Abstract
China Journal of Accounting Studies, 2014 Vol. 2, No. 2, 137–159, http://dx.doi.org/10.1080/21697213.2014.926854 Dividend policy and cash flow manipulation: Evidence from China ,1 Donghua Zhou* and Yujie Zhao Jiangxi University of Finance and Economics, Nanchang, 330013, China The China Securities Regulatory Commission (CSRC) issued a dividend policy in 2006 which set minimum payout levels for firms wishing to issue seasoned equity offerings. This paper investigates how listed Chinese firms may inflate reported cash flow from operations in order to reach the threshold set by the dividend policy. The results indicate a high level of cash flow manipulation among firms issuing seasoned equity offerings after 2006, the year when the dividend policy was implemented, suggesting that this manipulation is undertaken in response to the dividend policy. We further find that firms issuing seasoned equity offerings inflate cash flow upwards through working capital items and by including tax refunds in their cash flows statements after the implementation of the dividend policy. The manipulation of real cash flow activities by firms issuing seasoned equity offerings eventually damages the value of their firm. Keywords: cash flow manipulation; dividend policy; firm value; seasoned equity offering 1. Introduction In China’s emerging capital market, paying a cash dividend is one of the mechanisms that protect the interests of minority shareholders. However, due to imperfect gover- nance mechanisms, some other listed companies do not have a sustainable and stable system for cash dividends, while they tend to distribute little or no cash dividends to their shareholders (Chen & Zhang, 2009). In the light of this limitation, the China Securities Regulatory Commission (CSRC) issued a new regulation obliging listed firms wishing to issue seasoned equity offerings (SEO) to pay cash dividends to their investors in 2006, which required the total dividends to be no less than 20% of the average annual distributable profits over the past three years for such firms (hereinafter, ‘the dividend regulation’). The dividend regulation requires a necessary dividend payment as one prerequisite to firms wishing to issue SEO, but firms can disobey the dividend regulation if they are not intending to issue SEO. By utilising this policy change, this paper examines the economic consequences of this dividend regulation on the manipulation of cash flows. Prior literature has indicated that firms inflate cash flow upwards to avoid reporting negative cash flow (Wu, Bo, & Wang, 2007; Zhang, 2009). For example, Lee (2012) provides evidence that firms are more likely to inflate cash flow upwards when their reported cash flow from operations (CFO) is particularly important to investors. Despite these studies, however, there remains limited research about whether and how firms *Corresponding author. Email: dhzhou@jxufe.edu.cn Paper accepted by Tong Yu. © 2014 Accounting Society of China 138 Zhou and Zhao manage their reported CFO to meet regulatory requirements. Literature relating to the Chinese market suggests that this corporate behaviour may be widespread given the existence of several government regulations operating in the securities market, espe- cially those relating to initial public offerings and SEO (Cai, Lee, & Zhang, 2003). It is in response to this suggestion, and an ongoing gap in evidence, that the relationship between cash flow manipulation and dividend policy is studied in the current paper. This topic is also of interest given that cash flow is an important determinant of a firm’s value (Dennis & Sibilkov, 2010; Jensen, 1986; Wu et al., 2007). SEO firms’ cash flow manipulation behaviour inflates the cash holding. However, there is limited research about whether firms’ management of their reported CFO may also damage the firm’s value. Therefore, the economic consequences of cash flow manipulation also present a relevant consideration. This paper examines cash flow manipulation from 2001 to 2011 in the Chinese cap- ital market. We find that the level of unexpected cash flow from operations is higher among SEO firms than non-SEO firms in the effort to fulfil the threshold of dividend policy. In addition to these findings, we also provide empirical evidence that the manip- ulation of real cash flow activities among SEO firms damages their firm value. Overall, these findings support the arguments that the dividend regulation has to face the limita- tion of the ‘regulatory paradox’ (Wang & Zhang, 2011). Our study contributes to the extant literature in several ways. First, our study is one of the few, if not the first, to examine cash flow manipulation affected by the dividend regulation, which is a significant regulation acting on the Chinese capital market. Our evidence on the positive association between cash flow manipulation and dividend pol- icy contributes to existing research on incentives around cash flow inflation. Second, our study highlights a negative relationship between cash flow manipulation and a firm’s value among SEO firms suggesting that SEO firms’ cash flow manipulation behaviour impairs a firm’s value, thus contributing to the evidence on the economic consequences of dividend regulation. Third, our study further investigates the mecha- nisms through which firms manage their reported CFO. Prior literature shows that firms undertake this by timing certain transactions such as delaying payments to suppliers or accelerating collections from customers (Lee, 2012). Our study shows that SEO firms inflate CFO through the adjustment of working capital and by acquiring tax refunds from local government. This finding contributes new knowledge to the existing litera- ture about CFO management mechanisms. The rest of the paper is organised as follows. Section 2 reviews the literature on cash flow manipulation. Section 3 describes the institutional background and develops our hypotheses. Section 4 discusses the research design and describes our data collec- tion process. Section 5 presents the main empirical results. Section 6 concludes. 2. Literature review According to signalling theory, a firm’s dividend payout provides positive signals to investors in terms of expectations of that firm’s future performance, and may convey more timely information than the firm’s actual earnings (Bhattacharya, 1980; Miller & Rock, 1985). Importantly, within this context, China’s institutional background distin- guishes Chinese listed firms from their counterparts in North America and Europe (Bradford, Chen, & Zhu, 2013). Lee and Xiao (2003) posit that return on equity (ROE)-enhancing incentives exert a significant influence over cash dividend policy in China because firms aiming to use cash dividend distributions to enhance their ROE China Journal of Accounting Studies 139 tend to distribute more cash dividends. Under the current accounting standard in China, dividend payment is recognized in the year when the profit is earned, not at the time of payment. As a result, paying a cash dividend can reduce the book value of equity and hence increase the ROE. Lee and Xiao (2003) find that when cash dividends push the reported ROE over the rights offering ROE threshold (with an average ROE of more than 6% in the previous three years), the probability of a firm paying out cash divi- dends increases, as does the payout ratio and also the dividend per share. Graham, Harvey, and Rajgopal (2005) provide evidence that managers regard cash flow as the second most important indicator for the market to evaluate a firm’s perfor- mance, but consider CFO more important than earnings when the firm is near financial distress. Therefore, investors pay more attention to the amount of cash flow, and the cash flow surplus is positively associated with the stock return (DeFond & Hung, 2003). Given the significant decline in average dividend payout ratios over recent years (Fatemi & Bildik, 2012), the stock price is particularly positively correlated with divi- dend increases (Fracassi, 2008). Roychowdhury (2006) suggests that cash flow can be influenced by the manipula- tion of certain real activities, such as offering price discounts to temporarily increase sales. Lee (2012) finds that managers often manipulate CFO by deciding when to dis- burse the cash outflow or receive the cash inflow; likewise, managers can increase reported CFO at the end of the year by delaying payments to suppliers and accelerating collections from customers. Prior literature provides evidence that firms are more likely to manage reported CFO upwards when: (1) they are in or near financial distress (DeFond & Hung, 2003; Graham et al., 2005; Sharma, 2001); (2) they are in or near investment/non-investment grade cutoff (Beaver, Shakespeare, & Soliman, 2006); (3) they are beating analysts’ cash flow forecasts (Brown, Huang, & Pinello, 2013; Zhang, 2008); or (4) their reported CFO has a strong correlation with stock returns (Call, 2008; Dechow & Ge, 2006). Lee (2012) suggests that firms manage reported CFO using classification and timing mechanisms. Classification refers to shifting items among the statement of cash flows categories, and holding earnings and aggregate cash flows constant. Timing refers to the adjustment of working capital in order to alter reported CFO. Unlike classifica- tion, timing involves real actions and reduces the chances of detection by auditors or the SEC (Lee, 2012). 