Which do second-generation heirs prefer in family firms: real investment or financial investment?

ABSTRACT In recent years, the financial investment level of enterprises particularly family firms has increased rapidly. This phenomenon has drawn intense attention from both government regulators and academia. In this study, we argue that the second-generation succession is an important reason for family firms’ preference for financial investment. Using 9,701 firm-year observations of Chinese family listed firms from 2007 to 2018, we find that the second-generation succession has a positive effect on financial investment in family firms, while successors with professional background have less preference for financial investment. Further, the positive effect exists mainly in both preparation and epistasis stages of second-generation succession. The number of family founders’ children will enhance successors' preference for financial investment. Finally, financial investment especially long-term one reduces real investment, and damages future performance.


Introduction
In recent years, China's economy exhibits a trend of transforming from real investment to financial investment. At the macro level, it means that a large amount of capital flows from the real economy to the virtual economy and makes the capital idle in the virtual economy. At the micro level, it means that real enterprises increase financial investment and reduce real investment, which leads to corporate financialization and the industrial hollowing problem (Epstein, 2005). This problem deviates from the basic rule that the virtual economy serves the real economy. It not only damages the creation and accumulation of social wealth but also pushes up the price of financial assets, aggravates financial risks and even results in serious economic crises (Peng & Huang, 2018;. Therefore, it has attracted considerable attention from the government. For example, the report of the Nineteenth National Congress of the Communist Party of China pointed out that 'we will deepen institutional reform in the financial sector and make it better serve the real economy'.
The problem of corporate financialization would shake the foundation of China as a manufacturing country. In fact, family firms are the most important component of China's manufacturing companies. The financialization of family firms has been increasingly obvious with the arrival of the peak of intergenerational inheritance. For example, Haixin Group, a private iron and steel enterprise, turned to concentrate on capital operation and financial investment after the second-generation successor, Mr. Zhaohui Li, taking over the family firm in 2003. However, the group entered the process of bankruptcy and reorganisation due to the capital chain rupture in 2004. Alternatively, instead of taking over the family business, Mr. Jianfeng He, the son of Midea Group's founder, chose to enter the capital market for capital operation, 1 displaying his specific preference for financial investment. In this context, on 1 November 2018, the President Xi Jinping presided over a symposium on private enterprises and emphasised that the new generations of family firms should inherit their fathers' entrepreneurship and focus on the family business, take steps to improve corporate competitiveness. Therefore, this study focuses on the financialization of family firms in China and discusses the reason for second-generation successors' preference for financial investment.
Using a sample of A-share family-controlled firms between 2007 and 2018, this study investigates the relation between the second-generation succession and financial investment in family firms. We find that as financial investment may help successors establish authority and legitimacy by acquiring high returns more quickly, family firms with secondgeneration successors have significantly higher levels of financial investment than their counterparts without succession. However, the professional background of secondgeneration successors is found to reduce their preference for financial investment. The results hold in a series of robustness tests including tests for endogeneity problems. Furthermore, we find that the effect of second-generation succession mainly exists in the preparation and epistasis stages of second-generation succession. The number of family founders' children will enhance the preference of second-generation successors for financial investment. In addition, we find that financial investment particularly longterm one will crowd out real investment, and ultimately damage family firms' future corn performance, displaying important economic consequences.
Our research contributes to the literature in several ways. First, this study enriches the literature of corporate financial investment. The existing literature focuses on the economic consequences of financial investment. Only a few studies explore influencing factors and determinants of financial investment, such as corporate governance and macroeconomic environment (Ke et al., 2019;Liu & Cao, 2018;Y. M. Hu et al., 2017;Wen & Ren, 2015;Yan & Chen, 2018;Z. Yang et al., 2019). In this study, we distinguish the nature of enterprises' property rights and focus on family firms' financial investment. We suggest and find that the intergenerational inheritance is an important factor in shaping financial investment decisions in family firms, thereby extending the literature on determinants of corporate financial investment. Second, we enrich the literature regarding consequences of intergenerational inheritance. As the peak of intergenerational inheritance arrived, many scholars have examined the consequences of intergenerational inheritance and found that intergenerational inheritance may affect corporate performance and market value (Bennedsen et al., 2015;N. Xu et al., 2015;Zhu & Lyu, 2019). The literature also finds that intergenerational inheritance will deeply shape family firms' various decisions including innovation, entrepreneurship, M&A and social responsibility (Q. F. Cai et al., 2019;Carney et al., 2019;Huang, Lu et al., 2018;Li et al., 2016;Pan et al. 2019;Zhao et al., 2015;Zhu et al., 2018). In this study, we are the first to examine the effect of intergenerational inheritance on family firms' financial investment, thereby enriching the literature on consequences of intergenerational inheritance. Finally, our study also provides important practical implications. On the one hand, our conclusions are conducive to deepening the government's understanding of family firms' financial activities and help the government formulate policies to guide firms to focus on their main business. On the other hand, as the peak of intergenerational inheritance arrived, our conclusions indicate that family firms should make the plan of intergenerational inheritance in advance and improve the communication with second-generation successors.

