A Stochastic Approach Analysing Enterprises ’ Investment following Financing Reform in China

This paper explores the change of enterprises ’ investment following the ﬁ nancing system reform through the established stochastic investment model. In this constructed model, ﬁ nancing property, market-oriented reform, and government intervention are regarded as a stochastic process. Furthermore, the modern China ’ s economic situation is interpreted to analyze the enterprises ’ investment by government intervention plan combined with he deduced proposition from the stochastic investment model. The results provide a depth understanding for characteristics of the enterprise in China that the steady capital of state-owned enterprises ’ investment is higher than that of non-state-owned enterprises without government intervention before completing ﬁ nancing reform. Although government intervention can increase the investment level of state-owned enterprises, doing so increases the turbulence of the market economy. Additionally, the impact of government-led ﬁ nancing reform on enterprises ’ investment is asymmetrical. Promoting market-oriented, clear-cut ﬁ nancing reform, and reducing government-led investment plans will improve enterprises ’ investment e ﬃ ciency and stabilize China ’ s economic development. The present paper provides a speci ﬁ c future orientation of China ’ s ﬁ nancing reform determining the level of enterprises ’ investment.


Introduction
China is a country with consumer power shortage, and its economic growth mainly depends on corporate investment. China can increase corporate investment through financing reform so as to get through the crisis smoothly, when the market environment is in a slump. The financing reform increased the potential output of enterprises and made economic development more resistant to economic shocks [1]. But financing reform also has some shortcomings such as placing some enterprises lacking political connection under huge financial burden due to unfair institutional contracts [2]. Yet, existing studies on enterprise investments have not identified the mechanism by which financing reform changes enterprise investment behavior.
Previous studies have shown that the financing reform by government might distort the level of enterprises' investment due to political intervention, resulting in inefficient investment [3][4][5]. Enterprise managers may abuse freer cash flow for overinvestment, leading to more inefficient investments [6,7]. Other studies have shown that the financing reform was valuable in promoting efficient enterprises' investment during the financial crisis [8][9][10][11], which can better remedy market failures and compensate for inefficient market allocation [12,13]. Financing reform can better ensure reform in enterprises' investment and economic growth faced with economic recession and external shocks, which may result in a national fiscal deficit in the short term, but improve the output of enterprises and employment in the medium term and generate a huge fiscal surplus in the long run [14]. With the development of a market-oriented system, the rewarding priority of reform would bring the knowledge transfer spillovers [15]. This paper first attempts to build a new stochastic investment model to evaluate the consequence of the governmentbased financing reform. When the economic institution is relatively simple, the level can be adjusted through financing reform if something goes wrong. However, when the economic institution is more complicated, this transaction cost of correcting course is too large. In the past, many economic problems and social contradictions have been solved by massive investment expansion. If the economic growth rate declines, many contradictions and problems were revealed due to declining capital returns from enterprises' investment. The enterprise investment is related to the state administrative institution [16]. So, due to the random volatility of China's policy plan, China's financing reform defers to the Markov process containing the correction mechanism with independent increments; that is, the factor of this reform consists of the institutional constant, correction mechanism, and white noise, making deterministic quantization of financing reform difficult.
An answer to identify the enterprises' investment efficiency is relevant to the considerations for China's external capital and the government intervention-oriented plan. Based on the assumption that external capital has infinite flexibility, enterprise financing has depended to some extent on its demand for debt [17]. Unfortunately, China's capital markets, especially the stock market and bond market, were relatively unsophisticated, and bank credit has been the main source of financing for Chinese enterprises [18,19]. If the Chinese central government intervened in enterprises' investment in response to economic depression, it would loosen the bank credit supply by implementing economic stimulus plans and release more liquidity to the market to remedy the failure of the capital market [9,10,11,20]. However, Liu et al. pointed out that this