3. Background and hypothesis development 3.1. Background It is generally accepted that firms in developed countries tend to have stable dividend polices (Adaoglu, 2000; Pandey, 2003). In contrast, most Chinese listed firms are in a rapid growth phase, with a focus on capital accumulation and expansion (He, Li, Shi, & Twite, 2009). Hence, fewer initiations and lower dividend payments are generally observed among Chinese firms (He et al., 2009; Shao & Lin, 2004). As a result, the CSRC has sought to encourage public firms to establish dividend policies with an aim to promote cash dividend payments. Since March 2001, the CSRC has published a series of regulations requiring listed firms intending to take on equity financing to pay cash dividends to their shareholders. These regulations aim to protect shareholders’ interests and motivate Chinese listed firms to pay more cash dividends by enforcing dividend policies. 140 Zhou and Zhao In March 2001, the CSRC issued ‘Administrative Measures for New Shares’ Issu- ance by Listed Companies’ and stated that the underwriters of common equity must pay close attention to refinancing applications (i.e. rights issues, seasoned offerings and convertible bond issues, as defined by the CSRC) by firms who failed to pay out divi- dends in the past three years, and whose Boards of Directors provided no justifiable explanation. This was the first time in China that the CSRC formally attempted to regu- late dividend payments (He et al., 2009). However, the CSRC did not specify the types of dividend (i.e. cash dividend or stock dividend) to be paid. More importantly, the payment of the dividend was not mandatory as long as the firms’ Boards of Directors provided satisfactory explanations for failing to pay dividends. In December 2004, the CSRC further notified that firm refinancing would not be approved if cash dividends had not been paid in the past three years. In May 2006, the CSRC issued another regulation ‘The Administrative Measures of the Securities Issuance by Listed Companies’, that requires the proportion of cumulated dividends (including cash dividends and stock dividends) paid in the last 3 years to average annual net income over the same period to be at least 20%, for those listed firms that intended to make seasoned equity offerings. In 2008, CSRC issued another regulation, ‘Decisions on Amending Some Provisions on Cash Dividends by Listed Companies’. This policy amended the previous regulation and increased the payout level to 30%. It also narrowed dividends type to cash dividends only. The dividend pol- icy of firms is ultimately influenced by the firms’ own financial decisions. Therefore, an interesting question arises: whether and how these institutional and regulatory changes have influenced dividend paying decisions. 3.2. Hypothesis development According to the ‘pecking order’ theory relevant to the USA and other developed capi- tal markets, it is argued that firms prefer internal to external finance (Myers & Majluf, 1984). However, this may not be the case in China. It has been found that Chinese listed firms have a strong preferences for external equity financing (Lu & Ye, 2004), because external equity financing is tightly controlled in China and represents a ‘lux- ury’, available only for a small group of firms. Being eligible for issuing seasoned offerings is, therefore, a valuable intangible asset for firms (Xue, Cai, & Guo, 2008). In China, most listed firms were formed by partial privatisation of state-owned bodies (Sun & Tong, 2003). Consequently, the Chinese government is typically the largest co-investor or controlling shareholder in these firms (Sun & Tong, 2003), and these firms have incentives to issue equity offerings in order to pursue the consumption of perquisites, such as new office buildings, or more leisure time benefiting managers at the expense of all shareholders (Xue et al., 2008). Therefore, Chinese listed firms have strong incentives to manipulate cash flow upwards to fulfil the threshold of divi- dend regulation when their real cash flow is insufficient. In addition, firms undertake an SEO in order to collect new capital for funding real growth opportunities or acquire other firms (DeAngelo, DeAngelo, & Stulz, 2010). The amount of capital collected by a firm at the time of an SEO depends on its stock price on the day of the offering. As such, managers have incentives to inflate cash flow in order to maximise SEO proceeds. Because cash flow has gained increasing importance in evaluating financial performance and assessing firm value (Xue et al., 2008), it can influence the stock price by changing market perception. Hence, managers have strong China Journal of Accounting Studies 141 incentives to inflate cash flow to increase the stock price. Accordingly, we have devel- oped our first hypothesis in this paper as follows. H1: Compared to non-SEO firms, SEO firms have stronger incentives to manipulate cash flow upwards. As discussed previously, in May 2006, CSRC issued a new regulation on cash divi- dends for firms wishing to issue seasoned equity offerings. This new regulation requires that firms which need to issue new shares have to maintain the average dividend payout ratio of at least 20% over the past three years. The policy amendment in 2008 man- dated firms to pay out no less than 30% of their profits in dividends. Moreover, the approval of equity financing in China obeys the Approval System, which means that the CSRC first verifies applications from firms and decides which firms should qualify to issue seasoned equity offerings. Not all applications are approved by the CSRC even if the firms satisfy all necessary requirements (Wang & Zhang, 2011). Given that the firm’s application is reviewed by the CSRC based on the cash divi- dend ratio, firms that need to raise capital from the equity market may have to increase their level of cash dividend to meet the ratio requirement. This reduces the free cash flow in those firms, and further forces those firms to access the capital market more fre- quently than before. Therefore, we argue that the increase in cash dividend due to divi- dend regulation decreases the level of firms’ cash holding. Anticipating more cash outflow to equity holders, firms will inflate cash flow upwards, signalling that their cash flows are sufficient to pay more cash dividend. Accordingly, we formulate our second hypothesis as follows. H2: SEO firms significantly inflate cash flows upwards following the implementation of the dividend regulation. 