Consequences of intergenerational inheritance
Succession planning is one of the most critical issues in family firms, which determines the development of family firms (Lee et al., 2003;De Massis et al., 2008;Zahra et al., 2004). As such, the consequence of intergenerational succession is one of the most important topics in the family business research. For example, Pérez-González (2006) finds that firms with family CEOs underperform in terms of operating profitability and market value. Similarly, Villalonga and Amit (2006) find that firm value reduced when secondgenerations served as CEOs in Fortune-500 firms. Meanwhile, Cucculelli and Micucci (2008) find similar results in small firms in Italy. In China, N. Xu et al. (2015) find that second-generation involvement enhances firm performance and reduces the firm's tunnelling behaviour. However, Bennedsen et al. (2015) and Zhu and Lyu (2019) find that intergenerational inheritance would damage firm performance. Given those conflicting conclusions, recent studies turn to explore potential situational factors including characteristics of family founders and heirs. For instance, birth order is found to be one potential factor that may affect firm performance in family firms. To appointing secondor subsequent-born child as the successor has a positive effect on post-succession firm performance (Calabrò et al., 2018;Kellermanns et al., 2012;Miller et al., 2014;Nicholson, 2008). In Japanese family firms, Mehrotra et al. (2013) find that adopted heirs' firms outperform blood heirs' firms.
In addition to firm performance, studies also explore to investigate the impact of intergenerational inheritance on corporate decisions. For example, Zhao et al. (2015) and Zhu et al. (2018) find that family firms will conduct significant strategic changes and choose differentiated strategies during the succession window. Q. F. Cai et al. (2019) and Li et al. (2016) point out that the trans-generational resource heterogeneity and the growth experience of second generations promote the portfolio entrepreneurship and the cross-industry mergers of family firms. In addition, Pan et al. (2018) find that family firms will proactively engage in social outreach activities and use more corporate philanthropy in the process of succession. Carney et al. (2019) and Huang, Lü et al. (2018) find that second-generation involvement has a positive effect on corporate innovation activities and innovation efficiency. Further analysis shows that this effect is more pronounced when the second generations are overseas returnees.

