Dynamic relationship between technological knowledge and products: diversification strategy for firm growth

ABSTRACT

This study has two separate analyses that examine (i) how a firm's patterns of technological and product diversification are determined through the coevolution of technological knowledge and products, and (ii) the combined role of technological and product diversification in shaping sales growth. We build a dataset of listed manufacturing firms in Korea over the period 1988–2014 by merging financial, patent, and product information. In the first analysis, we find that specialised knowledge in the firm plays an important role not only in improving products in existing market segments but also in developing products in new segments. The second analysis observes that firms with diversified product portfolios based on specialised knowledge show higher sales growth compared to other firms. Our findings suggest that firms need to accumulate specialised knowledge and apply it to products in diverse market segments.

KEYWORDS:

• Technological knowledge and products
• diversification
• firm growth
• coevolution

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 If for firms with more than 10 segments, the 10th segment contains the rest. In our sample, there are 3.5 market segments per firm on average.

2 Lybbert and Zolas (2014) use NAICS 2007. Thus, this study transforms NAICS2007 to SIC codes.

3 For example, A21D and A01F belong to different knowledge categories according to the 4-digit IPC codes. However, they are in the same SIC code category (0111), and thus, are categorized in the same technological fields in this study.

4 $\begin{array}{c}\begin{array}{rl}\left(i\right)c_i=& s_0+s_{1}KS\mathrm{\_}Inn{o}_{i0}+s_{2}PS\mathrm{\_}Inn{o}_{i0}+s_3\overline{X_i}+k_{1i}\\ h_i=& r_0+r_{1}KS\mathrm{\_}Inn{o}_{i0}+r_{2}PS\mathrm{\_}Inn{o}_{i0}+r_3\overline{X_i}+k_{2i}\end{array}\\ \left(ii\right){\rho }_{k1,k2}=corr\left(k_{1i},k_{2i}\right)\end{array}$

5 When KS_Existing (t) and KS_New (t) are analyzed as dependent variables, respectively, strategic choices in PS (t-1) do not affect any strategic choices in KS (t).

6 We expect that the reason why product innovation does not persist in Table 3 is that the two effects have been offset.

7 Our sample contains 446 technological fields (4-digit IPC) and 549 market segments (4-digit SIC).

8 Descriptive statistics and correlations of variables are shown in Table A1.

9 The IMR is calculated from a selection equation, including R&D intensity, debt ratio, profit ratio, firm size, age, and year and industry dummies. The result is reported in Table A2.

10 We split the sample into two groups of firms divided by the median of Tech_DIV. The positive impact of Prod_DIV on sales growth is only observed in a group with lower than the median but it is not significant in another group.

11 RTA measures the relative advantage of firm i at time t in the technological field k as follows: $RT{A}_{i,k,t}=\frac{\left(P_{i,k,\left[t-3,t\right]}/P_{k,\left[t-3,t\right]}\right)}{\left(P_{i,\left[t-3,t\right]}/P_{\left[t-3,t\right]}\right)}$where P is the number of patents.

Additional information

Funding

This work has supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) [grant number 2017R1A2B4009376].

Notes on contributors

Taewon Kang

Taewon Kang is a Research Fellow at Scuola Superiore Sant'Anna. His research interests include economics of innovation,industrial dynamics, and entrepreneurship.

Chulwoo Baek

Chulwoo Baek is an Associate Professor at Duksung Women’s University. His research interests include innovation policy, industrial dynamics, and productivity analysis.

Jeong-Dong Lee

Jeong-Dong Lee is a Professor at Seoul National University. His research interests include data envelopment analysis (DEA), innovation theory, and industrial dynamics.