본 논문은 한국거래소 유가증권시장(KRX)을 대상으로 호가단위가 유동성 및 시장의 질적 수준에 미치는 영향을 분석한 후, 현행 호가단위의 적절성을 파악한다. 기존 연구와는 달리, 호가단위와 시장깊이의 관계에 초점을 두며 모든 분석에서 내생성 문제를 고려한다. 분석결과는 다음과 같다. 첫째, 호가단위를 축소할 경우 스프레드율과 거래활동 측면에서 분석한 유동성 증가로 시장의 질적 수준은 개선되는 반면, 시장깊이 측면에서 분석한 유동성 감소로 인해 시장의 질적 수준은 악화된다. 둘째, 호가단위 축소시 시장깊이가 감소하는 것은 “주문크기”가 감소하고 “최우선호가보다 우선한 주문비율”이 증가하는 데 그 원인이 있다. 셋째, 호가단위를 축소하더라도 호가단위가 점프하는 부근(10,000원 제외)에서 거래되는 종목의 시장깊이는 감소하지 않아, 시장의 질적 수준은 악화되지 않는다. 넷째, 현행 상대호가단위의 상한이 0.2%인 가격대의 경우, 0.2% 상한은 더 이상 축소할 필요성이 없는 최소한의 상대호가단위임을 시사한다.
To determine the appropriate tick size for the Korean Exchange (KRX), we analyze the effect of changing the minimum tick size (hereafter “tick size”) on, first, the liquidity, measured by relative spread, market depth, and volume and secondly, on the overall market quality. The results from these analyses shed light on the optimal tick size for the KRX. Our finding shows an ambiguous effect of changing the tick size on market quality: reducing the tick size increases relative spread and volume but decreases market depth. Also, compared to the U.S. exchanges whose tick size is one cent regardless of the stock price, the tick size of the KRX is a step function of the stock price. This characteristic allows us to study the effect of changing the tick size by comparing stocks with prices just above and just below the level at which the tick size changes. Unlike previous studies, our analyses focus on the effect of changing the tick size on market depth among liquidity variables, by, among others, addressing the endogeneity problem among the variables. Other specific methods used in our analyses are as follows. First, as a preliminary study, we examine whether our results regarding the effects of changing tick size on three variables of liquidity--relative spread, market depth, and volume--in the KRX are the same as those from previous studies on domestic and overseas exchanges. Second, we investigate the effects of changing tick size on investors’ order submission patterns to find the reasons and implications of the adverse effects of reducing tick size on market depth. For both analyses, we conduct panel-data analyses, which are robust even with the endogeneity problem. Rogers’ (1993) standard errors are incorporated in order to adjust for the clustering effect. Third, we test dynamic relations between tick size and market depth in order to obtain detailed results indicating the effects of reducing tick size on market depth. This test is designed to measure the extent to which the KRX tick size could be reduced. For this test, we use the sample stocks whose tick sizes change during the sample period. We, first, sort them into 9 categories based on their upper and lower price limits and, then, conduct “endogeneous” event studies. Using all 689 common stocks listed on the KRX during our sample period of 122 trading days from January 3 to June 30 in 2005, we apply both TAQ (Trades and Quotes) and daily data to each stock. The data are provided by the KRX. The test results for our seven null hypotheses are as follows. •Results from Analyses for the Effects of Changing Tick Size on Liquidity: Reducing tick size has a favorable effect on relative spread and volume, but an unfavorable effect on market depth. This finding of trade-off among the liquidity variables confirms similar results from previous studies; however our finding is more robust, since the endogeneity problem among the variables is taken into account in the regression analyses. Our study also has the practical implication that policy makers must heed unfavorable effect on market depth when considering reducing the tick size as a way to reduce transaction costs. •Results from Analyses Using Investors’ Order Submission Patterns: As tick size is reduced, order size decreases and the ratio of cancellation and correction orders increases. Market depth, in turn, is decreased and market quality deteriorates. As tick size is reduced, the ratio of “order-more-improved to best-order” also increases. This increase leads market depth and relative spread to decrease; as a result, it can either improve or deteriorate market quality due to the trade-off between them. As tick size is reduced, the market-order ratio slightly increases, decreasing liquidity and, in turn, resulting in a slight deterioration of market quality. All together, a decrease in order size and an increase in the ratio of cancellation and correction orders are, at least, principal causes for decreasing market depth as tick size reduces. •Results from Dynamic Relation between Tick Size and Market Depth: Reducing tick size does not decrease the market depths of the stocks whose prices are around the points at which the tick size jumps, except at 10,000 won. This implies that the tick size can be reduced at most price levels without deterioration of market depth. The price range jumping around 10,000 won is the only exception at which market depth decreases with statistical significance as the tick size decreases. This phenomenon is closely related to the large change in the relative tick size (from 0.1% to 0.5%) at this price range, which is a very abrupt change compared to that at other price ranges (from 0.1% to 0.2%, in particular for 5,000 won and 50,000 won price ranges). Therefore, if we set our argument to stocks under 100,000 won, the KRX can reduce the relative tick size 0.2 percent without having significant adverse effects on market depth.