Research on corporate financial investment
In recent years, with the substantial investment of real enterprises in financial assets and the increasing profit-making through financial channels (Demir, 2009;Duchin et al., 2017;Krippner, 2005;, scholars have conducted extensive studies on determinants of corporate financial investment. There are two theoretical views in the literature. The profit-driven view believes that as financial investment has a higher rate of return than real investment and insiders including managers and controlling shareholders have tendency to pursue short-term performance and obtain excess return, focal firms will increase investment in virtual areas such as financial industry (Demir, 2009;Du et al., 2017;Wen & Ren, 2015). Liu and Cao (2018) provide evidence that institutional investors drive the financialization of real enterprises due to their myopia. Meanwhile, Yan and Chen (2018) find that firms with poor governance are more likely to hold financial assets. However, the precautionary saving view, proposed by Keynes, suggests that as real investment has the characteristics of long periods, low liquidity and high risk, it is a forward-looking strategy to deal with future uncertainties for real enterprises to hold financial assets with strong liquidity (Du et al., 2017;Duchin et al., 2017;Keynes, 1936;Zhang & Zheng, 2018). Y. M. Hu et al. (2017) find that firms invest financial assets based on the reservoir motivation and macroeconomic factors will affect the allocation of financial assets. It should be noted that the motivations of corporate financial investment are complex and diverse (Ke et al., 2019;Song & Lu, 2015).
Meanwhile, the consequences of corporate financial investment have been widely discussed. Scholars find that financial investment has a 'crowding out' effect, that is, it reduces corporate innovation investment and inhibit innovation activities (Seo et al., 2012;Xie et al., 2014;Xu & Zhu, 2017;Ya et al., 2018). Du et al. (2017) also find that corporate financialization crowds out physical investments, and finally damages the corporate future core performance. Moreover,  find that firm financialization leads to a significant increase in the likelihood of stock price crashes. Conversely, however, Liu (2017) argues and finds that financial assets play a role of 'reservoir' and promote firm innovation.
Overall, prior studies focus on exploring the roles of macroeconomic and corporate governance factors in shaping corporate financial investment behaviour. However, the literature has largely ignored fundamental differences between state-owned enterprises and non-state-owned enterprises particularly family firms. What is more, few studies, if any, have considered the potential effect of intergenerational inheritance on financial investment in family firms. As the peak of intergenerational inheritance arrived, family firms in China would conduct significant strategy changes including financial investment strategy around the succession window (Cai & Luo, 2015;Huang, Lü et al., 2018;Zhao et al., 2015). Therefore, this study intends to explore the association between the secondgeneration succession and financial investment in family firms, thereby making contributions to the literature.

Hypothesis development
According to the authority theory proposed by Max Weber, rule and domination need legitimacy. Legitimacy means the recognition for the person who has authority and the obedience to their commands. Therefore, when the second-generation successors enter the family firms, they should establish the legitimacy for taking over the family firms (Li et al., 2015).
In the process of intergenerational inheritance, both internal and external environment will change dramatically in family firms. As to internal environment, the organisational structure inertia makes the organisation achieve a dynamic equilibrium within family firms before the intergenerational inheritance. However, the dynamic equilibrium will be broken after the involvement of the second generation. The power will be redistributed within the firm caused by the adjustment and replacement between incumbent and fresh employees. As to external environment, family firms carry out business activities in complex networks with different stakeholders. The occurrence of intergenerational inheritance will attract stakeholders' attention to the second-generation and make stakeholders revalue the family firm (Eesley & Lenox, 2006;Zhao et al., 2015). During this period, the functional authority of second-generation successors is much valuable and helpful. Because the functional authority is composed of competence authority and personal authority. Unlike formal authority, functional authority cannot be passed along with the founder's wealth and authority to successors. If second-generation successors lack the functional authority, they will be hard to make their employees obey and implement their decisions in their family firms, thereby suffering from the embarrassment that 'young lords can hardly let others be convinced'. Moreover, secondgeneration successors face competition from their brothers and/or sisters in the intergenerational inheritance process in family firms. Therefore, they have the pressure to make a quick achievement to defeat their succession competitors (N. H. Xu et al., 2019;Li et al., 2015).
Under such environment, to invest in financial assets seems a better choice for second-generation successors to establish their functional authority, as financial investment has the characteristics of short-term high returns and strong liquidity (Demir, 2009;Du et al., 2017;H. J. Wang et al., 2017). What is more, financial investment is a completely new field, which makes it difficult for stakeholders to compare it with the main business of family firms. It means that second-generation successors can bypass the reference point which is set up by their founders and help to establish their personal authority better (Li et al., 2015). In this regard, family founders would also help their second-generation successors make more financial investment for the sake of building successors' functional authority and smoothing the transition of intergenerational succession, thereby protecting their family socioemotional wealth (Liu et al., 2021;Zhu et al., 2018). In short, secondgeneration successors will have preferences for financial investment activities for the sake of their own benefits.
In addition, with the continuous turbulences of global political and economic environment, China and their firms have to change traditional growth mode. It is particularly true for family firms when their intergenerational inheritance inevitably aggravates their operational risk (H. J. Cai et al., 2021;Liu et al., 2021;Zhao et al., 2015). In China, most family firms are in traditional manufacturing industries and face severe financing constraints. Considering the volatility of future cash flow, they are more likely to hold financial assets as precautionary savings in the process of intergenerational inheritance. Furthermore, Chinese family firms are also struggling to restructure and upgrade during the period of intergenerational inheritance . Financial investment turns to be a fine choice with high liquidity and returns. Taken together, we argue that second-generation successors will prefer for making more financial investment than real investment due to both subjective and objective reasons, thereby increasing the level of financial investment in family firms. Therefore, we hypothesise: Hypothesis 1: Ceteris paribus, family firms with second-generation successors have higher levels of financial investment than their counterparts without succession.
As we know, a main reason why second-generation successors cannot establish functional authority is that they lack professional knowledge and experience related to the family business. According to the survey of Hurun Rich List, more than 50% of second-generation successors are unwilling to take over the family business from their parents. Many successors are reluctant to accumulate professional knowledge and skills related to the family business. The Chinese family business report published by China Europe International Business School also points out that second-generation successors are more interested in financial activities than their family business. Meanwhile, a fact reveals that many family successors major in finance or accounting when they study overseas. In other words, many successors lack professional background for facilitating them to establish functional authority after taking over the family firm. In contrast, it is easier for successors with professional background to build up their functional authority and gain the legitimacy for taking over the family firms, thereby reducing their need to invest in financial assets for achieving functional authority.
Meanwhile, successors with professional background tend to have more interpersonal networks in the field of family business. They can use their social relational capital to identify and obtain more investment opportunities in the field of family business. Therefore, they are more likely to make real investment (C. M. Chen & Sun, 2008;S. Y. Chen et al., 2010;Datta & Guthrie, 1994). In addition, according to the prospect theory and anchoring and adjustment heuristic theory, second-generation successors with professional background tends to choose the investment fields that they are more familiar with when they face uncertain environment. Therefore, we posit that professional background can help second-generation successors establish their functional authority and find more investment opportunities in real economy, thereby reducing their preference for financial investment. In general, we hypothesise: Hypothesis 2: Ceteris paribus, second-generation successors with professional background have less preference for financial investment than those successors without professional background.

