International Review of Economics and Finance 64 (2019) 412–426

Contents lists available at ScienceDirect

International Review of Economics and Finance

journal homepage: www.elsevier.com/locate/iref

Volatility information trading in the index options market: An

intraday analysis
Heejin Yang a, Ali M. Kutan b, Doojin Ryu c, *
a
Department of Global Economics and Commerce, College of Management and Economics, Dongguk University Gyeongju Campus, Gyeongsangbuk-do,
South Korea
b
School of Business, Southern Illinois University Edwardsville, USA
c
College of Economics, Sungkyunkwan University, 25-2, Sungkyunkwan-ro, Jongno-gu, Seoul, 03063, South Korea

A R T I C L E I N F O A B S T R A C T

JEL classification: By analyzing intraday volatility information trading according to the demand for options, we
G10 determine the types of investors that are informed about future spot market volatility and conduct
G14 volatility information trading in a highly liquid options market. Although the overall aggregate
Keywords: options demand does not predict intraday market volatility, the vega-weighted net demand of
Foreign investment firm foreign investment firms conveys significant information about future volatility dynamics. By
Market microstructure
tracking the positions of all options market participants according to option moneyness, we find
Implied volatility
that foreign investment firms conduct volatility trading using highly levered options and that their
Intraday volatility
Volatility information intraday volatility information superiority is more prominent when they close their existing
positions.

1. Introduction

Many studies in financial economics have analyzed the return predictability of information embedded in options trades. These
studies examine, in various settings, whether options trades predict spot market returns and, consequently, whether options investors
have directional information. Although most empirical studies provide evidence supportive of directional informed trading in global
options markets (Cao, Chen, & Griffin, 2005; Chang, Hsieh, & Lai, 2013; Easley, O'Hara, & Srinivas, 1998; Ge, Lin, & Pearson, 2016; Hu,
2014; Johnson & So, 2012; Lin, Tsai, Zheng, & Qiao, 2018; Pan & Poteshman, 2006; Roll, Schwartz, & Subrahmanyam, 2010; Ryu &
Yang, 2018), a significant number of studies reach the opposite conclusion (Chan, Chung, & Fong, 2002; Fahlenbrach & Sandås, 2010;
Ryu, 2015; Schlag & Stoll, 2005). Thus, the information role of options trading remains an empirical question and merits further analysis
from a different perspective.
Investors with superior information about the dynamics of market volatility may benefit from options trading more than investors
with directional information. By predicting underlying asset price movements, the “directional informed traders” can profit from both
equity and options trading. However, the “volatility informed traders” would profit only from trading nonlinear derivative securities,
such as options, rather than stocks or futures contracts with linear payoff structures. Furthermore, options investors often delta-hedge
their exposures using the underlying assets and hedge their exposure to directional movements of the underlying assets. These hedging
activities might have a confounding effect when the directional information content of options trading is measured.
As such, several recent options market studies have focused on the options demands of investors with volatility information and

* Corresponding author.
E-mail address: sharpjin@skku.edu (D. Ryu).

https://doi.org/10.1016/j.iref.2019.07.006
Received 1 August 2018; Received in revised form 13 July 2019; Accepted 24 July 2019
Available online 26 July 2019
1059-0560/© 2019 Elsevier Inc. All rights reserved.
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