4. Research design 4.1. Testing cash flow manipulation using unexpected cash from operations Dechow, Kothari, and Watts (1998) model a firm’s cash-generating process at the firm level and empirically estimate firm-specific parameters using firms with at least ten years of annual data. Lee (2012) uses a firm-level estimation of the model over the prior ten years to derive expected CFO for each firm year. Roychowdhury (2006) expresses nor- mal cash flow from operations as a linear function of sales and change in sales in the cur- rent period, running the cross-sectional regression for every industry and year. Similar to Roychowdhury (2006) and Zhang (2008), we estimate real activities manipulation by running regression on the following equation by each two-digit SIC industry-year group: CFO 1 Sale DSale i;t i;t i;t ¼ k þ k þ k þ k þ e (1) 0 1 2 3 i;t TA TA TA TA i;t1 i;t1 i;t1 i;t1 We used the residuals ε in equation (1) as the proxy for cash flow manipulation, it which is measured as unexpected cash flow from operations (UCFO), where CFO is i,t the cash flow from the operations of firm i for the period t, TA is the total assets of i,t–1 firm i at the end of period t–1, Sale and △Sale are the sales and change in sales of i,t i,t firm i during period t. We use the parameter estimates from equation (1) to generate expected CFO, with unexpected CFO subsequently being the difference between actual and expected CFO. 142 Zhou and Zhao 4.2. Empirical design To test H1and H2, we estimate the following regression: UCFO ¼b þ b SEO þ b Dummy þ b SEO Dummy þ b CFO 01 þ b EPS þ b MB 0 1 2 3 4 5 6 X X þ b Size þ b Growth þ b Debt þ year dummy + ind dummy þ e ð2Þ 7 8 9 where SEO is an indicator set to 1 if a firm has issued seasoned equity offerings, and 0 otherwise. Dummy equals 1 if the sample observations pertain to 2006–2011 (post- implementation of dividend regulation in 2006), and 0 otherwise. As previously dis- cussed, we expect SEO firms to inflate cash flow significantly upwards, implying that β is expected to be positive. To test H2, we interact SEO with Dummy; in this case, H2 predicts that β will be positive. We control for other common determinants of unexpected CFO based on prior research. CFO_01 is an indicator set to 1 if firm reported its positive operating cash flows to be less than 1% after being scaled by total assets at the beginning of the year, and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, measured as debt divided by total assets. MB refers to the market-to-book ratio of equity. Industry dummy variables as well as year dummies are also included as explanatory variables. All variables are defined in the Appendix. 4.3. Data and sample selection Chinese public firms have been required to disclose cash flow statements in their annual financial reports since 1998 (Clinch, Dishu, & Sin, 2002). In order to make the financial data consistent across our sample firms, we collected data from 2001, the year in which the 2001 Chinese Accounting System for Business Enterprises (ASBE) was implemented, which provided a broad guideline and summary of general principles in accounting for various accounting issues such as assets impairments, contingencies and income taxes. Our sample includes firms listed in China’s capital market from 2001 to 2011. Cash dividends and other financial data were taken from the China Securities Markets and Accounting Research Database (CSMAR), a widely cited professional database in China. In order to minimise data selection bias, such as that concerning missing data, we further examined the sampled firms’ annual reports to minimise the problem of missing observations. We eventually compiled 14,871 firm-years observa- tions to enable us to estimate the degree of upward inflations of cash flow. The sample was drawn up according to various criteria, including certain key exclusions. Our initial sample consists of 17,686 firm-year observations. First, 1,118 firms that only issued B or H shares for foreign investors were excluded. This is because these firms use different accounting standards to those used by other firms that issue shares for domestic investors; their corporate governance is also different from that of other listed firms. Second, 283 banks and financial institutions were excluded from the sam- ple because their structure and accounting practices differ substantially from non-finan- cial firms. Third, since it was necessary to calculate the unexpected operating cash flow (UCFO) based on the sales data from the previous year, we excluded 953 firms that were listed as operating on the stock market for less than one year. Fourth, we excluded 461 firms whose accounting information contained fewer than 10 observations for each industry-year grouping. China Journal of Accounting Studies 143 Combined, these selection methods and exclusions resulted in a sample of 14,871 firm years over the period 2001–2011, including 21 industries and 1,876 individual firms. Our panel data includes both a cross-sectional and a time-series dimension. The unexpected CFO was found to differ between SEO and non-SEO firms. There was also a difference in CFO management between the pre-implementation and post-implemen- tation periods of the dividend regulation. Our unbalanced panel data-set can be seen as useful in resolving the endogeneity problem (Hermalin & Weisbach, 1991; Hsiao, 1985). The sample selection process is summarized in Table 1. Table 2 shows the annual and industrial distribution of the sample. Panel A presents the annual distribution of the sample, demonstrating an increase in the number of the sampled firms year by year. For example, the observation in 2001 is of 1,050 firms, 7.06% of the total sample, which increases to 1,861 by 2011, and 12.51% of the total sample. Table 2, Panel B, presents the industrial distribution of the sample. We classi- fied all the firms into 21 industries in accordance with the divisions set by the CSRC industry code. The least frequently occurring industry is timber and furnishings, at 0.3% of the total sample, while the most frequently occurring is machinery, at 15.30% of the total sample. Given these variations present in terms of both the annual and industrial distribution of the sample, in the subsequent regression model constructed we controlled the year dummy variables and industry dummy variables in order to control the year effect and industry effect. 5. Empirical results 5.1. Descriptive statistics Table 3, Panel A, presents descriptive statistics for the variables used in the main analy- ses. The mean and median UCFO in the sample are 0.01% and 0.04% respectively; the standard deviation of UCFO is 12.9%, with an extreme maximum value of 3.9126 and an extreme minimum value of –4.2386. Approximately 6.5% of the sampled firms have issued SEO in China’s emerging capital markets. The mean of CFO_01 is 0.0475, which implies that 4.75% of the sample firms reported positive operating cash flows to be less than 1% after being scaled by total assets at the beginning of the year. The mean and median Size in the sample are 21.4200 and 21.3100 respectively. The mean and median Growth are 2.100 and 0.1500, respectively, with an extreme maximum value of 14,865.9000. The mean and median EPS in the sample are 21% and 17% respectively. The mean and median Debt in the sample are 64% and 50% respectively. The mean and median MB in the sample are 0.7281 and 0.7561 respectively. The regression variables are winsorized at the top and bottom 1% of their respective distri- butions to eliminate the influence of outliers. Table 1. Sample selection process. Initial observations for the period 2001–2011 17,686 Less: Firm-year observations that only issue B or H shares for foreign investors 1,118 Firm-year observations that pertain to banks and financial institutions 283 Firm-year observations listed on the stock market for less than 1 year 953 Industry-year observations comprise fewer than 10 observations 461 Total firm-year observations 14,871 144 Zhou and Zhao Table 2. Annual and industrial distribution. Panel A: Annual distribution 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Total Obs. 1,050 1,056 1,123 1,189 1,287 1,296 1,354 1,477 1,542 1,636 1,861 14,871 Percent (%) 7.06 7.10 7.55 8.00 8.65 8.71 9.10 9.93 10.37 11.00 12.51 100 Panel B: Industrial distribution Firm-industry Percent (%) Obs. Agriculture, forestry, livestock 287 1.93 farming, fishery Mining 374 2.51 Food & Beverage 681 4.58 Textiles & Apparel 520 3.50 Timber & Furnishings 44 0.30 Paper & Printing 251 1.69 Petrochemicals 1,521 10.23 Electronics 647 4.35 Metal & Non-metallic mineral products 1,245 8.37 Machinery 2,275 15.30 Pharmaceuticals 1,021 6.87 Other manufacturing 121 0.81 Utilities 688 4.63 Construction 261 1.76 Transportation 590 3.97 IT 782 5.26 Wholesale and retail trade 1,062 7.14 Real estate 1,324 8.90 Social Services 451 3.03 Communications and Cultural Industry 181 1.22 Comprehensive 545 3.66 Total sample 14,871 100.00 China Journal of Accounting Studies 145 Table 3. Descriptive statistics. Variable Mean Maximum Minimum 25% Median 75% Std. Dev. Panel A: Full sample (N=14,871) UCFO 0.0001 3.9126 -4.2386 -0.0495 0.0004 0.0466 0.1290 SEO 0.0650 