Economic Cycle, Uncertainty of Economic Policy and Cash Holding of Listed Companies

Based on the systematic review and economic analysis of the theoretical literature, we consider not only the impact of the economic cycle or the economic policy uncertainty (EPU) on the cash holding ratio, but also the com-prehensive impact of the two on the cash holding rate. We raise the research hypothesis by using the data from 2004-2015 in the A-share listed companies which are listed in Shanghai and Shenzhen securities exchange as research samples. The empirical results show that: 1) There are respectively negative correlation between the economic cycle and the cash holding of listed companies, and positive correlation between the EPU and the cash holding. 2) During the boom, the cash holdings are significantly positive with the current EPU and the last stage; during the recession, the cash holdings of listed companies is significantly negatively correlated with the current EPU, while positive with the last stage. 3) We further examine the role of the economic cycle and the EPU on the cash holding value, and find that EPU will reduce the cash holding value. 4) When the economy is booming, the increase in EPU will reduce the market value of corporate cash holdings, but it is not significant. During recession, the increase in EPU will increase the market value of cash holdings.


Introduction
Cash is equivalent to the company's "blood", and holding liquid assets such as cash help companies to obtain valuable investment opportunities in the future [1]. In the theoretical range, scholars' research on corporate cash holdings has been enduring. As early as 1936, Keynes [1] proposed that the liquidity of cor-the cash holdings of listed companies. EPU mainly affects the company's cash holding decisions in two ways: on the first hand the higher EPU, the more careful management will make investment decisions, and the total amount of corporate investment will fall [6], which will raise the level of corporate cash holdings from the outflow perspective; On the other hand, the increase in EPU will enable financial institutions to treat corporate financing needs more rigorously, increase corporate financing constraints, and enable companies to increase cash holdings for preventive motives.
There is a lot of debate about the relationship between economic cycles and EPU. Most scholars have found that EPU is countercyclical. Johannsen [7] found that fiscal policy uncertainty will lead to a sharp decline in consumption, investment and output, and the economic cycle will decline. Mumtaz et al. [8] determined that the time-varying monetary policy uncertainty was determined by stochastic volatility, and found that when the uncertainty increases, the nominal interest rate, output growth, and inflation fall. But Lee [9] found that as long as the market guarantees that productive companies can survive, EPU will promote exploratory research and innovation, and contribute to overall economic growth.
When the macroeconomic environment changes, the government will carry out macroeconomic regulation and control in order to maintain market stability and achieve governance goals. For example, during recession, stable economic policies can reduce the impact of economic environment on the financing for enterprises, and weakening the impact of the economic cycle on the corporate cash holdings, and vice versa.
The purpose of this paper is to study the impact of macroeconomic cycle and EPU on the cash holding level of listed companies. Compared with other existing literatures, the innovation of this paper lies in the research on the combined impact of the macroeconomic cycle and EPU on the company's cash ratio. This paper theoretically enriches the mechanism and the economic consequences of the interaction of macroeconomic factors affecting business decision-making.
The significance of this paper is to remind policy makers to distinguish the economic cycle from the stage of economic cycle when introducing policies. When the economy is booming, it should play more roles in the market and reduce the intervention in the market. In the recession period, appropriate policy intervention is beneficial to the whole market.
The rest of the paper is structured as follows: the second part is literature review and hypothesis development, the third part is research design, the fourth part is empirical analysis, and finally the conclusion and limitation. The existing literature on cash holding decision-making focuses on the impact of the company's financial status [10], governance structure [11] and industry competition [12] on the micro range. Most of the research on the impact of the macro environment on corporate cash holding decisions is related to financing constraints. Fazzari [13] and others used various investment models to test that the investment of companies with large financing constraints is more sensitive to cash flow. The study by Opler et al. [14] also proves financial constraints and the relevance of the decision of corporate cash holdings.