analyzed the resulting volatility information trading in options markets. For example, a series of studies on the US options market
(Bollen & Whaley, 2004; Gharghori, Maberly, & Nguyen, 2017; Holowczak, Hu, & Wu, 2014; Lai, 2017; Le & Zurbruegg, 2016; Ni, Pan,
& Poteshman, 2008) and Asian markets (Chang, Hsieh, & Wang, 2010; Chen, Chung, & Yuan, 2014) show that net demand for volatility
partly predicts the underlying assets’ future volatilities, whereas other studies find contrasting results (Chiang, Chung, & Louis, 2017).
Rourke (2014) finds that vega-informed near-the-money options trading is more commonplace than delta-informed near-the-money
options trading in the US options market. However, the literature on volatility information in the options market is relatively scarce.
Furthermore, previous studies focus on specific options trading strategies and fail to address the characteristics of volatility information
trading systematically. Because market volatility characteristics are essential determinants of option price dynamics and the trading
motives of option market participants, volatility information trading should be further examined using high-quality and high-frequency
data.
Thus, we examine whether intraday options trades provide information beyond directional movements, especially volatility in-
formation, and enable sophisticated trading based on this information. We use a microstructure dataset from the KOSPI200 index
options to conduct a detailed analysis. The KOSPI200 options market is highly liquid with little friction and low transaction costs, and its
underlying spot market is more volatile than those in developed markets.1 As such, volatility informed investors have a strong incentive
to use options contracts as their trading vehicle. Further, previous studies detect volatility information in options trades under restrictive
conditions (e.g., specific volatility strategies, such as straddle and opening trades), because they use data with a lower frequency, such as
daily trades. However, in highly speculative and informationally efficient index options markets, informed investors could not
consistently outperform others over the time horizon of a day (Chung, Park, & Ryu, 2016; Ryu, 2011; Ryu & Yang, 2017). Thus, we
analyze the intraday, high-quality dataset of KOSPI200 options, the use of which enables the exact classification of each buy/sell order,
which is necessary to construct the net volatility demand measure. Furthermore, the detailed information in the trade and quote data,
such as the full identification of investors and their real-time options positions, allows us to examine aspects that previous studies fail to
uncover.
Our study contributes to the volatility information trading literature by exploiting a high-quality, informative dataset. The identi-
fication of options market participants and their options positions per order and trade allow us to determine who in the options market
has volatility information and, by examining their net demand for options, how superior investors act on volatility information. We
calculate intraday options demand for volatility and investigate whether net options demand predicts intraday market volatility. Most
previous studies, including Bae and Dixon (2018), Chang et al. (2010), Ni et al. (2008), and Ryu and Yang (forthcoming), consider only
daily net demand and volatility. Professional investors can be informed by processing market-wide and/or public information faster
than their index derivative competitors (Ahn, Kang, & Ryu, 2008; Ryu, 2016), but they must quickly act on volatility information to
make significant profits. The inconsistent evidence of volatility information trading in previous studies may be caused by their use of
lower-frequency measures and lower-quality datasets. To address this, we conduct intraday analyses and clarify which intraday trades
predict future market volatility.
Previous studies also use imprecise volatility measures. Chang et al. (2010), Ni et al. (2008), and Ryu, Ryu, and Yang (forthcoming)
define daily volatility of an underlying asset as the highest intraday price minus the lowest intraday price during each trading day.
However, this daily volatility measure, which is a range-based volatility proxy, and other potential volatility measures, such as
five-minute intraday return changes, fail to characterize future market volatility or capture actual asset price variances. Use of these
realized volatility constructs in our intraday analyses requires using higher-frequency observations in the calculations because
tick-by-tick price changes must be used to construct the intraday volatility series. This yields a noisy volatility measure because of
microstructural biases such as bid-ask bounce problems and temporary illiquidity. Consequently, we employ the intraday time series of
model-free implied volatility, which is similar to the US volatility index (VIX). The intraday implied volatility is more appropriate for
examining volatility information conveyed by options trades, because it more clearly reflects market participants’ aggregate expecta-
tions and opinions of future states and volatility, and its changes are directly related to the intraday flow of volatility information
(Banerjee, Doran, & Peterson, 2007; Giot, 2005; Guo & Whitelaw, 2006; Lee & Ryu, 2019). Substantial information about future realized
volatility is contained in the implied volatility index, which can effectively predict future volatility dynamics. As with options volume,
the volatility index provides critical option-implied information (Busch, Christensen, & Nielsen, 2011; Jiang & Tian, 2005; Song, Ryu, &
Webb, 2018).
To the best of our knowledge, no study has rigorously examined intraday volatility information trading using an options market
microstructure dataset, despite its importance and implications. The vega-weighted net demand for volatility is calculated by investor
type, option moneyness, and open-interest changes using option investors' net positions to examine whether quantity information
contained in options trades can predict the intraday volatility dynamics of the underlying assets. Our empirical findings from the an-
alyses on the KOSPI200 options intraday dataset are as follows. First, the aggregate options market demand for volatility does not
predict the underlying asset's future volatility, indicating that volatility information contained in aggregate options volume is noisy or
contaminated. Second, domestic investors generally have little volatility information, but trades by foreign institutional investors
(especially foreign investment firms) significantly predict intraday volatility dynamics, indicating foreign investors are more sophis-
ticated and better informed than their domestic counterparts. Third, foreign investment firms use their intraday volatility information
advantage to trade highly levered options, which are speculative and liquid contracts. Fourth, while previous research finds directional
information is more prominent in opening trades, orders by skillful foreign institutions to close existing positions convey volatility
information, suggesting closing trades are initiated by more sophisticated, strategic orders. These findings are robust when the strong