1.0000 0.0000 0.0000 0.0000 0.0000 0.2465 CFO_01 0.0475 1.0000 0.0000 0.0000 0.0000 0.0000 0.2128 Size 21.4200 28.2700 10.8400 20.6600 21.3100 22.0800 1.2100 Growth 2.1000 14,865.9000 -1.0000 -0.0100 0.1500 0.3400 127.8300 EPS 0.2100 8.4400 -21.8600 0.0400 0.1700 0.3700 0.5700 Debt 0.6400 877.2500 0.0000 0.3500 0.5000 0.6400 7.4500 MB 0.7281 5.6780 -0.1304 0.5380 0.7561 0.9298 0.2685 Panel B: SEO sample during pre-implementation of dividend regulation (N=216) UCFO -0.0100 0.7850 -0.6307 -0.0687 -0.0033 0.0487 0.0844 CFO_01 0.0370 1.0000 0.0000 0.0000 0.0000 0.0000 0.1893 Size 21.5425 25.6600 19.5800 21.0600 21.4450 22.0450 0.8492 Growth 0.2576 2.5900 -0.7400 0.0100 0.1800 0.3900 0.4262 EPS 0.2925 1.1712 -0.4405 0.1598 0.2500 0.3940 0.2259 Debt 0.4143 0.7779 0.0868 0.3073 0.4253 0.5141 0.1412 MB 0.7836 1.2072 0.2215 0.6692 0.8011 0.8937 0.1678 Panel C: SEO sample during post-implementation of dividend regulation (N=751) UCFO 0.0101 3.9126 -2.2438 -0.0534 0.0034 0.0681 0.1227 CFO_01 0.0333 1.0000 0.0000 0.0000 0.0000 0.0000 0.1795 Size 22.2925 26.2000 18.8300 21.4700 22.1600 22.9100 1.1572 Growth 25.9829 1486.9000 -0.9000 0.1300 0.3100 0.6100 546.8043 EPS 0.5098 5.5300 -1.2500 0.2100 0.3900 0.6700 0.5756 Debt 0.4985 1.6161 0.0465 0.3694 0.5014 0.6255 0.1769 MB 0.6883 2.0098 0.0000 0.4941 0.6830 0.8870 0.2525 Panel D: Non-SEO sample during pre-implementation of dividend regulation (N=5,489) UCFO -0.0008 1.2052 -0.7666 -0.0443 -0.0001 0.0420 0.0972 CFO_01 0.0508 1.0000 0.00000 0.0000 0.0000 0.0000 0.2197 Size 21.1203 26.9800 17.1100 20.5100 21.0600 21.6800 0.9576 Growth 0.4725 398.0700 -1.0000 -0.0100 0.1500 0.3500 6.7120 EPS 0.0912 2.3700 -14.0800 0.0300 0.1300 0.2700 0.5475 Debt 0.5339 43.0754 0.0081 0.3566 0.4916 0.6214 0.7703 MB 0.8438 5.6780 0.0000 0.7148 0.8666 0.9858 0.2186 Panel E: Non-SEO sample during post-implementation of dividend regulation (N=8,415) UCFO -0.0002 1.8988 -4.2386 -0.0520 0.0006 0.0486 0.0910 CFO_01 0.0469 1.0000 0.0000 0.0000 0.0000 0.0000 0.2115 Size 21.5493 28.2700 10.8400 20.7200 21.4200 22.2500 1.3149 Growth 1.0937 3,733.6600 -1.0000 -0.0100 0.1400 0.3100 46.3218 EPS 0.2652 8.4400 -21.8600 0.0500 0.1971 0.4400 0.5847 Debt 0.7363 877.2560 0.0000 0.3566 0.5176 0.6607 9.8945 MB 0.6548 2.0956 -0.1304 0.4437 0.6549 0.8656 0.2749 Note: UCFO is the residual of the regression model in equation (1). SEO equals 1 if the firm has issued sea- soned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t, and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. 146 Zhou and Zhao Panels B and C show the statistics for the SEO sample before and after the imple- mentation of the dividend regulation. The mean and median UCFO in the sample are 1.01% and 0.34% respectively during the post-implementation phase of the dividend regulation, while the mean and median UCFO in the sample are –1.00% and –0.33% respectively during the pre-implementation phase. Panels D and E highlight the statis- tics for the non-SEO sample before and after the implementation of the dividend regu- lation. The mean and median UCFO in the latter sample are –0.02% and 0.06% respectively during the post-implementation phase, and –0.08% and –0.01% respec- tively during the pre-implementation phase. Table 4 presents the difference between SEO firms and non-SEO firms in terms of CFO management. Panel A shows that the mean UCFO of SEO firms is 0.55%, which is significantly higher than the mean UCFO –0.05% of non-SEO firms (supported at a 5% level). In addition, it was found that SEO firms report receiving more tax refunds from the government than non-SEOs (0.24% versus –0.02%). We further classified the total sample into two sub-samples. Panel B reports the difference between SEO firms and non-SEO firms prior to the implementation of the dividend regulation during 2001–2005, while panel C reports the difference between SEO firms and non-SEO firms following the regulation implementation during the time period 2006–2011. The results show that the mean UCFO of SEO firms is only significantly higher than those of non-SEO firms in the post-implementation sub-sample. This finding indicates that SEO firms manage their reported CFO upwards in order to fulfil the cash dividend threshold of dividend regulation in China. Table 5 reports the Pearson (upper triangle) and Spearman (lower triangle) correla- tions among the variables. Consistent with the hypotheses, SEO is positively correlated with UCFO. One potential explanation for this positive univariate correlation is that SEO firms may manipulate CFO upwards to meet the threshold of dividend regulation; specifically, CFO_01 is positively correlated with UCFO, which implies that firms that reported very small positive cash flows may manipulate CFO upwards to avoid report negative cash flows. We also find that the control variables, Size and Growth, are negatively correlated with the variable, UCFO, indicating that firms with larger scale Table 4. UCFO difference: SEO versus non-SEO enterprises. SEO non-SEO SEO-non-SEO Mean Median Mean Median Mean difference t-value Median difference z-value Panel A: Full sample 0.0055 0.0021 -0.0005 0.0003 0.0060 2.43** 0.0018 3.24*** Panel B: Pre-implementation of dividend regulation -0.0100 -0.0033 -0.0008 -0.0001 -0.0092 -1.32 -0.0032 -1.47 Panel C: Post-implementation of dividend regulation 0.0101 0.0034 -0.0002 0.0006 0.0095 3.98*** 0.0028 3.29*** Note: UCFO is the residual of the regression model in equation (1). *** ** and indicate 0.01 and 0.05 significance levels, respectively. China Journal of Accounting Studies 147 Table 5. Correlation matrix. UCFO SEO CFO_01 Size Growth EPS Debt MB UCFO 0.028** 0.058** −0.008 0.1870** −0.139** 0.001 0.065** SEO 0.011 −0.017* 0.151** 0.037** 0.113** −0.006 −0.018* CFO_01 0.132** −0.017* −0.066** −0.003 −0.064** 0.045** −0.020* Size −0.025** 0.160** −0.059** 0.019* 0.286** −0.098** 0.357** Growth −0.039** 0.121** −0.034** 0.161** 0.012 0.000 0.013 EPS −0.225** 0.163** −0.101** 0.387** 0.366** −.039** −0.098** Debt 0.126** −0.030** 0.030** 0.199** 0.017* −0.203** −0.040** MB 0.062** −0.025** −0.017* 0.353** 0.003 −0.066** 0.189** Note: UCFO is the residual of the regression model in equation (1). SEO equals 1 if the firm has issued sea- soned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. ** and * indicate significance levels of 0.05 and 0.10 respectively. and higher growth have lower level UCFO. Most independent variables are not strongly correlated, suggesting that multicollinearity is not a major concern. 5.2. Empirical results Table 6 reports the results of the panel data model that examines how the dividend regulation influences the cash flow manipulation of listed firms in China. As seen in Column 1, the coefficient of SEO is significantly positive (0.0146, t = 4.66), which sup- ports the hypothesis that SEO firms have stronger incentives to manipulate cash flow upwards compared with non-SEO firms. In Column 2, the coefficient on SEO*Dummy is significantly positive (0.0206, t = 2.82), which suggests that SEO firms inflate their cash flow upwards after the implementation of dividend regulation. We further divided the full sample into two sub-samples based on the implementa- tion phases of the dividend regulation. As seen in Columns 3 and 4, the SEO coeffi- cient is insignificant prior to the policy implementation, but significantly positive after the implementation. This suggests that SEO firms mainly began to manage their cash flow from operations upwards following the implementation of the dividend regulation. We also divided the full sample into two sub-samples based on firm characteristics. As seen in Column 5, the Dummy coefficient is significantly positive (0.0462, t=2.44), suggesting that SEO firms began to upward manage their cash flow after the implemen- tation of the dividend regulation. In all regressions in Table 6, control variables are estimated to be consistent with findings in prior literature. For example, the CFO_01 coefficient is significantly posi- tive, suggesting that firms manage their cash flows upwards in order to avoid reporting negative cash flows. The EPS coefficient is significantly negative, suggesting that firms with better performance will generate more cash flows from real activities, resulting in less cash flow manipulation. The coefficient on MB is significantly positive, indicating that firms with a high market-to-book ratio, such as glamour stocks and high-growth firms, are more likely to manage CFO upwards (Lee, 2012). The coefficient on Size is significantly negative, suggesting that larger firms are less likely to manage CFO upwards. 