Literature Review and Hypothesis
Bernanke and Gertler [15] found that the external financing ability of enterprises is largely plagued by economic cycle fluctuations. Since capital markets often have a series of barriers that are considered uncontrollable. External financing costs are common in terms of financing costs, and higher than internal financing costs, resulting in financing constraints. Baum et al. [16] found that when macroeconomic uncertainty is high, companies will increase their cash holdings and improve their ability to cope with future risks and investments for preventive motives. Generally speaking, during the economic boom period, external financing is more harmonious, and the company will reduce the company's cash holdings and increase the scale of transactions. However, when the economic cycle enters a tightening period, the company's financing capacity declines, and the company's management team will increase the company's cash holdings, actively respond to possible financial distress, or ensure the steady growth of the company's investment level. Therefore, compared with the period of economic expansion, the company's cash holdings during the economic contraction period show that enterprises have higher cash holding levels, and Jiang Long and Liu Xiaosong [17], Ni Huiping and Zhao Shan [18] also support the opinion. Based on the above analysis, we propose the following hypothesis: H1: Economic cycle is negatively correlated with the cash holding level of listed companies.

EPU and Cash Holding
EPU will increase the uncertainty of the environment in which the company is located, and increase the systemic risks faced by enterprises [19], which will affect the company's cash holdings in two ways: cash inflows and cash outflows.
Bai et al. [20] explored the global financial crisis and found that the uncertainty of enterprises in the crisis period increased significantly, the interaction between financial friction and increasing uncertainty at the enterprise level led to a sharp decline in credit, corporate financing difficulty and financing costs. The sharp increase has led to a greater restriction on cash inflows, which has reduced corporate cash holdings. On the other hand, EPU have a greater impact on corporate cash outflows. Bloom [21] believes that uncertainty increases the size of po-  [6] have found that EPU and corporate investment levels are negative correlation, that is, economic policy uncertainty will reduce the company's cash outflow, thereby increasing the level of corporate cash holdings.
Generally speaking, when enterprises face high EPU, the uncertainty of future expectations of enterprises will increase, and enterprises cannot accurately predict the possibility of cash shortage in the future. For preventive motivation, enterprises will increase cash holdings. On the other hand, when EPU increases, it is easy for management or major shareholders to increase the company's cash flow in order to facilitate their own gains, which will be difficult to be captured by regulators, thereby further increasing the company's cash holding level. We therefore hypothesize the following: H2: EPU is positively correlated with the cash holding level of listed companies.

Economic Cycle, EPU and Cash Holding
Mitchell & Buns [22] defined the economic cycle as a volatility in a country's overall economic activity, including four continuous and recurring phases: recovery, expansion, boom, and depression. This paper adopts the above division The government's introduction of economic policies can play a multiplier role if it is compatible with the economic cycle. The impact of EPU on the cash holding level of listed companies is asymmetric. When the economy is booming, the active market will increase the amount of investment and reduce the level of cash holdings. At this time, with the increase of EPU, the level of investment in enterprises will decline, and the investment efficiency will increase, so that the level of cash holdings of enterprises will also be rise. On the other hand, during the boom period, the market is in an expanding and active state, and the flow of funds is accelerating. The financing pressure faced by enterprises is relatively small. The cash holdings of enterprises themselves are at a low level. At this time, the increase in EPU will cause enterprises to rapidly increase cash holdings with preventive motives, which is reflected in the immediate response of cash holdings to EPU.
When the economic cycle is in recession, the financing constraints faced by enterprises are relatively large. At this time, the increase of EPU will further increase the financing constraints of enterprises in the short term. Enterprises are subject to an objective financing environment and EPU. It will also increase the amount of corporate investment. In general, the level of corporate cash holdings will decrease in the short term, which is reflected as negative correlation between economic policy uncertainty and cash holdings. After a period of time, the market will gradually absorb the impact and gradually reach a new financing equilibrium. The increase in investment volume brought about by EPU will shift the cash shortage cost curve of enterprises. Enterprises will increase cash holding for preventive motive. Based on this, we present the following hypothesis: H3a: When economic cycle in prosperous, the EPU is positively correlated with the cash holding in the short term. The EPU is still positively correlated with the cash holding in the subsequent stage.
H3b: When economic cycle in recession, the EPU is negatively correlated with the cash holding in the short term. The EPU is positively correlated with the cash holding in the subsequent stage. On the one hand, it provides a reasonable excuse for management and major shareholders to maintain a large amount of cash for personal gain. On the other hand, the increase in EPU also makes the supervision of agency problems more difficult to implement. It is expected that EPU will reduce the value of corporate cash holdings when the economy is in a prosperous phase. When the economic cycle is in recession, the whole market is in a downturn, and the production and operation of enterprises can only be maintained at a low level. When EPU increases, the company increases cash holdings and reduces bankruptcy due to preventive motives, thereby enhancing corporate value. And in the recession stage, the company's shortage cost reduction is greater than the increase in holding costs, thereby increasing the value of corporate cash holdings. Based on the above analysis, this paper proposes the following hypothesis: H5a: When economic cycle in prosperous, EPU is negatively related to the value of corporate cash holdings.

