Price Discovery and Foreign Participation in the Republic of Korea's Government Bond Cash and Futures Markets
Abstract
EconStor is a publication server for scholarly economic literature, provided as a non-commercial public service by the ZBW.
Full text
Park, Cyn-Young; Mercado, Rogelio; Choi, Jaehun; Lim, Hosung Working Paper Price Discovery and Foreign Participation in the Republic of Korea's Government Bond Cash and Futures Markets ADB Economics Working Paper Series, No. 427 Provided in Cooperation with: Asian Development Bank (ADB), Manila Suggested Citation: Park, Cyn-Young; Mercado, Rogelio; Choi, Jaehun; Lim, Hosung (2015) : Price Discovery and Foreign Participation in the Republic of Korea's Government Bond Cash and Futures Markets, ADB Economics Working Paper Series, No. 427, Asian Development Bank (ADB), Manila, https://hdl.handle.net/11540/3301 This Version is available at: https://hdl.handle.net/10419/109509 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by-nc-nd/3.0/igo/
ASIAN DEVELOPMENT BANK AsiAn Development BAnk 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets The authors assess the impact of foreign participation in Korean Treasury Bond (KTB) cash and futures markets and their role in the price discovery process. Using daily data from the over-the-counter market for cash and the Korea Exchange for futures transactions, the results show that foreign trading in the KTB futures market leads the price discovery process for the underlying bonds. Specifically, foreigners’ daily net long positions in the futures market exert significant influence in both KTB cash and futures prices. The empirical findings also indicate that it is the unexpected component of foreign investors’ net long futures positions that explains a significant share of the pricing effects. About the Asian Development Bank ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries reduce poverty and improve the quality of life of their people. Despite the region’s many successes, it remains home to approximately two-thirds of the world’s poor: 1.6 billion people who live on less than $2 a day, with 733 million struggling on less than $1.25 a day. ADB is committed to reducing poverty through inclusive economic growth, environmentally sustainable growth, and regional integration. Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. PRiCe DisCoveRy AnD FoReiGn PARtiCiPAtion in the RePuBliC oF KoReA’s GoveRnMent BonD CAsh AnD FutuRes MARKets Cyn-Young Park; Rogelio Mercado, Jr.; Jaehun Choi; and Hosung Lim adb economics working paper series no. 427 march 2015
ADB Economics Working Paper Series Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets Cyn-Young Park; Rogelio Mercado, Jr.; Jaehun Choi; and Hosung Lim No. 427 | 2015 Cyn-Young Park ([email protected]) is Assistant Chief Economist of the Asian Development Bank; Rogelio Mercado, Jr. ([email protected]) is a Graduate Student/Research Associate of Trinity College Dublin; Jaehun Choi ([email protected]) is Senior Economist, of The Bank of Korea; and Hosung Lim ([email protected]) is Economist of The Bank of Korea. The authors are deeply indebted to Sungjin Park, Choong Won Park, and Eun Yeong Song for valuable research inputs and discussions during the course of the study. The authors also wish to thank the seminar participants at the Bank of Korea for the midterm review of this study in November 2014. ASIAN DEVELOPMENT BANK
Creative Commons Attribution-NonCommercial-NoDerivs 3.0 IGO license (CC BY-NC-ND 3.0 IGO) © 2015 Asian Development Bank and the Bank of Korea Asian Development Bank Bank of Korea 6 ADB Avenue, Mandaluyong City 39, Namdaemun-ro, Jung-gu, Seoul 1550 Metro Manila, Philippines 100-794, Republic of Korea Tel +63 2 632 4444; Fax +63 2 636 2444 Tel +82 2 759 4114; Fax +82 2 759 4060 www.adb.org; openaccess.adb.org http://eng.bok.or.kr Some rights reserved. Published in 2015. Printed in the Philippines. ISSN 2313-6537 (Print), 2313-6545 (e-ISSN) Publication Stock No. WPS157139-2 The views expressed in this publication are those of the authors and do not necessarily reflect the views and policies of the Bank of Korea and the Asian Development Bank (ADB) or its Board of Governors or the governments they represent. When reporting or citing this paper, the authors' names should always be stated explicitly. ADB and the Bank of Korea do not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by ADB and the Bank of Korea in preference to others of a similar nature that are not mentioned. By making any designation of or reference to a particular territory or geographic area, or by using the term “country” in this document, ADB and the Bank of Korea do not intend to make any judgments as to the legal or other status of any territory or area. This work is available under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 IGO license (CC BY-NC-ND 3.0 IGO) http://creativecommons.org/licenses/by-nc-nd/3.0/igo/. By using the content of this publication, you agree to be bound by the terms of said license as well as the Terms of Use of the ADB Open Access Repository at openaccess.adb.org/termsofuse This CC license does not apply to non-ADB copyright materials in this publication. If the material is attributed to another source, please contact the copyright owner or publisher of that source for permission to reproduce it. ADB cannot be held liable for any claims arising as a result of your use of the material. Attribution—In acknowledging ADB and the Bank of Korea as the source, please be sure to include all of the following information: Author. Year of publication. Title of the material. © Asian Development Bank and the Bank of Korea. https://openaccess.adb.org. Available under a CC BY-NC-ND 3.0 IGO license. Please contact [email protected] or publicatio[email protected] if you have questions or comments with respect to content, or if you wish to obtain copyright permission for your intended use that does not fall within these terms, or for permission to use the ADB logo. Note: In this publication, “$” refers to US dollars. The ADB Economics Working Paper Series is a forum for stimulating discussion and eliciting feedback on ongoing and recently completed research and policy studies undertaken by the Asian Development Bank (ADB) staff, consultants, or resource persons. The series deals with key economic and development problems, particularly those facing the Asia and Pacific region; as well as conceptual, analytical, or methodological issues relating to project/program economic analysis, and statistical data and measurement. The series aims to enhance the knowledge on Asia’s development and policy challenges; strengthen analytical rigor and quality of ADB’s country partnership strategies, and its subregional and country operations; and improve the quality and availability of statistical data and development indicators for monitoring development effectiveness. The ADB Economics Working Paper Series is a quick-disseminating, informal publication whose titles could subsequently be revised for publication as articles in professional journals or chapters in books. The series is maintained by the Economic Research and Regional Cooperation Department.
CONTENTS TABLES AND FIGURES iv ABSTRACT v I. INTRODUCTION 1 II. LITERATURE REVIEW 3 III. DATASET, DESCRIPTIVE STATISTICS, AND DECOMPOSITION OF NET PURCHASES AND POSITIONS 5 A. Dataset 5 B. Descriptive Statistics 7 C. Decomposition of Net Purchases and Positions 12 IV. EMPIRICAL SPECIFICATION AND RESULTS 20 A. Empirical Specification 20 B. Analysis of Results 21 V. SUMMARY AND POLICY IMPLICATIONS 37 REFERENCES 39
TABLES AND FIGURES TABLES 1 Descriptive Statistics for Korean Treasury Bond and Futures Returns 7 2 Trading Volume 8 3 Average Net Buy and Net Long Position in Cash and Futures Market 10 4a Determinants of Net Purchases and Net Long Positions (Precrisis Period) 14 4b Determinants of Net Purchases and Net Long Positions (Crisis Period) 15 4c Determinants of Net Purchases and Net Long Positions (Postcrisis Period) 16 5 Expected Component of Net Purchases by Investor Groups 18 6 Full Sample (Own-Market Estimates) 22 7 Full Sample (Market Interaction Estimates) 24 8 Full Sample (Expected and Unexpected Component Estimates) 25 9 Precrisis Estimates 26 10 Precrisis Estimates (Expected and Unexpected) 27 11 Crisis Estimates 28 12 Crisis Estimates (Expected and Unexpected) 29 13 Postcrisis Estimates 30 14 Postcrisis Estimates (Expected and Unexpected) 31 15 Full Sample by Composition of Domestic Finance Institutions 33 16 Full Sample by Composition of Domestic Finance Institutions (Expected and Unexpected) 34 FIGURES 1a Net Purchases of Foreign and Domestic Finance Institutions in the 3-Year Treasury Cash Market 11 1b Net Positions of Foreign and Domestic Finance Institutions in the 3-Year Treasury Futures Market 12 2a Unexpected Component of Net Purchases by Foreign and Domestic Finance Institutions in the 3-Year Treasury Cash Market 19 2b Unexpected Component of Net Positions by Foreign and Domestic Finance Institutions in the 3-Year Treasury Futures Market 19
ABSTRACT This paper examines the impact of foreign participation in Korean Treasury Bond (KTB) futures and its role in price discovery for KTBs, using daily transactions data from the over-the-counter market for KTBs and from the Korea Exchange for the futures. Our analysis suggests that foreign trading in the KTB futures market leads the price discovery process for the underlying bonds. Empirical results show that foreigners’ daily net long positions in the futures market exert significant influence in KTB and KTB futures prices. We also find that it is the unexpected component of foreign investors’ net long futures positions that explains a significant share of the pricing effects, suggesting that how foreign trading responds to news carries additional information content. Keywords: price discovery, emerging market bonds, foreign participation JEL Classification: G10, G13, G14
I. INTRODUCTION The fast growth of local currency bond markets, combined with the wave of financial globalization, boosted foreign participation in many emerging Asian markets. Local currency bonds outstanding for nine emerging Asian markets reached $8.0 trillion in September 2014 from about $0.8 trillion in December 2000, up nearly tenfold. Data from Asian Bonds Online show a clear upward trend in foreign participation across emerging Asian local currency government bond markets since the mid-2000s.1 Growth accelerated even more following a dip in late 2008 associated with the global financial crisis, as emerging Asia’s economic resilience, in contrast to the financial turmoil in the United States (US) and the eurozone economies, made their local currency government bonds relatively more attractive to global investors. The Republic of Korea has the second largest local currency bond market in emerging Asia,2 with total bonds outstanding at $1.8 trillion. Starting with the announcement of the Government Bond Market Stimulus Plan in August 1998, a number of policy reforms have been undertaken, including the introduction of the primary dealer system, interdealer market, and government bond futures in 1999. The Korea Exchange introduced the cash-settled, 3-year Korean Treasury Bond (KTB) futures contract on 29 September 1999. Foreign holdings of Korean government bonds are now nearly 15%, up from less than 1% in the mid-2000s. Foreign interest in KTB futures has also been strong, effectively accelerating the growth of the KTB market. Foreigners find it easier to trade futures than cash bonds due to taxation, leverage, and liquidity issues. Foreign investors have played an important role in the Korean bond markets. Their active participation has helped boost market liquidity, depth, and sophistication. There are also concerns, however, about its potentially destabilizing effects during financial turmoil. As foreign participation grows, the local bond markets seem to respond more sensitively to global financial conditions as herding behavior among global funds takes hold (on top of domestic macroeconomic conditions). Similarly, while the introduction of financial futures facilitates price discovery of the underlying financial assets, some market observers note that greater foreign investor participation in the futures market might raise market volatility with relatively large, one-way transaction volumes compared to those of domestic investors. For instance, some note that just several foreign investors can take substantially large positions in the futures market, moving prices and subsequently affecting cash prices. Domestic players also closely monitor foreign traders, who are often viewed as better informed and more sophisticated, which in turn influences domestic trading behavior. The Korean case merits some discussion in this regard. Following the Asian financial crisis, Korean policy makers actively sought the development of domestic bond markets as an alternative source of funding to bank lending. Since then, the market has seen tremendous growth, in no small part, due to the establishment of KTB futures and the proactive promotion of foreign participation, as many observers note. But do foreign investors help trigger excessive futures price movements and transmit this instability to the cash market? This concern reflects a lack of understanding about the trading 1 Data available at http://asianbondsonline.adb.org/regional/data/bondmarket.php?code=Foreign_Holdings 2 Emerging Asia in this paper refers to the nine economies in East and Southeast Asia covering the People’s Republic of China; Hong Kong, China; Indonesia; the Republic of Korea; Malaysia; the Philippines; Singapore; Thailand; and Viet Nam.
