Parent-subsidiary investment layers and the value of corporate cash holdings

Figure 1 maps the organizational structure of Asus Corp., a famous multinational corporation in Taiwan, based on “Quanxi Business Operation Report,” Asus Corp.’s 2009 annual report, as follows: ASUS Corp. (TW), at layer zero, the top of the investment structure, indirectly controls the lowest-tier firm, Tubesonic Technology Ltd. (China), at layer six through Pegatron Co. (layer one), HuaWei Investing Corp. (layer two), Kinsus Interconnect Technology Corp. (layer three), Kinsus Holding (Samoa) (layer four), and Kinsus Holding Ltd. (Cayman) (layer five). Thus, the number of investment layers for Asus Corp is six.

Kusnadi (2011) uses a sample of firms listed in Malaysia and Singapore and finds that pyramid firms with a single leadership structure (i.e. the CEO and the Chairman of the Board are the same person) hold more cash than those with more effective governance, consistent with the notion that entrenched managers of pyramid firms have more discretion to hoard cash reserves. However, Kusnadi (2011) uses only an indicator variable to capture the pyramidal structure of ultimate family controllers. They do not include detailed features such as the number of investment layers and the deviation between cash flow rights and voting rights.

For firms with multiple chains in investment structures, we focus on the longest chain, i.e., the chain with the largest number of intermediate layers. All pyramidal firms can be handled in the same manner and processed using the TEJ database.

However, we are not able to rule out the possibility that the value of cash holdings may increase along with the number of investment layers for the following reason. A firm with a large number of investment layers has higher agency costs, which can make it difficult for external funding providers to assess and monitor its operations. Consistent with this argument, Hsu et al. (2015) find that the number of investment layers for FDI (foreign direct investment) in China increases agency costs, which reduces creditors’ willingness to provide capital. Chan and Hsu (2013) directly document a positive association between the number of investment layers and the cost of debt. Myers and Majluf (1984) argue that external finance is costly and cash provides a safe buffer. Corporate cash holdings enable firms to make additional investments without raising external capital, helping companies avoid high financing costs. Thus, as external finance tends to be more expensive for pyramid firms due to their organizational complexity, each additional dollar of cash holdings by such firms may have a higher value.

To calculate the benchmark portfolio’s value-weighted return, we use 25 Fama and French (1993) portfolios based on size and book-to-market. In particular, for each fiscal year, we assign each firm into one of 25 portfolios based on its size and book-to-market ratio.The benchmark portfolios are designed to offset the expected return component of stock i due to its size and book-to-market ratio at the beginning of the fiscal year. As noted in Dittmar and Mahrt-Smith (2007), specification of the returns in excess of the benchmark portfolio controls for risk and discount rate, which may affect both cash holdings and stock returns other than cash holdings and their interaction.

Following some other studies, such as Pinkowitz et al. (2006), we also use Fama and French’s (1998) valuation regression as a robustness check, where the market value of the firm is regressed on cash holdings, the variables of interest, and a set of control variables. The inferences from the untabulated results do not change.

Instead of using cash and cash equivalents to measure a firm’s cash holdings, we also include marketable securities as cash holdings, as do Faulkender and Wang (2006), because marketable securities can also be converted to cash easily; the results are qualitatively the same.

For U.S. studies, the G-index of Gompers et al. (2003) is usually used to proxy for governance; it is constructed using the U.S. database, Investor Responsibility Research Center (IRRC). IRRC tracks 28 distinct corporate governance provisions, including four provisions that intend to delay hostile takeover bidders, six provisions that protect directors and officers from legal liability and job termination, six provisions that deal with shareholder voting rights, six provisions that address state takeover laws, and six provisions that are related to other takeover defenses. For every firm, Gompers et al. (2003) add one point for every provision that reduces shareholder rights and construct a "Governance Index" as a proxy for the balance of power between shareholders and managers. However, in Taiwan, merger and acquisition is not so common and shareholder activism does not rise as much as in the U.S. Following Gul et al. (2017), we construct an index that adds one point for each governance feature that protects shareholder rights as a proxy for the strength of corporate governance. We believe this index better fits the Taiwan business environment.