policy mechanism is of great significance to the investment of state-owned enterprises, but not significant in non-state-owned enterprises' investment [21]. Despite increasing the investment in state-owned enterprises, the financing system may fail to maintain the effectiveness of these investments from stateowned ones, thus leading to overcapacity from expanding production due to government intervention. In China, there are double atrophy of output and investment in state-owned enterprises and double rise in output and investment in nonstate-owned enterprises [22] Also, the paper differentiates the state-and non-stateowned enterprises' investment function from macroenterprise investments. In China, non-state-owned enterprise is subject to unfair discrimination compared to state-owned one with which the state power has a good relationship. Although individual non-state-owned enterprises maintain a relationship with state-owned banks or governments, their ability to obtain information is less than that of state-owned enterprises; they also suffer from the discriminatory treatment of bank credit, in which banks are reluctant to lend to non-state-owned enterprises for investment [23,24]. Even if non-state-owned enterprises is lent based on reliable business judgments [25], depending on their reputation, most of the credit goes to state-owned enterprises [21,26]. In the case of incomplete social information, state-owned enterprises increase their level of investment, and more investments are based on private interests or the trend to crony capitalism rather than economies of scale. Wang et al. pointed out that the government may use political power to control state-owned enterprises to achieve private goals [27]. Government intervention creates market distortions under certain circumstances and leads to improper allocation of market resources [28,29].
This paper relaxes the assumption for the certainty of the economic situation, namely, the accurate known and predicted economic status, and views the financing as a Brownian movement condition. There are inherent unobservable noneconomic fluctuations [30,31]; that is, an economic entity cannot predict the fluctuation of economic conditions caused by disturbances in the process of economic development, and such disturbances change the information set of enterprises for future economic investment output. Truly, there are macroeconomic uncertainties in the process of financing institution reform. If the reform decision-making adjustment cost function is nonconvex, some irreversible government-led investment projects will increase the real option value of enterprise investment [32]. The uncertainty in institutional reforms, namely, the unknown and unpredicted economic consequence resulting from financing reform, can lead to an increase in the market risk premium, so that a rise in risk premiums increases corporate financing costs or difficulties and has a significant negative impact on enterprise investment [33]. In addition, the return on capital also increases with the rise in the financing cost, so the drive to perform the investment will decline, and cash holdings will tend to increase, resulting in a liquidity preference or ambiguity aversion behavior [34,35].
This paper contributes to the growing body of literature on enterprise investment following financing reform by providing novel insights into the mechanism of enterprises' investment. The enterprises' investment effect by financing reform is reconstructed to investigate the role of government intervention through the built stochastic investment model with the consideration for the uncertainty from the economic consequence produced by institutional arrangement. Existing researches are based on the influence mechanism of financing reform and enterprise investment in developed Western capitalist countries and rarely consider the relationship between the financing reform in this particular economy and enterprise investment. The central role of China's state-owned enterprises means that its economic development is very different from western developed capitalist countries. Opportunistic behaviors such as political corruption and rent-seeking are inevitable, and distortion of market resource allocation may lead to biased corporate investment. Therefore, this paper focuses on the financing system and the random fluctuation of the reform process 2 Advances in Mathematical Physics and explores the relationship between the financing reform and the enterprise investment level in China. The rest of the paper is organized as follows. The second part establishes the stochastic investment model following the financing reform; the third part solves the model and obtains the investment level of the enterprise in the steady state and the enterprise investment in the process of financing reform. The fourth part further explores changes in levels of investment in combination with China's economic development and provides a deeper deconstruction of the "structural imbalance" of the current depressed Chinese economy using propositions derived from the third part. The fifth part is the conclusion.