2 | ADB Economics Working Paper Series No. 427 behaviors of foreign investors and their role in financial asset pricing in comparison to those of domestic investors. Price discovery inevitably involves increased market volatility. But if this higher market volatility really only reflects the increase in information as more heterogeneous groups of investors enter the market, with its diverse sources of information, it may not be a concern in itself. Therefore, the real question should be what type of foreign investors does the futures market attract? Are they different from the type of foreign investors participating in the cash market? And does their trading behavior show more inclination toward herding and speculation, leading to excessive market volatility? Better understanding of foreign investors’ trading behaviors in futures markets and their role in market efficiency and volatility is critical for assessing the risk of financial liberalization and for designing a macroprudential policy framework appropriate during rapid financial market development. To our knowledge, no study looks at the impact on price discovery of sovereign bonds of foreign participation in the futures market. This is especially true for emerging market economies, given a lack of quality trading data. Research in this unexplored area could have a profound impact on financial market development, as the findings could guide policy makers in crafting regulatory guidelines that win the benefits of foreign participation in developing futures markets while avoiding the potentially adverse consequences. This paper investigates the price impact of foreign participation in local currency bond futures from the emerging market perspective. The paper also introduces some important new elements. First, similar to Brandt, Kavajecz, and Underwood (2007),3 we examine the trading patterns of different investor groups, specifically looking into the impact of the net transactions of foreign and domestic finance institutions (DFI) on KTB futures and cash prices. We also examine in which market price discovery takes place. Second, following Richards (2005), we decompose net buys and net long positions into expected and unexpected components to examine how the price discovery happens, but add the analysis of the different types of investor groups to assess whether the type of investor matters in price discovery and, if so, by how much. And third, we look at the discovery process by investor groups at three subperiods—precrisis, crisis, and postcrisis—in our sample for any marked changes in the trading behaviors of different investor groups that have influenced price discovery across the subperiods. Specifically, the paper aims to answer the following three questions: (i) Which market (cash or futures) leads price discovery in Korean treasury bonds? (ii) What types of investor groups drive this process? (iii) Which component (expected or unexpected) of net purchases and net long positions influences price movements in both markets? In summary, our findings confirm that price discovery takes place in the futures market, in that our model fit substantially increased when we took the interaction between net transactions in both cash and futures markets into account. This is as opposed to considering the cash market only. We also find that foreigners’ net long positions in the futures market are highly associated with futures and cash returns, suggesting that they drive price discovery. Finally, the unexpected component of foreigners’ net long positions (their private information or idiosyncratic response to news) seems to exert significant influence on prices, although slightly less than their expected component; while it is the expected component that matters overwhelmingly more for the price impact of DFIs’ cash trading. 3 Brandt, Kavajecz, and Underwood (2007) examine four different investor groups based on whom Chicago Board of Trade (CBOT) members trade for in their accounts.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 9 Table 2 continued Crisis (16 Sep 2008 to 31 Dec 2009) Mean 33,483 147,361 … … Standard deviation 17,823 59,806 … … Skewness 0.28 1.78 … … Kurtosis 3.41 8.10 … … Observation 315 315 … … Postcrisis (01 Jan 2010 to 31 Dec 2013) Mean 29,584 239,465 29,352 71,398 Standard deviation 12,678 110,322 19,116 47,535 Skewness 0.84 1.80 1.20 0.12 Kurtosis 4.88 7.92 5.58 2.11 Observation 962 962 962 766 … = not available, KRW = Korean won. Source: Authors’ calculations. Table 3 presents the average net purchases (cash market) and net long position (futures market) by investor group across subperiods. For the cash market, both foreigners and DFIs are net buyers. Because KTB issuance increased over time, all market participants may accumulate net buys, reflecting market demand growth. In the futures market, however, net long positions have to be met by net short positions. It appears that in the futures market, foreign investors are initiating trading, while DFIs are accommodating the net positions of foreigners. For instance, DFIs are selling their holdings in the 3-year futures market as foreigners purchase those assets, but in the 10-year market, foreigners are taking short positions, while DFIs are taking long positions. We also find that foreigners increased their net purchases in the 3-year cash market during the crisis and postcrisis periods and reduced their net long positions in the futures market. In the cash market, most DFI subgroups are net buyers, except for banks, which are net sellers of 10-year KTBs. We note that foreigners purchased more in the 10-year than the 3-year cash market before the crisis, but this pattern was reversed during the crisis and postcrisis periods. This may be in line with ongoing deleveraging in advanced economies, where foreign investors adjusted their portfolios in favor of shorter tenors. In addition, we also find that among the DFI subgroups, domestic banks have larger net purchases or net long positions compared to other institutions. For the futures market, banks, among the DFIs, are again the largest players. For 3-year futures, domestic banks, asset management companies, funds, and securities companies are net sellers, although both funds and securities were net buyers during the global financial crisis. For 10-year futures, banks and securities companies are net buyers, while asset management firms and funds are net sellers. Except for banks, other DFIs have smaller positions in the futures market, compared to foreigners. This supports the view that foreign participation instigates price discovery in the futures markets. Foreigners also show a tendency of herding9 and trend-chasing,10 amassing a large net position compared to DFIs at times. 9 A similar type of investor following a similar investment strategy may make a common investment decision; for example, simultaneously buying or selling the same securities. 10 Buying financial securities after a recent upward trend in prices and selling after a recent downward trend. For example, investors can make a consistent one-directional investment for some time based on their projections of the trend in the policy rate.
10 | ADB Economics Working Paper Series No. 427 Table 3: Average Net Buy and Net Long Position in Cash and Futures Market 3-Year 10-Year Foreigners Domestic Finance Institutions Banks Asset Management Funds Securities Foreigners Domestic Finance Institutions Banks Asset Management Funds Securities Average Net Buy in Cash Market Full sample 0.700 1.149 0.733 0.143 0.273 … 0.941 6.142 –0.652 0.553 6.241 … Precrisis 0.421 0.840 0.469 –0.048 0.419 … 1.127 9.091 –1.198 0.738 9.551 … Crisis 0.933 0.679 0.431 0.101 0.147 … 0.661 5.416 –0.556 0.520 5.452 … Postcrisis 0.952 1.665 1.141 0.381 0.143 … 0.813 2.914 –0.043 0.347 2.610 … Average Net Long Position in Futures Market Full sample 0.089 –0.068 –0.038 –0.009 –0.006 –0.015 –0.034 0.060 0.046 –0.032 –0.011 0.057 Precrisis 0.138 –0.090 –0.045 –0.009 –0.004 –0.033 … … … … … … Crisis 0.031 –0.058 –0.083 –0.006 0.001 0.033 … … … … … … Postcrisis 0.052 –0.045 –0.016 –0.009 –0.010 –0.009 –0.034 0.060 0.046 –0.032 –0.011 0.057 … = not available. Note: Values in percent of cash and futures trading volumes. Source: Authors’ calculations.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 11 We also look into how different investor groups react to the monetary policy of the Bank of Korea (BOK) in both cash and futures markets.11 Figures 1a and 1b show the net purchases and net long positions of foreign and DFIs for 3-year government bonds and futures.12 Foreigners tend to reduce their net long positions in the futures market during BOK monetary policy actions (either easing or tightening stance), leading to the reverse positions of DFIs. In the cash market, DFIs are more active traders than foreigners, although both foreigners and DFIs seem to respond to BOK policy changes by adjusting their net purchases. These results imply that foreign investors who initiate trading activity in the futures market are different from the foreign investors in the cash market, and their trading decisions are independent of each other. In addition, DFIs are followers in the futures market by adjusting their net positions as a counterpart to the foreign investors’ positions, but they act as an independent major player in the cash market. Figure 1a: Net Purchases of Foreign and Domestic Finance Institutions in the 3-Year Treasury Cash Market DFI = domestic finance institution, GFC = global financial crisis. Note: Futures price index scaled by 100. Source: Authors’ calculations. 11 Policy stance refers to episodes when the BOK increased or decreased the base rate by at least 25 basis points in relation to the previous three interest rate policy decisions. 12 We do not show the 10-year government bonds due to a shorter sample. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 –2.0 –1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Jan-04 Jul-05 Jan-07 Jul-08 Jan-10 Jul-11 Jan-13 Rate cut Rate hike GFC Foreign DFIs
12 | ADB Economics Working Paper Series No. 427 Figure 1b: Net Positions of Foreign and Domestic Finance Institutions in the 3-Year Treasury Futures Market DFI = domestic finance institution, GFC = global financial crisis. Note: Futures price index scaled by 100. Source: Authors’ calculations. C. Decomposition of Net Purchases and Positions Foreign investors have emerged as important players of increasing influence in the KTB markets, and their trading is closely watched by other market participants. Many observers have also noted that net purchases of foreigners in the KTB futures market seem to be associated with price changes in the KTB and its futures markets. Do these net purchases by foreigners represent additional information to net investor demand in these markets? To understand the price impact of foreign net purchases in KTB cash and futures markets, we try to estimate the new information content of net purchases by foreigners and other investor groups. The trading decisions of foreign investors are presumably based on an information set that may be different from that of domestic investors. Their motivation can be also different from domestic investors, reflecting differences in their overall investment portfolios. For example, foreign investors may be able to extract information from global returns about the future prospects of emerging markets, or they are simply increasing their investment allocations to emerging market assets for other portfolio benefits. Richards (2005) presented a model to explain what drives net purchases of foreign investors in emerging market equity and futures. Using a model in which the net purchases of foreign investors are regressed on lagged returns of various markets and other lagged variables, such as net flows, he suggests foreign investors tend to respond to price movements in various markets or the information that drives those movements. Such similar trading patterns could be also interpreted as a form of herding by these investors, even if it is unconscious. Several studies have also looked into the difference between expected and unexpected components of order flows and transaction volumes in financial futures markets. For instance, Bessembinder and Seguin (1993) examined the expected and unexpected components of futures 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 –2.0 –1.5 –1.0 –0.5 0.0 0.5 1.0 1.5 Jan-04 Jul-05 Jan-07 Jul-08 Jan-10 Jul-11 Jan-13 Rate cut Rate hike GFC Foreign DFIs
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 13 trading volume, and found that unexpected volume shocks have greater impact on volatility. Drawing on the regression methodology in Richards (2005), we decompose net purchases of different investor groups into expected and unexpected components. It would be interesting to see which component (expected or unexpected) of net purchases and net long positions drives the discovery process. If the net purchase of foreign investors is largely a response to news and provides additional information to the market, it could be that the unexpected component plays a bigger role in price discovery. Richards (2005) tested this hypothesis in his study of foreign participation in Asian equity markets and found both expected and unexpected components to be significant. However, the coefficients and variance for the unexpected component are relatively large, suggesting that the majority of the contemporaneous impact of flows on returns can be attributed to the unexpected component. Following Richards (2005), we decompose the expected and unexpected components by: 10 3 (1) where NP is the net purchase (net long position) of investor groups, RETURNSt-1 refers to the lagged bond returns (either cash or futures), UST10Yt-1 is the lagged US 10-year Treasury bond yield used as proxy for global interest rates, KTB3Yt-1 is the lagged yields of 3-year Korean government bonds used as proxy for domestic interest rates, VIX t-1 is the lagged VIX used to measure investor risk appetite, ΔFXt-1 is the lagged change in nominal exchange rate to account for exchange rate movements, and εt is the error term. We sourced daily 10-year US Treasury bond yields, 3-year Korean bond yields, nominal exchange rate (the Korean won per US dollar) from the BOK and the VIX from the Chicago Board Options Exchange. We estimate Equation 1 for foreigners and DFIs (including its subgroups) for both cash and futures bond markets. We derived the unexpected component by estimating Equation 1 and then assuming that the residuals (εt) correspond to the unexpected component, while the expected component corresponds to the fitted values of the above regression. Both expected and unexpected components derived from Equation 1 are also estimated for the three subperiods (precrisis, crisis, postcrisis). We do so to account for the changing investor expectations during the three sample periods. Tables 4a–c present the regression results on the determinants of net purchases and net long positions for precrisis, crisis, and postcrisis periods, respectively. We note several findings. First, lagged net purchases are significant for all sample periods. In the model, the lag length is set at five, as Richards (2005) suggests.13 Net purchases seem to show positive autocorrelation, as investors may build their positions gradually (to mitigate the market impact of their trading) or investors of similar types may respond to new information in similar ways, but with different speeds. Second, net purchases are significantly correlated with lagged returns, which is more pronounced for several investor groups. The lagged returns have significant, negative effects on DFIs’ net purchases in the 3year KTB cash market and on the net long positions of domestic funds and securities companies in the 3-year futures market during the precrisis period. The effects of lagged returns on net purchases in the cash market become rather insignificant during the crisis and postcrisis periods, but the lagged 13 We also tried to fit the model using one lag, but the R-squared was lower. Using five lags improved the model fit. Although the R-squared is low, our model specification seems to be adequate as the residuals do not show any distinct patterns, which may suggest biases in our estimates.
