation: From research to practice. Review of Accounting Studies 6: 109–154.

37.

Orhangazi

(2008) Financialisation and capital accumulation in the non-financial corporate sector: A theoretical and empirical investigation on the US economy: 1973–2003. Cambridge Journal of Economics 32: 863–886.

38.

Orhangazi

(2019) The role of intangible assets in explaining the investment–profit puzzle. Cambridge Journal of Economics 43: 1251–1286.

39.

Penman

(2013) Financial Statement Analysis and Security Valuation. New York: McGraw-Hill.

40.

Porta

Lopez-de-Silanes

Shleifer

, et al. (1998) Law and finance. Journal of Political Economy 106: 1113–1155.

41.

Restuccia

Rogerson

(2017) The causes and costs of misallocation. Journal of Economic Perspectives 31: 151–174.

42.

Schumpeter

(1936) The General Theory of Employment, Interest and Money. New Delhi, India: Atlantic Publishers and Distributors.

43.

Seo

Kim

(2012) Financialization and the slowdown in Korean firms’ R&D investment. Asian Economic Papers 11: 35–49.

44.

Shleifer

Vishny

(1997) A survey of corporate governance. The Journal of Finance 52: 737–783.

45.

Singh

Davidson

IIIWN

(2003) Agency costs, ownership structure, and corporate governance mechanisms. Journal of Banking & Finance 27: 793–816.

46.

Stulz

(1990) Managerial discretion and optimal financing policies. Journal of Financial Economics 26: 3–27.

47.

Theurillat

Corpataux

Crevoisier

(2010) Property sector financialization: The case of Swiss pension funds (1992–2005). European Planning Studies 18: 189–212.

48.

Tornell

(1990) Real vs. financial investment: Can Tobin taxes eliminate the irreversibility distortion? Journal of Development Economics 32: 419–444.

49.

Wang

Cao

Yang

(2017) Whether financialization of real firms promotes or inhibits corporate innovation: An empirical study based on listed companies in China’s manufacturing industry. Nankaimanagement Review 20: 155–166. [In Chinese].

50.

Wang

(2018) Accrual manipulation and annual report text message tone manipulation research. Accounting Research 4: 51–45. [In Chinese]

51.

Fang

(2021) 30 years of China’s capital market: Exploration and change. Finance and Trade Economics 42: 20–36. [In Chinese].

52.

Xiao

Bai

Qin

, et al. (2021) Financial slack and inefficient investment decisions in China. Managerial and Decision Economics 42: 920–941.

53.

Xiong

Xiao

Bai

, et al. (2023) The religion effect on corporate cash holding in China: Buddhism and Taoism. International Journal of Finance & Economics 28: 4420–4457.

54.

Yang

Xiao

Xiong

, et al. (2024) Is “collaborative governance” or “competitive collusion” the effect of cross-ownership on corporate innovation? China Accounting and Finance Review. Epub ahead of print 15 November. DOI: 10.1108/CAFR-08-2023-0090.

55.

Zhang

Hao

(2022) Investment and financing maturity mismatch and corporate bond default risk. Research on Financial Issues 02: 63–71.

56.

Zhou

Bai

, et al. (2022) The COVID-19 outbreak and corporate cash-holding levels: Evidence from China. Frontiers in Psychology 13: 942210.

57.

Zhou

Zheng

(2021) How digital finance helps corporate innovation: Based on the perspective of financing constraints and information constraints. Industrial Economics Review 12: 49–65. [In Chinese]