Sample selection
Our sample includes all family firms listed on China's Shanghai and Shenzhen stock exchanges during the period of 2007-2018. Following prior literature (Faccio & Lang, 2002;La Porta et al., 1999), we define family firms as those in which the controller of the family firm can be traced to a person or a family with at least 10% voting rights. Our initial sample is comprised of 15,484 firm-year observations. In order to alleviate the influence of abnormal samples, we exclude: (1) firms in financial industry; (2) firms under the status of special transaction (ST or *ST); (3) firms with debt exceeding asset value; and (4) firms with missing data. The final sample includes 9,701 firm-year observations.
The information about second-generation succession of family firms and the characteristics of the founders and the successors is manually collected from multiple channels including IPO prospectuses and annual reports. The financial data and corporate governance information are from CSMAR database. The macro statistical data are from the National Bureau of Statistics.

Dependent variable
Following previous studies (Demir, 2009;song & Lu, 2015), we measure the extent of corporate financial investment (FA) as the sum of financial assets divided by total assets. Specifically, financial assets include trading financial assets, derivative financial assets, net short-term investments, available-for-sale financial assets, held-to-maturity investments, net long-term debt investments, net investment properties, financial equity investments, entrust loans and financial management products and trusts.

Independent variable
Following previous studies (Carney et al., 2019;Huang, Lü et al., 2018;Zhu et al., 2018), the fact that the family successor joins the family firm as the top manager or the board member, represents a strong willingness of the founder to carry on the intergenerational inheritance. Specifically, we construct a dummy variable (Success) to capture and define the second-generation succession from two levels. Success equals one if the founder's second generation (e.g. child, daughter-in-law, son-in-law) enters the board of directors or the top management team of a focal family firm or joins its parent firm as the chairman, CEO, vice chairman, vice CEO or director, and zero otherwise.