14 | ADB Economics Working Paper Series No. 427 Table 4a: Determinants of Net Purchases and Net Long Positions (Precrisis Period) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) Variables for3 fin3 fin3bk fin3am fin3fd fin3sc spotfb3 spotfinb3 spotfinb3bk spotfinb3am spotfinb3fd NP (-1) 0.0576 0.0639 0.0520 –0.0176 –0.0482 0.0126 0.120** 0.0970** 0.0626 0.149*** 0.0783* (0.0502) (0.0497) (0.0417) (0.0453) (0.0385) (0.0390) (0.0590) (0.0481) (0.0420) (0.0367) (0.0423) NP (-2) 0.122*** 0.122*** 0.00360 –0.100** –0.0374 0.0349 0.0536 0.112** 0.0756* 0.0721* 0.0463 (0.0364) (0.0364) (0.0393) (0.0443) (0.0344) (0.0348) (0.0438) (0.0437) (0.0445) (0.0369) (0.0337) NP (-3) 0.0247 0.0287 –0.0160 0.0436 0.0157 0.000330 –0.123** 0.0133 –0.0332 0.0344 0.103*** (0.0453) (0.0439) (0.0400) (0.0434) (0.0414) (0.0342) (0.0533) (0.0367) (0.0395) (0.0416) (0.0327) NP (-4) 0.0427 0.0369 0.0202 0.0101 –0.0397 –0.0509 0.0385 –0.0219 –0.0768** –0.0137 0.0351 (0.0395) (0.0387) (0.0363) (0.0380) (0.0298) (0.0396) (0.0521) (0.0363) (0.0369) (0.0364) (0.0357) NP (-5) 0.0416 0.0204 0.0716** –0.0424 –0.0234 0.0241 0.0500 –0.0520 0.0861** –0.0316 –0.0552 (0.0425) (0.0409) (0.0348) (0.0528) (0.0442) (0.0403) (0.0526) (0.0420) (0.0393) (0.0393) (0.0392) Returns (-1) 0.860 –0.497 1.040 0.584* –0.272** –2.301*** 0.194 –6.253*** –2.272** –1.830 –2.195*** (1.242) (1.240) (0.849) (0.324) (0.135) (0.569) (0.566) (2.032) (1.106) (1.200) (0.799) UST 3-Yr (-1) –0.279 0.271 0.359** 0.0167 0.0359 –0.185 0.290 –0.498 –0.0172 –0.390 –0.0899 (0.203) (0.205) (0.177) (0.0643) (0.0290) (0.116) (0.215) (0.386) (0.289) (0.250) (0.169) KTB 3-Yr (-1) 0.167 –0.194 –0.267 0.0747 –0.0809 0.195 –0.147 –0.217 –0.239 0.254 –0.207 (0.286) (0.283) (0.234) (0.103) (0.0504) (0.177) (0.176) (0.465) (0.314) (0.307) (0.199) VIX (-1) –0.0213 0.0256 0.0434 –0.00209 0.00575 –0.0397 0.142*** –0.0941 0.0368 –0.105* –0.0177 (0.0428) (0.0424) (0.0362) (0.0137) (0.00620) (0.0276) (0.0394) (0.0841) (0.0575) (0.0556) (0.0334) ΔFX (-1) 0.453 –0.500* –0.0422 –0.0946 –0.0366 –0.213 –0.00394 0.0135 –0.379 0.0137 0.333* (0.282) (0.277) (0.224) (0.105) (0.0447) (0.178) (0.148) (0.433) (0.280) (0.271) (0.185) Constant 0.797 –0.713 –0.912 –0.448 0.168 0.339 –2.257*** 5.107** 0.984 1.958 1.891** (1.209) (1.209) (1.023) (0.478) (0.236) (0.725) (0.803) (2.100) (1.405) (1.390) (0.913) Observations 832 832 832 832 832 832 773 773 773 773 773 R-squared 0.033 0.031 0.016 0.020 0.014 0.028 0.080 0.062 0.030 0.051 0.055 … = not available, DFI = domestic finance institution. Notes: for = foreign futures; FIN = DFIs futures; _bk = DFI banks futures/spot; _am DFI asset management future/spot; _fd = DFI funds futures/spot; _sc = DFI securities futures; spotfb = foreign spot; spotfin = DFI spot. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 15 Table 4b: Determinants of Net Purchases and Net Long Positions (Crisis Period) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) Variables for3 fin3 fin3bk fin3am fin3fd fin3sc spotfb3 spotfinb3 spotfinb3bk spotfinb3am spotfinb3fd NP (-1) 0.215*** 0.183*** –0.00428 0.0191 –0.0279 0.0437 0.236** 0.0831 –0.0601 0.0385 0.0468 (0.0695) (0.0701) (0.0595) (0.0721) (0.0707) (0.0621) (0.108) (0.0507) (0.0707) (0.0633) (0.0642) NP (-2) 0.107* 0.134** 0.0624 –0.0224 0.0466 –0.0170 –0.0318 –0.0411 –0.0130 –0.0158 –0.0241 (0.0630) (0.0573) (0.0546) (0.0553) (0.0744) (0.0565) (0.0618) (0.0504) (0.0623) (0.0672) (0.0758) NP (-3) –0.0335 –0.0416 0.00241 –0.00906 –0.123* –0.0157 0.0297 –0.00465 –0.0779 0.0655 –0.0735 (0.0607) (0.0574) (0.0553) (0.0551) (0.0738) (0.0640) (0.0549) (0.0595) (0.0596) (0.0844) (0.0637) NP (-4) 0.151** 0.163*** –0.104 0.124*** 0.134* –0.0444 0.100 0.0338 0.00252 –0.0201 0.0870 (0.0616) (0.0611) (0.0664) (0.0454) (0.0754) (0.0653) (0.0721) (0.0604) (0.0537) (0.0679) (0.0635) NP (-5) 0.0569 0.0549 0.0227 0.0636 0.0623 0.131* 0.0624 0.103* 0.0501 –0.0202 –0.00806 (0.0638) (0.0632) (0.0626) (0.0497) (0.0649) (0.0675) (0.0641) (0.0545) (0.0610) (0.0576) (0.0628) Returns (-1) 0.921 –0.820 –1.340** –0.462*** 0.0924 0.595 –0.872* 1.149 1.175 –0.327 –0.255 (0.705) (0.721) (0.540) (0.172) (0.125) (0.476) (0.484) (1.142) (0.834) (0.611) (0.458) UST 3-Yr (-1) –1.911** 1.847** 0.551 0.349* 0.174 1.411** –0.663 1.540 0.669 0.389 1.521** (0.927) (0.917) (0.743) (0.186) (0.214) (0.700) (0.640) (1.337) (1.003) (0.483) (0.620) KTB 3-Yr (-1) 0.491 –0.553 –0.127 –0.0650 –0.0535 –0.286 0.340 –1.018 –0.515 –0.340 –0.584** (0.424) (0.416) (0.360) (0.0989) (0.102) (0.323) (0.343) (0.650) (0.494) (0.518) (0.273) VIX (-1) –0.00297 0.00376 0.00682 0.000843 –0.00206 –0.00170 –0.0191** –0.00723 0.00691 –0.00155 –0.00872 (0.0118) (0.0116) (0.00924) (0.00313) (0.00306) (0.00842) (0.00936) (0.0178) (0.0131) (0.0108) (0.00698) ΔFX (-1) 0.120* –0.155** –0.0731 0.0159 –0.0113 –0.0584 –0.0442 0.0154 0.0435 0.0436 –0.0185 (0.0660) (0.0688) (0.0770) (0.0263) (0.0212) (0.0668) (0.0664) (0.143) (0.145) (0.0666) (0.0611) Constant 0.991 –0.684 –0.715 –0.281 0.0416 –0.810 0.817 2.711 1.436 0.234 0.587 (1.022) (1.059) (1.103) (0.250) (0.279) (0.926) (1.126) (2.181) (1.684) (1.544) (0.780) Observations 253 253 253 253 253 253 253 253 253 253 253 R-squared 0.187 0.178 0.052 0.068 0.052 0.048 0.129 0.037 0.028 0.017 0.063 … = not available, DFI = domestic finance institution. Notes: for = foreign futures; FIN = DFIs futures; _bk = DFI banks futures/spot; _am DFI asset management future/spot; _fd = DFI funds futures/spot; _sc = DFI securities futures; spotfb = foreign spot; spotfin = DFI spot. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
16 | ADB Economics Working Paper Series No. 427 Table 4c: Determinants of Net Purchases and Net Long Positions (Postcrisis Period) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) Variables for3 fin3 fin3bk fin3am fin3fd fin3sc spotfb3 spotfinb3 spotfinb3bk spotfinb3am spotfinb3fd for10 fin10 fin10bk fin10am fin10fd fin10sc spotfb10 spotfinb10 spotfinb10bk spotfinb10am spotfinb10fd NP (-1) 0.304*** 0.304*** –0.0525 0.00572 –0.153** 0.129*** 0.128 0.0813* 0.0619 0.168*** 0.0246 0.0842 0.0541 –0.119 0.0270 0.0610 –0.246*** 0.160** 0.178** –0.0529 0.108*** 0.196*** (0.0493) (0.0491) (0.0372) (0.0451) (0.0634) (0.0433) (0.0982) (0.0427) (0.0390) (0.0377) (0.0869) (0.0806) (0.0797) (0.0785) (0.0584) (0.114) (0.0639) (0.0699) (0.0694) (0.0482) (0.0402) (0.0662) NP (-2) 0.125*** 0.113** –0.0925** 0.0316 –0.0408 0.0451 –0.00525 0.0476 0.0221 0.0710* 0.0217 –0.0327 –0.0203 –0.0977 0.109 0.0601 –0.0905 0.0628 0.149** 0.0536 0.0497 0.0887 (0.0456) (0.0458) (0.0372) (0.0416) (0.0579) (0.0393) (0.0538) (0.0382) (0.0367) (0.0418) (0.0364) (0.151) (0.132) (0.0685) (0.0699) (0.0861) (0.0644) (0.0689) (0.0614) (0.0533) (0.0474) (0.0732) NP (-3) 0.117*** 0.115*** –0.0227 0.00414 –0.00124 0.0417 0.0139 0.0368 0.00483 0.102** –0.00898 0.0363 0.0779 0.00539 –0.00580 –0.0440 –0.0140 0.166*** 0.00338 0.0780* –0.00661 0.0903 (0.0425) (0.0424) (0.0370) (0.0358) (0.0563) (0.0430) (0.0263) (0.0451) (0.0372) (0.0411) (0.0398) (0.0495) (0.0597) (0.0820) (0.0564) (0.0740) (0.0606) (0.0561) (0.0408) (0.0455) (0.0419) (0.0700) NP (-4) 0.00275 0.0107 –0.0211 –0.0601 –0.0348 0.0658* 0.0159 –0.000967 –0.0223 –0.00682 0.0699 –0.138 –0.125 –0.0234 0.0445 –0.0870 –0.0252 