This study examines how the offshore listing of Korean equity index derivatives is associated with the domestic KOSPI 200 market. We focus on the launch of Morgan Stanley Capital International (MSCI) Korea Index Futures on Eurex Exchange on July 14, 2025, which created a regulated offshore venue for international investors to trade Korean equity market exposure. Using daily data from January 2024 to May 2026, we analyze changes in the domestic KOSPI 200 market through short-run event-window tests and long-run interrupted time-series regressions. The results show that domestic trading activity declines immediately after the offshore listing, but this short-run decline does not coincide with a significant deterioration in liquidity. Over the longer horizon, KOSPI 200 trading activity and liquidity improve, suggesting that offshore listing may strengthen cross-market linkages and domestic market depth. However, the KOSPI 200's relative market share declines, indicating a redistribution of trading activity across related domestic benchmark products. We also find changes in investor participation: institutional participation increases around the event, while foreign investors' domestic trading share declines over time. Finally, KOSPI 200 volatility becomes more sensitive to the Cboe Volatility Index (VIX) after the offshore listing, suggesting stronger transmission of global volatility shocks. Overall, offshore listing appears to improve some dimensions of domestic market quality while reshaping market structure and increasing exposure to global risk conditions.
1. Introduction
The globalization of derivatives markets has increasingly blurred the boundary between domestic financial infrastructure and offshore trading venues. As equity index derivatives linked to domestic markets become available on foreign exchanges, international investors can hedge, speculate, and adjust country-specific exposures outside the jurisdiction and trading hours of the home market. This development can improve global market access and enhance the international visibility of domestic financial assets. At the same time, it raises important concerns for domestic exchanges and regulators because offshore listings may divert order flow, fragment liquidity, alter investor composition, and strengthen the transmission of global financial shocks to the home market.
These issues are particularly important for emerging and internationally integrated equity markets. Theoretically, offshore derivatives can generate two competing effects. On the one hand, market fragmentation and order-flow migration theories suggest that an offshore contract may compete with domestic products by offering foreign investors an alternative trading venue, potentially reducing domestic trading activity and weakening liquidity concentration in the home market (Domowitz et al., 1998; Economides and Siow, 1988; Barclay and Hendershott, 2004; Degryse et al., 2015). On the other hand, market integration theories suggest that offshore derivatives may broaden global investor attention, improve cross-market hedging opportunities, tighten arbitrage links, and ultimately enhance domestic liquidity and price discovery (Heston et al., 1995; Xu and Fung, 2002; Foucault and Menkveld, 2008; Akbari and Ng, 2020). The net effect of offshore listing on the domestic market is therefore theoretically ambiguous and must be evaluated empirically.
This study examines these issues using the offshore listing of MSCI Korea Index Futures on Eurex Exchange as a novel institutional setting. On July 14, 2025, Eurex launched futures on the MSCI Korea Index under the ticker FMKR. The contract is a US dollar-denominated instrument linked to the MSCI Korea benchmark and provides international investors with a regulated offshore venue for trading Korean equity market exposure. This event is particularly useful for empirical analysis because it represents a clearly dated offshore listing of Korean equity index derivatives and occurred during a period in which the Korea Exchange and policy authorities were actively seeking to improve global accessibility to Korean capital markets while remaining concerned about potential liquidity outflows from the domestic derivatives market.
Although the offshore contract is based on the MSCI Korea Index rather than the KOSPI 200, the domestic KOSPI 200 market provides a natural setting for evaluating the domestic market implications of the offshore listing. Both indices represent broad Korean equity market exposure and share substantial overlap in underlying large-cap constituents. Prior studies show that closely related spot, futures, and cross-listed markets are linked through information transmission, arbitrage, and price discovery channels (Chung, 1991; Chen and Knez, 1995; Craig et al., 1995; Lee, 2012; Joo et al., 2016). Trading activity and price discovery in the offshore MSCI Korea futures market may therefore be transmitted to the domestic KOSPI 200 market through information flows, arbitrage activity, hedging demand, and global investor rebalancing. At the same time, because the MSCI Korea Index and KOSPI 200 are not identical benchmarks, the offshore listing may also redistribute trading activity across related domestic equity index derivatives, including KOSDAQ 150 and KRX 300.
We investigate four dimensions of domestic market adjustment. First, we examine whether the offshore listing is associated with domestic trading activity and liquidity in the KOSPI 200 market. Second, we analyze whether the relative market share of KOSPI 200 changes within the broader domestic equity index derivatives ecosystem. Third, we examine changes in investor participation, focusing on institutional investors, foreign investors, and foreign net buying behavior. Fourth, we test whether the offshore listing changes the sensitivity of domestic market volatility to global volatility conditions, measured by VIX. This multidimensional approach allows us to assess whether offshore listing is associated with the domestic market through order-flow migration, market integration, investor reallocation, or volatility transmission channels.
Empirically, we use daily data from January 2, 2024 to May 29, 2026 and employ an interrupted time-series (ITS) framework centered on the July 14, 2025 offshore listing event. This design allows us to distinguish immediate level shifts around the event from gradual changes in post-listing trends. We complement short-run event-window tests with long-run interrupted time-series regressions estimated using Newey-West standard errors. The main outcome variables include KOSPI 200 trading value, an Amihud-based liquidity measure, KOSPI 200 market-share measures, investor participation variables, GARCH-based volatility, and VIX-related volatility spillover measures.
The empirical results reveal a nuanced pattern. In the short run, domestic trading activity declines after the offshore listing, consistent with temporary adjustment frictions or partial order-flow diversion. However, this decline does not coincide with a significant deterioration in domestic liquidity. Over the longer horizon, KOSPI 200 trading activity and liquidity improve, suggesting that offshore listing is associated with stronger cross-market linkages and greater domestic market depth over time. These findings indicate that offshore listing does not simply weaken domestic market quality.
At the same time, the results show that offshore listing changes the structure of the domestic market. KOSPI 200's relative market share declines both in the short-run event windows and in the long-run trend analysis. Since absolute KOSPI 200 trading activity and liquidity improve over the post-listing period, the decline in relative market share is better interpreted as a redistribution of trading activity across related domestic benchmark products rather than as a deterioration of the KOSPI 200 market. Investor participation also changes. Institutional participation increases around the event, while foreign investors' domestic trading share declines over the longer horizon. Foreign net buying increases immediately after the listing but weakens over time, suggesting a gradual adjustment in foreign investors' domestic trading behavior rather than an abrupt withdrawal from the domestic market.
Finally, we find that offshore listing is associated with stronger volatility transmission from global markets to the domestic market. Prior research suggests that financial integration can strengthen cross-market volatility spillovers, particularly when global investors use related instruments to adjust risk exposures across markets (Xu and Fung, 2002; Dasilas and Molyneux, 2009; Horvath and Yang, 2024; Adeloye and Olawoyin, 2025). Consistent with this view, the sensitivity of KOSPI 200 GARCH volatility to VIX increases after the offshore listing. This result suggests that greater global accessibility comes with a trade-off: offshore derivatives may support liquidity and market integration, but they may also make domestic volatility more responsive to global risk conditions. Thus, the offshore listing appears to deepen the integration of the Korean equity index derivatives market with global financial conditions.
This study contributes to the literature in three ways. First, it provides new evidence on the domestic market consequences of offshore derivatives listings, a setting that has received less attention than equity cross-listings or dual-listed stocks. Second, it contributes to the literature on market fragmentation and integration by showing that offshore listing can simultaneously improve domestic market quality and reduce the relative market share of the primary domestic benchmark. Third, it adds to the literature on volatility spillovers by documenting that offshore derivatives can strengthen the transmission of global volatility shocks to the home market.
The findings also have important policy implications. For the Korea Exchange and financial regulators, offshore listings should not be viewed solely as a threat to the domestic derivatives market. Instead, they should be managed as part of a broader market-access and market-integration process. A phased opening strategy, combined with close monitoring of liquidity outflows, investor composition, cross-market arbitrage, and global volatility transmission, can help preserve domestic market quality while enhancing the international competitiveness of Korean capital markets. The central policy challenge is therefore to ensure that domestic market infrastructure and regulatory surveillance evolve together with the expanding offshore trading environment.
2. Institutional background and hypothesis development
2.1 Offshore listing of MSCI Korea index futures
On July 14, 2025, Eurex Exchange launched futures on the MSCI Korea Index, trading under the ticker symbol FMKR. The contract is a US dollar-denominated financial instrument based on the MSCI Korea Net Total Return Index. This launch represents an important institutional development because it introduced a regulated offshore venue through which international investors can trade Korean equity market exposure outside the domestic derivatives market.
Prior to the launch of FMKR, international investors already had access to Korean equity index derivatives outside regular Korean trading hours through the Eurex/KRX link. However, this access channel was decommissioned shortly before the FMKR launch. KOSPI derivatives were no longer available for trading at Eurex from June 5, 2025, with the last trading day at Eurex on June 4, 2025, following KRX's introduction of extended trading hours on its own platform from June 9, 2025. The pre-existing Eurex/KRX link was primarily designed as an after-hours trading channel for KRX-linked KOSPI 200 products and remained closely connected to the domestic KRX market infrastructure. In contrast, FMKR represents a distinct offshore futures contract listed on Eurex. It is based on the MSCI Korea Net Total Return Index rather than the KOSPI 200 Index, is denominated in US dollars, and is traded and cleared within the Eurex market infrastructure. Therefore, the launch of FMKR should not be interpreted as the first offshore access channel for Korean equity index derivatives in a broad sense. Rather, it represents an incremental but economically meaningful expansion of offshore market access because it provides global investors with a separate MSCI-based, USD-denominated, Eurex-listed instrument for trading Korean equity exposure.
This distinction is important for our empirical setting. Because FMKR is linked to the MSCI Korea benchmark, it is more directly aligned with the benchmark framework used by many global institutional investors for Korean equity exposure. Its US dollar denomination may also reduce the need for investors to transact directly in a won-denominated domestic contract when managing Korea-related index exposure. In addition, the listing on Eurex places Korean equity exposure within a broader global derivatives trading and clearing environment. These institutional features suggest that FMKR may be associated with changes in domestic market conditions through investor attention, hedging demand, cross-market arbitrage, and volatility transmission channels, particularly because it was introduced after the decommissioning of the pre-existing Eurex/KRX link.
The economic significance of this event is closely related to the systemic role of South Korea in global equity portfolios. As a major constituent of global emerging market equity indices, the Korean stock market is widely benchmarked by international institutional investors. The launch of MSCI Korea Index Futures therefore provided an additional instrument for international asset managers, asset owners, investment banks, and other institutional participants to hedge Korean equity exposure, mitigate portfolio risks, and express index-linked views on Korea within a global derivatives ecosystem. Prior studies on Korean derivatives markets also show that KOSPI 200-related products play an important role in market quality, price discovery, and volatility transmission in Korea (Lee, 2012; Kang et al., 2014; Choi, 2023).
