Jump to ContentJump to Main Navigation
Multinational Firms in ChinaEntry Strategies, Competition, and Firm Performance$

Sea-Jin Chang

Print publication date: 2013

Print ISBN-13: 9780199687077

Published to Oxford Scholarship Online: January 2014

DOI: 10.1093/acprof:oso/9780199687077.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 23 May 2017

(p.180) Appendix 4 Performance Implications of Conversion of Joint Ventures to Wholly Owned Subsidiaries

(p.180) Appendix 4 Performance Implications of Conversion of Joint Ventures to Wholly Owned Subsidiaries

Source:
Multinational Firms in China
Publisher:
Oxford University Press

Two research colleagues and I studied the impact of foreign subsidiary on performance when a firm switches from joint venture to wholly owned subsidiary (Chang, Chung, and Moon 2013). In order to single out the performance improvement associated with the changes from joint venture to wholly owned subsidiary, we compare the performance differentials between joint ventures that convert with comparable joint ventures that do not convert. In order to find the most comparable firms in each group, we use a technique called propensity score matching. The propensity score matching technique calculates the ex ante likelihood of a joint venture converting to a wholly owned subsidiary. If two firms have similar propensity scores, their ex ante likelihoods of conversion are the same, even though one converted while its match did not.

In order to maximize the contrast between pre-conversion and post-conversion, we focus on minority joint ventures in which the foreign partner owns between 25 and 50% of the equity.2 We identify 31, 435 minority joint ventures that appear between 1998 and 2006. Of these, 2991 converted to wholly owned subsidiaries and 28, 444 remained joint ventures, either until they disappeared from the database or until 2006, the end of our period of analysis. Because we measure changes in performance from year t through year t+3 and because we include lagged explanatory variables when we calculate propensity scores, we restrict our sample to firms for which we have at least five consecutive years of (p.181) observations, i.e., from year t-1 through year t+3, with firm conversion occurring in year t. Thus, we consider only the cases in which conversion occurred by 2003. Among the 19, 557 minority joint ventures lasting for at least five consecutive years, 840 firms converted to wholly owned subsidiaries. We match each of these 840 firms with the most similar minority joint venture that did not convert during the sample period within the same 3-digit SIC industry and year.

We include other variables in the conversion model. We use return on assets (ROA), defined as net income divided by total assets, as our measure of financial performance. In order to address the concern that ROA may be sensitive to financial leverage or non-operating income, e.g., asset sales and tax payments, we also use operating return on assets (operating ROA), which is defined as operating income divided by total assets. The results are consistent. It is important to note that financial performance indicators based on subsidiary-level accounting data are susceptible to potential biases from transfer pricing, scope economies, and cross-selling. As far as transfer pricing is concerned, our research design poses a conservative test against any downward bias, as we expect higher profitability improvement in joint ventures-turned-wholly-owned subsidiaries compared to continuing joint ventures. It is usually harder to repatriate profits via transfer pricing schemes in joint ventures compared to wholly owned subsidiaries, as joint venture partners can monitor multinational partners’ transfer pricing schemes.3 We discuss potential income shifting activities of multinational firms in China in Chapter 6 and in Appendix 10.

We include several firm indicators to predict ownership conversion. To control for any size-related factors leading to conversion, we measure firm size as the logarithm of assets. We operationalize firm age as the calendar years passed since a firm’s establishment. This controls for local knowledge and/or resources gained from a firm’s joint venture experiences (Hennart 1991). Leverage is defined as total debt divided by total assets. We measure export ratio as export sales divided by total sales, which reflects a firm’s export orientation. In calculating propensity score, we include ROA, leverage, and export ratio of a joint venture at one year (p.182) prior to conversion to control for any real option values (Kogut 1991; Kogut and Chang 1996). For instance, a multinational parent is more likely to acquire a joint venture with higher profitability, a sound balance sheet, and export platform, as if it exercises a call option (option to acquire).

Intangible assets ratio measures the relative importance of a firm’s intangible assets, defined as intangible assets divided by total assets. We believe that the intangible assets item on a foreign subsidiary’s balance sheet is a key indicator of the transfer of intangible assets from multinational parents to foreign subsidiaries. R&D or advertising-related activities carried out in individual foreign subsidiaries would be regarded as current expenditures on subsidiaries’ income statements, but technology, brands, and trademarks developed by multinational parents elsewhere and then transferred to their foreign subsidiaries are usually treated as investments in intangible assets that can be amortized over several years (Kieso, Weygandt, and Warfield 2010; Chapter 12). Fixed assets ratio, defined as fixed assets divided by total assets, indicates the firm’s capital intensity. This variable also reflects technology transfer in the form of more sophisticated machinery or equipment, which shows up on the balance sheet as an increase in fixed assets. These intangible and fixed assets are important considerations in the joint venture conversion decisions. They also serve as key strategic indicators in observing changes after conversion.