0.0178 0.0276 0.0954** –0.0101 –0.0506 (0.0407) (0.0406) (0.0385) (0.0429) (0.0482) (0.0395) (0.0272) (0.0391) (0.0331) (0.0357) (0.0459) (0.131) (0.117) (0.0638) (0.0346) (0.0742) (0.0596) (0.0513) (0.0447) (0.0421) (0.0611) (0.0531) NP (-5) –0.0890** –0.0867** –0.000215 0.00679 0.0548 –0.0353 0.0207 –0.00607 –0.0378 –0.0354 –0.0441 0.0137 –0.0326 –0.0368 0.0334 0.117 –0.0619 0.0672 –0.0294 –0.0315 –0.0363 0.0942* (0.0392) (0.0390) (0.0371) (0.0367) (0.0484) (0.0384) (0.0311) (0.0401) (0.0349) (0.0420) (0.0465) (0.0428) (0.0513) (0.0845) (0.0548) (0.0974) (0.0635) (0.0414) (0.0442) (0.0472) (0.0398) (0.0568) Returns (-1) –1.655** 1.656** –1.462*** 0.242* –0.211 1.849*** 0.750 –2.207 –1.998 0.280 –0.512 –0.232 0.170 0.403 0.113** 0.0110 –0.265 0.292 –0.280 0.436 0.136 –0.559 (0.726) (0.709) (0.530) (0.143) (0.200) (0.698) (0.991) (2.011) (1.393) (0.868) (0.750) (0.192) (0.226) (0.351) (0.0558) (0.111) (0.347) (0.283) (0.895) (0.515) (0.215) (0.855) UST 3-Yr (-1) 0.337 –0.302 –0.542* –0.0228 –0.0226 0.145 0.523 0.689 0.807 0.230 –0.274 –0.277 0.363 0.475* 0.0100 0.129** –0.144 –0.184 0.778 –0.291 –0.0428 0.980 (0.385) (0.381) (0.298) (0.0638) (0.147) (0.386) (0.374) (0.828) (0.568) (0.312) (0.315) (0.217) (0.224) (0.261) (0.0334) (0.0644) (0.247) (0.179) (0.671) (0.375) (0.142) (0.643) KTB 3-Yr (-1) –0.0633 0.0674 0.494* –0.0656 –0.123 –0.286 –0.497 –1.670* –1.549*** –0.358 0.203 0.477 –0.515* 0.00744 –0.0986 –0.415** –0.0699 0.323 0.789 0.348 0.634** –0.254 (0.384) (0.377) (0.267) (0.0597) (0.132) (0.384) (0.356) (0.852) (0.588) (0.325) (0.313) (0.298) (0.305) (0.409) (0.0624) (0.193) (0.413) (0.303) (1.116) (0.612) (0.246) (1.140) VIX (-1) 0.0247 –0.0236 –0.0406*** –0.00190 0.00154 –0.00141 0.0293 –0.0579 0.00145 –0.0137 –0.0502*** –0.0155 0.0160 –0.00856 0.00156 0.0370* –0.0168 –0.0166 –0.0480 –0.0186 –0.0108 –0.00935 (0.0160) (0.0161) (0.0138) (0.00301) (0.00687) (0.0167) (0.0303) (0.0379) (0.0309) (0.0158) (0.0143) (0.0132) (0.0137) (0.0190) (0.00396) (0.0191) (0.0266) (0.0141) (0.0531) (0.0303) (0.0131) (0.0475) ΔFX (-1) 0.182 –0.200 0.104 –0.0462* –0.0542 –0.231* –0.454*** –0.460 –0.304 –0.0419 –0.119 0.0294 –0.0358 0.324 –0.00893 –0.0468 –0.335 –0.0511 –1.420*** –0.694** –0.0329 –0.606 (0.151) (0.154) (0.110) (0.0274) (0.0479) (0.139) (0.173) (0.322) (0.233) (0.133) (0.130) (0.126) (0.140) (0.221) (0.0459) (0.106) (0.254) (0.176) (0.485) (0.279) (0.0897) (0.382) Constant –0.434 0.380 –0.476 0.253* 0.376 0.849 1.448* 7.665*** 5.703*** 1.569* 0.650 –0.635 0.570 –0.840 0.230 0.358 0.856 –0.0380 –1.762 –0.401 –1.425*** 0.186 (0.982) (0.964) (0.611) (0.144) (0.252) (0.963) (0.868) (2.218) (1.576) (0.835) (0.764) (0.554) (0.584) (0.940) (0.160) (0.382) (0.986) (0.619) (2.061) (1.196) (0.492) (1.973) Observations 720 720 720 720 720 720 720 720 720 720 720 564 564 564 564 564 564 595 595 595 595 595 R-squared 0.157 0.151 0.038 0.022 0.034 0.034 0.032 0.048 0.028 0.061 0.029 0.031 0.032 0.043 0.036 0.048 0.078 0.096 0.111 0.036 0.044 0.114 … = not available, DFI = domestic finance institution. Notes: for = foreign futures; FIN = DFIs futures; _bk = DFI banks futures/spot; _am DFI asset management future/spot; _fd = DFI funds futures/spot; _sc = DFI securities futures; spotfb = foreign spot; spotfin = DFI spot. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 17 returns have significant effects on net long positions of foreign investors in the futures market. The negative coefficients in the cash market suggest lower returns lead to higher net purchases, as opposed to the finding in Richards (2005). This may suggest that there is some underlying demand for KTBs attracting investors to buy when the price falls. In the earlier years, DFIs have increased their holdings of KTBs for various reasons other than pure investment purposes. This pattern has changed over time, as their KTB holdings may have reached adequate levels. On the other hand, the effects of lagged returns have become significant in the 3-year futures market, with foreigners buying on previous day low returns. In the postcrisis period, the effects of lagged returns are more significant for 3than for 10year KTBs, implying that for long-term investors, daily returns may not be as important as they are for short-term investors. Third, the muted effects of lagged returns in more recent periods may also suggest that investors extract information from more broadly based information sets and the lagged returns may provide only little new information. What is noticeable is that, the explanatory power of other information variables such as the US and Korean interest rates, risk aversion, and exchange rate changes has become more significant over time. This could imply that investors extract more information from global and domestic economic conditions and potential risk factors for their trading decisions. Table 5 shows the summary statistics for the expected component of net purchases, the fitted values derived from the above regression results. Overall, we find that net purchases in the cash markets are better fitted by the model than the net long positions in the futures, with explanatory variables, including lagged net purchases, lagged returns, and other economic variables. The model seems to explain DFIs’ net purchases in the cash market better than foreign investors’ net purchases. In the futures market, the net positions of both foreign and domestic investors seem to be driven by news or private information, as suggested by the relatively low value of the expected component. Foreign investors and DFI groups take the opposite positions in the futures market. It seems foreign investors have been net buyers of 3-year futures contracts, while DFIs have been net sellers. By the nature of the regression model, the unexpected component of net purchases is averaged zero as they are residuals from Equation 1. However, when we examine the pattern of unexpected components in relation to the BOK monetary policy stance, several observations can be made. Figures 2a and 2b present the unexpected component of net buy and net long positions of various investor groups in both cash and futures markets in relation to the BOK monetary policy stance. Overall, the unexpected components of foreign investors’ net purchases and positions are negative in both the cash and futures markets, while those of DFIs are more positive. In the cash market, the unexpected component of foreigner investors’ net purchases was more positively responsive to the monetary policy changes during the precrisis period, but became more negative through the global crisis and in the postcrisis period right after the crisis (Figure 2a). On the other hand, the unexpected component of DFIs’ net purchases increased during the precrisis period running up to the crisis, but decreased during the crisis and postcrisis periods until very recently. We note that for 3-year bond futures, the unexpected component of foreigners’ net positions is generally negative in response to the BOK’s aggressive policy actions (either increasing or decreasing the base rate), while DFIs tend to take net long positions in the futures markets as the counterparts to the foreigners’ positions during those periods (Figure 2b).