1
The detailed characteristics of the KOSPI200 options market and the properties of options trading are explained in the next section.

413
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

serial correlation of the intraday implied volatility series is controlled.

The rest of this paper proceeds as follows. Section 2 describes the characteristics of the KOSPI200 options market and explains why it
is chosen for this study; we also discuss informed trading issues in the options market. Section 3 describes the construction of the study's
sample and the predictive regression model used to examine the information content of vega-weighted volumes by investor and trade
type. Section 4 presents our empirical findings and robustness checks. Finally, Section 5 concludes the study.

2. KOSPI200 options market and informed trading

The most representative index derivatives of the Korea Exchange (KRX), KOSPI200 options were introduced about 20 years ago; its
options market is one of the most liquid and popular options markets worldwide (Yang, Ryu, & Ryu, 2018). Although its trading volume
has decreased due to recent interventions by both the Korean government and the administrative office that controls speculation and
noisy trading, trading by global and local derivatives traders remains active. The market's small bid-ask spreads, absence of brokerage
and exchange fees, low trading and capital gains taxes, and substantial market depth result in low transaction costs and few market
frictions, creating a market with high liquidity and active investor participation. The quoted bid-ask spreads usually equal the minimum
tick size, and there is sufficient market depth to promptly absorb adverse price movements and temporary price impacts caused by large
trade orders. Informed investors are encouraged by its low fees and tax rates to select KOSPI200 options as a profitable trading vehicle.
These properties increase the reliability of our estimation results, which are based on a sample of intraday quotes and trades.
The investor composition of the KOSPI200 options market enables us to investigate the trading behaviors of diverse investor types.
There is balanced investor participation in this market, while other developed derivatives markets are dominated by institutional in-
vestors. Local and foreign traders, as well as individual and institutional investors, take balanced positions in KOSPI200 options trading.
Options investors are classified in our raw microstructure dataset as individual investors, financial investment companies (including
securities and futures companies), insurance companies, investment trusts, banks, pension funds, and government-owned firms. We use
this classification and investor nationality to categorize options market participants into five investor groups: foreign investment firms,
other (non-investment) foreign institutions, domestic investment firms, other (non-investment) domestic institutions, and domestic
individuals. Table 1 reports the trading activities in the KOSPI200 options market by investor type; trades by domestic individual in-
vestors account for about one-third of the options market trading volume during the sample period of 2010–2014. The table also shows
that financial investment companies conduct most of the domestic institutional trades in the KOSPI200 options market. Trades by other
domestic institutional investors, such as insurance companies, investment trusts, banks, pension funds, and government-owned firms,

Table 1
Trading volume by investor type.
Foreign Invest. Foreign Non-invest. Domestic Invest. Domestic Non-invest. Individuals