148 Zhou and Zhao Table 6. Dividend regulation and cash flow manipulation. Full sample 2001–2005 2006–2011 SEO Non-SEO SEO 0.0146*** −0.0012 0.0005 0.0220*** (4.66) (−0.19) (0.09) (3.08) Dummy 0.0067* 0.0462** 0.0039 (1.84) (2.44) (0.84) SEO*Dummy 0.0206*** (2.82) CFO_01 0.0292*** 0.0292*** 0.0307*** 0.0283*** 0.0225 0.0313*** (8.25) (8.24) (6.14) (5.83) (0.57) (6.87) EPS −0.0431*** −0.0430*** −0.0523*** −0.0406*** −0.0493*** −0.0310*** (−19.97) (−19.95) (−14.28) (−14.97) (−3.33) (−17.06) MB 0.0561*** 0.0561*** 0.0443*** 0.0621*** −0.0121 0.0587*** (12.27) (12.26) (5.40) (10.96) (−0.28) (10.58) Size −0.0026*** −0.0026*** −0.0034** −0.0024** −0.0066 −0.0037*** (−2.84) (−2.88) (−2.21) (−2.08) (−0.76) (−3.47) Growth 0.0005 0.0003 0.0023 −0.0008 0.0002*** 0.0001*** (0.42) (0.21) (1.28) (−0.50) (12.43) (2.58) Debt 0.0151*** 0.0150*** 0.0081 0.0171*** 0.1480*** −0.0001 (4.99) (4.97) (1.62) (4.45) (2.87) (−0.47) Intercept 0.0098 0.0120 0.0381 −0.0079 0.0983 0.0382* (0.54) (0.66) (1.26) (−0.35) (0.56) (1.75) Year FE YES YES YES YES YES YES Industry FE YES YES YES YES YES YES Adjusted R 5.75% 5.76% 6.15% 5.69% 18.23% 3.69% F-value 25.53*** 24.92*** 13.06*** 17.75*** 6.98*** 15.78*** N 14,871 14,871 5,705 9,166 967 13,904 Note: The dependent variable is unexpected CFO (UCFO), UCFO is the residual of the regression model in equation (1). SEO equals 1 if the firm has issued seasoned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t, and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. *** ** * , and indicate 0.01, 0.05, and 0.10 significance levels, respectively. China Journal of Accounting Studies 149 Table 7. Dividend regulation and cash flow manipulation: paired sample. Full sample 2001–2005 2006–2011 SEO Non-SEO SEO 0.0189** −0.0067 0.0062 0.0227** (2.30) (−0.39) (0.63) (2.22) Dummy 0.0215 0.0462** 0.0106 (1.04) (2.44) (0.59) SEO*Dummy 0.0158*** (2.81) CFO_01 0.0331 0.0326* 0.0379* 0.0297 0.0225 0.0401** (1.61) (1.98) (1.73) (1.11) (0.57) (2.14) EPS −0.0435*** −0.0437*** −0.0386* −0.0453*** −0.0493*** −0.0410*** (−4.90) (−4.92) (−1.77) (−4.50) (−3.33) (−4.31) MB 0.0481** 0.0474* 0.0276 0.0510* −0.0121 0.0964*** (1.98) (1.95) (0.64) (1.79) (−0.28) (4.03) Size −0.0034 −0.0033 −0.0113 −0.0016 −0.0066 −0.0033 (−0.69) (−0.67) (−1.29) (−0.28) (−0.76) (−0.66) Growth 0.0002*** 0.0002*** 0.0000 0.0002*** 0.0002*** 0.0008 (15.79) (15.78) (0.02) (14.30) (12.43) (0.97) Debt 0.0599*** 0.0598*** 0.0639* 0.0601** 0.1480*** 0.0215 (2.60) (2.60) (1.90) (2.18) (2.87) (1.09) Intercept −0.0110 −0.0155 0.2090 −0.0530 0.0983 0.0018 (−0.11) (−0.15) (1.21) (−0.43) (0.56) (0.02) Year FE YES YES YES YES YES YES Industry FE YES YES YES YES YES YES Adjusted R 15.99% 15.98% 4.80% 16.54% 18.23% 7.18% F-value 10.95*** 10.67*** 1.72** 10.29*** 6.98*** 3.08*** N 1,934 1,934 432 1,502 967 967 Note: The dependent variable is unexpected CFO (UCFO), UCFO is the residual of the regression model in equation (1). SEO equals 1 if the firm has issued seasoned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t, and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. *** ** * , and indicate 0.01, 0.05, and 0.10 significance levels, respectively. 150 Zhou and Zhao 5.3. Robustness tests 5.3.1. Matched sample We also constructed matching samples in which each SEO firm was paired with a peer non-issuance firm that had the closest size value and the same two-digit SIC code in the same year. This matching process resulted in paired samples with 1,934 firm-years’ observations, comprising 967 SEO firms and 967 firms that did not issue equity offer- ings from 2001 to 2011. Table 7 presents the regression results. It shows that SEO firms have stronger incentives to manipulate cash flow upwards compared with non- SEO firms. SEO firms mainly inflate their cash flow upwards in order to fulfil the threshold of the dividend regulation, after these policies were introduced. 5.3.2. Alternative measure: previous year’s unexpected cash flow from operations (UCFO ) t–1 China’s dividend regulation demands that a SEO firm fulfil the condition that dividends distributed accumulatively in the last three years shall not be less than a percentage of the average annual distributable profits realised in the last three years. Given the policy uncertainty in China, listed firms are unlikely to wait in circumstances when it becomes possible to meet the requirements for SEO, as deferring this might imply forfeiting those rights (Lee & Xiao, 2003). Hence, listed firms have a strong incentive to inflate their cash flow early on in order to meet the rights of the SEO threshold. This incentive means that we might expect SEO firms to manage their cash flow upwards not only in the issuing seasoned equity offerings year, but also in the previous year. In this paper, we substitute the previous year’s unexpected cash flow from operations (UCFO )as t–1 the dependent variable in the regression analysis. Table 8 presents the results. The coefficients on SEO and SEO*Dummy are still sig- nificantly positive in Column 1 and Column 2 respectively, which are consistent with the results above. The results indicate that SEO firms have stronger incentives to manipulate cash flow upwards compared with non-SEO firms after the implementation of dividend regulation. 5.3.3. Alternative definition of unexpected cash flow from operations (UCFO) The dividend policy regulation seems to have had an influence on the cash dividend policy of SEO firms, given that some firms have been found to manipulate their cash flows upwards to pay out cash dividends to fulfil the threshold of dividend regulation. In our robustness check, we present an alternative way of measuring unexpected CFO. Our implementation of the Roychowdhury (2006) model follows his theoretical model of the firm’s cash-generating process at a firm-industry level. As a robustness check, we implement a cross-sectional variation of the Dechow et al. (1998) and Lee (2012) models. They model a firm’s cash-generating process at the firm level, and empirically estimate firm-specific parameters using firms with at least ten years of annual data. In order to derive expected CFO for each firm-year, we use a firm-level estimation of the model over the prior ten years. For each firm, unexpected CFO is the firm’s CFO devi- ation from the difference between actual and expected CFO. This process yielded 11,199 firm-years observations. The results are consistent with our regression findings. China Journal of Accounting Studies 151 Table 8. Dividend regulation and previous years’ cash flow manipulation. Full sample 2001–2005 2006–2011 SEO Non-SEO SEO 0.0015** 0.0045 −0.0039 0.0008** (2.35) (0.42) (−0.44) (2.17) Dummy 0.0119** 0.0089** 0.0120 (2.48) (2.42) (0.44) SEO*Dummy 0.0036** (2.31) CFO_01 0.0031 0.0030 0.0158** −0.0038 0.0114 0.0025 (0.63) (0.63) (2.43) (−0.58) (0.50) (0.50) EPS −0.0275*** −0.0275*** −0.0231*** −0.0306*** −0.0355*** −0.0272*** (−14.64) (−14.64) (−8.47) (−12.28) (−4.41) (−14.10) MB 0.0435*** 0.0435*** 0.0281*** 0.0510*** −0.0145 0.0449*** (7.60) (7.59) (3.08) (6.98) (−0.60) (7.58) Size −0.0023** −0.0023** −0.0040** −0.0019 −0.0123** −0.0024** (−2.09) (−2.09) (−2.19) (−1.34) (−2.58) (−2.06) Growth −0.0000 −0.0000 0.0023*** −0.0000 −0.0000 0.0000 (−0.68) (−0.68) (10.99) (−0.90) (−1.31) (1.28) Debt 0.0001 0.0001 0.0008 0.0001 0.1760*** 0.0001 (0.49) (0.49) (0.44) (0.56) (6.04) (0.49) Intercept 0.0154 0.0156 0.0589 −0.0003 0.2170** 0.0149 (0.68) (0.69) (1.60) (−0.01) (2.23) (0.63) Year FE YES YES YES YES YES YES Industry FE YES YES YES YES YES YES Adjusted R 2.24% 2.24% 4.50% 2.35% 4.47% 2.24% F-value 9.28*** 9.03*** 7.93*** 7.44*** 2.16*** 8.93*** N 12,995 12,995 4,416 8,579 867 12,128 Note: The dependent variable is unexpected CFO (UCFO), UCFO is the residual of the regression model in equation (1). SEO equals 1 if the firm has issued seasoned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t, and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. *** ** * , and indicate 0.01, 0.05, and 0.10 significance levels, respectively. 