18 | ADB Economics Working Paper Series No. 427 Table 5: Expected Component of Net Purchases by Investor Groups Variable 3-Year KTB 10-Year KTB Precrisis Crisis Postcrisis Postcrisis Obs. Mean Std. Dev. Min. Max. Obs. Mean Std. Dev. Min. Max. Obs. Mean Std. Dev. Min. Max. Obs. Mean Std. Dev. Min. Max. Foreign future 867 0.29 0.74 –2.83 2.90 259 0.24 1.16 –3.93 2.69 749 0.06 1.08 –4.40 3.84 588 –0.06 0.32 –2.42 2.61 DFI future 867 –0.28 0.71 –2.93 2.80 259 –0.26 1.12 –2.79 3.62 749 –0.05 1.05 –3.78 4.29 588 0.08 0.34 –2.38 2.74 DFI bank future 867 –0.15 0.42 –1.54 1.28 259 –0.22 0.47 –1.39 1.73 749 –0.01 0.36 –1.37 0.88 588 0.13 0.53 –2.43 3.18 DFI AM future 867 –0.06 0.20 –1.14 0.97 259 –0.03 0.18 –0.52 0.54 749 –0.02 0.06 –0.22 0.21 588 –0.04 0.08 –0.64 0.34 DFI funds future 867 0.01 0.07 –0.28 0.24 259 0.00 0.15 –0.45 0.39 749 –0.02 0.13 –0.67 0.56 588 0.02 0.28 –0.89 2.08 DFI securities future 867 –0.07 0.43 –2.01 1.88 259 0.00 0.42 –1.25 1.26 749 0.01 0.45 –1.47 1.58 588 –0.04 0.77 –4.12 4.77 Foreign spot 802 0.49 0.76 –1.57 5.62 259 0.92 0.79 –1.39 3.84 749 0.88 0.53 –2.05 4.76 619 0.45 0.62 –2.14 3.72 DFI spot 802 0.78 1.55 –5.91 7.76 259 0.51 0.78 –2.05 2.44 749 1.72 1.23 –2.77 6.26 619 2.48 2.07 –4.01 14.23 DFI bank spot 802 0.42 0.74 –1.92 3.30 259 0.47 0.51 –1.12 1.89 749 1.16 0.71 –0.71 4.16 619 –0.39 0.68 –3.18 1.92 DFI AM spot 802 –0.02 0.91 –3.80 4.13 259 0.02 0.27 –0.93 0.83 749 0.40 0.57 –1.49 2.85 619 0.33 0.29 –0.40 2.14 DFI funds spot 802 0.38 0.68 –2.06 3.61 259 0.02 0.42 –1.68 1.39 749 0.16 0.35 –1.32 1.59 619 2.55 1.91 –1.49 16.39 AM = asset management, DFI = domestic finance institutions, KTB = Korean Treasury Bond. Source: Authors' calculations.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 25 Table 8: Full Sample (Expected and Unexpected Component Estimates) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Variables sreturns3 sreturns10 sreturns3 sreturns10 sreturns3 sreturns10 freturn3 freturn10 freturn3 freturn10 freturn3 freturn10 Unexpected futures foreign 0.01273*** 0.00335 0.01119*** –0.00017 0.01561*** –0.00107 0.01401*** –0.00788 (0.00117) (0.00693) (0.00124) (0.00638) (0.00132) (0.00801) (0.00141) (0.00758) Lag unexpected futures foreign –0.00357*** –0.00118 –0.00448*** 0.00271 –0.00497*** –0.00358 –0.00612*** 0.00506 (0.00137) (0.00765) (0.00140) (0.00834) (0.00152) (0.00954) (0.00153) (0.01018) Expected futures foreign 0.02458*** –0.01663 0.02602*** –0.00625 0.02823*** –0.01781 0.03059*** –0.00736 (0.00603) (0.04789) (0.00624) (0.04251) (0.00678) (0.05636) (0.00697) (0.05250) Lag expected futures foreign –0.01130** 0.01448 –0.01325** –0.01496 –0.01310** 0.02480 –0.01516** –0.01221 (0.00571) (0.03464) (0.00566) (0.04534) (0.00634) (0.04249) (0.00626) (0.05758) Unexpected spot foreign 0.00068 –0.00426 0.00010 –0.00347 0.00099 –0.00415 0.00040 –0.00424 (0.00149) (0.00770) (0.00166) (0.00753) (0.00170) (0.00908) (0.00187) (0.00899) Lag unexpected spot foreign –0.00113 0.00592 –0.00075 0.00593 –0.00174 0.01043 –0.00141 0.00981 (0.00151) (0.00931) (0.00161) (0.00757) (0.00172) (0.01251) (0.00183) (0.00970) Expected spot foreign 0.00595 0.01389 0.00653 0.01476 0.00982 0.00523 0.01102 0.00831 (0.00823) (0.04488) (0.00808) (0.03354) (0.00935) (0.05671) (0.00914) (0.04189) Lag expected spot foreign –0.00437 –0.00980 –0.00598 –0.01714 –0.00584 –0.00289 –0.00753 –0.01355 (0.00834) (0.03174) (0.00824) (0.02739) (0.00951) (0.04082) (0.00941) (0.03440) Unexpected futures DFIs –0.01109*** 0.00224 –0.00117 0.00010 –0.01384*** 0.00937 –0.00225 0.03670 (0.00123) (0.00566) (0.00630) (0.03758) (0.00139) (0.00680) (0.00710) (0.05030) Lag unexpected futures DFIs 0.00434*** –0.00115 0.01970*** –0.05408** 0.00594*** –0.00300 0.02533*** –0.09296*** (0.00142) (0.00614) (0.00630) (0.02117) (0.00155) (0.00793) (0.00700) (0.02598) Expected futures DFIs –0.02739*** 0.00041 –0.07883** –0.02906 –0.03252*** 0.00554 –0.11144*** –0.02794 (0.00646) (0.03171) (0.03212) (0.10737) (0.00714) (0.03963) (0.03501) (0.13595) Lag expected futures DFIs 0.01456** 0.00701 0.03654 0.10030 0.01716*** 0.00414 0.05195 0.08346 (0.00589) (0.02987) (0.03164) (0.08931) (0.00646) (0.03911) (0.03556) (0.11546) Unexpected spot DFIs –0.00409*** –0.01323*** –0.00415*** –0.01280*** –0.00437*** –0.01647*** –0.00443*** –0.01575*** (0.00066) (0.00225) (0.00065) (0.00223) (0.00074) (0.00273) (0.00073) (0.00271) Lag unexpected spot DFIs 0.00258*** 0.00117 0.00272*** 0.00100 0.00303*** 0.00252 0.00322*** 0.00192 (0.00069) (0.00240) (0.00070) (0.00236) (0.00078) (0.00302) (0.00078) (0.00298) Expected spot DFIs –0.02094*** –0.00324 –0.02109*** –0.00472 –0.02269*** 0.00155 –0.02292*** –0.00040 (0.00497) (0.00917) (0.00495) (0.00934) (0.00556) (0.01152) (0.00553) (0.01164) Lag expected spot DFIs 0.01279*** –0.00149 0.01303*** –0.00061 0.01434*** –0.00800 0.01487*** –0.00641 (0.00425) (0.00737) (0.00428) (0.00731) (0.00473) (0.00942) (0.00475) (0.00932) Lag cross market returns 0.01273 –0.01879 –0.01316 –0.01843 –0.00203 –0.01890 –0.00199 0.02310 –0.03128 0.03285 –0.01293 0.02659 (0.03103) (0.04040) (0.03495) (0.03890) (0.03482) (0.04003) (0.03780) (0.06358) (0.04312) (0.06102) (0.04307) (0.06204) Constant 0.00052 0.00559 0.01102* 0.02015 0.01034 0.02213 0.00329 0.01819 0.01549** 0.03586 0.01242 0.03936 (0.00581) (0.01756) (0.00575) (0.01843) (0.00788) (0.02351) (0.00638) (0.02191) (0.00641) (0.02244) (0.00860) (0.02926) Observations 1,645 528 1,645 528 1,645 528 1,645 528 1,645 528 1,645 528 R-squared 0.08933 0.00638 0.12707 0.09938 0.13326 0.11676 0.10351 0.00679 0.13760 0.10300 0.14610 0.13013 DFIs = domestic finance institutions. Notes: sreturns3 = spot returns 3-year. sreturns10 = spot returns 10-year. freturn3 = future returns 3-year. freturn10 = future returns 10-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
26 | ADB Economics Working Paper Series No. 427 the other hand, DFI transactions in the 3-year cash market are more predictable, provided that their assessment of overall economic conditions is expected. This finding holds true for futures returns as well. Fourth, compared to Table 7, the subperiod estimates show several noteworthy issues. During the precrisis period (Table 9), the effects of lagged transaction variables are insignificant, while the effects of lagged cross-market returns are significant. The effects of lagged, cross-market returns disappear during crisis and postcrisis periods (Tables 11 and 13). The results seem to suggest that during the precrisis period, investors did not pay as much attention as in the later periods to the information content of trading. Rather, they tried to extract information from observing prices in other markets and exploited arbitrage opportunities. Over time, investors have become much more active in their information gathering from intraday price movements. As information transmission quickens to nearly instant, the effects of lagged returns are not important anymore. This reasoning seems to have worked well in recent years, whereby cash and futures trading has been much more closely watched, and there is now instant and intraday price transmission between the two markets. Table 9: Precrisis Estimates (1) (2) (3) (4) (5) (6) Variables sreturns3 sreturns3 sreturns3 freturn3 freturn3 freturn3 Futures foreign 0.01124*** 0.01075*** 0.01317*** 0.01270*** (0.00119) (0.00121) (0.00130) (0.00133) Lag futures foreign –0.00081 –0.00081 –0.00144 –0.00150 (0.00127) (0.00130) (0.00142) (0.00145) Spot foreign –0.00161 –0.00251 –0.00194 –0.00288 (0.00187) (0.00198) (0.00209) (0.00220) Lag spot foreign –0.00231 –0.00180 –0.00262 –0.00204 (0.00167) (0.00190) (0.00197) (0.00221) Futures DFIs –0.01070*** –0.00177 –0.01265*** –0.00332 (0.00120) (0.00598) (0.00131) (0.00662) Lag futures DFIs 0.00092 0.00549 0.00164 0.00844* (0.00128) (0.00426) (0.00143) (0.00492) Spot DFIs –0.00233*** –0.00253*** –0.00244*** –0.00269*** (0.00067) (0.00067) (0.00074) (0.00075) Lag spot DFIs 0.00088 0.00075 0.00112 0.00097 (0.00069) (0.00069) (0.00079) (0.00079) Lag cross market returns 0.10453*** 0.10265*** 0.09440** 0.11010** 0.11121** 0.09996** (0.03733) (0.03708) (0.03697) (0.04669) (0.04673) (0.04648) Constant –0.00141 –0.00191 0.00036 0.00057 –0.00034 0.00225 (0.00433) (0.00424) (0.00438) (0.00483) (0.00474) (0.00489) Observations 920 920 920 919 919 919 R-squared 0.11615 0.12214 0.12932 0.11985 0.12501 0.13263 DFIs = domestic finance institutions. Notes: sreturns3 = spot returns 3-year. freturn3 = future returns 3-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 