Economic Cycle, EPU and Cash Holding Value
H5b: When economic cycle in recession, EPU is positively correlated with the value of corporate cash holdings.

Model Design and Variable Definition
Considering that the research on the cash holding level of listed companies involves a variety of influencing factors, this paper builds the following foundations based on the models set by Opler et al. [14], Jiang Long and Liu Xiaosong [17]. Inspection model: The above model is mainly used to observe the degree of impact of EPU on the cash holdings. Since the response of enterprises to macro factors usually has a certain degree of delay, this paper also verifies the impact of one order lag EPU on the cash holding level of listed companies. The model is as follows:  represents the regression coefficient, i represents the company i, and t represents the year t. Since the model is mostly a period variable, in order to eliminate the possible deviation in data extraction, this paper use annual variable in this model. "t" is the current period, The main variables involved in several of the above models are described below.

1) Interpreted variables
Cash holding level (CASH). This paper mainly explores the extent to which the cash holding level of listed companies is affected by the macroeconomic environment. Considering the difference between companies' scale, we use size-adjusted cash holding level, which is cash holding of company to the total assets. .In addition to monetary funds, the cash in this indicator also includes short-term investments (pre-2006) and trading financial assets (post-2006).
Corporate value (V). We use the sum of the stock market price of the SFC algorithm and the book value of the debt. The calculation of the stock market price distinguishes between A shares (including AB shares) and B shares, if "(Circulating A shares + restricted A shares) Not equal to 0", then the total market listed shares = A share closing price × (total number of shares − H − overseas shares − B shares) + B shares closing price × RMB exchange rate * B shares. If "(Circulation A shares + Restricted A shares) = 0" and "B shares total not equal to 0", then the total market value = B shares closing price × RMB exchange rate × B shares total shares.
2) Explanatory variables Economic Policy Uncertainty (EPU). Economic Policy Uncertainty is one of the systemic risks faced by enterprises, and it will have different degrees of impact on various decisions of enterprises. This paper draws on the data of China's economic policy uncertainty developed by Baker et al. [29] and uses the calculation method of Rao Pingui et al. [6] to calculate the uncertainty of quarterly This paper represents the economic cycle by GDP growth rate.

4) Control variables
Refer to Opler et al. [14], Lu Zhengfei and Zhu Jigao [4], Jiang Long and Liu Xiaosong [17] and other documents. In the basic model, we introduces company scale (SIZE), main business income (MBI), assets and liabilities (LEV), net working capital ratio (NWC), short-term debt ratio (SD), industry, etc. The control variables involved in the cash value model include operating cash flow ratio (CFO), non-cash asset ratio (NA), interest expense ratio (I), cash dividend payout ratio (D), and capital expenditure ratio (CAPEX) ( Table 1).

Sample Selection and Data Source
We takes the listed companies of Shanghai and Shenzhen A-shares as the re-   [29]. We use STATA 14.0 and EXCEL to process data.

Test Results of the Impact of the Economic Cycle on the Cash Holding
Column (1) of Table 3 presents the baseline regression without the main terms.
The R 2 is 0.4535, more than 0.45, which indicates that the overall model fit well.
The coefficients of the net working capital ratio (NWC) and the company scale (SIZE) are negative and significant at the 1% level respectively, indicating that the higher net working capital ratio, or the larger company size, the lower cash holding ratio. There was a significant negative correlation between the financial leverage and the corporate cash-holding ratio. The cash holding level was significantly positively correlated with the income level of the main business, and was significantly negatively correlated with the short-term bank loan interest rate.
Column (2) Table 3. Regression analysis of the first model.  will invest more, thereby reducing the cash holding level. These results support H1.