From a theoretical standpoint, this offshore listing carries two competing economic implications for the domestic Korean market. On the one hand, according to the liquidity externalities and market fragmentation theories (Economides and Siow, 1988; Barclay and Hendershott, 2004), an offshore futures contract may divert order flow away from the home market by offering international investors an alternative, lower-friction venue for Korean equity exposure. This could lead to a reduction in domestic trading volume or a fragmentation of market liquidity (Degryse et al., 2015). On the other hand, the introduction of the offshore contract may enhance global investor attention (Xu and Fung, 2002), improve cross-market hedging efficiency (Sundararajan and Balasubramanian, 2025), and tighten arbitrage bounds between offshore and domestic markets (Nagano et al., 2007). While these integration mechanisms can stimulate domestic trading activity and alleviate transactional friction (Ramos and Thadden, 2008) (liquidity enhancement), they may simultaneously accelerate the cross-border transmission of global volatility shocks into the domestic financial system (Adeloye and Olawoyin, 2025).
This institutional setting provides a useful empirical opportunity to examine how the introduction of an offshore derivatives market is associated with changes in the home market. In the subsequent empirical analysis, we formally investigate whether the inception of MSCI Korea Index Futures was followed by post-listing adjustments in domestic trading activity, market liquidity, market-share redistribution across competing derivatives, investor composition, and overall market stability.
2.2 Linkages to the domestic KOSPI 200 market
Although the offshore derivative contract examined in this study is anchored to the MSCI Korea Index rather than the KOSPI 200, the domestic KOSPI 200 market provides a useful empirical setting to evaluate the domestic market implications of this offshore listing. The MSCI Korea Index and the KOSPI 200 serve as the two most prominent benchmarks for Korean equity exposure, both predominantly comprised of the same high-market-cap, systemic corporations. Because of this substantial structural overlap in underlying economic exposure and basket constituents, trading activity and price discovery in the offshore MSCI Korea futures market may be transmitted to closely related domestic equity index markets (Chung, 1991; Chen and Knez, 1995; Fremault, 1991). Accordingly, the empirical relevance of FMKR arises not from the creation of offshore access per se, but from the introduction of a separate MSCI-based, USD-denominated offshore futures contract that may interact with the domestic KOSPI 200 market through overlapping benchmark exposure, hedging demand, and cross-market arbitrage.
The informational and structural linkages between FMKR and the domestic KOSPI 200 market may operate through several channels. First, the offshore Eurex contract offers global market participants an auxiliary, additional offshore venue to trade Korean equity risk outside domestic trading hours. Consequently, information and global macro shocks reflected in offshore futures prices may be incorporated into domestic spot and derivatives markets when the Korean market reopens (Lee, 2012; Joo et al., 2016). Second, the inception of FMKR may alter cross-market hedging and arbitrage dynamics (Gromb and Vayanos, 2018). International asset managers holding physical portfolios can utilize the offshore contract for risk management (Topaloglou et al., 2020), while sophisticated arbitrageurs eliminate transient price discrepancies across jurisdictions by executing simultaneous trades in both offshore and domestic instruments (Maldonado and Saunders, 1983). These dual mechanisms—information integration and arbitrage bound-enforcement—may influence domestic trading volume, liquidity supply, and conditional volatility dynamics (Richie et al., 2008).
Third, the offshore listing can trigger a reallocation of trading activity across various segments of the domestic equity index derivatives market. Because the constituent criteria of the MSCI Korea Index are not perfectly congruent with those of the KOSPI 200, offshore listing may not translate into a simple one-for-one shift in KOSPI 200 activity. Instead, investors may adjust their thematic exposures across multiple domestic benchmarks, including the KOSPI 200, KOSDAQ 150, and KRX 300. Therefore, to ensure a comprehensive identification, it is important to investigate not only the absolute levels of KOSPI 200 trading activity and liquidity but also the evolution of KOSPI 200's relative market share within the broader domestic derivative ecosystem.
Finally, the offshore listing may affect the transmission elasticity of global volatility shocks into the home market (Xu and Fung, 2002; Dasilas and Molyneux, 2009). By establishing an additional offshore channel for international capital, the listing potentially increases the sensitivity of domestic equity volatility to global systematic risk factors (Xu and Fung, 2002; Horvath and Yang, 2024) (such as the VIX). In this regard, the KOSPI 200 market provides a useful domestic setting to test whether the offshore derivative listing intensifies the integration between Korean capital markets and global financial conditions.
For these reasons, our empirical framework focuses on the domestic KOSPI 200 market as the primary observation window, while treating the Eurex listing of MSCI Korea Index Futures as the key offshore listing event. This research design allows us to examine whether the introduction of an offshore derivative market is associated with changes in domestic trading activity, liquidity provision, market-share redistribution, and international volatility spillovers.
2.3 Expected effects on domestic market outcomes
The offshore listing of MSCI Korea Index Futures is theoretically postulated to alter the domestic Korean equity index derivatives ecosystem through several distinct economic mechanisms. First, the introduction of the offshore venue exerts a dual, competing influence on domestic trading activity and liquidity dynamics. On the one hand, according to the “order flow migration” hypothesis (Domowitz et al., 1998), an offshore futures contract may cannibalize domestic order flow by providing international investors with an alternative, potentially lower-friction platform for Korean equity exposure. This structural displacement could result in a contraction of domestic trading activity and a fragmentation of liquidity. On the other hand, the “market integration” hypothesis (Heston et al., 1995; Akbari and Ng, 2020) suggests that the offshore contract can amplify global investor attention, broaden sophisticated hedging matrices, and intensify cross-market arbitrage efficiency (Craig et al., 1995). These complementary channels can induce positive liquidity externalities (Hendershott and Mendelson, 2000; Levine, 2001), thereby stimulating domestic trading volume and narrowing bid-ask spreads (Foucault and Menkveld, 2008; Barclay and Hendershott, 2004). Consequently, the net structural effect on domestic trading activity and liquidity remains theoretically ambiguous, dictating a rigorous empirical determination.
Second, the offshore listing may lead to a redistribution of trading activity across domestic equity index derivatives. Given that the offshore contract tracks the MSCI Korea Index rather than the KOSPI 200 directly, institutional investors may dynamically rebalance their portfolio exposure across a spectrum of domestic benchmarks, including the KOSPI 200, KOSDAQ 150, and KRX 300. As a result, the offshore listing could alter not only the absolute trading infrastructure of the KOSPI 200 but also its relative market share within the broader domestic derivative landscape. Crucially, a potential variation in KOSPI 200's market share does not inherently signify market deterioration; rather, it reflects a nuanced structural realignment of trading demand across interrelated index-linked instruments.
Third, the offshore listing may structurally reshape the heterogeneous composition of domestic investor participation. The availability of an offshore trading framework may diminish the reliance of certain global macro funds on direct domestic market access. Conversely, it may heighten trading incentives for foreign and domestic institutional market makers through enhanced cross-border hedging and risk-transfer mechanics (Sundararajan and Balasubramanian, 2025). These behavioral shifts are highly likely to diverge across investor clienteles, as foreign, institutional, and retail investors possess asymmetric access to offshore derivatives, heterogeneous hedging mandates (Pennings and Garcia, 2010), and varying investment horizons (Kim and Yi, 2015; Weston and Ciccotello, 2018). Thus, the reallocation of investor clienteles serves as a vital transmission channel through which the offshore listing impacts the home market.
Finally, the offshore listing may be associated with domestic market stability and the sensitivity of volatility to global shocks. While offshore derivatives can enhance global risk-sharing capabilities and stabilize the home market during normal periods, they simultaneously construct a persistent conduit for the propagation of global systemic shocks (Adeloye and Olawoyin, 2025). If the Eurex market becomes a dominant informational hub for processing global systematic risk, the conditional variance of the domestic KOSPI 200 is expected to exhibit heightened sensitivity to global uncertainty vectors (such as the VIX) in the post-listing era (Kim et al., 2015). Therefore, our empirical framework rigorously dissects both the absolute shifts in domestic conditional volatility and the structural changes in its transmission elasticity to global macro shocks.
Furthermore, these micro-structural alterations in foreign participation are fundamentally expected to scale up to dictate aggregate macro-liquidity and ecosystem-wide market shares. By accommodating international cross-border liquidity demand and minimizing cross-market information asymmetry, the institutionalization of the offshore linkage is hypothesized to activate continuous multi-asset arbitrage trading and informational efficiency. These mechanisms suggest that offshore listing may be associated with domestic market quality and the allocation of trading activity across related index products, but the direction of the net effect remains an empirical question.
In sum, the theoretical predictions regarding the externalities of offshore listings are inherently non-monotonic. Offshore listing may either compromise domestic market quality through liquidity fragmentation and order diversion, or fortify it via enhanced global integration, investor attention, and risk-allocation efficiency. The subsequent empirical analysis is dedicated to resolving these competing theoretical tensions by investigating the net impacts on domestic trading activity, liquidity provision, market-share redistribution, investor clienteles, and volatility spillover dynamics.
3. Data and methodology
3.1 Data and variables
This study uses daily data on the KOSPI 200 index and related Korean equity index markets to examine the effects of offshore Korean equity index derivatives listing on the domestic market. The sample period begins on January 2, 2024 and ends on May 29, 2026. We obtain data on the KOSPI 200 index, trading volume, trading value, investor-type trading activity, and corresponding trading activity for KOSDAQ 150 and KRX 300 from the Korea Exchange Data Marketplace. These data are used to construct measures of domestic trading activity, liquidity, market share, investor participation, and market volatility. We collect VIX data from Investing.com to control for global volatility conditions and to examine volatility spillovers from global markets to the domestic Korean market.
The main event date is July 14, 2025, when Eurex Exchange launched MSCI Korea Index Futures under the ticker FMKR. We use this launch as the offshore listing event because it represents the first major offshore listing of Korean equity index derivatives during the sample period. To avoid potential contamination from subsequent market-access expansions and additional offshore listings, the main empirical analyses are centered on this first event.
To evaluate the domestic market consequences of the offshore listing, we construct outcome variables that capture four dimensions of market adjustment. First, we examine changes in domestic trading activity and liquidity using KOSPI 200 trading value and an Amihud-based liquidity measure. Second, we analyze whether trading activity is reallocated across major Korean equity index derivatives by constructing KOSPI 200 market-share measures based on trading volume and trading value. Third, we examine changes in investor participation using investor-type trading shares and net buying measures for foreign, institutional, and individual investors. Fourth, we examine market stability and volatility transmission using GARCH-based volatility and its sensitivity to global volatility conditions measured by VIX. These variables allow us to test whether offshore listing is associated with the domestic market through trading activity and liquidity, market-share redistribution, investor participation, and volatility spillovers.
To measure domestic trading activity, we use , defined as the natural logarithm of daily KOSPI 200 trading value. To capture market liquidity, we employ a modified version of the illiquidity measure originally proposed by Amihud (2002). Specifically, Liquidity is defined as the negative logarithm of the absolute return-to-trading-value ratio:
where is a small positive constant added to avoid undefined values when daily returns are zero. In the baseline analysis, we set (). This transformation converts the conventional Amihud illiquidity ratio into a liquidity-oriented measure. Higher values of indicate greater market liquidity, while lower values indicate lower liquidity. Therefore, a positive coefficient in the liquidity regressions indicates an improvement in domestic market liquidity.