In order to capture the characteristics of foreign parents and local joint venture partners, we look at joint venture ownership structure. Conventional local parent refers to joint ventures in which local parent firms are conventional local firms like state-owned enterprises or collectives, as opposed to modernized local firms like private enterprises or incorporated firms. HMT foreign parent captures the types of foreign parent: ethnic Chinese investors from Hong Kong, Macao, and Taiwan (HMT) or non-ethnic Chinese foreign multinational parents. Finally, we control for the level of foreign ownership by including foreign share, which is the percentage share held by foreign parents, and its squared term. We expect that the likelihood of conversion to wholly owned (p.183) subsidiary will have an inverted-U relationship with the level of foreign ownership, as higher levels of foreign ownership obviate the need for conversion, while lower levels of foreign ownership render conversion more challenging.

Appendix Table 4.1 displays the probit regression results of the conversion decision, describing which joint ventures switch to wholly owned subsidiaries. The first column reports the coefficients from the regression using the full sample; the remaining four columns report coefficients from subsamples of high/low R&D and advertising-intensive industries. We conduct matching separately for each subsample of industries with high/low R&D or advertising intensity. Firm size is negatively significant, suggesting that joint ventures with more assets are less likely to convert, as the foreign partner requires more capital to buy out the local partner. Firm age is negatively significant, suggesting that older joint ventures are less likely to be converted. Conversely, smaller joint ventures and younger joint ventures are more likely to be converted to wholly owned subsidiaries. Joint ventures that are more profitable are less likely to be converted, as it would be expensive to buy out local partners from these profitable joint ventures. Financial leverage influences the conversion decision positively only in the case of low advertising industries. Export ratio is positively associated with the likelihood of conversion in the whole sample, as well as in the high R&D and low advertising-intensive industry samples, reflecting the real option value of an export platform (Kogut and Chang 1996). Ratios of intangible assets to total assets are positively associated with the likelihood of conversion in the high R&D-intensive industry, in line with transaction cost theory (Hennart 1982). This suggests that the higher the proportion of intangible assets, the more likely a venture in a high R&D-intensive industry will convert. The ratio of fixed assets to total assets also demonstrates a significantly positive association with the likelihood of conversion in all samples except for the low advertising-intensive industry subsample, thereby suggesting that firms that require a high fixed asset investment are more likely to be converted to wholly owned subsidiaries. (p.184) (p.185)

Appendix Table 4.1. Conversion decision from joint venture to wholly owned subsidiary

(1)

(2)

(3)

(4)

(5)

JV TO WOS CONVERSION

ALL FIRMS

HIGH R&D

LOW R&D

HIGH ADV.

LOW ADV.

Firm size (t-1)

-0.07***

-0.09***

-0.06**

-0.07***

-0.07***

(0.02)

(0.02)

(0.02)

(0.03)

(0.02)

Firm age (t-1)

-0.02***

-0.01***

-0.02***

-0.02***

-0.01**

(4.0 x 10-3)

(5.0 x 10-3)

(6.0 x 10-3)

(6.0 x 10-3)

(5.0 x 10-3)

ROA (t-1)

-4.5 x 10-3 **

-5.1 x 10-3 *

-4.3 x 10-3

-9.8 x 10-3 ***

0.3 x 10-3

(2.0 x 10-3)

(3.0 x 10-3)

(3.0 x 10-3)

(3.0 x 10-3)

(3.0 x 10-3)

Leverage (t-1)

0.5 x 10-3

1.1 x 10-3

-0.4 x 10-3

-1.9 x 10-3

2.6 x 10-3 **

(1.0 x 10-3)

(1.0 x 10-3)

(1.0 x 10-3)

(1.0 x 10-3)

(1.0 x 10-3)

Export ratio (t-1)

0.08*

0.13*

0.06

0.01

0.17**

(0.05)

(0.07)

(0.07)

(0.07)

(0.07)

Intangible assets ratio (t-1)