Model
2.1. The Basic Model of Enterprises' Investment. Enterprises in China, especially state-owned enterprises, have different business objectives due to the special arrangement of the property rights system, which goes beyond the maximization of profits (i.e., the sales income of enterprises minus the operating cost of raw materials) as in the West, or the maximization of outputs. It is a common practice to expand investment scale and production capacity to win political benefits because of the objectives of state-owned enterprise management (since the equilibrium prices of market commodities were unknown during China's planned economy era, the SOEs chose a quantitative index with the maximizations of their output instead of a market profit or income. Furthermore, SOE managers with correlated administrative levels at the company level sought a strong incentive to maximize productions due to the returns on companies' control property rights. For the SOE, expanding the investment scale is the common method winning the political asset (Li, 2006)). Considering that state-owned enterprises with output maximization as their quantitative objective have the dual objective of profits and production, the profit function of state-owned enterprises expressed as an instantaneous profit function is pf ðK, LÞ − CðK, L, IÞ, where f ð⋅Þ and Cð⋅Þ represent production function and cost function, with p, K, L, and I representing relative prices, capital, labor, and investment related to product production. ðθ − 1Þpf ðK, LÞ represents the production objective of enterprises. θ ≥ 1 measures the impact of property rights on firms' operations. When θ = 1, property rights reform is completed. The bigger θ is, the lower the marketization from property rights reform. The actual financing cost can be characterized by the relative price of finished goods to capital goods implicit in China. We modeled χ (χ ≥ 1) as the commercialization of banks. When χ = 1, the commercialization of banks is complete, and soft constraints are gone. The bigger the χ, the lower the commercialization of banks and the stronger the impact of soft constraints.
Of course, every aspect of economic life in China, especially dependence on government's top-down planning, involves government intervention. The government in China, according to its judgment of the economic situation, exercises its management power over state-owned enterprises and intervenes in their operation at its discretion either through administration or direction. State-owned enterprises in China are at present still subordinate to the administrative management of SASAC (State-owned Assets Supervision and Administration Commission), so the intervention of government administrative power will inevitably affect the operation of enterprises. If the government wishes to pursue rapid national economic development and requires the cooperation of state-owned enterprises, stateowned enterprises need to accomplish certain quantitative tasks, although doing so may violate the interests of stateowned enterprises. λ (λ > 0) is used for the measurement of government intervention. If λ > 1, the government hopes to expand its capacity; otherwise, the government imposes limitations over the operation of enterprises. Therefore, we have the following decision objective: The difference in state-and non-state-owned enterprises is their business objectives and financing costs, so the specific modeling must be carried out in different areas for state-and non-state-owned enterprises. If two types of enterprises are assumed to have the same production technology, the investment behavior of non-state-owned enterprises could be regarded as a special case of the stateowned enterprise model; for instance, λ = χ = θ ≡ 1. Because non-state-owned enterprises are not administratively regulated by the Chinese government, which holds surplus control over many banks, they without the benefits of lowinterest bank credit have only one profit goal to make market profits rather than political profits. Therefore, we discuss non-state-owned enterprises as a special case of state-owned enterprises, when λ = χ = θ ≡ 1.

The Stochastic Process Setting.
A continuous time uncertain investment model is usually characterized by a random process. This paper divides the randomness into two parts: the stochastic process setting and its correlation.
It is necessary to analyze the characteristics of China's economic transformation, among which gradual reform is China's overall reform idea, including reform of the property rights system and the commercialization of banks. The miracle of China in the past 40 years of reform and opening-up demonstrates that the most worthwhile reform approach is at exploring methods guided by asymptotic marketoriented reform through government intervention. China's reform process, in fact, includes random disturbances called stochastic deviations. The benchmark, under the condition of complete marketization with the relative price of finished goods to capital goods, property rights reform, and bank commercialization reform through government intervention, is analyzed in accordance with the following geometric Brownian motion equations.
3 Advances in Mathematical Physics where μ p , μ θ , μ χ , and μ λ and σ p , σ θ , σ χ , and σ λ are constants that represent drift rate and variance rate concerning relative price, property design, commercialization of bank reform, and government intervention. dz p , dz θ , dz χ , and dz λ are set in accordance with standard Brownian motion. It is worth noting that there may be uncertainties about labor productivity in a real-life economy, but to simplify this model, stochastic changes are reflected in the random process of relative price of finished goods to capital goods variable p. It is a kind of "real price" with labor productivity or that of "relative price" compared with cost price. If China's economic transformation still adheres to reform and opening up over the long term, μ θ and μ χ should be less than 0, which indicates that property rights reform for stateowned enterprises has improved in the direction of the modern enterprise system, with state-owned commercial banks more and more commercialized but less affected by institutional factors. The government has no policy rules in the early stages of reform, only depending on history in the formulation and operation of policies according the actual situation, so μ λ can be arbitrary, positive or negative, and its sign can be heterogeneous at different times. The profits of enterprises in Equation (1) mainly come from the following three aspects: (1) θχ, the institutional arrangement  (2)-(5) can be classified into four categories. On one hand, assuming that dz λ , dz χ , and dz θ have no correlation with dz p , if X ∈ ðλ, χ, θÞ, with Eðdz X Þ = Eðdz p Þ ≡ 0 and var ðdz X Þ = var ðdz p Þ ≡ dt due to dz in accordance with standard Brownian motion. On another hand, given that the property rights system will affect the dual goal decision-making of state-owned enterprises, and government intervention is defined as the direct administrative meddling of the government in enterprises, there is a relationship between voluntary decision-making and administrative order with ρðdz θ , dz λ Þ = 0, but it is possible that state-owned enterprises remain "loyal" to the government with ρðdz θ , dz λ Þ > 0, and we do not set ρðdz θ , dz λ Þ > 0 to avoid reinforcing the latter conclusion. There are still some positive correlations between λ and χ due to government management between state-owned enterprises and state-owned banks. If the outcome of the government's discretion is to stimulate investment and the economy, it would require the cooperation of state-owned enterprises and state-owned banks. When state-owned enterprises hope to stabilize product returns with the increase of λ, the state-owned banks will provide more credit under government pressure and also will be motivated to ease credit for state-owned enterprises with the increase of χ, so ρðdz λ , dz χ Þ > 0. The current state of the Chinese economy in the Sino-US trade war proves that the central bank has lowered its benchmark to implement a looser monetary policy, with the main credit flowing to state-owned enterprises.
(1) θ depends on the arrangement of the property rights of state-owned enterprises and the behavior characteristics of state-owned enterprises themselves (2) χ depends on the mechanism of "rotten meat in the pot" and state-owned banks themselves In general, relaxing this assumption does not affect the conclusion of mathematical derivation in this paper, and we can conclude the following:

Model Solution and Related Propositional Deduction
3.1. Model Solution. Consider a representative state-owned enterprise whose production function is Cob-Douglas, F s L a s K 1−a s , while the cost of investment is CðIÞ = γI β and β > 1 with no correlation between investment cost and investment stock, C K ′ = 0. Following Equations (1)-(6), let F ≡ λθχp. Given state variable K t (capital stock) and F t (including current institutional arrangement θ t χ t , government intervention λ t , and pure economic factor p t ), state-owned enterprises make investment decisions to maximize the expected discounted present value Vð·Þ of the income stream or investment under their dual objectives. Therefore, the maximum value function in the case of uncertainty is where w is wages and r is interest rate. δ is capital discount rate, and E t is the expectation operator of this value function. The stochastic Bellman equation for this optimization problem is Advances in Mathematical Physics Using Ito's Lemma, we obtain V K and V and V KK and V F F are the first and second derivatives of subscript variables, and V FK = ∂ 2 V/ð∂F∂KÞ.

Advances in Mathematical Physics
First order of Equation (23) related to K t : Equation (24) is the nonlinear second-order differential equation, called the Euler Equation, whose general solution is: where B, A 1 , and A 2 are the undetermined parameters, and η 1 > η 2 are the two roots of r + δ − ðμ − ρ λ,χ σ λ σ χ Þη − ð1/2Þð σ 2 + 2ρ λ,χ σ λ σ χ Þηðη − 1Þ = 0. In fact, q is the shadow price of capital, and we make sure this equation does not diverge, If Equation (26) is substituted into Equation (24), we obtain: If Equation (27) is substituted into Equation (26), we obtain: With Equation (18) and q t ≡ V K , we can obtain the optimal investment level of state-owned enterprises: With the differential of Equation (28) based on Ito's Lemma, we obtain: With the differential of Equation (29) based on Ito's Lemma, we obtain: With Equation (30) substituted into Equation (31), we obtain:

Advances in Mathematical Physics
With Equation (14) and E t dz X = E t dz p = 0, we obtain: With Equation (6), Equation (33), and ðdtÞ 2 = ðdtÞðdz X Þ = 0, we obtain: With Equation (33) and Equation (34) substituted into Equation (32), we obtain the expected growth rate of investment in state-owned enterprises: where F t ≡ λ t θ t χ t p t , μ ≡ μ λ + μ θ + μ χ + μ p , and σ 2 ≡ σ 2 λ + σ 2 θ + σ 2 χ + σ 2 p . The investment function of non-state-owned enterprises is only a special case of the function of state-owned enterprises when λ t = θ t = χ t ≡ 1. The optimal investment level of non-state-owned enterprises is: The expected growth rate of investment in non-stateowned enterprises is:

Government Intervention and Investment of Enterprises
Proposition 3. Government intervention increases investment turbulence when it effectively regulates the investment of state-owned enterprises.
In other words, government intervention can regulate the investment level of state-owned enterprises efficiently and also increase investment turbulence, consistent with the conclusion of Gulen et al. (2016), which is not conducive to the sustainability of investment by state-owned enterprises.