To examine changes in the relative importance of KOSPI 200 within the domestic equity index market, we construct two market-share variables. Volume Share is defined as KOSPI 200 trading volume divided by the combined trading volume of KOSPI 200, KOSDAQ 150, and KRX 300:
Value Share is defined analogously using trading value. These variables allow us to examine whether the offshore listing changes the relative allocation of trading activity across major Korean equity indices, rather than only the absolute level of trading activity in KOSPI 200.
Third, to examine investor participation, we construct investor-type trading variables using daily trading value data to capture both the intensity and direction of institutional and international participation in the domestic KOSPI 200 market. For domestic institutional investors, we focus on their relative trading dominance by constructing the institutional trading share. For foreign investors, whose capital flows are intrinsically linked to cross-border risk-sharing, we evaluate both their relative market share and their net directional purchasing intensity.
The investor-type trading value share is defined as follows:
where denotes the aggregate buy-and-sell trading value of investor type i (institutional or foreign investors) on day t, and denotes total trading across all investor types on day t.
Symmetrically, to capture the directional urgency and net capital flows specific to foreign investors, we calculate the foreign net buying intensity, defined as foreign net purchases scaled by total market turnover:
where and denote the buy-side and sell-side trading value of foreign investors on day t, respectively. A positive value indicates net purchasing by foreign investors, whereas a negative value signifies net liquidation. Crucially, while relative trading shares reflect intra-market clienteles and structural asset realignments, this foreign net buying behavior directly measures the net capital flows of global funds. This unique directional property renders the foreign net buying intensity distinctively sensitive to global risk-appetite shifts and macroeconomic uncertainty vectors (such as the CBOE VIX), justifying its distinct econometric specification in our long-run analysis.
Finally, for the volatility analysis, we estimate the time-varying daily volatility of KOSPI 200 returns, GARCH Volatility, using the conditional volatility obtained from a GARCH(1,1) model. Following the ARCH/GARCH framework of Engle (1982) and Bollerslev (1986), the return and conditional variance equations are specified as follows:
where denotes daily KOSPI 200 returns, μ is the conditional mean return, is the return innovation, and is the conditional variance. We define GARCH Volatility as the conditional standard deviation, . This variable captures the time-varying volatility of the domestic KOSPI 200 market. To examine global volatility spillovers, we use the VIX as a proxy for US and global market uncertainty. [1]
Table 1 reports the summary statistics for our formalized sample of 585 daily observations spanning the quasi-experimental event horizon. The domestic KOSPI 200 trading activity, logs an average ln(Trading Value) of 16.177, alongside a market-wide Liquidity mean of 21.069. Within the broader Korean index derivatives ecosystem, the KOSPI 200 maintains a dominant posture, commanding a mean Volume Share of 0.351 and a Value Share of 0.394. In terms of key investor participation metrics, domestic institutional investors exhibit a stable presence with a mean Institutional Value Share of 0.237 (S.D. = 0.035), whereas international participants display more dynamic allocation vectors, with the Foreign Value Share averaging 0.335 and spanning a wide operational band from 0.206 to 0.631. Symmetrically, the Foreign Net Buying centers tightly around the zero-bound with a sample mean of −0.004, yet demonstrates substantial intertemporal variation ranging from −0.119 to 0.106. Finally, the baseline risk environment reflects realistic market conditions; the conditional variance logs a mean GARCH Volatility of 0.017, while the contemporary spot Return is centered near zero (0.002). The forward-looking domestic and global uncertainty metrics, VKOSPI and VIX, register sample averages of 27.010 and 17.660, respectively, providing a robust variance baseline to clear our policy parameters from confounding macroeconomic shocks.
Summary statistics
| Variables | N | Mean | S.D. | Min | Max |
|---|---|---|---|---|---|
| Trading activity variables | |||||
| ln(Trading Value) | 585 | 16.177 | 0.517 | 15.107 | 18.117 |
| Liquidity | 585 | 21.069 | 1.123 | 18.756 | 25.734 |
| Market share variables | |||||
| Volume Share | 585 | 0.351 | 0.041 | 0.150 | 0.431 |
| Value Share | 585 | 0.394 | 0.035 | 0.288 | 0.467 |
| Investor participation variables | |||||
| Institutional Value Share | 585 | 0.237 | 0.035 | 0.139 | 0.393 |
| Foreign Value Share | 585 | 0.335 | 0.059 | 0.206 | 0.631 |
| Foreign Net Buying | 585 | −0.004 | 0.030 | −0.119 | 0.106 |
| Daily KOSPI 200 volatility variable | |||||
| GARCH Volatility | 585 | 0.017 | 0.009 | 0.009 | 0.070 |
| Control variables | |||||
| Return | 585 | 0.002 | 0.020 | −0.127 | 0.094 |
| VKOSPI | 585 | 27.010 | 13.856 | 15.070 | 80.370 |
| VIX | 584 | 17.660 | 4.752 | 11.860 | 52.330 |
| Variables | N | Mean | S.D. | Min | Max |
|---|---|---|---|---|---|
| Trading activity variables | |||||
| ln(Trading Value) | 585 | 16.177 | 0.517 | 15.107 | 18.117 |
| Liquidity | 585 | 21.069 | 1.123 | 18.756 | 25.734 |
| Market share variables | |||||
| Volume Share | 585 | 0.351 | 0.041 | 0.150 | 0.431 |
| Value Share | 585 | 0.394 | 0.035 | 0.288 | 0.467 |
| Investor participation variables | |||||
| Institutional Value Share | 585 | 0.237 | 0.035 | 0.139 | 0.393 |
| Foreign Value Share | 585 | 0.335 | 0.059 | 0.206 | 0.631 |
| Foreign Net Buying | 585 | −0.004 | 0.030 | −0.119 | 0.106 |
| Daily KOSPI 200 volatility variable | |||||
| GARCH Volatility | 585 | 0.017 | 0.009 | 0.009 | 0.070 |
| Control variables | |||||
| Return | 585 | 0.002 | 0.020 | −0.127 | 0.094 |
| VKOSPI | 585 | 27.010 | 13.856 | 15.070 | 80.370 |
| VIX | 584 | 17.660 | 4.752 | 11.860 | 52.330 |
Note(s): This table presents summary statistics for the daily variables used in the analysis. The sample covers the period from January 2, 2024 to May 29, 2026
3.2 Empirical design
To examine the effects of offshore listing on the domestic KOSPI 200 market, we employ an interrupted time-series (ITS) framework. This approach is appropriate for our setting because the offshore listing represents a clearly dated market event that may generate both an immediate shift in domestic market conditions and a gradual change in post-event dynamics. The interrupted time-series design allows us to compare the pre-event trajectory with the post-event trajectory and to evaluate whether the offshore listing is associated with changes in the level or trend of domestic market outcomes.
Our baseline specification is as follows:
where denotes the outcome variable of interest, including trading value, liquidity, market-share measures, investor participation measures, or GARCH volatility. is a linear time trend, is an indicator equal to one for trading days on or after the offshore listing date and zero otherwise, and is the interaction between the post-event indicator and the time elapsed since the event. In the empirical analysis, and are scaled by 100 trading days to facilitate economic interpretation.
In this specification, captures the pre-event trend in the outcome variable, while captures the immediate level shift after the offshore listing. The coefficient of primary interest is , which captures the change in the post-event trend relative to the pre-event trend. A statistically significant indicates that the trajectory of the domestic market outcome changes after the offshore listing.
The vector includes control variables that account for contemporaneous market conditions. In the trading activity, liquidity, market-share, and investor participation regressions, we include daily KOSPI 200 returns (Return) and KOSPI volatility index (VKOSPI) to control for domestic return movements and domestic market uncertainty. In the volatility regressions, however, we do not include Return because GARCH Volatility is estimated directly from KOSPI 200 returns. We also do not include VKOSPI as a main control in the volatility specification because VKOSPI is derived from the KOSPI 200 options market and shares a closely related information set with the dependent variable. Instead, we include VIX to capture global volatility conditions.
To examine volatility spillovers, we extend the baseline specification by including VIX and its interaction with the post-event indicator.
In this model, captures the pre-event sensitivity of KOSPI 200 volatility to global volatility conditions, while captures the change in this sensitivity after the offshore listing. Thus, the post-event VIX sensitivity is given by . This specification allows us to test whether the offshore listing is associated with stronger volatility transmission from global markets to the domestic KOSPI 200 market.
All regressions are estimated using Newey-West standard errors with five lags to account for heteroskedasticity and serial correlation in daily time-series data. [2]
4. Empirical results
4.1 Impact on trading activity and market liquidity
In this section, we examine the empirical validity of the competing “trading migration” and “market integration” hypotheses formulated in Section 2.3.
4.1.1 Short-run effects
We first examine the short-run effects of offshore Korean equity index derivatives on domestic trading activity and liquidity. To capture immediate market responses around the offshore listing event, we compare pre- and post-event values within symmetric event windows.
Table 2 reports the univariate event study results for the short-run changes in market dynamics surrounding the offshore listing, evaluating both the (−20, +20) and (−30, +30) trading-day event windows in Panels A and B, respectively.
Short-run effects on trading activity and liquidity
| Variables | Pre | Post | Diff (pre–post) | t-stat |
|---|---|---|---|---|
| Panel A. (−20, +20) Event window | ||||
| ln(Trading Value) | 16.320 | 16.114 | 0.206 | 3.89*** |
| Liquidity | 21.462 | 21.605 | −0.143 | −0.374 |
| Panel B. (−30, +30) Event window | ||||
| ln(Trading Value) | 16.283 | 16.049 | 0.234 | 4.73*** |
| Liquidity | 21.343 | 21.408 | −0.065 | −0.211 |
| Variables | Pre | Post | Diff (pre–post) | t-stat |
|---|---|---|---|---|
| Panel A. (−20, +20) Event window | ||||
| ln(Trading Value) | 16.320 | 16.114 | 0.206 | 3.89*** |
| Liquidity | 21.462 | 21.605 | −0.143 | −0.374 |
| Panel B. (−30, +30) Event window | ||||
| ln(Trading Value) | 16.283 | 16.049 | 0.234 | 4.73*** |
| Liquidity | 21.343 | 21.408 | −0.065 | −0.211 |
Note(s): This table reports mean differences in ln(Trading Value) and Liquidity around the offshore listing event. Panel A uses the (−20, +20) trading-day event window, and Panel B uses the (−30, +30) trading-day event window. The pre-event period includes trading days before the event date, while the post-event period includes trading days after the event date. Diff is calculated as the pre-event mean minus the post-event mean. ln(Trading Value) is measured as the natural logarithm of daily trading value. Liquidity is measured using an Amihud-based liquidity measure. Higher values of Liquidity indicate greater market liquidity. t-statistics are reported for tests of mean differences. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
The short-run empirical evidence offers preliminary baseline support for the “trading migration” or “order flow diversion” hypothesis immediately following the listing event. As shown in Panel A, the mean of ln(Trading Value) significantly declines from 16.320 during the pre-event period to 16.114 in the post-event period. This reduction translates into a mean difference (Pre minus Post) of 0.206, which is statistically significant at the 1% level (t-stat = 3.89). A highly consistent structural pattern is observed in Panel B within the wider (−30, +30) window, where aggregate trading value drops from 16.283 to 16.049, yielding a significant mean difference of 0.234 (t-stat = 4.73). This immediate and robust contraction in trading value suggests that in the ultra-short run, the introduction of the Eurex contract may have temporarily diverted some trading activity away from the domestic market.