6.0 x 10-3

1.2 x 10-2 **

-1.3 x 10-3

3.0 x 10-3

7.2 x 10-3

(4.0 x 10-3)

(6.0 x 10-3)

(6.0 x 10-3)

(6.0 x 10-3)

(5.0 x 10-3)

Fixed assets ratio (t-1)

2.7 x 10-3 **

3.2 x 10-3 **

2.5 x 10-3 *

2.8 x 10-3 *

2.5 x 10-3

(1.0 x 10-3)

(1.6 x 10-3)

(1.4 x 10-3)

(1.8 x 10-3)

(2.0 x 10-3)

Conventional local

-0.20***

-0.23***

-0.18***

-0.21***

-0.19***

parent (t-1)

(0.04)

(0.06)

(0.06)

(0.06)

(0.06)

HMT foreign

0.04

0.02

0.04

0.09*

-0.02

parent (t-1)

(0.04)

(0.06)

(0.05)

(0.06)

(0.05)

Foreign share (t-1)

13.28***

12.94***

14.26***

16.76***

10.70***

(1.78)

(2.62)

(2.50)

(2.80)

(2.23)

Foreign share

-15.80***

-15.79***

-16.68***

-20.27***

-12.42***

squared (t-1)

(2.33)

(3.47)

(3.25)

(3.69)

(2.90)

Industry fixed effects

Yes

Yes

Yes

Yes

Yes

Year fixed effects

Yes

Yes

Yes

Yes

Yes

Region fixed effects

Yes

Yes

Yes

Yes

Yes

Pseudo R-squared

0.14

0.14

0.15

0.18

0.11

Chi-squared

953.37***

446.40***

538.85***

619.31***

376.14***

Observations

19,557

9,195

9,603

8,757

10,262

Note: Standard errors in parentheses.

(***) p 〈 0.01,

(**) p 〈 0.05,

(*) p 〈 0.1.

This table is reprinted with permission from the Wiley (Strategic Management Journal).

(p.186)

Appendix Table 4.2. Financial performance (ROA%) of JV-turned wholly owned subsidiary vs. continuing JV

Whole Sample

YEAR

t-1

t

t+1

t+2

t+3

JV-to-WOS

4.20

4.86

5.73

5.49

5.31

JV

4.24

4.92

4.67

4.34

3.82

ATT

1.12**

1.21**

1.55**

S.E.

0.51

0.61

0.68

# Matches

799

799

799

799

799

HIGH R&D INTENSITY INDUSTRIES

LOW R&D INTENSITY INDUSTRIES

YEAR

t-1

t

t+1

t+2

t+3

year

t-1

t

t+1

t+2

t+3

JV-to-WOS

3.96

4.43

5.65

5.63

4.92

JV-to-WOS

4.43

5.21

5.68

5.28

5.55

JV

4.44

4.68

4.40

4.27

3.41

JV

4.91

5.05

5.08

4.32

4.00

ATT

1.49**

1.60**

1.76**

ATT

0.44

0.80

1.39

S.E.

0.75

0.77

0.89

S.E.

0.75

0.96

0.97

# Matches

388

388

388

388

388

# Matches

410

410

410

410

410

HIGH ADVERTISING INTENSITY INDUSTRIES

LOW ADVERTISING INTENSITY INDUSTRIES

Year

t-1

t

t+1

t+2

t+3

year

t-1

t

t+1

t+2

t+3

JV-to-WOS

3.44

4.08

5.63

5.37

5.11

JV-to-WOS

5.16

5.54

5.78

5.45

5.54

JV

3.41

4.60

3.79

4.05

3.54

JV

5.39

5.77

5.15

4.56

4.94

ATT

2.36***

1.84**

2.08**

ATT

0.86

1.12

0.83

S.E.

0.71

0.85

0.93

S.E.

0.66

0.78

0.88

# Matches

414

414

414

414

414

# Matches

386

386

386

386

386

Note: Standard errors in parentheses.

(***) p 〈 0.01,

(**) p 〈 0.05,

* p 〈 0.1. ATT stands for average treatment effects for the treated. SE stands for standard error. This table is reprinted with permission from the Wiley (Strategic Management Journal).