Panel A. Call options

2010 Buy 23,030,845 329,211,014 267,011,720 15,406,462 204,543,074
Sell 22,879,772 331,391,337 276,175,730 17,901,461 224,679,792
2011 Buy 52,890,241 416,549,685 239,973,039 7,635,513 215,709,851
Sell 53,609,561 424,224,240 249,940,109 8,357,366 241,843,749
2012 Buy 43,251,093 173,131,203 92,460,211 3,024,838 78,424,476
Sell 43,511,001 175,846,805 96,908,122 3,195,617 88,252,543
2013 Buy 17,954,637 63,500,357 29,969,358 1,220,056 30,728,772
Sell 17,802,778 64,316,863 31,509,784 1,166,174 34,987,632
2014 Buy 3,793,739 22,962,479 13,036,054 373,520 11,709,023
Sell 3,708,392 23,407,609 13,777,025 404,832 13,698,091
Total 282,432,059 2,024,541,592 1,310,761,152 58,685,839 1,144,577,003
(5.86%) (41.99%) (27.19%) (1.22%) (23.74%)

Panel B. Put options

2010 Buy 21,160,032 386,940,268 199,214,565 21,063,969 200,766,162
Sell 21,114,476 392,673,438 210,104,625 24,349,749 221,900,132
2011 Buy 52,779,567 408,319,307 158,424,685 6,656,399 182,664,491
Sell 53,222,517 415,764,161 166,704,457 7,534,811 204,651,263
2012 Buy 49,171,067 165,188,832 65,024,791 3,346,547 73,122,225
Sell 48,836,825 167,335,674 69,474,861 3,506,315 81,800,011
2013 Buy 20,772,131 59,439,567 20,142,715 1,171,004 28,468,924
Sell 20,360,571 60,012,473 22,028,171 1,205,857 32,240,316
2014 Buy 4,624,758 24,870,073 9,568,046 367,369 11,163,509
Sell 4,359,601 25,001,396 10,394,861 388,655 12,945,723
Total 296,401,545 2,105,545,189 931,081,777 69,590,675 1,049,722,756
(6.66%) (47.29%) (20.91%) (1.56%) (23.58%)

Note: This table presents the trend in trading volumes of KOSPI200 call and put options for five investor types: foreign investment firms (Foreign
Invest.), foreign non-investment firms (Foreign Non-invest.), domestic investment firms (Domestic Invest.), domestic non-investment firms (Domestic Non-
invest.), and domestic individuals (Individuals). The sample period is January 2010 to June 2014. Buy and Sell denote buy and sell options volumes,
respectively. Trading volume is presented as the number of options contracts. Figures in parentheses are percentages.