Scenario variable
In order to test H2, we use a dummy variable (Professional) to capture the background of second-generation successor (Datta & Guthrie, 1994;H. Yang et al., 2015). Professional equals one if the second-generation successor has educational background and/or work experience in outside companies related to the industry of his/her family firm, and zero otherwise.

Statistical models
We construct the following models for testing our hypotheses: where FA is financial investment, Success represents the second-generation succession of family firms, Professional represents the professional background of second-generation successor, Control is a vector of control variables. The variable of interest is Success. According to Hypothesis 1, we predict β to be significantly positive in Model (1). We examine Hypothesis 2 in the subsample of family firms that have second-generation

Variables
Definitions and Measurements FA The sum of financial assets divided by the total assets.

Success
Dummy variable, which equals one if the second-generation successor enters the board of directors or the top management team of a focal family firm or joins its parent firm as the chairman, CEO, vice chairman, vice CEO or director and zero otherwise. Professional Dummy variable, which equals one if the second-generation successor has educational background or work experience related to the family business, and zero otherwise.

SIZE
The natural logarithm of the sum of total assets plus 1.

DEBT
The ratio of interest-bearing debt to total assets.

ROA
The ratio of net income to total assets.

CFO
The operating cash flow divided by total assets.

CURRENT
The ratio of current assets to current liabilities.

RELATION
The ratio of long-term loans to total assets.

GROWTH
The ratio of increase in operating revenue to the operating revenue of previous year TOBINQ The ratio of market value to total assets.

AGE
The natural logarithm of the sum of listing years plus 1.

CFR
The ultimate cash flow rights owned by the ultimate controller.

BOARD
The natural logarithm of the sum of the number of directors plus 1.

INDSIZE
The percentage of independent directors in the board of directors. TURNOVER Dummy variable, which equals one if a focal firm experiences the change of chairman or CEO, and zero otherwise.

DUAL
Dummy variable, which equals one if a focal firm's chairman also serves as the CEO, and zero otherwise.

EAST
Dummy variable, which equals one if a focal firm is registered in Beijing, Tianjin, Hebei, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Hainan or other eastern provinces, and zero otherwise.

M2
The growth rate of money supply and quasi-money supply (M2).

GDP
The annual growth rate of GDP.

Year
A series of dummy variables to represent each sample year.
successors. We predict β to be significantly negative in Model (2). T-statistics from regression analyses are processed by the heteroscedasticity adjustment. We also control for firm and year fixed effects and winsorise all continuous variables at the 1% and 99% levels.  Table 3 presents the regression results for examining the effect of second-generation succession on family firms' financial investment. As Table 3 shows, the coefficients on Success are positive and significant at the 5% level regardless of including control variables or not (Model 1: β = 0.010, T = 2.028; Model 2: β = 0.010, T = 2.014).

Multivariate regression analyses
According to the results of Model 2, family firms with second-generation succession have 1.0% higher level of financial investment than their counterpart without second- generation succession, accounting for 19.23% (=0.010/0.052) of the sample mean with a considerable economic significance. These results provide solid support for our hypothesis H1. Table 4 reports regression results for examining Hypothesis 2 in the subsample of family firms with second-generation successors. As it shows, the coefficients on Professional are negative and significant at the 1% level no matter of control variables are included or not (Model 1: β = −0.282, T = −43.734; Model 2: β = −0.269, T = −17.865). It indicates that second-generation successors with professional background will be more easily to be recognised by their family firms' employees and establish functional authority, thereby mitigating their preference for financial investment. Our hypothesis H2 is also supported.

Robustness test
This study may suffer from potential endogeneity problems. Specifically, we may suffer from the problems of reverse causality and omitted variables. In order to address this concern, we will employ the propensity score matching plus difference-in-differences (PSM-DID) method and the instrumental variable two-stage regression (IV-2SLS) model. As we may also face the problem of sample selection bias, we will use the PSM method to address this problem. Notes: (1) ***, ** and * represent significance at 1%, 5% and 10% levels, respectively, two-tailed; (2) T-statistics, based on heteroscedasticity adjustment, are in parentheses.