27 Table 10: Precrisis Estimates (Expected and Unexpected) (1) (2) (3) (4) (5) (6) Variables sreturns3 sreturns3 sreturns3 freturn3 freturn3 freturn3 Unexpected futures foreign 0.0102*** 0.0101*** 0.0123*** 0.0122*** (0.00122) (0.00127) (0.00137) (0.00143) Lag unexpected futures foreign –0.00191 –0.00241* –0.00263* –0.00330** (0.00133) (0.00137) (0.00151) (0.00154) Expected futures foreign 0.0173 0.0199* 0.0182 0.0211 (0.0107) (0.0115) (0.0124) (0.0135) Lag expected futures foreign –0.00417 –0.00173 –0.00380 –0.00130 (0.00975) (0.00880) (0.0112) (0.0101) Unexpected spot foreign –0.00273 –0.00337 –0.00289 –0.00361 (0.00209) (0.00217) (0.00237) (0.00247) Lag unexpected spot foreign –0.00413* –0.00354 –0.00435 –0.00378 (0.00229) (0.00244) (0.00267) (0.00283) Expected spot foreign 0.0179 0.0171 0.0197 0.0193 (0.0119) (0.0118) (0.0139) (0.0138) Lag expected spot foreign –0.0187 –0.0187 –0.0215 –0.0212 (0.0132) (0.0133) (0.0157) (0.0158) Unexpected futures DFIs –0.0101*** –0.00837 –0.0121*** –0.00954 (0.00129) (0.00654) (0.00145) (0.00737) Lag unexpected futures DFIs 0.00229* 0.0110* 0.00313** 0.0151** (0.00133) (0.00581) (0.00150) (0.00671) Expected futures DFIs –0.0222* –0.0386 –0.0236* –0.0476 (0.0122) (0.0430) (0.0143) (0.0485) Lag expected futures DFIs 0.00299 0.000503 0.00298 0.00478 (0.00909) (0.0394) (0.0104) (0.0446) Unexpected spot DFIs –0.00210*** –0.00235*** –0.00230*** –0.00260*** (0.000729) (0.000714) (0.000827) (0.000814) Lag unexpected spot DFIs 0.00167 0.00163 0.00174 0.00174 (0.00106) (0.00104) (0.00123) (0.00121) Expected spot DFIs –0.00968 –0.00962 –0.00863 –0.00894 (0.00756) (0.00696) (0.00894) (0.00810) Lag expected spot DFIs 0.00787 0.00713 0.00775 0.00718 (0.00606) (0.00584) (0.00693) (0.00660) Lag cross market returns 0.0954** 0.0596 0.0612 0.103* 0.0715 0.0723 (0.0437) (0.0587) (0.0546) (0.0565) (0.0821) (0.0746) Constant –0.00322 –0.00364 –0.00238 –0.00152 –0.00325 –0.00165 (0.00585) (0.00605) (0.00770) (0.00654) (0.00699) (0.00879) Observations 726 726 726 726 726 726 R-squared 0.118 0.126 0.137 0.123 0.129 0.141 DFIs = domestic finance institutions. | Notes: sreturns3 = spot returns 3-year. freturn3 = future returns 3-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates. In the crisis period (Table 11), the effects of foreign net positions in the futures market are significant and large compared to pre and postcrisis periods. However, during 2008–2009, the share of foreign net long positions in 3-year KTB futures was not very big in absolute terms, although compared to DFIs, their transactions were relatively larger. During the postcrisis period (Table 13), the effect of
28 | ADB Economics Working Paper Series No. 427 the DFIs’ net long positions in the futures market is greater than the effect of foreign investors’ net long positions. The price effect of DFI trading increased significantly after the crisis, reflecting their increased size and sophistication. Table 11: Crisis Estimates (1) (2) (3) (4) (5) (6) Variables sreturns3 sreturns3 sreturns3 freturn3 freturn3 freturn3 Futures foreign 0.03362*** 0.03032*** 0.04173*** 0.03875*** (0.00551) (0.00584) (0.00607) (0.00651) Lag futures foreign –0.01326*** –0.01378*** –0.01838*** –0.01867*** (0.00423) (0.00459) (0.00462) (0.00487) Spot foreign 0.00766 0.00053 0.00608 –0.00078 (0.00482) (0.00602) (0.00543) (0.00671) Lag spot foreign –0.00019 0.00012 –0.00068 –0.00018 (0.00559) (0.00590) (0.00587) (0.00622) Futures DFIs –0.02922*** –0.00483 –0.03754*** –0.00944 (0.00539) (0.02489) (0.00604) (0.02658) Lag futures DFIs 0.01349*** –0.01423 0.01819*** –0.01585 (0.00434) (0.03031) (0.00459) (0.03000) Spot DFIs –0.01385*** –0.01355*** –0.01338*** –0.01328*** (0.00382) (0.00425) (0.00415) (0.00460) Lag spot DFIs –0.00257 –0.00287 –0.00220 –0.00258 (0.00334) (0.00350) (0.00349) (0.00368) Lag cross market returns –0.04208 –0.04061 –0.04761 –0.08516 –0.08468 –0.09363 (0.06496) (0.06670) (0.06694) (0.07446) (0.07629) (0.07647) Constant –0.00525 0.01036 0.00954 0.00746 0.02050 0.02108 (0.01640) (0.01357) (0.01797) (0.01739) (0.01468) (0.01952) Observations 288 288 288 288 288 288 R-squared 0.14829 0.19642 0.20214 0.18954 0.22507 0.23176 DFIs = domestic finance institutions. Notes: sreturns3 = spot returns 3-year. freturn3 = future returns 3-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates. In terms of the expected and unexpected components subperiods, Tables 10, 12, and 14 show that the unexpected foreign net long position in the futures market seems to consistently exert price influence across subperiods. However, our results also confirm the significance of the expected component, with its larger price effect, but the results are not consistent across periods and specifications. The greater impact of the expected component, albeit inconsistent across specifications and periods, might be due to increased transparency, and access to information has raised the effectiveness of expected net long positions, and therefore increased their effect on returns relative to unexpected net long positions.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 29 Table 12: Crisis Estimates (Expected and Unexpected) (1) (2) (3) (4) (5) (6) Variables sreturns3 sreturns3 sreturns3 freturn3 freturn3 freturn3 Unexpected futures foreign 0.03024*** 0.02670*** 0.03984*** 0.03642*** (0.00663) (0.00690) (0.00732) (0.00770) Lag unexpected futures foreign –0.02577*** –0.01894* –0.03375*** –0.02679** (0.00902) (0.00998) (0.00953) (0.01091) Expected futures foreign 0.09851*** 0.06878** 0.11288*** 0.08663*** (0.02835) (0.02902) (0.02976) (0.03125) Lag expected futures foreign –0.05696** –0.04478* –0.06612** –0.05555** (0.02575) (0.02487) (0.02741) (0.02733) Unexpected spot foreign 0.00911 0.00156 0.00725 –0.00025 (0.00559) (0.00685) (0.00645) (0.00770) Lag unexpected spot foreign 0.01387 0.01153 0.01548 0.01347 (0.00956) (0.01039) (0.01009) (0.01118) Expected spot foreign –0.04301 –0.04099 –0.05183 –0.05093 (0.03440) (0.03297) (0.03580) (0.03549) Lag expected spot foreign 0.02041 0.02355 0.02705 0.03091 (0.02880) (0.02868) (0.03062) (0.03130) Unexpected futures DFIs –0.02548*** 0.00754 –0.03464*** 0.00591 (0.00633) (0.02676) (0.00709) (0.03012) Lag unexpected futures DFIs 0.02208*** 0.02886 0.02942*** 0.02629 (0.00779) (0.03301) (0.00800) (0.03416) Expected futures DFIs –0.07663*** –0.09797 –0.09408*** –0.10225 (0.02478) (0.09078) (0.02600) (0.09849) Lag expected futures DFIs 0.05404** 0.04283 0.06496*** 0.04044 (0.02276) (0.07993) (0.02458) (0.08772) Unexpected spot DFIs –0.01125*** –0.01064** –0.01102** –0.01067* (0.00430) (0.00499) (0.00478) (0.00549) Lag unexpected spot DFIs 0.00291 0.00327 0.00296 0.00318 (0.00312) (0.00340) (0.00344) (0.00379) Expected spot DFIs –0.06336*** –0.06481*** –0.05699*** –0.05800** (0.02092) (0.02215) (0.02131) (0.02286) Lag expected spot DFIs –0.00267 0.00197 0.00303 0.00794 (0.01820) (0.01990) (0.01984) (0.02159) Lag cross market returns –0.15983* –0.04137 –0.07315 –0.22654** –0.08679 –0.13908 (0.08399) (0.08151) (0.09352) (0.08963) (0.08807) (0.09901) Constant 0.02272 0.03643* 0.05034 0.03431 0.04066* 0.05624 (0.03578) (0.02198) (0.03857) (0.03655) (0.02390) (0.04124) Observations 243 243 243 243 243 243 R-squared 0.16715 0.22449 0.23670 0.21385 0.24787 0.26147 DFIs = domestic finance institutions. Notes: sreturns3 = spot returns 3-year. freturn3 = future returns 3-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