Proposition 4. The impact of government intervention behavior on enterprises' investment level is asymmetric in terms of promotion or restraint.
Proof. See Appendix A.4.
Given in Proposition 3, no matter whether government intervention is one-time or continuous, dz λ > 0, μ λ = 0, or dz λ > 0, μ λ > 0, the sustainability of state-owned enterprises' investment will increase. In addition, if the reform of financing institutions is complete, Edθ t /θ t = μ θ = 0, and Edχ t /χ t = μ χ = 0, Equation (38) can be transformed: The difference in investment growth rate between stateowned enterprises and non-state-owned enterprises is (Equations (37) and (38)) If β < ð2 − aÞ/ð1 − aÞ, the latter item in Equation (28) would be greater than 0. No matter whether μ λ > 0 or μ λ = 0, the investment growth rate of state-owned enterprises is higher than that of non-state-owned enterprises because E t g It − E t g I _ t > 0. Therefore, we obtain Proposition 5.

Advances in Mathematical Physics
Proposition 5. If the effect of government intervention is positive, the investment drive and investment growth rate of state-owned enterprises are higher than those of non-stateowned enterprises. However, if the reform of financing institutions is stopped with β < ð2 − aÞ/ð1 − aÞ, the investment growth rate of state-owned enterprises is increasingly higher than that of non-state-owned enterprises. The emergence of a Sino-US trade war in 2018 has had a devastating impact on China's enterprises, especially for non-state-enterprise operations, due to a sharp drop in foreign demand with the decrease of μ p . Moreover, the absolute level and growth rate of business investment will decrease, which will cause a greater economic decline such as the thunderstorm on current P2P platform in accordance with Equations (29), (35), (36), and (37). The central government in China may decide to adopt a large-scale positive economic stimulus package such as the "New 0.682 Billion Dollars Economic Stimulus Plan" announced in August 2018 to deal with the economic decline the Sino-US trade war triggered, when dz λ > 0 and maybe μ λ > 0. At the same time, when local governments face a decline in their economies, they carried out 1 billion dollar leading investment plan in September 2018, and thus, dz λ and μ λ would be on the rise continuously. This economic stimulus package implemented in 2018, which points that state-owned enterprises and nonstate-owned enterprises with politically connected and local governments must be saved, is a directional rescue compared with the plan in 2008. Revenue from local stateowned enterprises has become the main financing source of local governments since the implementation of fiscal and tax decentralization in 1994. Local governments' economic stimulus packages mainly benefit local state-owned enterprises controlled by local political power, which to a certain extent distorts resource allocation. Of course, stateowned banks also take the initiative to develop robust positive loan policies, such as the targeted cuts to required reserve ratios that release 1,081 billion dollars in credit in October 2018, when ρ λ,χ > 0 brings dz χ > 0 and dχ/ðχdtÞ  1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 e proportion of output or investment (%)

