Certification Benefits in High-Tech Industry: Evidence from Supply Contracts in the Korean Market

Abstract

We examine the certification benefits of supply contracts in the Korean market, with special attention to high-tech and KOSDAQ supplier firms by analyzing the influences of supply contract announcements on supplier firms’ valuation and operating performances. We find that the announcement of supply contracts have a significantly positive effect on the supplier firm value. This positive effect is greater for the suppliers in the high-tech industry especially in the KOSDAQ market. We also find that the enhancements of operating performances are greater for high-tech firms traded in KOSDAQ. These results indicate stronger certification benefits of supply contracts by high-tech suppliers.

KEY WORDS:

• certification
• high-tech
• product market relation
• supply contracts

ORCID

Noolee Kim

0000-0001-7821-1703

Kyoung-Min Kwon

0000-0001-9732-5570

Notes

^1. As explained in the data section, our sample supply contracts account for more than 10 percent of the previous year’s total sales of supplier firms. Therefore, the certification effects of supply contracts we discuss in the article are, to be exact, the certification benefits of supply contracts with large customers.

^2. The high-tech industry has been the main interest in many studies due to its unique characteristics. Please see Tsai and Chang (2008) for the limited accounting discretion issue and Lin, Zhao, and Guan (2014) for the R&D issue.

^3. These studies in capital-raising literature provide empirical evidence that high-tech firms have higher underpricing in IPOs and higher discount in SEOs due to their higher level of uncertainty.

^4. DART is an electronic disclosure system that allows companies to submit disclosures online, where they become immediately available to investors and other users.

^5. Please note that our entire empirical results should be interpreted as results for large customers given this data construction procedure.

^6. Even though DART provides a web-based data retrieving system, most of the collecting/compiling of data should be done manually with researchers’ discretion, and therefore the sample construction process costs tremendous time and labor. Because our sample cannot be obtained from any established database, this makes our data highly unique. To the best of our knowledge, no previous study employs a data set similar to ours.

^7. FnGuide Database is a database serviced by FnGuide Inc., which provides the stock returns and financial information data of Korean firms. In Korea, it is one of the most popular databases for academic research in the finance area.

^8. The list of KSIC for high-tech industry is available from the authors on request.

^9. The related literature generally employs one of the two criteria to define a high-tech industry. The first is to use the classification of the SDC database (Lowry and Schwert 2002; Lowry and Shu 2002). The high-tech industry in this criteria covers biotech, computer equipment, electronics, communications, and general technology. The second is to follow the definition of Loughran and Ritter (2004) (Ljungqvist and Wilhelm 2003; Mola and Loughran 2004; Ritter and Welch 2002). This definition based on SIC codes covers computer hardware, communications equipment, electronics, navigation equipment, measuring and controlling devices, medical instruments, telephone equipment, communications services, and software. The coverage of the high-tech industry in this article is generally consistent with the two criteria.

^10. We should be careful, however, in interpreting these results because the statistics are based on the numbers unadjusted for industry effects. We address this issue in our univariate analyses (Table 4).

^11. We use industry-adjusted measures, and therefore this classification scheme is the same as dividing the entire sample according to whether suppliers’ sales growth rate and ROA changes are above or below their respective industry medians.

^12. The basic intuition behind this second-best strategy is as follows. We have two kinds of variables, one for the valuation effects and the other for the change of operating performance to measure the effects of supply contracts, which include the certification effects and the effects of deal quality together. Because the effect of deal quality is mostly harvested in a short term, we can expect that the short-term change of operating performance reflects most of the value of the contract at hand and some certification effects. Although this strategy is not flawless because the variables for the change of operating performance contain some certification benefits, the control of these variables in the multivariate regression makes our estimate of certification benefits more conservative.

^13. Here, we regard the year 2000 as the ending point of the dot-come bubble period for the following two reasons. First, there is a general consensus among previous literature on the dot-com bubble in the U.S. market about the ending point of the period, which is the year 2000 (Griffin et al. 2011; Ljungqvist and Wilhelm 2003; Ofek and Richardson 2003; Ritter and Welch 2002). Second, we examine the trend of several market indexes for the Korean stock market (KOSPI, KOSDAQ, and KOSDAQ-IT indexes) and find that the dot-com bubble in the Korean stock market exhaustively burst before the end of the year 2000.

^14. In unreported tests, we estimate the same regression specifications using probit regressions instead of logistic regressions and get qualitatively similar results.

^15. We still do not know fully about the determinants of making supply contracts, especially for KOSPI supplier firms.

Additional information

Funding

This work was supported by the research fund of Hanyang University (HY-2011-G).