(p.187)

Among the parent type variables, joint ventures with conventional local firms are less likely to convert, as the foreign partner finds it harder to sever its relationship with state-owned enterprises or collectives because of their government backing. It may therefore be easier to convert joint ventures with private firms. On the other hand, foreign parent type—HMT foreign firm or non-HMT foreign firm—does not appear to influence the conversion decision. The foreign share variable is significant and positive, while its squared term is significant and negative, suggesting that, as predicted, the relationship between the level of foreign ownership and likelihood of conversion takes an inverted-U shape. Lastly, we include industry, region, and year fixed effects to control for their respective shocks. All unknown rational or irrational factors are commonly held by all firms and so treated as the error term in the probit model.4

With propensity scores calculated from this probit regression, we match these two groups of firms within the 3-digit SIC industry and within the same year to ensure a tight match. This matching procedure generates matches for 799 of the 840 converted wholly owned subsidiaries, with the same number of continuing joint ventures from the full sample of 19, 557 minority joint ventures that remained in our database for at least five years. Appendix Table 4.2 shows the value of ROA for both converted and remaining joint ventures over time. This is based on the difference-in-differences estimation, which denotes the differences in the performance measure estimates for the 799 converted wholly owned subsidiaries relative to the performance of their counterparts in the control group. The average treatment effect on the treated (ATT) measures the difference in cumulative changes in ROA since the year of conversion (at time t) between the two groups. The results demonstrate that, for each of the first three years after conversion, converted wholly owned subsidiaries demonstrate 1.12, 1.21, and 1.55 average percentage point greater increase in ROA than firms that remained joint ventures with similar observed characteristics. Appendix Table 4.2 also displays the performance differences between the converted wholly owned subsidiaries and continuing joint ventures in subsamples of (p.188) high/low R&D and advertising-intensive industries. This is analogous to De Loecker (2007), Girma and Görg (2007), and MacGarvie (2006), who present separate results between two subgroups. The difference in ROA increase of the converted wholly owned subsidiaries compared to continuing joint ventures seems to be present only in high R&D or advertising-intensive industries. In low R&D or advertising-intensive industries, the differences in ROA are insignificant. These results support transaction cost theory.

In order to identify potential sources for performance improvement, we trace the changes of key strategic indicators like sales, intangible assets, and fixed assets. Converted wholly owned subsidiaries show substantial increases in sales and intangible assets compared to continuing joint ventures. We expected that foreign parents would bring in more sophisticated technology or brands to their converted wholly owned subsidiaries. If a technology transfer or brand introduction is associated with conversion, there should be an increase in subsidiary-level intangible assets on the balance sheet. We find that fixed assets grow faster for converted wholly owned subsidiaries than joint ventures, which captures new investment in physical capital, e.g., the transfer of technology that is embedded in sophisticated machinery or equipment, especially in technology-intensive industries.5 Please refer to Chang, Chung, and Moon (2013) for more details.

Notes:

(2) . Gatignon and Anderson (1988) use these criteria to distinguish among minority and majority joint ventures and wholly owned subsidiaries. We include 50:50 joint ventures in the category of minority joint venture.

(3) . While it is also possible to have upward bias in performances of wholly owned subsidiaries via transfer pricing (i.e., multinationals using transfer pricing to shift profits from their multinational headquarters to their Chinese subsidiaries), it is unlikely given the high corporate tax rates and regulations on profit repatriation in China.

(4) . The explanatory variables included in our probit model may not capture every factor affecting the conversion decision. For instance, profitability, leverage, and export ratios may not fully indicate the option value of a joint venture. Similarly, firm age may be an imperfect measure for differential learning among partners. Some multinational parents may simply favor wholly owned subsidiaries due to their organizational move toward more centralization or changes in their business portfolios (Franko 1971; Killing 1983). There are also some unknown rational reasons. For example, even though multinational parents with high R&D/advertising intensity seek to convert joint ventures to wholly owned subsidiaries, the negotiation may stalemate if they have difficulty finding local managers to oversee operations or if local partners demand an exorbitant price in return for the breakup. Sometimes, the conversion decision can be driven by irrational factors. For instance, some joint ventures, particularly those in industries with low intensity in R&D and advertising, may favor conversion out of imitative or herd behavior, despite no clear benefits (DiMaggio and Powell 1983; Yiu and Makino 2002). (p.232) Other joint ventures, on the other hand, may not convert despite being natural cases for conversion, like those in industries with high R&D and advertising.

(5) . We also trace export ratio and financial leverage for each time period in order to see whether the improved performance of converted wholly owned subsidiaries might be attributable to financial restructuring or an increase in multinational parents’ export sales after conversion. We find no difference between these two groups. Results are available upon request.