Test Results of the Impact of Economic Cycle and EPU on the Cash Holding
Columns (3) and (4) of Table 3 report the regressions of the current period EPU and one order lag EPU on the listed company's cash holding level. According to the regression results, the coefficient of EPU in both the current period and the lag period is significantly positive, indicating that when the EPU increases, the We divides the economic cycle into four groups by descending order of GDP growth rate to study the impact of macroeconomic policies on the cash holding of listed companies under different economic cycle periods. The first quarter is divided into the prosperity group, and the last quarter is divided into the recession group. Columns (5)-(8) of Table 3 report the impact of the EPU on the cash holdings of listed companies in different stages of the economic cycle. In general, the four regressions in Table 3 are all have large R 2 , which is more than 0.32, indicating that the model fits well. Columns (5) and (6) report the regression results of EPU and the level of cash holdings of enterprises in the period of economic prosperity. It can be seen from the regression results that the impact of economic policies on corporate cash holdings during the boom period has a certain lag, and the coefficient of the first-order lag variable is significantly larger than the coefficient of the original variable, indicating the impact of EPU on the current period of the enterprise is significantly smaller, compared with the next phase. So this paper only considers the first-order lag variable in the subsequent cash value model. When the economic cycle is in prosperous stage, the EPU in the short term is positively correlated with the cash holding level of listed companies. The EPU in the subsequent stage is still positively correlated with the cash holding level of listed companies, which is consistent with Hypothesis 3a.
Columns (7) and (8) represent the regression of EPU and the level of cash holdings of enterprises in the period of economic recession. From the regression results, it can be seen that when the economy in recession, the EPU will further increase the financing pressure of enterprises in the short term, which reflects the decline of cash holdings of enterprises. However, after a period of time, companies further increase cash holdings for preventive motivation. In the economic recession, the EPU in the short term is negatively correlated with the cash holding level of listed companies. The EPU in the subsequent stage is positively correlated with the cash holding level of listed companies, which is consistent with the setting of hypothesis 3b.

Robustness Test
Since the economic cycle and EPU are macroeconomic factors, they are highly exogenous to enterprises, so we do not consider the endogeneity of the model. In this paper, the robustness test is carried out in the following way, and the regression results are almost indistinguishable from the corresponding regression results in the text above. 2) Change period: Since the economic cycle is usually divided into 10 years, and the cash value model needs to lag two periods. Therefore, this paper selects the data for the 12-year period from 2004 to 2015, and further increases the research period in the robustness regression to 2004-2017. It was found that there was almost no change in the regression results; In the test of Hypothesis 2, only the one-stage lag variable was used for the relationship between EPU and the late cash holding level of listed companies, and more was adopted in the robustness regression. We found that there is no longer significant impact on the cash holdings after the two periods of lag, so the lag effect of economic policy uncertainty may only exist in the one order lag period.

Conclusions and Recommendations
This terprises has greater financing constraints brought by EPU, which is reflected in the decline of cash holdings in the short-term, and after the market has absorbed the impact of the financing constraints, the impact of EPU on cash holdings is more manifested in expectations theory.
This paper further examines the role of economic cycle and EPU on the value of cash holdings of listed companies. The empirical results show that EPU will reduce the cash holding value of listed companies. When the economy is booming, the increase in EPU will reduce the market value of cash holdings, but it is not significant. When the economy is in recession, the increase in EPU will increase the market value of corporate cash holdings. Therefore, the significance of this paper is to remind policy makers to distinguish the economic cycle from the stage of economic cycle when introducing policies. When the economy is booming, it should play more roles in the market and reduce the intervention in the market. In the recession period, appropriate policy intervention is beneficial to the whole market.

Conflicts of Interest
The author declares no conflicts of interest regarding the publication of this paper.