Advances in Mathematical Physics
> 0. When the central government grants property rights to state-owned enterprises, it also gives them political advantages, scale advantages, and market advantages. Stateowned bank loan credit might give first place to stateowned enterprises, especially central enterprises, due to their national political power from their soft constraints on credit, meaning that the investment level of state-owned enterprises may increase. The result may be another example of "the state advancing and the people retreating." The outcome of excessive accumulation of capital and rapid investment of state-owned enterprises would further strengthen the inefficiency of their investment, based on Propositions 1 and 5. When the government realized the problem of inefficient investment, it decided to "restructure" to maintain efficiency. The difficulty and complexity of hindering enterprises' investment and economic growth largely outweigh its positive ability, unless it again deepens the reform of financing institutions, due to the asymmetry explained by Proposition 4 and Proposition 5. According to Proposition 5, although the past evolution of the investment system would lead to "the state retreating and the people advancing," the reverse would happen with massive government intervention, especially if the financing reform is suspended. We also contribute to the current debate on whether "the state is advancing and the people retreating" or not: the development in China is bound to be "the state retreating and the people advancing" as stated by Liu et al. (2016), in contrast to before 1978, but it is "the state advancing and the people retreating" surely in response to recent economic circumstances. Hence, the "state advancing and the people retreating" due to government intervention has a number of costs, including the low efficiency of state-owned enterprise investment and greater instability in the national economy. We thus obtain Proposition 6. Proposition 6. The implementation of market-oriented reform of investment institutions and the reduction of government-led economic stimulus plans will improve the efficiency of investments and reduce volatility.
To be clear, we are not opposed to a stimulus package to deal with the recent economic decline, but we need to know its purpose. The implementation of a robust positive economic stimulus plan in regard to this depression would further aggravate the inefficiency of state-owned enterprises' investment level, as suggested by Proposition 5. Although the current administrative and investment intervention of the Chinese government may achieve good results in the short term, it could also be accompanied by huge costs, such as the low efficiency of investment, repeated construction (the supply of products or services whose quality is not as good as the existing products or services is greater than the social demand) in the long term, and nonperforming loans. The government should focus on domestic demand such as increasing household income and guiding household consumption, which may be more and more effective than the leading positive financing policy. σ 2 and ρ λ,χ σ λ σ χ in Equation (29) provide the impact for uncertainty (second order) of financing reform on stateowned enterprises' investment. If uncertainty about the financing reform (σ θ σ χ σ λ ) increases, state-owned enterprises' investment level will increase as the denominator in Equation (29) decreases. In addition, an increase in uncertainty of reform of financial institutions will improve the expected investment growth rate of state-owned enterprises only with β < ð2 − aÞ/ð1 − aÞ. That is, if the central government in China carries out regular and normative reform of financial institutions (conforming to objective criteria defining the scope of powers and responsibilities of local government and enterprises), the variance rate (σ θ , σ λ , and σ χ ) in Equations (3)-(5) will drop, which means that the investment and expected growth rate of state-owned enterprises will decline. Hence, we obtain Proposition 7.

Proposition 7.
Poor financing reform may increase the investment level of state-owned enterprises, but regular and normative reforms that are fair and just can reduce economic volatility.
Using enterprises' investment to stimulate economic growth has become our "conventional weapon" for macrocontrol. Because the financing reform involves a great deal of uncertainty, it may result in excessive investment among state-owned enterprises and thus have consequences such as waste of resources, environmental pollution, and economic overheating. When these consequences emerge, unless social planners under the financing reform (which could cause economic upheaval), it will be much more difficult to regulate the economic system, in accordance with propositions (5) and (6). Proposition 7 indicates that the evolution of systemic reform and financing reform will reduce price fluctuation and the overinvestment of stateowned enterprises, which will stabilize market supply and alleviate the "distortion" of economic structure.