In contrast, the short-run response of market liquidity exhibits a distinct, non-synchronous behavior. In Panel A, the modified Amihud liquidity index (Liquidity) marginally increases from 21.462 to 21.605, resulting in a negative mean difference of −0.143, which fails to reach statistical significance at conventional levels (t-stat = −0.374). Similarly, the (−30, +30) window in Panel B shows a statistically indistinguishable difference of −0.065 (t-stat = −0.211).
Taken together, these short-run univariate findings imply that the inception of the offshore derivative listing was followed by an immediate outward migration of transactional volume from the home market, yet it did not concurrently impair domestic transactional efficiency or fragment localized market liquidity. Therefore, while the short-run evidence is consistent with a short-run decline in domestic trading activity, it does not support the argument that such migration immediately deteriorates the liquidity quality of the domestic spot infrastructure.
4.1.2 Long-run effects
The short-run results suggest that offshore listing is associated with a temporary decline in domestic trading activity, while liquidity remains broadly unaffected in the immediate event window. However, short-run event-window tests may not fully capture gradual market adjustment following the introduction of offshore derivatives. The effects of offshore listing may emerge over time as investors adjust their trading venues, market participants incorporate the new offshore instrument into their trading strategies, and liquidity providers respond to changes in cross-market trading opportunities. Therefore, we next examine the long-run effects using an interrupted time-series specification that allows for both an immediate level shift and a change in the post-listing trend.
To examine the structural changes in the long-run trajectories of the domestic market, we conduct interrupted time-series (ITS) regressions. Table 3 presents the estimated parameters for trading activity and liquidity, and Figure 1 visualizes these long-run patterns.
Long-run effects on trading activity and liquidity
| Variables | ln(Trading Value) | Liquidity | ||
|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |
| Trend | 0.0308 | 0.0060 | 0.0614 | 0.1392** |
| (1.072) | (0.211) | (1.055) | (2.408) | |
| Post | −0.1406 | 0.0002 | −0.0177 | −0.4519* |
| (−1.407) | (0.002) | (−0.082) | (−1.666) | |
| Post × Trend | 0.7149*** | 0.4191*** | 0.0184 | 0.9319*** |
| (10.645) | (3.962) | (0.125) | (3.313) | |
| Return | 0.8736 | −4.8780** | ||
| (1.274) | (−2.070) | |||
| VKOSPI | 0.0131*** | −0.0401*** | ||
| (3.280) | (−4.489) | |||
| Constant | 15.8598*** | 15.6406*** | 20.8885*** | 21.5572*** |
| (351.266) | (182.412) | (205.242) | (121.577) | |
| Observations | 585 | 585 | 585 | 585 |
| Adjusted R2 | 0.733 | 0.749 | 0.004 | 0.036 |
| Variables | ln(Trading Value) | Liquidity | ||
|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |
| Trend | 0.0308 | 0.0060 | 0.0614 | 0.1392** |
| (1.072) | (0.211) | (1.055) | (2.408) | |
| Post | −0.1406 | 0.0002 | −0.0177 | −0.4519* |
| (−1.407) | (0.002) | (−0.082) | (−1.666) | |
| Post × Trend | 0.7149*** | 0.4191*** | 0.0184 | 0.9319*** |
| (10.645) | (3.962) | (0.125) | (3.313) | |
| Return | 0.8736 | −4.8780** | ||
| (1.274) | (−2.070) | |||
| VKOSPI | 0.0131*** | −0.0401*** | ||
| (3.280) | (−4.489) | |||
| Constant | 15.8598*** | 15.6406*** | 20.8885*** | 21.5572*** |
| (351.266) | (182.412) | (205.242) | (121.577) | |
| Observations | 585 | 585 | 585 | 585 |
| Adjusted R2 | 0.733 | 0.749 | 0.004 | 0.036 |
Note(s): This table reports interrupted time-series (ITS) regression results for the long-run effects of offshore Korean equity index derivatives on domestic trading activity and liquidity. Models 1 and 2 use ln(Trading Value) as the dependent variable, while Models 3 and 4 use Liquidity as the dependent variable. ln(Trading Value) is measured as the natural logarithm of daily trading value. Liquidity is measured using an Amihud-based liquidity measure. Higher values of Liquidity indicate greater market liquidity. Post is an indicator equal to one for trading days on or after the offshore listing event and zero otherwise. Trend and Post × Trend are scaled by 100 trading days. Post × Trend captures the change in the post-event trend. Return and VKOSPI are included as control variables in Models 2 and 4. Newey-West standard errors with five lags are used. t-statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
Two scatter plots with fitted lines depict trends in trading value and liquidity over time. Panel A shows a scatter plot with fitted lines for the natural logarithm of trading value. The x-axis represents event time in days, ranging from -300 to 200, and the y-axis represents the natural logarithm of trading value, ranging from 15 to 18. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date. Panel B shows a scatter plot with fitted lines for liquidity. The x-axis represents event time in days, ranging from -300 to 200, and the y-axis represents liquidity, ranging from 18 to 28. Gray dots denote daily observations. The vertical dashed line indicates the offshore listing event date.Fitted trends in trading activity and liquidity. This figure plots the long-run fitted patterns around the offshore listing event. Panel A presents ln(Trading Value), and Panel B presents Liquidity. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date. Liquidity is measured using an Amihud-based liquidity measure, with higher values indicating greater market liquidity
Two scatter plots with fitted lines depict trends in trading value and liquidity over time. Panel A shows a scatter plot with fitted lines for the natural logarithm of trading value. The x-axis represents event time in days, ranging from -300 to 200, and the y-axis represents the natural logarithm of trading value, ranging from 15 to 18. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date. Panel B shows a scatter plot with fitted lines for liquidity. The x-axis represents event time in days, ranging from -300 to 200, and the y-axis represents liquidity, ranging from 18 to 28. Gray dots denote daily observations. The vertical dashed line indicates the offshore listing event date.Fitted trends in trading activity and liquidity. This figure plots the long-run fitted patterns around the offshore listing event. Panel A presents ln(Trading Value), and Panel B presents Liquidity. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date. Liquidity is measured using an Amihud-based liquidity measure, with higher values indicating greater market liquidity
Models 1 and 2 in Table 3 report the results for ln(Trading Value). While the short-run analysis in Table 2 indicates an immediate, transitory contraction in trading value, the long-run ITS results provide evidence of a reversal in the post-listing trend of trading activity. As shown in Model 2, which includes control variables for both KOSPI 200 returns (Return) and the KOSPI volatility index (VKOSPI), the coefficient on Post × Trend is positive and statistically significant (0.4191, t = 3.962). Because Post × Trend is scaled by 100 trading days, this coefficient indicates that the post-listing trend in ln(Trading Value) increases by approximately 0.419 log points over a 100-trading-day horizon relative to the counterfactual continuation of the pre-event trend. Since the dependent variable is measured as the natural logarithm of trading value, this estimate corresponds to an approximately 52% increase in the underlying trading value over 100 trading days.[3] This economically meaningful increase supports the market integration hypothesis and suggests that the offshore contract is associated with a gradual expansion in KOSPI 200 trading activity over the longer horizon. Notably, the immediate level-shift coefficient, Post, is statistically insignificant (0.0002, t = 0.002), indicating that the long-run effect is driven by a change in the post-listing trend rather than by a one-time level shift. This pattern is consistent with Panel A of Figure 1, where the fitted post-event trend rises above the counterfactual continuation of the pre-event trend.
Models 3 and 4 present the long-run effects on market liquidity using our Amihud-based Liquidity measure. In Model 4, after controlling for Return and VKOSPI, the coefficient on Post is negative and marginally significant (−0.4519, t = −1.666), while the coefficient on Post × Trend is positive and statistically significant (0.9319, t = 3.313). This pattern indicates an initial decline in Liquidity after the offshore listing, followed by a gradual improvement in the post-listing trend. Since Post × Trend is scaled by 100 trading days, the coefficient of 0.9319 indicates that Liquidity increases by approximately 0.932 units over a 100-trading-day horizon. Because Liquidity is an Amihud-based transformed index rather than a raw level of market liquidity, we interpret this coefficient as a change in the liquidity measure, not as a percentage change.[4] This pattern is consistent with Panel B of Figure 1, where the fitted post-event line initially shifts downward but subsequently rises relative to the counterfactual continuation of the pre-event trend.
The adjusted R2 in Model 3 is low, which is not unexpected because this specification is a parsimonious ITS model without market-condition controls. Daily liquidity is affected by short-run market conditions that are not fully captured by the baseline time-trend specification. When Return and VKOSPI are included in Model 4, the explanatory power improves and the positive Post × Trend coefficient remains statistically significant. Thus, the low adjusted R2 in Model 3 does not materially alter the interpretation of the post-listing liquidity pattern.
Overall, these long-run empirical findings strongly suggest that the offshore listing of Korean equity index derivatives did not compromise or destabilize the domestic KOSPI 200 market. Instead, domestic KOSPI 200 trading activity expanded progressively over time, and market liquidity exhibited a robust, structural improvement after rigorously controlling for underlying market conditions. Taken together, these results are highly consistent with the nuanced view that while the inception of an offshore derivative venue may initially generate short-run adjustment frictions and transitory order diversion, it ultimately fosters greater global market participation and enhances long-run liquidity provision in the domestic market.
4.2 Market share effects
We next examine whether the offshore listing catalyzed a structural realignment in the relative market prominence of the KOSPI 200 within the broader domestic multi-index ecosystem. While our previous findings in Section 4.1 confirm that the Eurex listing expanded the absolute informational and transactional capacity of the KOSPI 200 infrastructure, its impact on relative market shares remains an open empirical question.
On one hand, under the centralized liquidity framework, the expansion of the offshore venue could further entrench the KOSPI 200's institutional hegemony as international participants concentrate their strategic allocations heavily into the primary anchor index. On the other hand, a compelling alternative hypothesis suggests an ecosystem-wide spillover effect wherein the enhanced global visibility and international capital inflows attracted by the offshore derivative listing are not confined solely to the treated asset. Instead, the heightened global attention toward Korean benchmarks could trigger a broader diversification strategy among institutional portfolios, channeling incremental liquidity into secondary indices such as the KOSDAQ 150 and KRX 300. Under this systemic capital redistribution framework, the absolute expansion of the KOSPI 200 could paradoxically coincide with a relative dilution of its domestic market share. To empirically adjudicate between these competing structural dynamics, we rigorously evaluate the dynamic trajectories of KOSPI 200's trading-volume and trading-value shares utilizing both univariate event tests and interrupted time-series (ITS) regressions.