PSM-DID method
First, in order to meet the requirement of parallel trends assumption in the DID model with multiple time periods (Blundell & Costa Dias, 2000;Heyman et al., 2007), we conduct a PSM. According to the initial year of second-generation succession, we take all control variables as matching variables and employ 1:1 nearest neighbour matching and year-byyear matching to pair each family firm with second-generation succession (treatment firm) with a family firm without succession (control firm), thereby creating a new matching sample. Second, we generate two dummy variables, i.e. TREAT and POST. TREAT equals one if a family firm is a treatment firm with second-generation successors, and zero otherwise. POST equals one if the observation year equals or is after the successors' succession year, and zero otherwise. For control firms, POST equals zero in all sample years. Then, we construct an interaction TREAT×POST. According to the principle of DID models, the interaction can help us estimate changes in the levels of financial investment before and after the intergeneration succession by compared with family firms without succession. The regression results are shown in Table 5.
As Table 5 shows, the coefficient of TREAT×POST is positive and significant at the 5% level. It indicates that compared to family firms without succession, the level of financial investment significantly increases after the intergeneration succession in family firms with second-generation succession. These results further support our hypothesis H1. Note: (1) Due to the serious lack of information on the second-generation successors' background, the sample of this regression models is comprised of 1,172 observations; (2) ***, ** and * represent significance at 1%, 5% and 10% levels, respectively, two-tailed; (3) T-statistics, based on heteroscedasticity adjustment, are in parentheses.
In addition, Table 6 displays the regression results for testing the parallel trend assumption. Specifically, BEFORE6, BEFORE5, BEFORE4, BEFORE3, BEFORE2 and BEFORE1 are dummy variables that equal one, respectively, if the year of a focal observation is 6th year, 5th year, 4th year, 3rd year, 2nd year or 1st year before the succession year, and zero otherwise. AFTER1, AFTER2, AFTER3, AFTER4, AFTER5 and AFTER6 are dummy variables that equal one, respectively, if the year of a focal observation is 1st year, 2nd year, 3rd year, 4th year, 5th year or 6th year after the succession year, and zero otherwise. The dummy variable CURRENT equals one if the year of the observation is the succession year, and zero otherwise.
firms with second-generation succession have higher levels of financial investment after succession than their counterparts without succession.

IV-2SLS model
According to prior studies (Huang, Lü et al., 2018;Wu et al., 2019), we use two instrumental variables: familism culture (Familism) and decline rate of the regional birth rate (Decline). Specifically, Familism is measured by the familism culture of the family founder's native place. Familism culture reflects the difference in people's trust between family members and outsiders. Founders with stronger familism beliefs are more likely to trust family members, and thus tend to pass down their family firms to their later generations (Bertrand & Schoar, 2006;Fukuyama, 1995;Wu et al., 2019). However, the familism culture would not have any direct impact on an individual family firm's financial investment. The World Value Survey (WVS) provides the data regarding the extent to what people in particular province trust family members and acquaintances. We choose the first 2007 survey and the latest 2017 survey and construct two variables, i.e. WVS07 and WVS17, respectively, to calculate the differences between the mean values of local people's trust in family members and outsiders in each survey and match the mean value of WVS07 and WVS17 of each province with the family founder's native place to measure the focal firm's familism culture (Bertrand & Schoar, 2006;Wu et al., 2019). Decline is calculated as the difference between the birth rates in 1989 and in 2018 divided by the birth rate in 1989. The regions with slower decline rate of birth rate relatively have more children be born, indicating that family firms located in this region are more likely to have secondgeneration succession. Moreover, due to the limited resources and more children as family candidates, family founders also need to identify the successors as soon as possible to prevent the conflicts within the families (N. H. Xu et al., 2019). However, the changes of regional birth rate have no direct impact on an individual family firm's financial investment. We obtain the provincial-level data of birth rate from the China Statistical Yearbook and match the variable Decline of each province with the family founder's native place. Table 7 reports the regression results of IV-2SLS. In the first-stage regression, consistent with our expectations, the coefficient of Familism is significantly positive and the coefficient of Decline is significantly negative. More importantly, in the second-stage regression, the coefficient of Success is positive and significant at the 5% level, which provides further support for our hypothesis H1. Therefore, the problem of reverse causality seems insignificant and does not change our main findings.