Conclusion
The financing reform in China in connection with enterprises' investment behavior is generally advancing in the direction of marketization. The reform is gradual, in which government intervention is of a "discretionary" nature. In this paper, we establish the stochastic investment to explore the change of enterprises' (especially state-owned enterprises) investment and interpret the modern China's economy. We can transform the evolution of the financing reform and government intervention into a stochastic process and implant it into a classical enterprise investment model to construct a stochastic theory investment model to analyze the trend of enterprises' investment level in China's transition period, which suggests the following conclusions. First, the advancement of property rights and the financing reform before 2008 may have reduced state-owned enterprises' investment level and outputs in contrast with nonstate-owned enterprises, to some extent bringing about "the state retreating and the people advancing." Second, if 9 Advances in Mathematical Physics the government does not intervene economically and the financing reform remains incomplete, the stable capital accumulation of state-owned enterprises is higher than that of nonstate enterprises. However, it could aggravate this economic instability, resulting in "the state advancing and the people retreating," though government intervention and random reform of financing institutions could increase state-owned enterprises' investment level. The promoting or restraining effect of government intervention on enterprises' investment level appears to be asymmetric. Third, the decrease of government-led investment and marketoriented or normative reform of financing institutions could improve the efficiency of enterprises' investment levels and reduce economic turmoil and promoting social welfare.
For the record, the regular and normative reform of financing institution this paper proposes is an updated version of the top-down design suggested by China's social planners in 2011. It is a pity that this design has not been carried out by China's local governments, and the policy has been implemented through mandatory monopoly plan directives such as the circuit breaker in 2016. Deepening reform of financing institutions has been mentioned in the "opinions on deepening financing reform" published by the CPC Central Committee and State Council in 2016 and in the report of the nineteenth national congress of the CPC in 2017. However, this statement only refers to a specific target, without mentioning a standard, specific implementation plan and specific institutional guarantee. "Shouting slogans" and "singing with high voices" will not develop China's economy and improve the efficiency of social investment. The market-oriented reform of financing institutions regulates the relationship between local government and enterprises, through contract law protected by legal institutions, though the antimonopoly law of 2007 and the law on unfair competition in commodities of 2018 were promulgated without exact criteria for "monopoly" and "unfair competition," and are thus unenforceable. A market-oriented reform of financing institutions needs fair, clear, and equitable rules to balance the interests of local governments and enterprises.
The article only discusses "what should be done," but the question of "how to do it" still needs more in-depth and systematic analysis. Its conclusion could help deepen the understanding of enterprises' investment behavior in response to the reform of financing institutions and current investment phenomena in China's transition period. Of course, this paper also has some shortcomings, which should be supplemented by future research. It has focused its analysis on China's state-owned enterprises and adopted a general setting for non-state-owned enterprises. Non-stateowned enterprises in China also have some irrational characteristics, which should be the focus of future research. In addition, the short-term and long-term effects of China's economic depression as a result of the Sino-US trade war have not been discussed in this paper. What intervention should the government make, such as intervening in the hiring practices or investment of enterprises, especially stateowned enterprises? This question might be the focus of our future research.
impact of drift rate μ θ or χ , and the change of expected growth rate. Above all, the impact of μ θ < 0 and μ χ < 0 changes the expectation of institutional design F t and that of the income discount in the enterprise objective function. Furthermore, institutional design F t and μ θ < 0 or μ χ < 0 both decrease the expected growth rate of enterprises, whereas μ θ or χ mostly decreases the absolute level of enterprises' investment. Therefore, the reform of China's marketization is effective at present.
A.3. Proof of Proposition 3. If the government decides to get involved in expanding or reducing all-over social production capacity, μ λ and dz λ in Equation (5) will change. If government intervention is short-term, dz λ > 0 and μ λ = 0; but if it continues for a long time, dz λ > 0, and μ λ > 0.
Whatever the government intervention, it will greatly affect the investment behavior of state-owned enterprises. If the government hopes that state-owned enterprises will expand their investment level, this will affect the real investment sustainability of state-owned enterprises. As long as the government engages in intervention behavior, dz λ > 0, dλ t /ðλ t dtÞ in Equation (5) will increase, because ρ λ,χ > 0, dz λ > 0, and dχ t /ðχ t dtÞ in Equation (4) will increase. I t in Equation (29) will increase as λ t and χ t increase. When dz λ > 0, μ λ > 0 will make the denominator in Equation (29) decrease and increase the investment level; in addition, E t g 1t increases. Thus, continuous government intervention changes the absolute investment level of enterprises and their expected growth rate of investment.
A.4. Proof of Proposition 4. The effect of continuous government intervention behavior on state-owned enterprises' investment level appears to be asymmetric. We obtain this result by substituting μ ≡ μ λ + μ θ + μ χ + μ p into Equation (35): ðA:4Þ The effect μ λ between positive change ðΔÞ and negative change ð−ΔÞ has the same strength over the expected investment growth rate of state-owned enterprises unless ϕ = 0 is different; that is, E t g It j μ λ =Δ ≠ −E t g It j μ λ =−Δ . Of course, ϕ = 0 is only a coincidental situation. Therefore, government intervention has an asymmetrical impact on the promotion or restraint of enterprises' investment levels.

Data Availability
The data used to support the findings of this study are included within the article.

Conflicts of Interest
The authors declare that they have no competing interest.