4.2.1 Short-run effects
Table 4 reports the univariate event study results capturing the short-run structural shifts in KOSPI 200 market shares surrounding the offshore listing event. Panels A and B evaluate the empirical dynamics within the (−20, +20) and (−30, +30) trading-day event windows, respectively.
Short-run effects on market share
| Variables | Pre | Post | Diff (pre–post) | t-stat |
|---|---|---|---|---|
| Panel A. (−20, +20) Event window | ||||
| Volume Share | 0.389 | 0.379 | 0.010 | 2.40** |
| Value Share | 0.433 | 0.420 | 0.013 | 4.12*** |
| Panel B. (−30, +30) Event window | ||||
| Volume Share | 0.392 | 0.381 | 0.011 | 2.87*** |
| Value Share | 0.431 | 0.418 | 0.013 | 4.40*** |
| Variables | Pre | Post | Diff (pre–post) | t-stat |
|---|---|---|---|---|
| Panel A. (−20, +20) Event window | ||||
| Volume Share | 0.389 | 0.379 | 0.010 | 2.40** |
| Value Share | 0.433 | 0.420 | 0.013 | 4.12*** |
| Panel B. (−30, +30) Event window | ||||
| Volume Share | 0.392 | 0.381 | 0.011 | 2.87*** |
| Value Share | 0.431 | 0.418 | 0.013 | 4.40*** |
Note(s): This table reports short-run mean differences in KOSPI 200 market share around the offshore listing event. Panel A uses the (−20, +20) trading-day event window, and Panel B uses the (−30, +30) trading-day event window. Diff is calculated as the pre-event mean minus the post-event mean. Volume Share is defined as KOSPI 200 trading volume divided by the combined trading volume of KOSPI 200, KOSDAQ 150, and KRX 300. Value Share is defined analogously using trading value. t-statistics are reported for tests of mean differences. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
The short-run evidence consistently points toward an immediate and statistically significant relocation of relative market shares. As detailed in Panel A, the KOSPI 200 Volume Share exhibits a reliable contraction, declining from 0.389 during the pre-event window to 0.379 in the post-event window. This adjustment yields a mean difference (Diff = Pre minus Post) of 0.010, which is statistically significant at the 5% level (t = 2.40). The contraction is even more pronounced when examining the Value Share specification, which drops from 0.433 to 0.420, translating into a mean difference of 0.013 that reaches statistical significance at the 1% level (t = 4.12).
This pattern of structural share dilution remains highly robust across the expanded temporal horizon reported in Panel B. Within the (−30, +30) day window, the KOSPI 200 Volume Share falls from 0.392 to 0.381, generating a highly significant mean difference of 0.011 (t = 2.87). Symmetrically, the Value Share registers an identical mean difference of 0.013, shifting from 0.431 to 0.418 with robust statistical significance (t = 4.40).
Taken together, these univariate findings provide strong preliminary validation for the asset realignment hypothesis. The immediate contraction in both volume and value shares indicates that the introduction of the offshore derivative platform induced an instantaneous structural rebalancing across domestic indices, temporarily tapering the relative concentration of trading activity within the primary benchmark immediately following the listing.
4.2.2 Long-run effects
To evaluate the longer-horizon market-share patterns of the domestic index ecosystem after the offshore listing, we conduct interrupted time-series (ITS) regressions. Table 5 reports the estimated parameters for the market share dynamics, and Figure 2 provides a visual illustration of these patterns.
Long-run effects on market share
| Variables | Volume Share | Value Share | ||
|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |
| Trend | 0.0257*** | 0.0234*** | 0.0227*** | 0.0204*** |
| (7.246) | (6.320) | (10.916) | (9.432) | |
| Post | −0.0398*** | −0.0261*** | −0.0214*** | −0.0077 |
| (−4.395) | (−2.857) | (−3.430) | (−1.322) | |
| Post × Trend | −0.0259*** | −0.0547*** | −0.0141*** | −0.0427*** |
| (−3.701) | (−4.551) | (−3.078) | (−5.183) | |
| Return | −0.0109 | 0.0096 | ||
| (−0.188) | (0.256) | |||
| VKOSPI | 0.0013*** | 0.0013*** | ||
| (2.758) | (3.790) | |||
| Constant | 0.3002*** | 0.2785*** | 0.3406*** | 0.3191*** |
| (32.789) | (23.053) | (65.689) | (42.196) | |
| Observations | 585 | 585 | 585 | 585 |
| Adjusted R2 | 0.289 | 0.315 | 0.498 | 0.534 |
| Variables | Volume Share | Value Share | ||
|---|---|---|---|---|
| Model 1 | Model 2 | Model 3 | Model 4 | |
| Trend | 0.0257*** | 0.0234*** | 0.0227*** | 0.0204*** |
| (7.246) | (6.320) | (10.916) | (9.432) | |
| Post | −0.0398*** | −0.0261*** | −0.0214*** | −0.0077 |
| (−4.395) | (−2.857) | (−3.430) | (−1.322) | |
| Post × Trend | −0.0259*** | −0.0547*** | −0.0141*** | −0.0427*** |
| (−3.701) | (−4.551) | (−3.078) | (−5.183) | |
| Return | −0.0109 | 0.0096 | ||
| (−0.188) | (0.256) | |||
| VKOSPI | 0.0013*** | 0.0013*** | ||
| (2.758) | (3.790) | |||
| Constant | 0.3002*** | 0.2785*** | 0.3406*** | 0.3191*** |
| (32.789) | (23.053) | (65.689) | (42.196) | |
| Observations | 585 | 585 | 585 | 585 |
| Adjusted R2 | 0.289 | 0.315 | 0.498 | 0.534 |
Note(s): This table reports interrupted time-series (ITS) regression results for the long-run effects of offshore Korean equity index derivatives on KOSPI 200 market share. Models 1 and 2 use Volume Share as the dependent variable, while Models 3 and 4 use Value Share. Volume Share is defined as KOSPI 200 trading volume divided by the combined trading volume of KOSPI 200, KOSDAQ 150, and KRX 300. Value Share is defined analogously using trading value. Post is an indicator equal to one for trading days on or after the offshore listing event and zero otherwise. Trend and Post × Trend are scaled by 100 trading days. Return and VKOSPI are included as control variables in Models 2 and 4. Newey-West standard errors with five lags are used. t-statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
The image contains two line graphs side by side, each representing different aspects of market share trends over time. The left graph, labeled Panel A, shows Volume Share, while the right graph, labeled Panel B, shows Value Share. Both graphs have the x-axis labeled as Event Time, ranging from -300 to 200, and the y-axis labeled as Volume Share and Value Share respectively, ranging from 0.10 to 0.45. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date. In both graphs, the blue lines show an upward trend before the event, while the red lines show a decline after the event. The graphs illustrate the impact of the offshore listing event on the market share dynamics of the KOSPI 200 index.Fitted trends in market share. This figure plots the long-run fitted patterns for KOSPI 200 market share around the offshore listing event. Panel A presents Volume Share, and Panel B presents Value Share. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date
The image contains two line graphs side by side, each representing different aspects of market share trends over time. The left graph, labeled Panel A, shows Volume Share, while the right graph, labeled Panel B, shows Value Share. Both graphs have the x-axis labeled as Event Time, ranging from -300 to 200, and the y-axis labeled as Volume Share and Value Share respectively, ranging from 0.10 to 0.45. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date. In both graphs, the blue lines show an upward trend before the event, while the red lines show a decline after the event. The graphs illustrate the impact of the offshore listing event on the market share dynamics of the KOSPI 200 index.Fitted trends in market share. This figure plots the long-run fitted patterns for KOSPI 200 market share around the offshore listing event. Panel A presents Volume Share, and Panel B presents Value Share. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date
Models 1 and 2 in Table 5 report the long-run results for Volume Share. In Model 2, which controls for contemporaneous market returns (Return) and systemic risk factors (VKOSPI), the coefficient on the post-listing trend interaction term (Post × Trend) is negative and highly significant at the 1% level (−0.0547, t = −4.551). Because the Trend variable is scaled by 100 trading days, this coefficient indicates that KOSPI 200's Volume Share declines by approximately 5.47 percentage points over a 100-trading-day horizon. This post-listing decline in relative volume share is illustrated in Panel A of Figure 2, where the fitted post-event line moves downward relative to the counterfactual continuation of the pre-event trend.[5]
Similarly, Models 3 and 4 report the empirical results for Value Share. In Model 4, the Post × Trend coefficient remains negative and statistically significant at the 1% level (−0.0427, t = −5.183). The coefficient indicates a decline of approximately 4.27 percentage points in KOSPI 200 value share over a 100-trading-day horizon. This decline is visually depicted in Panel B of Figure 2. The immediate level-shift coefficient, Post, is statistically insignificant (−0.0077, t = −1.322), suggesting that the decline in Value Share is driven mainly by a change in the post-listing trend rather than by a one-time level shift.
While a simple interpretation might view the decline in KOSPI 200's relative market share as a weakening of the primary benchmark, the results should be interpreted together with the absolute trading activity and liquidity results documented in Section 4.1. The KOSPI 200 market does not exhibit a deterioration in absolute trading activity or liquidity over the longer horizon. Instead, the evidence suggests that the offshore listing is associated with a broader redistribution of trading activity across domestic benchmark products.
To examine the redistribution interpretation more directly, we also estimate the post-listing trading activity patterns of the other domestic benchmark products included in the market-share denominator. The results are reported in Appendix C. The Post × Trend coefficients are positive and statistically significant for both KOSDAQ 150 and KRX 300 trading activity, indicating that trading activity in related domestic benchmark products increased over the post-listing period. Therefore, the decline in KOSPI 200's relative market share is consistent with a redistribution of trading activity across domestic benchmark products rather than a deterioration of the KOSPI 200 market itself.
4.3 Investor participation effects
This section investigates the micro-structural shifts in investor clienteles and order flow dynamics to illuminate the behavioral transmission channels underlying the macro-structural changes documented in Section 4.2. As cross-border financial derivative links are established, the relative participation and directional order flows of foreign, institutional, and retail investors may undergo a profound realignment due to their heterogeneous hedging mandates, varying investment horizons, and asymmetric access to offshore liquidity venues. Our approach allows us to rigorously adjudicate between the “order flow migration” and “market integration” hypotheses by capturing both the immediate liquidity shocks and the subsequent long-run structural asset reallocations across different market participants.