30 | ADB Economics Working Paper Series No. 427 Table 13: Postcrisis Estimates (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Variables sreturns3 sreturns10 sreturns3 sreturns10 sreturns3 sreturns10 freturn3 freturn10 freturn3 freturn10 freturn3 freturn10 Futures foreign 0.01289*** 0.00126 0.00598*** 0.00010 0.01574*** –0.00326 0.00799*** –0.00498 (0.00170) (0.00589) (0.00199) (0.00578) (0.00187) (0.00689) (0.00216) (0.00731) Lag futures foreign –0.00231* –0.00213 –0.00129 –0.00245 –0.00366** –0.00487 –0.00228 –0.00309 (0.00138) (0.00642) (0.00150) (0.00657) (0.00154) (0.00842) (0.00167) (0.00897) Spot foreign 0.00135 0.00156 0.00262** 0.00095 0.00207 0.00326 0.00349** 0.00283 (0.00117) (0.00661) (0.00127) (0.00612) (0.00140) (0.00781) (0.00142) (0.00735) Lag spot foreign –0.00149 0.00698 –0.00181 0.00501 –0.00147 0.00940 –0.00189 0.00728 (0.00134) (0.00573) (0.00133) (0.00488) (0.00149) (0.00798) (0.00147) (0.00745) Futures DFIs –0.00660*** –0.00062 0.04708*** –0.02195 –0.00871*** 0.00417 0.05548*** –0.02214 (0.00199) (0.00550) (0.01539) (0.01940) (0.00217) (0.00706) (0.01730) (0.03164) Lag futures DFIs 0.00080 0.00255 0.02041 0.01354 0.00172 0.00309 0.01942 –0.00153 (0.00150) (0.00625) (0.01618) (0.03038) (0.00167) (0.00811) (0.01706) (0.04692) Spot DFIs –0.00652*** –0.01377*** –0.00644*** –0.01381*** –0.00724*** –0.01679*** –0.00716*** –0.01672*** (0.00091) (0.00195) (0.00093) (0.00196) (0.00100) (0.00237) (0.00101) (0.00242) Lag spot DFIs 0.00102 0.00391** 0.00132* 0.00412** 0.00141* 0.00623*** 0.00172** 0.00646*** (0.00075) (0.00164) (0.00073) (0.00161) (0.00083) (0.00209) (0.00083) (0.00205) Lag cross market returns –0.02168 –0.00561 –0.03110 –0.00304 –0.04225 –0.00458 –0.03693 0.03190 –0.04550 0.05498 –0.05952 0.05019 (0.03417) (0.03866) (0.03421) (0.03881) (0.03313) (0.03840) (0.04367) (0.05812) (0.04391) (0.06012) (0.04327) (0.05999) Constant 0.00382 0.00349 0.01390*** 0.03241*** 0.01359*** 0.02943** 0.00772* 0.01035 0.01907*** 0.04246*** 0.01809*** 0.03753** (0.00382) (0.01076) (0.00413) (0.01139) (0.00428) (0.01204) (0.00428) (0.01342) (0.00465) (0.01406) (0.00478) (0.01517) Observations 859 679 859 679 859 679 859 680 859 680 859 680 R-squared 0.08984 0.00269 0.16544 0.10198 0.19006 0.10616 0.10317 0.00487 0.17379 0.10052 0.20137 0.10464 DFIs = domestic finance institutions. Notes: sreturns3 = spot returns 3-year. sreturns10 = spot returns 10-year. freturn3 = future returns 3-year. freturn10 = future returns 10-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 31 Table 14: Postcrisis Estimates (Expected and Unexpected) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Variables sreturns3 sreturns10 sreturns3 sreturns10 sreturns3 sreturns10 freturn3 freturn10 freturn3 freturn10 freturn3 freturn10 Unexpected futures foreign 0.01343*** 0.00335 0.00637** –0.00017 0.01626*** –0.00107 0.00855*** –0.00788 (0.00207) (0.00693) (0.00251) (0.00638) (0.00226) (0.00801) (0.00271) (0.00758) Lag unexpected futures foreign –0.00720* –0.00118 –0.00226 0.00271 –0.01237*** –0.00358 –0.00840* 0.00506 (0.00380) (0.00765) (0.00421) (0.00834) (0.00422) (0.00954) (0.00458) (0.01018) Expected futures foreign 0.02808** –0.01663 0.01135 –0.00625 0.04334*** –0.01781 0.02963** –0.00736 (0.01149) (0.04789) (0.01278) (0.04251) (0.01278) (0.05636) (0.01397) (0.05250) Lag expected futures foreign –0.01429 0.01448 –0.00763 –0.01496 –0.02417** 0.02480 –0.01963* –0.01221 (0.00886) (0.03464) (0.00935) (0.04534) (0.00990) (0.04249) (0.01026) (0.05758) Unexpected spot foreign 0.00225 –0.00426 0.00290 –0.00347 0.00347** –0.00415 0.00429** –0.00424 (0.00149) (0.00770) (0.00183) (0.00753) (0.00176) (0.00908) (0.00207) (0.00899) Lag unexpected spot foreign 0.00357 0.00592 0.00225 0.00593 0.00297 0.01043 0.00124 0.00981 (0.00220) (0.00931) (0.00225) (0.00757) (0.00249) (0.01251) (0.00253) (0.00970) Expected spot foreign –0.02250* 0.01389 –0.01085 0.01476 –0.01749 0.00523 –0.00384 0.00831 (0.01345) (0.04488) (0.01440) (0.03354) (0.01494) (0.05671) (0.01594) (0.04189) Lag expected spot foreign 0.00861 –0.00980 0.00029 –0.01714 0.00444 –0.00289 –0.00491 –0.01355 (0.01015) (0.03174) (0.00989) (0.02739) (0.01150) (0.04082) (0.01094) (0.03440) Unexpected futures DFIs –0.00728*** 0.00224 0.05042*** 0.00010 –0.00937*** 0.00937 0.05345*** 0.03670 (0.00250) (0.00566) (0.01569) (0.03758) (0.00270) (0.00680) (0.01790) (0.05030) Lag unexpected futures DFIs 0.00387 –0.00115 0.03263 –0.05408** 0.00864** –0.00300 0.04254 –0.09296*** (0.00385) (0.00614) (0.02716) (0.02117) (0.00418) (0.00793) (0.03128) (0.02598) Expected futures DFIs –0.01755 0.00041 0.06854 –0.02906 –0.03256*** 0.00554 0.02153 –0.02794 (0.01097) (0.03171) (0.06942) (0.10737) (0.01206) (0.03963) (0.08224) (0.13595) Lag expected futures DFIs 0.01358* 0.00701 0.05536 0.10030 0.02371*** 0.00414 0.06659 0.08346 (0.00824) (0.02987) (0.06638) (0.08931) (0.00898) (0.03911) (0.07779) (0.11546) Unexpected spot DFIs –0.00600*** –0.01323*** –0.00602*** –0.01280*** –0.00664*** –0.01647*** –0.00669*** –0.01575*** (0.00106) (0.00225) (0.00109) (0.00223) (0.00115) (0.00273) (0.00117) (0.00271) Lag unexpected spot DFIs 0.00268** 0.00117 0.00235** 0.00100 0.00299** 0.00252 0.00277** 0.00192 (0.00107) (0.00240) (0.00113) (0.00236) (0.00119) (0.00302) (0.00124) (0.00298) Expected spot DFIs –0.02613*** –0.00324 –0.02196** –0.00472 –0.02455*** 0.00155 –0.02138** –0.00040 (0.00843) (0.00917) (0.00929) (0.00934) (0.00933) (0.01152) (0.01005) (0.01164) Lag expected spot DFIs 0.01364* –0.00149 0.01131 –0.00061 0.01174 –0.00800 0.01015 –0.00641 (0.00750) (0.00737) (0.00858) (0.00731) (0.00830) (0.00942) (0.00916) (0.00932) Lag cross market returns 0.04906 –0.01879 –0.02726 –0.01843 –0.04295 –0.01890 0.06957 0.02310 –0.00301 0.03285 –0.01981 0.02659 (0.04677) (0.04040) (0.04714) (0.03890) (0.05210) (0.04003) (0.05863) (0.06358) (0.05888) (0.06102) (0.06504) (0.06204) Constant 0.01686 0.00559 0.02791*** 0.02015 0.03406*** 0.02213 0.02112 0.01819 0.03322*** 0.03586 0.03812*** 0.03936 (0.01189) (0.01756) (0.00807) (0.01843) (0.01254) (0.02351) (0.01370) (0.02191) (0.00927) (0.02244) (0.01427) (0.02926) Observations 676 528 676 528 676 528 676 528 676 528 676 528 R-squared 0.11216 0.00638 0.18809 0.09938 0.21397 0.11676 0.12992 0.00679 0.19921 0.10300 0.22587 0.13013 DFIs = domestic finance institutions. Notes: sreturns3 = spot returns 3-year. sreturns10 = spot returns 10-year. freturn3 = future returns 3-year. freturn10 = future returns 10-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
32 | ADB Economics Working Paper Series No. 427 What these findings imply is that, over the sample period, the trading patterns of investor groups have evolved. The pricing effect of foreign participation is consistently significant and positive in both 3-year cash and futures markets across all subperiods, and particularly high during the crisis period. During the latter, both the expected and unexpected components of foreign transactions exert strong influence on prices, but the impact of the expected component is substantially higher than that of the unexpected, and visibly increases compared to pre and postcrisis periods. This suggests very active foreign trading in response to announcements in global and domestic macroeconomic conditions. On the other hand, foreign participation in the cash market has a negligible, if any, price effect. In addition, the pricing effect of foreign participation does not seem to be significant for 10-year cash and futures returns, reflecting relatively inactive foreign trading in the 10-year KTB futures. Fifth, across subgroups of DFIs (Tables 15 and 16), different groups show different trading patterns in both markets. In futures markets, apart from the significant effect of foreigners’ net positions, net positions of asset management funds and DFIs played an important role in pricing with independent trading decisions. Banks and securities companies seem to react to foreign investors' trading, as their coefficients are significant only when DFI subgroups are run, but their effects disappear if foreign trading is included. In futures markets, domestic securities companies account for the largest share (about 30% in 2004 and 60% in 2013), while banks are also sizable.18 But when compared to foreign investors,19 DFIs are fragmented; that is, they consist of a large number of smaller accounts, which dilutes their collective influence on pricing. This could explain the weaker price effects of banks and security companies in the futures market compared to foreigners. In contrast, banks and funds appear to be dominant players in the cash markets, and their transactions have a significant impact on cash prices. In addition, the pricing effects of different DFI subgroups are different for different bond tenors. Domestic asset management and funds are longterm investors, exerting a significant, large influence on 10-year KTBs, and then affect the futures prices through hedging. In the 10-year KTBs, the expected components of these subgroups are very important pricing factors. However, these effects are neither very consistent nor significant across specifications (Table 16). 18 Banks have hedging demand for their cash bond holdings and are likely to take a large short position in the futures market, based on a similar market view through relatively long-term market analysis; similarly, domestic securities companies may take a short position for hedging. 19 Usually, it is only several large foreign investors who make very large investments and drive market prices.
Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets | 33 Table 15: Full Sample by Composition of Domestic Finance Institutions (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Variables sreturns3 sreturns10 sreturns3 sreturns10 sreturns3 sreturns10 freturn3 freturn10 freturn3 freturn10 freturn3 freturn10 Futures foreign 0.01359*** 0.00126 0.01785*** –0.01915 0.01629*** –0.00326 0.01997*** –0.02535 (0.00105) (0.00589) (0.00549) (0.01735) (0.00117) (0.00689) (0.00601) (0.02930) Lag futures foreign –0.00191* –0.00213 0.00183 0.01918 –0.00304*** –0.00487 0.00261 0.00449 (0.00098) (0.00642) (0.00483) (0.02727) (0.00109) (0.00842) (0.00529) (0.04001) Spot foreign 0.00080 0.00156 0.00055 –0.00232 0.00093 0.00326 0.00069 –0.00174 (0.00115) (0.00661) (0.00137) (0.00666) (0.00132) (0.00781) (0.00152) (0.00775) Lag spot foreign –0.00161 0.00698 –0.00130 0.00393 –0.00169 0.00940 –0.00137 0.00527 (0.00101) (0.00573) (0.00112) (0.00508) (0.00113) (0.00798) (0.00124) (0.00740) Futures bank –0.00903*** 0.00252 0.00858 –0.01416 –0.01100*** 0.00793 0.00871 –0.01498 (0.00146) (0.00566) (0.00570) (0.01661) (0.00164) (0.00702) (0.00624) (0.02797) Lag futures bank 0.00102 0.00439 0.00270 0.02180 0.00167 0.00691 0.00409 0.01090 (0.00127) (0.00552) (0.00481) (0.02694) (0.00141) (0.00692) (0.00526) (0.03903) Futures asset management –0.00612 0.00366 0.01141* –0.01175 –0.00888** –0.01992 0.01074 –0.04163 (0.00382) (0.02693) (0.00677) (0.03103) (0.00408) (0.03351) (0.00727) (0.04440) Lag futures assets management 0.00635** 0.01171 0.00795 0.02939 0.00927*** 0.02997 0.01163** 0.03452 (0.00273) (0.02559) (0.00535) (0.03758) (0.00298) (0.03264) (0.00589) (0.05257) Futures funds –0.04754*** –0.02035** –0.02969*** –0.03735** –0.05369*** –0.01888 –0.03369*** –0.04159 (0.00544) (0.00907) (0.00766) (0.01816) (0.00567) (0.01179) (0.00813) (0.03002) Lag futures funds –0.00202 0.01172 0.00050 0.02918 –0.00005 0.01826* 0.00329 0.02264 (0.00407) (0.00825) (0.00634) (0.02710) (0.00444) (0.01024) (0.00691) (0.03982) Futures securities –0.01023*** 0.00084 0.00765 –0.01567 –0.01286*** 0.00656 0.00713 –0.01538 (0.00135) (0.00588) (0.00569) (0.01682) (0.00150) (0.00824) (0.00627) (0.02932) Lag futures securities 0.00146 0.00529 0.00355 0.02133 0.00247 0.00746 0.00535 0.01055 (0.00141) (0.00628) (0.00510) (0.02670) (0.00150) (0.00824) (0.00551) (0.03998) Spot banks –0.00419*** –0.00802*** –0.00411*** –0.00814*** –0.00436*** –0.00714** –0.00429*** –0.00758** (0.00079) (0.00282) (0.00082) (0.00281) (0.00087) (0.00347) (0.00090) (0.00347) Lag spot banks 0.00039 0.00057 0.00042 0.00057 0.00061 0.00180 0.00063 0.00174 (0.00071) (0.00261) (0.00071) (0.00260) (0.00080) (0.00355) (0.00081) (0.00351) Spot asset management –0.00164 –0.00919 –0.00168 –0.00996 –0.00159 –0.01013 –0.00163 –0.01050 (0.00107) (0.00812) (0.00108) (0.00800) (0.00117) (0.00944) (0.00119) (0.00938) Lag spot asset management 0.00078 0.01769** 0.00069 0.01784** 0.00111 0.02112** 0.00101 0.02111** (0.00103) (0.00752) (0.00103) (0.00757) (0.00113) (0.00916) (0.00113) (0.00926) Spot funds –0.00871*** –0.01697*** –0.00866*** –0.01715*** –0.00956*** –0.02184*** –0.00950*** –0.02186*** (0.00155) (0.00225) (0.00155) (0.00227) (0.00168) (0.00283) (0.00167) (0.00289) Lag spot funds 0.00166 0.00369** 0.00153 0.00400** 0.00196 0.00635*** 0.00181 0.00669*** (0.00124) (0.00184) (0.00123) (0.00180) (0.00136) (0.00236) (0.00134) (0.00230) Lag cross market returns 0.01718 –0.00561 0.01109 0.00043 0.00680 0.00084 –0.00031 0.03190 –0.00276 0.05738 –0.00837 0.05619 (0.02843) (0.03866) (0.02894) (0.03941) (0.02892) (0.03916) (0.03488) (0.05812) (0.03610) (0.05933) (0.03610) (0.05950) Constant 0.00072 0.00349 0.00538* 0.03592*** 0.00589* 0.03518*** 0.00497 0.01035 0.00971*** 0.04937*** 0.01014*** 0.04771*** (0.00328) (0.01076) (0.00326) (0.01175) (0.00339) (0.01271) (0.00361) (0.01342) (0.00363) (0.01432) (0.00376) (0.01566) Observations 2,067 679 2,067 679 2,067 679 2,066 680 2,066 680 2,066 680 R-squared 0.08978 0.00269 0.15199 0.13429 0.15748 0.13843 0.10154 0.00487 0.16178 0.14618 0.16729 0.14877 Notes: sreturns3 = spot returns 3-year. sreturns10 = spot returns 10-year. freturn3 = future returns 3-year. freturn10 = future returns 10-year. Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1. Source: Authors' estimates.
34 | ADB Economics Working Paper Series No. 427 Table 16: Full Sample by Composition of Domestic Finance Institutions (Expected and Unexpected) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Variables sreturns3 sreturns10 sreturns3 sreturns10 sreturns3 sreturns10 freturn3 freturn10 freturn3 freturn10 freturn3 freturn10 Unexpected futures foreign 0.01273*** 0.00335 0.01332** 0.00889 0.01561*** –0.00107 0.01555** 0.02865 (0.00117) (0.00693) (0.00589) (0.03197) (0.00132) (0.00801) (0.00658) (0.04269) Lag unexpected futures foreign –0.00357*** –0.00118 0.00437 –0.04515* –0.00497*** –0.00358 0.00494 –0.07797** (0.00137) (0.00765) (0.00575) (0.02522) (0.00152) (0.00954) (0.00645) (0.03072) Expected futures foreign 0.02458*** –0.01663 0.01889** 0.01015 0.02823*** –0.01781 0.02038* 0.05166 (0.00603) (0.04789) (0.00938) (0.04870) (0.00678) (0.05636) (0.01056) (0.05995) Lag expected futures foreign –0.01130** 0.01448 –0.00223 –0.07664* –0.01310** 0.02480 –0.00116 –0.12023** (0.00571) (0.03464) (0.00918) (0.04332) (0.00634) (0.04249) (0.01046) (0.05255) Unexpected spot foreign 0.00068 –0.00426 0.00035 –0.00834 0.00099 –0.00415 0.00064 –0.00881 (0.00149) (0.00770) (0.00167) (0.00771) (0.00170) (0.00908) (0.00189) (0.00926) Lag unexpected spot foreign –0.00113 0.00592 –0.00066 0.00951 –0.00174 0.01043 –0.00131 0.01341 (0.00151) (0.00931) (0.00170) (0.00768) (0.00172) (0.01251) (0.00193) (0.00989) Expected spot foreign 0.00595 0.01389 0.00499 –0.00374 0.00982 0.00523 0.00914 –0.01418 (0.00823) (0.04488) (0.00902) (0.03501) (0.00935) (0.05671) (0.01017) (0.04331) Lag expected spot foreign –0.00437 –0.00980 –0.00984 –0.02324 –0.00584 –0.00289 –0.01105 –0.01852 (0.00834) (0.03174) (0.00836) (0.02769) (0.00951) (0.04082) (0.00948) (0.03426) Unexpected futures banks –0.00827*** 0.00513 0.00496 0.01285 –0.01058*** 0.01190* 0.00480 0.03797 (0.00160) (0.00565) (0.00608) (0.03075) (0.00185) (0.00711) (0.00682) (0.04115) Lag unexpected futures banks 0.00236 0.00162 0.00705 –0.04728* 0.00336** 0.00274 0.00847 –0.07622** (0.00155) (0.00811) (0.00574) (0.02756) (0.00171) (0.00926) (0.00645) (0.03372) Expected futures banks –0.02251* –0.01258 –0.00890 –0.04556 –0.02711** –0.00004 –0.01075 –0.02594 (0.01182) (0.04954) (0.01290) (0.06194) (0.01271) (0.06196) (0.01402) (0.07655) Lag expected futures banks 0.01915* 0.00931 0.02213* –0.02014 0.01978 0.00401 0.02351 –0.05100 (0.01150) (0.03206) (0.01299) (0.03726) (0.01263) (0.03904) (0.01445) (0.04510) Unexpected futures asset management –0.00708* 0.01899 0.00618 0.01947 –0.00991** –0.00387 0.00556 0.01101 (0.00410) (0.03060) (0.00710) (0.04307) (0.00442) (0.03767) (0.00779) (0.05594) Lag unexpected futures asset management 0.00894*** 0.05889 0.01308** 0.01382 0.01289*** 0.09351** 0.01743** 0.01976 (0.00292) (0.03692) (0.00633) (0.04673) (0.00321) (0.04638) (0.00717) (0.05801) Expected futures asset management 0.00027 –0.22616 0.01026 –0.18510 0.01164 –0.57024** 0.02115 –0.49447* (0.02467) (0.22493) (0.02561) (0.23863) (0.02666) (0.25127) (0.02786) (0.26724) Lag expected futures asset management 0.00798 –0.04250 0.01706 –0.06975 –0.00044 0.00827 0.00848 –0.05308 (0.02344) (0.16501) (0.02411) (0.16426) (0.02629) (0.19763) (0.02703) (0.19580) Unexpected futures funds –0.04266*** –0.01406 –0.02930*** –0.00862 –0.04846*** –0.01190 –0.03290*** 0.01184 (0.00532) (0.00968) (0.00772) (0.03234) (0.00558) (0.01233) (0.00839) (0.04326) continued on next page
ASIAN DEVELOPMENT BANK AsiAn Development BAnk 6 ADB Avenue, Mandaluyong City 1550 Metro Manila, Philippines www.adb.org Price Discovery and Foreign Participation in the Republic of Korea’s Government Bond Cash and Futures Markets The authors assess the impact of foreign participation in Korean Treasury Bond (KTB) cash and futures markets and their role in the price discovery process. Using daily data from the over-the-counter market for cash and the Korea Exchange for futures transactions, the results show that foreign trading in the KTB futures market leads the price discovery process for the underlying bonds. Specifically, foreigners’ daily net long positions in the futures market exert significant influence in both KTB cash and futures prices. The empirical findings also indicate that it is the unexpected component of foreign investors’ net long futures positions that explains a significant share of the pricing effects. About the Asian Development Bank ADB’s vision is an Asia and Pacific region free of poverty. Its mission is to help its developing member countries reduce poverty and improve the quality of life of their people. Despite the region’s many successes, it remains home to approximately two-thirds of the world’s poor: 1.6 billion people who live on less than $2 a day, with 733 million struggling on less than $1.25 a day. ADB is committed to reducing poverty through inclusive economic growth, environmentally sustainable growth, and regional integration. Based in Manila, ADB is owned by 67 members, including 48 from the region. Its main instruments for helping its developing member countries are policy dialogue, loans, equity investments, guarantees, grants, and technical assistance. PRiCe DisCoveRy AnD FoReiGn PARtiCiPAtion in the RePuBliC oF KoReA’s GoveRnMent BonD CAsh AnD FutuRes MARKets Cyn-Young Park; Rogelio Mercado, Jr.; Jaehun Choi; and Hosung Lim adb economics working paper series no. 427 march 2015