152 Zhou and Zhao 5.4. Additional tests 5.4.1. The mechanisms used by SEO firms to manipulate CFO It is usually not straightforward for listed firms to increase their true cash flow genera- tion capacity because they may already be operating in the market at their optimal level, with little improvement possible. Therefore, increasing their true cash flow may not be an effective way to meet the threshold of the dividend regulation. We posit that an alternative approach to increasing their cash flow is that of opportunistically inflating this through real activities manipulation (Roychowdhury, 2006). Generally, managers have some discretion over the timing of CFO through influencing when to disburse the cash outflow or receive the cash inflow; managers can also increase the reported CFO at the end of the year by delaying payments to suppliers and accelerating collections from customers (Lee, 2012). Specifically, these real activities include: sales acceleration via the introduction of price discounts, and the reduction of discretionary expenditures to increase cash flow. Public firms may accelerate the recovery of their accounting receivables by using additional price discounts or lenient credit terms, or by suspending the payments of money owed, with the permission of creditors, for several days. Man- agers may also adopt a policy of reducing inventories to inflate cash flow when their real cash flow is insufficient. Therefore, it can be surmised that, following the imple- mentation of the dividend regulation, SEO firms began to significantly inflate their cash flows upwards through the manipulation of real working capital items, including their Inventory (INV), Accounts Receivable (AR), and Accounts Payable (AP). This finding is compounded by the fact that local government has a greater incen- tive than central government to support listed firms for the benefit of the local economy and other social liabilities in China (Bradford et al., 2013; Chen, Chen, & Zhang, 2005). Friedman, Simon, and Mitton (2003) find that entrepreneurs use their private funds to prop up their firms, which is common among these special treated (ST) firms in order to maintain or obtain control rights. When listed firms are in financial distress, their controlling shareholders are more likely to conduct related party transactions to prop up their listed firms (Bai, Liu, & Song, 2004). As the Chinese government is usu- ally the largest co-investor or controlling shareholder in these firms, local government will help firms to fulfil the threshold of the dividend regulation in case they are unable to inflate their cash flow by manipulating normal transactions. For example, tax refunds are frequently used by local government to support listed firms. We therefore predict that local government will support public firms to fulfil the condition of the dividend regulation by providing tax refunds. We estimate the following regression to investigate the mechanisms used by SEO firms to manipulate cash flow: UCFO ¼ b þ b SEO þ b Dummy þ b INV þ b AR þ b AP þ b Tax þ b SEO 0 1 2 3 4 5 6 7 Dummy þ b INV Dummy þ b AR Dummy þ b AP Dummy þ b Tax 8 9 10 11 Dummy þ b SEO INV þ b SEO AR þ b SEO AP þ b SEO Tax 12 13 14 15 þb SEO INV Dummy þ b SEO AR Dummy þ b SEO AP Dummy 16 17 18 þb SEO Tax Dummy þ b CFO 01 þ b EPS þ b MB + b Size 19 20 21 22 23 X X þb Growth þ b Debt þ year dummy + ind dummy þ e ð3Þ 24 25 where INV is the amount of inventory decrease in the cash flow statement scaled by the beginning of the year total assets; AR is the amount of the decrease in operating China Journal of Accounting Studies 153 Table 9. Dividend regulation and cash flow manipulation mechanisms. Full sample 2001–2005 2006–2011 SEO Non-SEO SEO 0.0181*** −0.0241*** 0.0232 0.0296*** (4.27) (−2.86) (1.03) (5.78) Dummy 0.0069 0.0123*** 0.0858*** 0.0105*** (1.41) (2.81) (2.63) (2.64) INV 0.0079*** −0.6569*** −0.6678*** −0.1728*** −0.6818*** −0.6608*** (4.26) (−33.16) (−60.80) (−33.07) (−3.70) (−37.60) AR −0.0056*** −0.7057*** −0.7140*** −0.3806*** −0.6658*** −0.7059*** (−5.91) (−39.21) (−72.31) (−37.83) (−3.48) (−44.23) AP −0.0159*** −0.6412*** −0.6491*** −0.2474*** −0.4981*** −0.6417*** (−8.43) (−34.82) (−63.98) (−37.74) (−3.33) (−39.29) Tax −0.0327 −0.0335 −0.0281 −0.0653 −0.8621 −0.0347 (−0.78) (−0.63) (−0.96) (−0.99) (−0.97) (−0.73) SEO*Dummy 0.0543*** (5.77) INV*Dummy 0.4834*** 0.6336*** 0.4757*** (23.90) (3.44) (26.47) AR*Dummy 0.3253*** 0.6931*** 0.3197*** (16.42) (3.62) (18.19) AP*Dummy 0.3938*** 0.5401*** 0.3826*** (20.53) (3.61) (22.43) Tax*Dummy −0.0278 0.4381** −0.0257 (−0.36) (2.47) (−0.38) SEO*INV −0.1797* −0.1853*** 0.0955*** (−1.93) (−3.65) (11.39) SEO*AR −0.0622 −0.0665 0.4238*** (−0.65) (−1.28) (38.73) SEO*AP 0.1368* 0.1386*** 0.3109*** (1.82) (3.39) (33.47) SEO*Tax 0.1834 0.1856 0.4490*** (1.62) (1.05) (2.66) SEO*INV*Dummy 0.2750*** (2.95) (Continued) 154 Zhou and Zhao Table 9. (Continued). Full sample 2001–2005 2006–2011 SEO Non-SEO SEO*AR*Dummy 0.4860*** (5.08) SEO*AP*Dummy 0.1740** (2.31) SEO*Tax*Dummy 0.2660** (2.57) CFO_01 0.0309*** 0.0214*** 0.0193*** 0.0234*** 0.0202 0.0213*** (6.41) (5.09) (5.34) (3.58) (0.52) (5.57) EPS −0.0311*** −0.0444*** −0.0450*** −0.0446*** −0.0524*** −0.0442*** (−16.16) (−26.29) (−28.03) (−17.68) (−3.49) (−28.72) MB 0.0585*** 0.0464*** 0.0161*** 0.0616*** −0.0077 0.0467*** (10.03) (9.13) (3.15) (8.40) (−0.18) (10.03) Size −0.0030*** −0.0025** −0.0036*** −0.0023* −0.0091 −0.0025*** (−2.67) (−2.57) (−3.51) (−1.66) (−1.07) (−2.81) Growth 0.0002*** 0.0001*** 0.0003** 0.0001*** 0.0001*** −0.0002*** (24.07) (4.70) (2.47) (3.94) (4.17) (−6.85) Debt −0.0001 −0.0000 0.0033*** −0.0000 0.1290** −0.0000 (−0.42) (−0.33) (2.94) (−0.27) (2.53) (−0.34) Intercept 0.0218 −0.0033 0.0366* −0.0089 0.1230 −0.0051 (0.95) (−0.16) (1.77) (−0.32) (0.71) (−0.28) Year FE YES YES YES YES YES YES Industry FE YES YES YES YES YES YES Adjusted R 6.88% 29.32% 30.15% 21.76% 21.34% 32.38% F-value 27.80*** 115.25*** 221.81*** 64.72*** 6.96*** 152.28*** N 14,871 14,871 5,705 9,166 967 13,904 Note: The dependent variable is unexpected CFO (UCFO), UCFO is the residual of the regression model in equation (1). SEO equals 1 if the firm has issued seasoned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. INV is the amount of decrease in inventories in the cash flow statement scaled by total assets at the beginning of the year. AR is the amount of decrease in operating accounts receivable in the cash flow statement scaled by total assets at the beginning of the year. AP is the amount of decrease in operating accounts payable in the cash flow statement scaled by total assets at the beginning of the year. Tax is the amount of refund of tax and fee received in the cash flow statement scaled by total assets at the beginning of the year. *** ** * , and indicate 0.01, 0.05, and 0.10 significance levels, respectively. China Journal of Accounting Studies 155 Table 10. Firm value and cash flow manipulation. Full sample 2001–2005 2006–2011 SEO Non-SEO UCFO 0.0323 −0.1000*** −0.0590*** −0.1028** 0.0266 (0.86) (–2.63) (–2.79) (–2.30) (0.85) SEO −0.0179* −0.0094 −0.0135 (–1.89) (–0.72) (–1.63) Dummy 0.0939*** 0.2609*** 0.0921*** (8.25) (3.32) (10.48) UCFO*SEO −0.1385 −0.0556 −0.0584** (–1.77) (–0.66) (–2.01) UCFO*Dummy −0.1515* −0.0094** −0.1503 (–2.31) (–2.03) (–1.19) SEO*Dummy 0.0063 (0.52) UCFO*SEO*Dummy −0.2438** (–2.25) CFO_01 −0.0099 0.0069 −0.0158 0.0012 −0.0116 (–1.07) (0.61) (–1.44) (0.01) (–1.37) EPS −0.0346 −0.0746 −0.0063 −0.0165 − 0.0367*** (–1.39) (–1.64) (–1.33) (–0.45) (–10.69) MB −1.9577*** −1.4378*** −2.1719*** −2.9957*** −1.9417*** (–22.39) (–7.99) (–62.96) (–27.96) (–15.84) Size −0.0161** −0.0370** −0.0062 −0.0390* −0.0182*** (–2.09) (–2.36) (–1.39) (–1.85) (–9.01) Growth 0.0000 0.0003 0.0000 0.0000 0.0000 (1.05) (0.96) (0.86) (0.97) (0.64) Debt −0.0048 0.0595* −0.0252* −0.0253 −0.0006** (–0.37) (1.85) (–1.92) (–0.20) (–2.55) Intercept 3.2808*** 3.3027*** 3.2991*** 4.4022*** 3.3139*** (29.64) (15.49) (42.86) (10.17) (80.95) Year FE YES YES YES YES YES Industry FE YES YES YES YES YES Adjusted R 21.32% 18.10% 21.32% 64.87% 21.40% F-value 710.06*** 158.51*** 773.56*** 47.94*** 249.18*** N 14,860 5,702 9,158 967 13,893 Note: The dependent variable is firm value, computed as Tobin q value. UCFO is the residual of the regres- sion model in equation (1). SEO equals 1 if the firm has issued seasoned equity offerings during period t, and 0 otherwise. CFO_01 equals 1 if the firm reported its positive operating cash flows less than 1% after scaling by the total assets at the beginning of year t and 0 