PSM method
We adopt the PSM method to address the problem of sample selection bias. Specifically, we use Success and Professional as dependent variables, and take all control variables as independent variables to calculate propensity scores acquired from logit regression model. Then, we employ 1:1 nearest neighbour matching to pair each treatment observation with an observation from the control group to create a new matched sample and rerun related regressions. The results are shown in Table 8.
As Table 8 shows, the coefficient of Success is positive and significant at the 5% level, and the coefficient of Professional is negative and significant at the 1% level. These results are consistent with those in Tables 3 and 4. Therefore, our main findings are robust to the problem of sample selection bias.
In addition, we also conduct several other robustness checks including (1) deleting observations during 2007-2008 for controlling for the effect of 2008 financial crisis; (2) excluding observations from the real estate industry due to its financial attribution in China (Du et al., 2017;Krippner, 2005;Zhang & Zheng, 2018); and (3) applying the Tobit regression model given that some values of FA are clustered at zero. The results (not reported here) consistently support our hypotheses.
construction of successors' authority are in dynamic changes during the period of intergenerational inheritance (Hauck & Prügl, 2015). Therefore, we further examine potential different effects of successors on financial investment at different stages of succession. According to prior studies (Fan et al., 2012;Liu et al., 2021;Zhu et al., 2018), we divide the process of second-generation succession into three crucial stages and construct three correspondent dummy variables, i.e. succession preparation stage (Success1), succession epistasis stage (Success2) and succession independent governance stage (Success3). Success1 equals one if the second-generation successor takes the position that is important but non-core in the focal family firm or its parent firm, such as director, vice chairman, vice CEO, supervisor and zero otherwise. Success2 equals one if the second-generation successor takes the corn position (e.g. chairman, CEO) in the focal family firm or its parent firm and the elder generation still works in the firm or its parent firm, and zero otherwise. Success3 equals one if the second-generation successor takes the corn position in the firm or its parent firm and the elder generation exits the firm and its parent firm, and zero otherwise. Then, we regress FA on Success1, Success2, Success3 respectively. The results are shown in Table 9.
As Table 9 shows, the coefficient of Success1 is significant and positive at the 10% level, the coefficient of Success2 is significant and positive at the 5% level, while the coefficient of Success3 is positive but insignificant. These results suggest that the positive effect of second-generation succession mainly exists in both preparation stage and epistasis stage but not in the independent governance stage. Since successors have stronger need to build up their functional authority in the early stages of succession, these results are consistent with theoretical logics in developing our hypotheses.

The impact of the number of family founders' children
In developing hypotheses, we argue that the competition from their brothers and/or sisters will force successors to make a quick achievement through financial investment to obtain legitimacy and defeat their succession competitors (N. H. Xu et al., 2019;M. L. Wang et al., 2014;Li et al., 2015). The competition is naturally related to the number of family founders' children. Therefore, we further investigate the effect of the number of family Notes: (1) ***, ** and * represent significance at 1%, 5% and 10% levels, respectively, two-tailed; (2) T-statistics, based on heteroscedasticity adjustment, are in parentheses.
founder's children. The number of family founder's children (Children) is calculated as the number of children owned by the founder or the actual controller of the family firm (Calabrò et al., 2018). In the subsample of family firms with second-generation successors, we regress FA on Children. The results are shown in Table 10. As Table 10 shows, the coefficient of Children is positive and significant at the 5% level. It indicates that as more succession candidates will intensify the competition that successors face in the process of intergenerational inheritance, successors are more likely to engage in financial activities for the sake of making achievements to stabilise their succession legitimacy, thereby increasing the level of financial investment in family firms.