4.3.1 Short-run effects
Table 6 reports the short-run mean differences in investor participation surrounding the offshore listing event across two symmetric event windows to capture immediate structural shocks. In the narrow (−20, +20) event window (Panel A), the Foreign Value Share decreases marginally from 0.363 to 0.358, but the difference (−0.005) is statistically insignificant (t = 0.56). This initial finding provides critical preliminary evidence against the “order flow migration” hypothesis outlined in Section 2.3; the opening of the independent offshore venue does not indicate an immediate decline in foreign trading participation away from the domestic platform.
Short-run effects on investor participation
| Variables | Pre | Post | Diff (pre–post) | t-stat |
|---|---|---|---|---|
| Panel A. (−20, +20) Event window | ||||
| Foreign Value Share | 0.363 | 0.358 | 0.005 | 0.56 |
| Institutional Value Share | 0.228 | 0.244 | −0.016 | −2.13** |
| Foreign Net Buying | −0.003 | 0.010 | −0.013 | −2.49** |
| Panel B. (−30, +30) Event window | ||||
| Foreign Value Share | 0.376 | 0.375 | 0.001 | 0.10 |
| Institutional Value Share | 0.228 | 0.242 | −0.014 | −1.77* |
| Foreign Net Buying | 0.004 | 0.004 | 0.000 | −0.06 |
| Variables | Pre | Post | Diff (pre–post) | t-stat |
|---|---|---|---|---|
| Panel A. (−20, +20) Event window | ||||
| Foreign Value Share | 0.363 | 0.358 | 0.005 | 0.56 |
| Institutional Value Share | 0.228 | 0.244 | −0.016 | −2.13** |
| Foreign Net Buying | −0.003 | 0.010 | −0.013 | −2.49** |
| Panel B. (−30, +30) Event window | ||||
| Foreign Value Share | 0.376 | 0.375 | 0.001 | 0.10 |
| Institutional Value Share | 0.228 | 0.242 | −0.014 | −1.77* |
| Foreign Net Buying | 0.004 | 0.004 | 0.000 | −0.06 |
Note(s): This table reports short-run mean differences in investor participation around the offshore listing event. Panel A uses the (−20, +20) trading-day event window, and Panel B uses the (−30, +30) trading-day event window. Diff is calculated as the pre-event mean minus the post-event mean. Foreign Value Share is defined as foreign investors' trading value divided by total trading value across investor types. Institutional Value Share is defined analogously for institutional investors. Foreign Net Buying is defined as foreign investors' net buying value divided by total trading value across investor types, where net buying value is buy-side trading value minus sell-side trading value. t-statistics are reported for tests of mean differences. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
In contrast, domestic institutional participation exhibits an immediate, positive level shift. The Institutional Value Share increases from 0.228 to 0.244, with the mean difference (−0.016) being statistically significant at the 5% level (t = −2.13). Symmetrically, the Foreign Net Buying experiences a reliable expansion, shifting from −0.003 to 0.010. This transition from a net-selling stance to a net-buying trajectory yields a mean difference (−0.013) that is statistically significant at the 5% level (t = −2.49).
When evaluating the wider (−30, +30) event window in Panel B, these immediate level-shift dynamics display an expected, short-lived dissipation. The Foreign Value Share remains virtually unaltered (0.376–0.375, t = 0.10). The Institutional Value Share retains its upward trajectory, increasing from 0.228 to 0.242, and remains statistically significant at the 10% level (t = −1.77). However, the Foreign Net Buying exhibits no meaningful variation within this expanded horizon (0.004–0.004, t-stat = −0.06).
Taken together, these short-run results provide evidence more consistent with the market integration channel than with an immediate migration of foreign order flow away from the domestic market. The immediate aftermath of the offshore financial innovation is characterized not by international capital flight, but by an elevated participation of domestic institutional arbitrageurs and a transient, directional capital injection by foreign investors. This immediate footprint suggests that domestic institutions and global asset managers rapidly utilized the newly established cross-border trading linkages to calibrate their localized risk exposures and enforce cross-market arbitrage bounds, laying the empirical groundwork for the long-run structural reallocations analyzed in the subsequent section.
4.3.2 Long-run effects
Table 7 and Figure 3 present the long-run interrupted time-series results for investor participation around the offshore listing event. The specification allows us to distinguish between immediate level shifts after the listing and changes in post-listing trends.
Long-run effects on investor participation
| Panel A. Institutional investor participation | ||
|---|---|---|
| Variables | Institutional Value Share | |
| Model 1 | Model 2 | |
| Trend | −0.0045 | −0.0072*** |
| (−1.646) | (−2.630) | |
| Post | 0.0088 | 0.0244*** |
| (1.061) | (2.894) | |
| Post × Trend | 0.0190*** | −0.0137 |
| (3.111) | (−1.484) | |
| Return | 0.1025 | |
| (1.389) | ||
| VKOSPI | 0.0014*** | |
| (3.746) | ||
| Constant | 0.2393*** | 0.2152*** |
| (47.739) | (26.409) | |
| Observations | 585 | 585 |
| Adjusted R2 | 0.008 | 0.122 |
| Panel A. Institutional investor participation | ||
|---|---|---|
| Variables | Institutional Value Share | |
| Model 1 | Model 2 | |
| Trend | −0.0045 | −0.0072*** |
| (−1.646) | (−2.630) | |
| Post | 0.0088 | 0.0244*** |
| (1.061) | (2.894) | |
| Post × Trend | 0.0190*** | −0.0137 |
| (3.111) | (−1.484) | |
| Return | 0.1025 | |
| (1.389) | ||
| VKOSPI | 0.0014*** | |
| (3.746) | ||
| Constant | 0.2393*** | 0.2152*** |
| (47.739) | (26.409) | |
| Observations | 585 | 585 |
| Adjusted R2 | 0.008 | 0.122 |
| Panel B. Foreign investor participation | |||
|---|---|---|---|
| Variables | Foreign Value Share | Foreign Net Buying | |
| Model 3 | Model 4 | Model 5 | |
| Trend | 0.0220*** | 0.0198*** | −0.0028* |
| (5.401) | (4.837) | (−1.714) | |
| Post | 0.0292** | 0.0420*** | 0.0171*** |
| (2.088) | (2.689) | (3.070) | |
| Post × Trend | −0.0448*** | −0.0716*** | −0.0140*** |
| (−5.436) | (−4.306) | (−3.410) | |
| Return | −0.0426 | 0.7611*** | |
| (−0.464) | (6.146) | ||
| VKOSPI | 0.0012** | ||
| (1.967) | |||
| VIX | −0.0015*** | ||
| (−5.115) | |||
| Constant | 0.2771*** | 0.2567*** | 0.0291*** |
| (36.230) | (19.494) | (5.863) | |
| Observations | 585 | 585 | 585 |
| Adjusted R2 | 0.260 | 0.270 | 0.402 |
| Panel B. Foreign investor participation | |||
|---|---|---|---|
| Variables | Foreign Value Share | Foreign Net Buying | |
| Model 3 | Model 4 | Model 5 | |
| Trend | 0.0220*** | 0.0198*** | −0.0028* |
| (5.401) | (4.837) | (−1.714) | |
| Post | 0.0292** | 0.0420*** | 0.0171*** |
| (2.088) | (2.689) | (3.070) | |
| Post × Trend | −0.0448*** | −0.0716*** | −0.0140*** |
| (−5.436) | (−4.306) | (−3.410) | |
| Return | −0.0426 | 0.7611*** | |
| (−0.464) | (6.146) | ||
| VKOSPI | 0.0012** | ||
| (1.967) | |||
| VIX | −0.0015*** | ||
| (−5.115) | |||
| Constant | 0.2771*** | 0.2567*** | 0.0291*** |
| (36.230) | (19.494) | (5.863) | |
| Observations | 585 | 585 | 585 |
| Adjusted R2 | 0.260 | 0.270 | 0.402 |
Note(s): This table reports the estimated coefficients from the long-run interrupted time-series (ITS) regression results for institutional and foreign investor participation around the offshore listing of MSCI Korea Index Futures. Panel A estimates the Institutional Value Share (Models 1–2), and Panel B evaluates the Foreign Value Share (Models 3–4) and Foreign Net Buying (Model 5). Post is an indicator equal to one for trading days on or after the offshore listing event and zero otherwise. Trend and Post × Trend are scaled by 100 trading days. In Model 5, the global CBOE VIX index uniquely substitutes VKOSPI to capture international risk-appetite bounds and global systematic fear metrics that directly condition directional cross-border capital flows, while simultaneously evading severe multi-collinearity within the localized uncertainty specification. Newey-West standard errors with five lags are used. t-statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
The image contains two scatter plots with fitted trend lines, each representing different aspects of foreign investor participation over time. Panel A on the left shows Foreign Value Share, with the y-axis labeled 'Foreign Value Share' and the x-axis labeled 'Event Time'. Gray dots represent daily observations. The solid blue line indicates the pre-event fitted trend, the dashed blue line shows the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. Panel B on the right displays Foreign Net Buying, with the y-axis labeled 'Foreign Net Buying' and the x-axis labeled 'Event Time'. Similar to Panel A, gray dots denote daily observations, and the same color scheme is used for the trend lines. The vertical dashed line in both panels marks the offshore listing event date. The plots illustrate how foreign investor participation changes before and after the event, with distinct trends observed post-event.Fitted trends in foreign investor participation. This figure plots the long-run fitted patterns for foreign investor participation around the offshore listing event. Panel A presents Foreign Value Share, and Panel B presents Foreign Net Buying. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date
The image contains two scatter plots with fitted trend lines, each representing different aspects of foreign investor participation over time. Panel A on the left shows Foreign Value Share, with the y-axis labeled 'Foreign Value Share' and the x-axis labeled 'Event Time'. Gray dots represent daily observations. The solid blue line indicates the pre-event fitted trend, the dashed blue line shows the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. Panel B on the right displays Foreign Net Buying, with the y-axis labeled 'Foreign Net Buying' and the x-axis labeled 'Event Time'. Similar to Panel A, gray dots denote daily observations, and the same color scheme is used for the trend lines. The vertical dashed line in both panels marks the offshore listing event date. The plots illustrate how foreign investor participation changes before and after the event, with distinct trends observed post-event.Fitted trends in foreign investor participation. This figure plots the long-run fitted patterns for foreign investor participation around the offshore listing event. Panel A presents Foreign Value Share, and Panel B presents Foreign Net Buying. Gray dots denote daily observations. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The vertical dashed line indicates the offshore listing event date
Panel A of Table 7 reports the results for Institutional Value Share. In Model 1, the coefficient on Post × Trend is positive and statistically significant (0.0190, t = 3.111), suggesting an increasing post-event trend in the baseline specification. However, once Return and VKOSPI are included in Model 2, the Post coefficient becomes positive and statistically significant (0.0244, t = 2.894), whereas the Post × Trend coefficient becomes negative and statistically insignificant (−0.0137, t = −1.484). This pattern indicates that institutional participation increases around the offshore listing, but the evidence for a persistent post-listing trend is weaker once domestic market conditions are controlled for. Therefore, the institutional participation results should be interpreted as evidence of short-run adjustment rather than a definitive long-run structural increase.