otherwise. Size is the natural logarithm of total assets. Growth is the sales growth rate. EPS is the net profit after tax divided by total equity shares. Debt is financial leverage, which is measured as the debt divided by total assets. MB equals market to book ratio of equity. *** ** * , and indicate 0.01, 0.05, and 0.10 significance levels, respectively. receivables in the cash flow statement scaled by the total assets at the beginning of the year; AP is the amount of the increase in operating payables in the cash flow statement scaled by the total assets at the beginning of the years; and Tax is the amount of the tax refund and fee received in the cash flow statement scaled by the total assets at the beginning of the year. Other variables are defined as previously. Table 9 presents the regression results. The coefficients of SEO*Dummy*INV (0.2750, t = 2.95), SEO*Dummy*AR (0.4860, t = 5.08) SEO*Dummy*AP (0.1740, t = 2.31) and SEO*Dummy*Tax (0.2660, t = 2.57) are significantly positive, suggesting that SEO firms inflate their cash flow upwards through decreasing operating accounts receivable (AR), increasing operating accounts payable (AP), decreasing inventories 156 Zhou and Zhao (INV), and recording tax refunds (Tax) in their cash flows statements, following the implementation of the dividend regulation. 5.4.2. Cash flow manipulation and firm value Modern finance theory argues that a firm’s value depends on its stream of future cash flows, and that the value of the firm is equal to the present discounted value of future free cash flows (Miller & Modigliani, 1961). While upward cash flow inflation can increase a firm’s reported cash flow in the current period, it ultimately decreases the future cash flow and thus the firm’s value. Roychowdhury (2006) finds that manipula- tion of real activities serves to reduce a firm’s value because manipulated actions taken in the current period have a negative effect on the cash flows of subsequent periods. For example, aggressive price discounts to increase sales volumes can set the precedent of leading customers to expect such discounts in future periods as well, resulting in a decrease of future cash flow. Therefore, we expect that SEO firms undertaking a manip- ulation of their cash flow through real activities show a negative correlation with firm value. This negative relationship can be seen to intensify following the implementation of the dividend regulation. We establish equation (4) as follows: TobinQ ¼ b þ b UCFO þ b SEO þ b Dummy þ b UCFO SEO þ b UCFO 0 1 2 3 4 5 Dummy þ b UCFO SEO Dummy þ b CFO 01 þ b EPS þ b MB 6 7 8 9 X X + b Size þ b Growth þ b Debt þ year dummy + ind dummy þ e 10 24 25 ð4Þ where TobinQ is firm value, calculated as the ratio of market value of equity and book value of debt to book value of assets; other variables are as defined previously. Table 10 presents the regression results. As seen in Column 1 of Table 10, the coefficients of UCFO*SEO and UCFO*- SEO*Dummy are significantly negative (–0.138, t = –1.77, –0.240, t = –2.25), which indicates that SEO firms undertaking cash flow manipulation are negatively correlated with increasing firm value. This negative correlation can be seen as being intensified following the implementation of the dividend regulation. We also divided the full sample into two sub-samples based on the implementation of the dividend regulation as above. As seen in Columns 2 and 3, the coefficient of UCFO*SEO is insignificant prior to the policy implementation, but becomes signifi- cantly negative once the policies are implemented. We also divided the full sample into two sub-samples based on firm characteristics. As seen in Columns 4 and 5, the coeffi- cient of UCFO*Dummy is significantly negative (–0.0094, t=–2.03), suggesting that firms’ cash flow manipulation results in a decrease of firm value following the imple- mentation of the dividend regulation. 6. Conclusion Cash flow from operations and earnings are complementary measures of a firm’s perfor- mance (Lee 2012). Nwaeze, Yang, and Yin (2006) find that CFO is more important than earnings in managers’ performance evaluation and the determination of their com- pensation and reward. One potential explanation for this is based on the assumption that CFO is different from earnings in that it is the reflection of the true cash inflow and outflow of a firm. It is, therefore, theoretically less susceptible to manipulation China Journal of Accounting Studies 157 because of its greater visibility and scrutiny (McInnis & Collins, 2011), which has led to the argument that CFO cannot be managed or manipulated in the same way as earn- ings, and is overall more reliable (Zhang 2009). However, recent studies have found that managers exercise discretion in financial reporting and in the timing of transactions in order to manage their reported CFO (Lee, 2012; Zhang, 2008); specifically, it has been found that firms may engage in the manipulation of their real activities in order to inflate their cash flow. We find that SEO firms have stronger incentives to inflate reported CFO, and that the level of unexpected cash flow from operations is higher for SEO firms than non- SEO firms. We also document that SEO firms began to manipulate cash flow to meet the threshold of cash dividends payout following the implementation of the dividend regulation in 2006, indicating that central government regulations still play a major role in the emerging capital markets. Chinese listed firms must cater to these regulations in order to qualify for equity financing, thus incentivising the decision to inflate cash flow upwards to meet the threshold of cash dividend payouts. In addition to these findings, we also provide empirical evidence that SEO firms manage CFO through working capital items and by obtaining tax refunds from local government. The manipulation of real cash flow activities by SEO firms damages the value of their firm. These findings support the arguments in prior literature that real activities manipulation may have significantly negative economic consequences. A major implication of these results is that dividend regulation has to face the limi- tation of the ‘regulatory paradox’ (Wang & Zhang, 2011). That is to say, the implemen- tation of the dividend policy aims to motivate SEO firms to pay higher cash dividends to protect the interests of minority shareholders. However, as shown in our empirical evidence, the policy has actually resulted in SEO firms inflating cash flow upwards in order to reach the threshold set by the dividend regulation, and the manipulation of real cash flow activities damages the value of their firm, which actual damages investor interests. This, then, creates a paradoxical situation where the existence of regulation leads to the negation of what that regulation is trying to achieve. This study has implications for global investors, who are increasingly investing in China’s emerging capital markets, and for other users of the financial reports of Chinese listed firms. Our results highlight the importance of understanding regulatory rules when analysing financial information in emerging markets. In order to understand the earnings management and potential CFO manipulation of Chinese firms and the Chinese stock market, one must first understand the role of government regulatory rules and their influence on the firms’ economic activities. Acknowledgements Donghua Zhou acknowledges financial support from the National Nature Science Foundation of China (NSFC-71262005). The authors would also like to thank Liansheng Wu, Tong Yu, and two anonymous referees for their useful suggestions that substantially helped to improve the paper. Notes 1. This ASBE functioned as a comprehensive financial reporting framework, emphasising principles such as substance over form, consistency, timeliness, accrual basis, prudence, materiality, and transparency. It was much more consistent with international practice than the prior ASBE in 1998. Later, China issued a series of new Accounting Standards for 158 Zhou and Zhao Business Enterprises (CAS) on 15 February 2006, comprising a basic standard and 38 spe- cific standards. The new standards cover nearly all of the topics of IFRS, and are substan- tially in line with IFRS except for a few modifications made for China’s unique environment (He, Wong, & Young, 2009). 