The effect of financial investment on real investment
Despite of two opposite theoretical views regarding the effect of financial investment on real investment, i.e. 'crowding out' effect and 'reservoir' effect, we may wonder what effect exactly financial investment has on real investment in our sample firms. Therefore, we further investigate the effect of financial investment on real investment in family firms. Specifically, real investment (Capinv) is measured as the ratio of Δ (net fixed assets + net construction in progress + construction material) to total assets (Du et al., 2017). Considering that there are substantial differences in characteristics such as liquidity between short-term financial assets and long-term financial assets (Huang, Wu et al., 2018), we further divide financial investment into short-term financial investment and long-term financial investment. The short-term financial investment (FAS) is measured as the ratio of (trading financial assets + derivative financial assets + net short-term investments + entrust loans + financial management products and trusts) to total assets. The longterm financial investment (FAL) is measured as the ratio of (available-for-sale financial assets + held-to-maturity investments + net long-term debt investments + net investment properties + financial equity investments) to total assets. Then, we regress Capinv on FA, FAS and FAL respectively. The results are shown in Table 11.
As Table 11 shows, the coefficient of FA is significant and negative at the 5% level. When including FAS and FAL together in the regression, the coefficient on FAS is insignificant but the coefficient on FAL is negative and significant at the 1% level. These results (2) ***, ** and * represent significance at 1%, 5% and 10% levels, respectively, two-tailed; (3) T-statistics, based on heteroscedasticity adjustment, are in parentheses.
suggest that financial investment particularly long-term financial investment reduces real investment, displaying the 'crowding out' effect in family firms.

The effect of financial investment on future core performance
Since financial investment crowds out real investment, we can naturally expect that financial investment will damage firm core performance. In order to examine this prediction, we further regress firm core performance on financial investment. According to prior studies (Du et al., 2017;C. H. Hu et al., 2015), we measure firm core performance (CPF) as the ratio of (operating profit -investment income -Loss and Gain in fair value + Income from joint venture and joint venture investment) to total assets. The regression results are shown in Table 12.
As Table 12 shows, the coefficients of FA and FAL are significant and negative, but the coefficient of FAS is insignificant. These results indicate that financial investment particularly long-term financial investment indeed damages future corn performance in family firms. It suggests that financial activities are more likely to be profit-driven, which deviates from the fundamental principle of virtual economy back-feeding real economy. Therefore, the transforming from real investment to financial investment has negative impact on the long-term development of family firms.

Conclusion
The real economy is the foundation for a country's long-term development, stability and prosperity. However, a large number of real enterprises are involved in financial industries and hold large amounts of financial assets in recent years. Among them, most are family firms. In particular, as the peak of the intergenerational inheritance arrived, many news and cases report that second-generation successors prefer financial industries and financial investment. Therefore, this study aims to explore the relationship between the second-generation succession and the financial investment in family firms.
Using a sample set of Chinese family listed firms during the period of 2007-2018, we find that there is a positive association between the second-generation succession and Notes: (1) ***, ** and * represent significance at 1%, 5% and 10% levels, respectively, two-tailed; (2) T-statistics, based on heteroscedasticity adjustment, are in parentheses.
financial investment. Second-generation successors with professional background related to their family business have less preference for financial investment. In addition, we find that the positive effect of second-generation succession is mainly concentrated in the preparation and epistasis stages of second-generation succession. Moreover, a large number of family founders' children will intensify successors' competition among family candidates and increase levels of financial investment in family firms. Finally, we find that financial investment particularly long-term financial investment will crowd out real investment and do hard to future core performance of family firms.
Our findings have important practical implications. First, our findings suggest that family firms should make the succession plan in advance for better cultivating succession candidates' professional background related to the family business and helping successors build functional authority. In this way, when successors take over the family firm, they can have enough functional authority and legitimacy to implement their strategies for pursuing for long-term development. Second, as a response to the spirit of symposium on private enterprises made by President Xi Jinping, second-generation successors of family firms should inherit and carry forward their parents' entrepreneurship. Family firms have a tradition of focusing on their main business in Japan and Germany. Our secondgeneration successors can learn from them and devote themselves to make their family business evergreen. Finally, since the transforming from real investment to financial investment will damage real economy and shake the country's manufacturing foundation, the government should put forward policies to motivate enterprises including family firms to invest more in real activities. In particular, the government should deepen the reform of financial structure and enhance the capability of finance to serve the real economy. Notes: (1) ***, ** and * represent significance at 1%, 5% and 10% levels, respectively, two-tailed; (2) T-statistics, based on heteroscedasticity adjustment, are in parentheses.