대주주지분율이 코스닥 IPO 시장의 저평가에 미치는 영향

Abstract

Technology-intensive and innovation-driven small and medium sized enterprises(SMEs) are the growth engine for the future economy. They, characterized by short history, financial unstability and uncertain future perspectives, are frequently facing financing difficulties which are the biggest obstacle to their sustained growth. KOSDAQ in Korea, opened on July in 1996, has been playing the pivotal role in pumping money through IPOs to financially sound SMEs, especially so-called venture firms. For the last 13 years, as the IPO markets of KOSDAQ have been too hot, underpricing was the one of most attractive topics to finance researchers. As shown in many IPO studies worldwide, past researches show that the underpricing tend to snowball as informational asymmetry increases and as the price stabilization policies enhanced for the stronger protection of investors. This paper introduces somewhat different view that incentive of managers with large share-holding to maximize firm value may affect the magnitude of underpricing. Founders of technology-intensive SMEs during IPOs used to remain as CEOs(managers thereafter) as well as large stockholders. The idea is that manager's shareholdings affect firm value in two different levels. As shown in <Figure 1>, at the zone from 0% to x1%, firm value increases as the agency cost decreases assuming that managers becomes more incentive-compatible with the remaining shareholders. But the effect will be marginally deceasing to the shareholding ratio(X-axis) and matures at X2%(Incentive Compatibility Hypothesis). But just after X2%, it jumps as the manager pay great efforts to increase firm value to maximize control premium which he/she get paid during M&A deals. In reality, managers with control may prefer to choose to sell his technology-intensive companies through M&As, when decides to exit, rather than to sell his all shares in the capital market, since the control premium can be significant(Control Premium Hypothesis). Here we assumes that X2% is around 50%. With this ideas in mind, as managers' shareholding effects on firms' value reveal after IPOs, they can be one of the driving forces to increase stock prices after IPOs. Whether the managers' shareholding is the meaningful explanatory variable for underpricing or not, and whether the effects differ by shareholding zones or not should be the empirical questions. For the empirical tests, this paper follow 3 steps. In the first step, we structure the regression model and find statistically significant explanatory variables through repetitive multiple regressions to test whether the managers' shareholding is the meaningful explanatory variable for underpricing or not. And explanatory variables other than shareholding variable chosen at this step will be used to control underpricing effects other than shareholding at step 3 for testing whether the effects differ by shareholding zones or not(two hypotheses testing). They are some financial variables such as intangible assets/total assets ratio, R&D/sales ratio, ROA, and policy variable such as price stabilization dummy, and finally market condition variables such as the premium rate(offering price/face price ratio), the competitive rate of subscription. The regression model finally determined in this step will be used again in third step analysis. In this step, we estimate CAR(Cumulative Abnormal Returns) of 225 corporations listed on the KOSDAQ in the period from 2004 to 2007 and find that the managers' shareholding ratios are strong explanatory variable that influence the underpricing of KOSDAQ IPO market with the statistic significance. Here we used three CARs such as CAR1 as cumulative abnormal returns at the day of IPO, CAR10 and CAR20 as cumulative abnormal returns both in ten business days and in twenty business days. In the second step, in order to get grand view of underpricing at each shareholding zone, we divide shareholding ratio at every 5% level and then calculated CAARs at each shareholding zone. Here we used CAAR1, CAAR10, CAAR20 that averages CAR1, CAR10 and CAR20 as the proxies CAARs. As shown in <Table 6> and <Figure 2>, we can find curving points of CAAR which are roughly 30% and 50%. From the view, we might assume the Incentive Compatibility Hypothesis works, while we do have no idea on Control Premium Hypothesis. In the third step, we used two methods of regressions. One is doing multiple regressions at each shareholding zone with divided data samples each. Other method is the piecewise non-linear regression analysis with full sample data. The first is simple and clear, but lacks of sample data. The second is complex, but adopt non-linearity as shown at <Figure 1> and use full data without losing the informational value. Interestingly, the common outcome was brought out despite the methodological differences. In the zone of less than 30% of large shareholders' share-holding ratio, the more large shareholders' share-holding ratio, the higher underpricing with the statistical significance. So the Incentive Compatibility Hypothesis was firmly supported. But In the section of more than 50% of large shareholders' share-holding ratio, the Control Premium Hypothesis was not proved because statistical significance was not induced in every model even though there showed tendency that the higher the underpricing increased, the more large shareholders' share-holding ratio. Regarding the non-linearity, the piecewise non-linear regression showed that the variables representing curved line turned to be negative with the statistical significance. This results are consistent to the predictionas shown at <Figure 1>. In addition, this paper again describes the relationship between the level of underpricing and the share-holding ratio visually with <Figure 3> and <Figure 4>. Y-axis of these figures are fitted CAARs using regression coefficients shown at <Table 8> and <Table 9> respectively. In calculating the fitted CAARs in each section, average values of explanatory variables except share-holding ratio had been put into the regression equations. As shown in <Figure 3>and <Figure 4>, it is interesting to see that the results of all samples(a) and the sample of venture firms(b) are very similar to those of <figure 1>, the figure of underlying theories. In <figure 4>, the steep increase of the projected CAAR after 65% seems to hold on the Control Premium hypothesis, though not having the statistical meaning. To sum, large shareholders' share-holding ratio has the positive impact on the underpricing in IPO market of KOSDAQ. And this can be largely explained by the Incentive Compatibility Hypothesis. We may attribute the reason why the Control Premium Hypothesis was not proved to the conjecture that large shareholders' incentive to increase control premium may affect stock prices in the very long term bases and thus it has little influence upon stock price just after IPOs. Additional works on the sample of venture firms among KOSDAQ listed firms, showed similar results with all samples of KOSDAQ listed firms in every step as explained above.