Panel B reports the results for foreign investor participation. In Model 4, Foreign Value Share shows a positive immediate level shift (0.0420, t = 2.690), but the coefficient on Post × Trend is negative and statistically significant (−0.0716, t = −4.310). This result indicates that foreign investors' domestic trading share increases immediately after the listing but declines over the post-listing period. As illustrated in Panel A of Figure 3, the fitted post-event trend for Foreign Value Share moves downward relative to the counterfactual continuation of the pre-event trend. Similarly, Model 5 shows that Foreign Net Buying increases immediately after the offshore listing, as indicated by the positive Post coefficient (0.0171, t = 3.070). However, the negative and statistically significant Post × Trend coefficient (−0.0140, t = −3.410) suggests that this initial increase gradually weakens over time.
Model 5 includes VIX because Foreign Net Buying directly captures directional cross-border capital flows and is therefore likely to be more sensitive to global risk appetite and international market uncertainty. Including VIX helps distinguish changes in foreign net buying associated with the offshore listing from those driven by global volatility conditions. As illustrated in Panel B of Figure 3, foreign investors appear to increase net buying in the domestic KOSPI 200 market around the listing, but this increase gradually weakens over the longer horizon.
Overall, the investor participation results are more consistent with gradual market integration than with an abrupt withdrawal of foreign investors from the domestic market. Institutional participation increases around the offshore listing, but the evidence for a persistent post-listing trend is weaker once domestic market conditions are controlled for. Foreign investors' domestic trading share increases immediately after the listing but declines over the post-listing period, while Foreign Net Buying shows an immediate increase followed by a gradual weakening. These patterns suggest that foreign and institutional investors adjusted their domestic trading behavior following the offshore listing. However, the evidence should be interpreted as post-listing association rather than definitive causal proof.
4.4 Market stability and volatility spillover effects
Finally, we examine whether the offshore listing of MSCI Korea Index Futures is associated with changes in domestic market stability and the transmission of global risk shocks. The market integration perspective suggests that expanded hedging opportunities may help stabilize domestic markets by improving risk-sharing and cross-market price discovery. At the same time, stronger cross-market linkages may increase the sensitivity of the domestic market to global volatility conditions. To examine these possibilities, we analyze the longer-horizon patterns of GARCH-estimated volatility and its sensitivity to global uncertainty, measured by VIX.
Table 8 reports the ITS regression results for GARCH Volatility. In Model 1, the immediate level-shift coefficient, Post, is negative and statistically significant (−0.0065, t = −4.315), suggesting a short-run decline in GARCH Volatility immediately after the listing. However, the post-listing trend shows a different pattern. Across all specifications, the Post × Trend coefficient is positive and statistically significant. For example, the coefficient is 0.0131 in Model 1 and 0.0105 in Model 3. Because the Trend variable is scaled by 100 trading days, these estimates indicate an increase in GARCH Volatility over the post-listing period. In Model 1, the coefficient implies an increase of approximately 0.013 units in GARCH Volatility over a 100-trading-day horizon. As illustrated in Panel A of Figure 4, the fitted post-event trend moves upward relative to the counterfactual continuation of the pre-event trend.
Long-run effects on market stability and volatility spillovers
| Variables | Model 1 | Model 2 | Model 3 |
|---|---|---|---|
| Trend | 0.0005 | −0.0013*** | −0.0007* |
| (1.507) | (−2.850) | (−1.706) | |
| Post | −0.0065*** | −0.0018 | −0.0218*** |
| (−4.315) | (−1.347) | (−2.945) | |
| Post × Trend | 0.0131*** | 0.0133*** | 0.0105*** |
| (7.077) | (8.421) | (6.990) | |
| VIX | 0.0007*** | 0.0005*** | |
| (4.338) | (3.686) | ||
| Post × VIX | 0.0012*** | ||
| (2.600) | |||
| Constant | 0.0128*** | 0.0036* | 0.0068*** |
| (17.458) | (1.672) | (4.427) | |
| Observations | 585 | 585 | 585 |
| Adjusted R2 | 0.540 | 0.647 | 0.694 |
| Variables | Model 1 | Model 2 | Model 3 |
|---|---|---|---|
| Trend | 0.0005 | −0.0013*** | −0.0007* |
| (1.507) | (−2.850) | (−1.706) | |
| Post | −0.0065*** | −0.0018 | −0.0218*** |
| (−4.315) | (−1.347) | (−2.945) | |
| Post × Trend | 0.0131*** | 0.0133*** | 0.0105*** |
| (7.077) | (8.421) | (6.990) | |
| VIX | 0.0007*** | 0.0005*** | |
| (4.338) | (3.686) | ||
| Post × VIX | 0.0012*** | ||
| (2.600) | |||
| Constant | 0.0128*** | 0.0036* | 0.0068*** |
| (17.458) | (1.672) | (4.427) | |
| Observations | 585 | 585 | 585 |
| Adjusted R2 | 0.540 | 0.647 | 0.694 |
Note(s): This table reports interrupted time-series (ITS) regression results for KOSPI 200 market volatility and volatility spillover effects. The dependent variable is GARCH Volatility, measured as the conditional volatility estimated from a GARCH(1,1) model using daily KOSPI 200 returns. Post is an indicator equal to one for trading days on or after the offshore listing event and zero otherwise. Trend and Post × Trend are scaled by 100 trading days. VIX captures global market volatility, and Post × VIX captures the change in VIX-related volatility spillovers after the offshore listing. Newey-West standard errors with five lags are used. t-statistics are reported in parentheses. ***, **, and * denote significance at the 1%, 5%, and 10% levels, respectively
Two line graphs depict GARCH volatility and VIX sensitivity around an offshore listing event. Panel A shows daily KOSPI 200 GARCH Volatility and the fitted interrupted time-series trends. The x-axis represents Event Time in days, ranging from -300 to 200, and the y-axis represents GARCH Volatility, ranging from 0.00 to 0.07. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The post-event trend shows an upward movement relative to the pre-event trend. Panel B shows the marginal sensitivity of KOSPI 200 GARCH Volatility to VIX before and after the offshore listing. The x-axis represents VIX, ranging from 10 to 30, and the y-axis represents the Change in GARCH Volatility, ranging from -0.01 to 0.02. The blue line represents the pre-event sensitivity, and the red line represents the post-event sensitivity.Fitted volatility trends and VIX sensitivity. This figure visualizes the volatility results around the offshore listing event. Panel A plots daily KOSPI 200 GARCH Volatility and the fitted interrupted time-series trends from Model 2 of Table 8. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. Panel B plots the marginal sensitivity of KOSPI 200 GARCH Volatility to VIX before and after the offshore listing, based on Model 3 of Table 8. VIX is centered at its sample mean, so the fitted lines represent changes in predicted GARCH Volatility relative to the average VIX level. The VIX range is restricted to the 5th–95th percentile range for visualization
Two line graphs depict GARCH volatility and VIX sensitivity around an offshore listing event. Panel A shows daily KOSPI 200 GARCH Volatility and the fitted interrupted time-series trends. The x-axis represents Event Time in days, ranging from -300 to 200, and the y-axis represents GARCH Volatility, ranging from 0.00 to 0.07. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. The post-event trend shows an upward movement relative to the pre-event trend. Panel B shows the marginal sensitivity of KOSPI 200 GARCH Volatility to VIX before and after the offshore listing. The x-axis represents VIX, ranging from 10 to 30, and the y-axis represents the Change in GARCH Volatility, ranging from -0.01 to 0.02. The blue line represents the pre-event sensitivity, and the red line represents the post-event sensitivity.Fitted volatility trends and VIX sensitivity. This figure visualizes the volatility results around the offshore listing event. Panel A plots daily KOSPI 200 GARCH Volatility and the fitted interrupted time-series trends from Model 2 of Table 8. The solid blue line represents the pre-event fitted trend, the dashed blue line represents the counterfactual continuation of the pre-event trend, and the red line represents the fitted post-event trend. Panel B plots the marginal sensitivity of KOSPI 200 GARCH Volatility to VIX before and after the offshore listing, based on Model 3 of Table 8. VIX is centered at its sample mean, so the fitted lines represent changes in predicted GARCH Volatility relative to the average VIX level. The VIX range is restricted to the 5th–95th percentile range for visualization
Model 2 adds VIX to control for global volatility conditions. The coefficient on VIX is positive and statistically significant, indicating that KOSPI 200 volatility is positively associated with global market uncertainty. The estimated coefficient implies that a one-point increase in VIX is associated with an increase of approximately 0.0007 in GARCH Volatility. The Post × Trend coefficient remains positive and statistically significant after controlling for VIX, suggesting that the post-listing volatility pattern is not fully explained by contemporaneous global volatility conditions.
Model 3 includes the interaction term Post × VIX to examine whether the sensitivity of domestic volatility to global volatility conditions changes after the offshore listing. The coefficient on Post × VIX is positive and statistically significant (0.0012, t = 2.600), indicating that the sensitivity of KOSPI 200 volatility to VIX increases after the offshore listing. Economically, the estimated VIX sensitivity increases from 0.0005 before the offshore listing to 0.0017 after the listing, based on the sum of the coefficients on VIX and Post × VIX.
This result suggests that the domestic KOSPI 200 market becomes more sensitive to global volatility conditions after the Eurex listing. Panel B of Figure 4 illustrates this pattern, as the post-event sensitivity line has a steeper slope than the pre-event baseline. Overall, the evidence is consistent with stronger integration between the domestic market and global risk factors, while also suggesting that offshore listing may be associated with stronger transmission of global volatility shocks to the domestic market.
Because GARCH Volatility is estimated in a first stage and then used as the dependent variable in the ITS regressions, the second-stage standard errors may not fully account for first-stage estimation uncertainty. To address this concern, we conduct an additional robustness check using Absolute Return as a model-free proxy for daily market volatility. Absolute Return is defined as the absolute value of the daily KOSPI 200 return and captures the magnitude of daily price movements regardless of direction. Unlike GARCH Volatility, Absolute Return does not require a first-stage volatility model.
The results are reported in Appendix D. The Post × Trend coefficients remain positive and statistically significant across all specifications, indicating that the post-listing increase in the volatility proxy is also observed when using a model-free measure of volatility. In Model 3, the coefficient on Post × VIX is also positive and statistically significant, suggesting that the increase in the sensitivity of domestic volatility to global volatility conditions is not driven solely by the use of estimated GARCH Volatility as the dependent variable. These results are qualitatively consistent with the baseline volatility analysis.
4.5 Robustness checks
We conduct two sets of robustness checks to assess the stability of our main findings. The first robustness check addresses a sample-period concern related to the 2026 semiconductor-driven market expansion. The second robustness check addresses a separate econometric concern regarding the choice of Newey-West lag length.
First, a potential concern in interpreting the longer-horizon interrupted time-series results is that the post-listing period overlaps with other market developments. In particular, from early 2026, the Korean equity market experienced a sector-specific expansion associated with the semiconductor industry. This development is unlikely to affect the short-run event-window tests centered on the July 14, 2025 offshore listing event, but it may influence the longer-horizon post-event ITS estimates. To address this concern, we re-estimate the main ITS specifications using a truncated sample ending on December 31, 2025. This test helps reduce the potential influence of market developments that occurred after the initial post-listing period.