2. B shares are issued to foreign investors by firms listed at Shanghai Stock Exchange and Shenzhen Stock Exchange, denominated in US dollars and Hong Kong dollars respectively. H shares are issued to foreign investors by firms listed at the Securities Exchange of Hong Kong (SEHK) and foreign markets, such as New York, Singapore, and London. Therefore, firms that only issued B or H shares adopt the International Financial Reporting Standards, while domestic firms adopt the Chinese Accounting Standards (CAS). The CAS have become more closely aligned with IFRS from 2006. 3. A firm becomes an ST firm if one of the following four conditions is satisfied: (1) a listed firm has negative net profits for two consecutive fiscal years; (2) the shareholders’ equity is lower than the registered capital (the par value of the shares); (3) while auditing a listed firm’s financial report, the auditors issue an adverse opinion or a disclaimer of opinion; (4) a firm’s operations have been stopped, and there is no hope of them being restored within three months due to a natural disaster or serious accident or if the firm is involved in a damaging lawsuit or arbitration (Bai, Liu, & Song, 2004). References Adaoglu, C. (2000). Instability in the dividend policy of the Istanbul Stock Exchange (ISE) corporations: Evidence from an emerging market. Emerging Markets Review, 1, 252–270. Bai, C. E., Liu, Q., & Song, F. M. (2004). Bad news is good news: Propping and tunneling evi- dence from China. Working paper. University of Hong Kong, China. Retrieved from http:// www.hiebs.hku.hk/working_paper_updates/pdf/wp1094.pdf Beaver, W., Shakespeare, C., & Soliman, M. (2006). Differential properties in the ratings of certi- fied versus non-certified bond-rating agencies. Journal of Accounting and Economics, 42, 303–334. Bhattacharya, S. (1980). Nondissipative signaling structures and dividend policy. Quarterly Jour- nal of Economics, 95(1), 1–24. Bradford, W., Chen, C., & Zhu, S. (2013). Cash dividend policy, corporate pyramids, and owner- ship structure: Evidence from China, International Review of Economics and Finance, 27, 445–464. Brown, L. D., Huang, K., & Pinello, A. S. (2013). To beat or not to beat? The importance of analysts’ cash flow forecasts. Review of Quantitative Finance and Accounting, 41, 723–752. Cai, X., Lee, C. W., & Zhang, W. G. (2003). A review of the empirical accounting research in China. China Accounting and Finance Review, 5, 155–215. Call, A. (2008). The implications of cash flow forecasts for investors’ pricing and managers’ reporting of earnings. Working paper. The University of Georgia, USA. Retrieved from http://www.kellogg.northwestern.edu/accounting/papers/call.pdf Chen, G. & Zhang, Z. Y. (2009). PRC-listed companies face new cash dividend requirements. China Law & Practice, Business Law Bulletin. Chen, S. K., Chen, X., & Zhong, R. (2005). The impact of government regulation and ownership on the performance of securities companies: Evidences from China. Global Finance Journal, 16,113–124. Clinch, G., Sidhu, B., & Sin, S. (2002). The usefulness of direct and indirect cash flow disclo- sures. Review of Accounting Studies, 7, 383–404. DeAngelo, H., DeAngelo, L., & Stulz, R. (2010). Seasoned equity offerings, market timing, and the corporate lifecycle. Journal of Financial Economics, 95, 275–295. Dechow, P., & Ge, W. (2006). The persistence of earnings and cash flows and the role of special items: Implications for the accrual anomaly. Review of Accounting Studies, 11, 253–296. Dechow, P., Kothari, S. P., & Watts, R. (1998). The relation between earnings and cash flows. Journal of Accounting and Economics, 25, 133–168. DeFond, M., & Hung, M. (2003). An empirical analysis of analysts’ cash flow forecasts. Journal of Accounting and Economics, 35,73–100. Dennis, D., & Sibilkov, V. (2010). Financial constraints, investment and the value of cash hold- ings. Review of Financial Studies, 23, 247–269. China Journal of Accounting Studies 159 Fatemi, A., & Bildik, R. (2012). Yes, dividends are disappearing: Worldwide evidence. Journal of Banking & Finance, 36, 662–677. Fracassi, C. (2008). Stock price sensitivity to dividend changes. Working paper. University of California Los Angeles, USA. Retrieved from http://faculty.mccombs.utexas.edu/cesare.fracas si/Price%20Sensitivity%20to%20Dividend%20Change%20-%20072808.pdf Friedman, E., Simon, J., & Mitton, T. (2003). Propping and tunneling. Journal of Comparative Economics, 31, 732–750. Graham, J., Harvey, C., & Rajgopal, S. (2005). The economic implications of corporate financial reporting. Journal of Accounting and Economics, 40,3–73. He, X., Li, M., Shi, J., & Twite, G. (2009). Determinants of dividend policy in Chinese firms: Cash versus stock dividends. Working paper. Australian National University, Australia, Bowl- ing Green State University, USA, RMIT University, Australia, and University of Texas at Austin, USA. Retrieved from http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1479053 Hermalin, B. E., & Weisbach, M. S. (1991). The effects of board composition and direct incen- tives on firm performance. Financial Management, 20, 101–112. Hsiao, C. (1985). Benefits and limitations of panel data. Econometric Review, 4, 121–174. Jensen, M. C. (1986). Agency costs of free cash flow, corporate finance, and takeovers. The American Economic Review, 76, 323–329. Lee, C. F., & Xiao, X. (2003). Cash dividends in China: Liquidating, expropriation and earnings management. Working paper. Tulane University, USA and Tsinghua University, China. Retrieved from http://www.bm.ust.hk/~acct/Incubator_Research_Camp/cashdividendsinChina %20Lee%20and%20Xiao.pdf Lee, L. F. (2012). Incentives to inflate reported cash from operations using classification and tim- ing. The Accounting Review, 87(1), 1–33. Lu, Z. F., & Ye, K. T. (2004). The puzzle of equity financing preference in China’s listed compa- nies. Economic Research Journal (Chinese), (4),50–59. McInnis, J., & Collins, D. (2011). The effect of cash flow forecasts on accrual quality and bench- mark beating. Journal of Accounting and Economics, 51, 219–239. Miller, M., & Rock, K. (1985). Dividend policy under asymmetric information. Journal of Finance, 40, 1031–1051. Miller, M. H., & Modigliani, F. (1961). Dividend policy, growth, and the valuation of Shares. Journal of Business, 34,411–433. Myers, S. C., & Majluf, N. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187–221. Nwaeze, E., Yang, S., & Yin, Q. J. (2006). Accounting information and CEO compensation: The role of cash flow from operations in the presence of earnings. Contemporary Accounting Research, 23, 227–265. Pandey, I. M. (2003). Corporate dividend policy and behaviour: The Malaysian evidence. Asian Academy of Management Journal, 8,17–32. Roychowdhury, S. (2006). Earnings management through real activities manipulation. Journal of Accounting and Economics, 42, 335–370. Shao, Z., & Lin, Y. (2004). A study on dividend policy of listed firms. Market Weekly, 3,45–47. Sharma, D. (2001). The role of cash flow information in predicting corporate failure: The state of the literature. Managerial Finance, 27,3–28. Sun, Q., & Tong, H. S. (2003). China share issue privatization: The extent of its success. Journal of Financial Economics, 70, 183–222. Wang, Z., & Zhang, W. (2011). Financial flexibility, financing option and catering strategy of Chinese listed firms. Working paper. Xiamen University, China. Retrieved from http://papers. ssrn.com/sol3/papers.cfm?abstract_id=1929726 Wu, L. S., Bo, X. H., & Wang, Y. P. (2007). To what extent have cash flows been managed? Evidence from China. Journal of Finance (Chinese), (3), 162–174. Xue, S., Cai, X., & Guo, H. (2008). Are cash flows managed? Empirical evidence from quarterly financial statements of listed Chinese firms. China Accounting and Finance Review, 10,69–121. Zhang, R. (2009). Cash flow management in the Chinese stock market: An empirical assessment with comparison to the U.S. market. Frontiers of Business Research in China, 3, 301–322. Zhang, W. N. (2008). Real activities manipulation to meet analysts’ cash flow forecasts. Working paper. Cheung Kong Graduate School of Business, China. Retrieved from http://papers.ssrn. com/sol3/papers.cfm?abstract_id=1013228
Journal
China Journal of Accounting Studies
– Taylor & Francis
Published: Apr 3, 2014
Keywords: cash flow manipulation; dividend policy; firm value; seasoned equity offering