Second, independently of the sample-period concern, we examine whether the statistical significance of the main results depends on the choice of Newey-West lag length. The baseline regressions use Newey-West standard errors with five lags. As a robustness check, we re-estimate the main specifications using alternative lag lengths of three, seven, and ten. This test examines whether our statistical inference is sensitive to the lag-length choice used to correct for heteroskedasticity and serial correlation in daily time-series data.
The detailed robustness results are reported in Appendix A and B. Appendix A presents the truncated-sample results ending on December 31, 2025, while Appendix B reports the results using alternative Newey-West lag lengths. The main coefficients of interest are Post and Post × Trend in the ITS specifications and Post × VIX in the volatility-spillover specification. Overall, the robustness checks show that the main findings remain qualitatively similar. The post-listing trend in ln(Trading Value) remains positive, and the post-listing trend in Liquidity remains positive in the controlled specification. The market-share variables continue to show negative post-listing trends, consistent with a redistribution of trading activity across domestic benchmark products. The investor participation results remain broadly consistent with the baseline interpretation: institutional participation shows evidence of an immediate level increase, while foreign investor participation exhibits a declining post-listing trend. Finally, the volatility results continue to indicate stronger sensitivity of KOSPI 200 volatility to global volatility conditions after the offshore listing.
These robustness checks should be interpreted as supporting evidence rather than as definitive proof of causality. They indicate that the main results are not driven solely by the inclusion of the 2026 semiconductor-driven period or by the specific Newey-West lag length used in the baseline analysis.
5. Policy implications
The empirical findings provide several policy implications for the Korea Exchange and financial regulators. A key policy message is that the offshore listing of Korean equity index derivatives should be understood as part of a broader market-access and market-integration strategy rather than simply as a competitive threat to the domestic derivatives market. Before allowing offshore listings of Korea-related index derivatives, the Korea Exchange had traditionally expressed concerns about potential liquidity outflows from the domestic derivatives market. More recently, however, the Korea Exchange has shifted toward a phased opening strategy, motivated by the necessity to improve global investors' access to Korean capital markets and enhance the international competitiveness of Korea-related index products. The results of this study provide empirical evidence that is broadly consistent with such a balanced policy perspective.
First, the findings suggest that offshore listing may generate short-run adjustment frictions but does not necessarily weaken domestic market quality over the longer horizon. Although domestic trading activity declines in the immediate event window, the long-run results show that KOSPI 200 trading activity and liquidity improve during the post-listing period. This pattern implies that offshore listings may be initially associated with domestic order flow, but they can also broaden global investor attention, strengthen cross-market linkages, and contribute to domestic liquidity over time. For the Korea Exchange, this suggests that the appropriate policy response should not be limited to preventing offshore competition or protecting existing domestic contracts. Rather, policy efforts should focus on strengthening domestic market quality, enhancing trading infrastructure, and improving connectivity between domestic and offshore trading environments.
Second, the decline in KOSPI 200's relative market share indicates that offshore listing can reshape the allocation of trading activity across domestic benchmark products. This does not necessarily imply deterioration in the KOSPI 200 market, because its absolute trading activity and liquidity improve over time. Instead, the evidence suggests that offshore listing may contribute to a broader redistribution of trading activity across related Korean equity index derivatives, including KOSDAQ 150 and KRX 300. From a market-development perspective, the Korea Exchange may therefore need to manage the domestic equity index derivatives market as an integrated product ecosystem rather than as a set of isolated contracts. Policies that improve liquidity provision, product complementarity, and hedging efficiency across multiple benchmark indices can help the domestic market absorb global demand generated by offshore listings while preserving the competitiveness of domestic trading venues.
Third, the investor participation results highlight the importance of monitoring how offshore listings are associated with different investor groups. Institutional investor participation increases around the event, while foreign investors' domestic trading share declines over the longer horizon. This pattern suggests that offshore markets may partly substitute for direct domestic trading by foreign investors, while domestic institutional investors may become more active in response to new hedging and arbitrage opportunities. These findings support the Korea Exchange's cautious approach of pursuing offshore listings in stages while closely monitoring market conditions. In particular, regulators and exchange operators should strengthen data-sharing arrangements with offshore exchanges and index providers, monitor investor composition and cross-market arbitrage activity, and assess whether offshore trading is complementing or substituting for domestic market activity.
Fourth, the volatility results suggest that offshore listing may increase the domestic market's sensitivity to global volatility shocks. The positive Post × VIX coefficient indicates that KOSPI 200 volatility becomes more responsive to global volatility conditions after the offshore listing. This does not imply that offshore listing simply destabilizes the domestic market. Rather, it suggests that greater global accessibility comes with a policy trade-off. Offshore listings may enhance the international visibility and liquidity of Korean equity markets, but they may also strengthen the transmission of global risk shocks to the domestic market. Therefore, financial regulators should develop surveillance and risk-monitoring systems that incorporate offshore trading activity, overnight price discovery, foreign futures positions, and global volatility indicators.
Finally, the results suggest that offshore listing policy should be coordinated with broader capital-market accessibility reforms. The policy objective should not be to prevent offshore market development, but to ensure that domestic market infrastructure, investor protection, liquidity provision, and volatility monitoring evolve together with the offshore trading environment. A phased opening strategy, combined with close communication with domestic market participants and continuous monitoring of liquidity outflows, can help balance the benefits of global access against the risks of market fragmentation. For the Korea Exchange and financial regulators, the central policy challenge is therefore to use offshore listings as an opportunity to enhance the global competitiveness of Korean capital markets while preserving domestic market quality and resilience.
6. Conclusion
This study examines the domestic market consequences of the offshore listing of Korean equity index derivatives by focusing on the launch of MSCI Korea Index Futures on Eurex Exchange. Using the July 14, 2025 listing as the key offshore listing event, we analyze how the introduction of a regulated offshore trading venue for Korean equity exposure is associated with changes in the domestic KOSPI 200 market. Specifically, we examine four dimensions of market adjustment: trading activity and liquidity, market-share redistribution, investor participation, and market stability through volatility spillovers.
The empirical evidence suggests that the offshore listing did not simply weaken the domestic market. In the short run, KOSPI 200 trading activity declines around the event, indicating that the introduction of the offshore venue may have generated temporary adjustment frictions or partial order-flow diversion. However, this short-run decline does not coincide with a significant deterioration in domestic liquidity. More importantly, the long-run interrupted time-series results show that both trading activity and liquidity improve during the post-listing period. These findings suggest that the offshore listing may have strengthened cross-market linkages and contributed to deeper domestic market conditions over time.
At the same time, the results show that offshore listing changes the structure of the domestic derivatives market. The KOSPI 200's relative market share declines after the offshore listing, even though its absolute trading activity and liquidity improve. This pattern indicates that offshore listing is better understood as a broader redistribution of trading activity across related Korean equity index derivatives rather than as a simple deterioration of the KOSPI 200 market. The investor participation results further support this interpretation. Institutional participation increases around the event, while foreign investors' domestic trading share declines over the longer horizon. Foreign Net Buying increases immediately after the listing but gradually weakens over time, suggesting a gradual adjustment in foreign investors' domestic trading behavior rather than an abrupt withdrawal from the domestic market.
The volatility results point to an additional trade-off associated with offshore market integration. The sensitivity of KOSPI 200 GARCH Volatility to VIX increases after the offshore listing, indicating that the domestic market becomes more responsive to global volatility conditions. This finding suggests that offshore derivatives may enhance the international integration and visibility of Korean equity markets, but they may also strengthen the transmission of global risk shocks to the domestic market. Thus, the net effect of offshore listing is not one-dimensional. It appears to improve some aspects of domestic market quality while simultaneously reshaping market structure and increasing exposure to global volatility conditions.
These findings have important implications for the Korea Exchange and financial regulators. Offshore listings should not be viewed solely as a threat to domestic derivatives markets. Rather, they should be managed as part of a broader market-access and market-integration process. A phased opening strategy, combined with close monitoring of liquidity outflows, investor composition, cross-market arbitrage, and volatility transmission, can help balance the benefits of global accessibility against the risks of market fragmentation and stronger external shock transmission. Strengthening domestic market infrastructure, improving product complementarity across benchmark indices, and enhancing surveillance of offshore-domestic linkages will be important as Korean equity index derivatives become more integrated with global markets.
This study also has limitations that suggest directions for future research. Because the offshore listing examined in this paper occurred relatively recently, the post-listing observation period is still limited. As more time passes, future studies will be able to evaluate whether the patterns documented in this study persist, strengthen, or reverse over a longer horizon. In addition, Korea's policy toward offshore listings of Korea-related derivatives is being implemented gradually. Therefore, future research should examine how subsequent stages of market opening, additional offshore listings, extended trading arrangements, or new cross-market agreements are associated with domestic liquidity, investor participation, and volatility transmission. Such evidence will be valuable for assessing the extent to which offshore listings complement domestic market development, substitute for domestic trading activity, or reshape the allocation of trading activity across domestic and offshore venues over the longer term.
The work in this paper was supported by the Korea Exchange Academic Research Support Program in 2025. The author used ChatGPT solely for editorial assistance purposes—such as grammar correction and clarity improvements. All research design, methodology, analysis, and intellectual contributions are solely those of the author. The author reviewed and edited the final content and takes full responsibility for this publication.
Notes
VIX is the Chicago Board Options Exchange (CBOE) Volatility Index, which measures the 30-day implied volatility of the S&P 500 index options. It is included as an exogenous control variable to capture time-varying global market uncertainty and macro-financial shocks that could be simultaneously associated with the domestic equity derivative market.
Our findings remain qualitatively unchanged when using alternative Newey-West lag lengths, including three, seven, and ten lags.
Since ln(Trading Value) is measured in natural logarithms and Post × Trend is scaled by 100 trading days, the percentage effect on the underlying trading value is calculated as , or approximately 52%.
As a robustness check, we re-estimate the liquidity specification using the conventional raw Amihud illiquidity measure, defined as (|Returnt|/Trading Valuet) × 1,000,000, and an alternative value of ε in the modified liquidity measure. When the raw Amihud illiquidity measure is used as the dependent variable, the controlled specification yields a negative and statistically significant Post × Trend coefficient (−0.0012, t = −3.794), indicating a decline in illiquidity over the post-listing period. When the modified liquidity measure is recalculated using ε = 0.004, the controlled specification yields a positive and statistically significant Post × Trend coefficient (0.9171, t = 3.551). These results are qualitatively consistent with the baseline liquidity results.
Since Volume Share is measured in levels and Post × Trend is scaled by 100 trading days, the absolute cumulative percentage-point change over a 100-trading-day horizon is calculated via a linear projection as (or a 5.47 percentage-point reduction).
The supplementary material for this article can be found online.