414
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

represent only about 1% of all options trading volume.2 Although options trading volumes have decreased since the 2012 market re-
form, Table 1 shows that the trading proportions of domestic individuals, domestic institutions, and foreign institutions are relatively
unchanged. Trades by individuals are uninformed and noisy, while those made by institutions are better informed and more sophis-
ticated; moreover, given the intense debate on whether domestic or foreign investors are informed, the investor composition in the
KOSPI200 options market provides an ideal setting for examining investor type differences.
The demand concentration and structure of the KOSPI200 index options market are also useful features for performing our
microstructure analyses. In Korea, individual options contracts are traded over-the-counter (OTC), and their trading volumes are quite
small. Although other options-like assets such as equity-linked securities (ELWs) are listed on the KRX, few have been traded since the
ELW market breakdown in 2010. Therefore, investor demands for derivative assets with non-linear payoff structures are concentrated in
the KOSPI200 options market, and investors with volatility information are likely to choose index options to take advantage of their
information. The KOSPI200 options market is purely order-driven, with no designated market makers. This simple market structure
makes it possible to trace the types of investors submitting orders, as differentiation between market makers and non-market makers is
unnecessary (Park & Ryu, 2019).
Motivated by the institutional features, structure, and high-quality dataset of the KOSPI200 options market, several studies inves-
tigate whether options market trading activities provide directional information about underlying asset returns. Although studies show
the existence of informed trading in this market (Ahn, Kang, and Ryu, 2010; Lee, Kang, & Ryu, 2015), no significant directional in-
formation content is detected in overall options volume. For example, Ryu (2015) compares the directional information content of
KOSPI200 futures and options trades and concludes that options trades are not informative, while futures trades contain significant
directional information, indicating that the aggregate options trading volume provides ambiguous information about the evolution of
spot returns. Bae and Dixon (2018) show that there is little information in options trades about expected spot returns and claim that
investors with directional information prefer futures market trading to the trading in the options market. Studies attribute the lack of
evidence of directional information to the noisy and uninformed traders in the emerging options market (Webb, Ryu, Ryu, & Han,
2016). These studies suggest that although a significant number of investors are informed, the directional information conveyed by
aggregate trading volume becomes noisy. However, in the index options market, information may be revealed through volatility in-
formation trading rather than directional information trading. To examine this previously unexplored possibility, we trace the “com-
plete” quote and transaction records of the KOSPI200 index options market and exploit the detailed investor identification information
contained in our dataset to determine which investors possess intraday volatility information.
The KOSPI200 options market opens at 9:00 a.m. and closes at 3:15 p.m. on regular trading days. The uniform pricing rule governs
transactions during the last 10 min of daily trading and 1 h immediately before the beginning of the daily session. During these two
periods (i.e., from 8:00 to 9:00 a.m. and from 3:05 to 3:15 p.m.), all orders submitted are accumulated in a centralized limit order book
and executed at a single market price during the last moments of each session. During the market's continuous trading session (from 9:00
a.m. to 3:05 p.m.), all submitted orders are electronically matched based on price and time priority. Though four different maturity
months classify each option series, only the nearest-maturity contracts are actively traded; longer-term contracts are rarely traded. A
“point” is the quoting unit of the KOSPI200 options market. Before the market reform in June 2012, one point corresponded to 100,000
Korean won (KRW) but increased fivefold to 500,000 KRW afterward.

3. Data and methodology

3.1. Sample data

We use model-free implied volatility (i.e., the VKOSPI) as a proxy for market volatility and extract a one-minute intraday sample.
Fig. 1 shows the sampled intraday patterns of the KOSPI200 spot index and the VKOSPI based on a one-minute sampling frequency.
Panel A (Panel C) exhibits their intraday dynamics of the first (last) trading day in our sample period (January 2010 through June 2014).
Panel B shows the dynamics at the first trading day of April 2010. Fig. 1 illustrates a negative relationship between spot index and
volatility, and also shows substantial intraday movements of the VKOSPI, reflecting the speed of information flows and sentiment
changes in the KOSPI200 spot and options markets.
Our analysis uses an ultra-fine microstructure dataset consisting of all quotes and trades in the KOSPI200 index options market from
January 2010 through June 2014. We use the following filtering procedure to alleviate concerns about market microstructural noises
and biases. First, to precisely identify buy and sell trades and consider post-trade volatility changes caused by incoming options orders,
transactions made and orders submitted during the continuous trading sessions are included in the final sample, and we exclude those
made during the pre-opening and closing call market periods. Second, options contracts with time-to-maturity of less than five calendar
days are excluded to mitigate the impact of the liquidity problem and the irregular trading behavior around options expiration dates.3
Third, options contracts with less than two contracts traded during any given day are excluded from that day's sample.4 Finally, all
options with quoted prices equal to the minimum tick size are removed to eliminate the effects of tick size restrictions.

2
When measuring investors' volatility information content in each moneyness or trade category, trades of these other domestic institutions
represent a tiny portion of total options trades. Therefore, we exclude these non-investment domestic firms from the analyses.
3
If the five calendar days include weekends and/or holidays, periods of fewer than five trading days are excluded. As a robustness check, we also
include these near-maturity periods and exclude only the exact maturity dates. However, the empirical results are qualitatively similar.
4
Our overall conclusions remain unchanged when this filtering process is not used.

415
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Fig. 1. Intraday patterns of the KOSPI200 and VKOSPI.

416
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

This dataset offers several advantages for our research. First, it contains an extensive and accurate time-stamped history of all quotes
and transaction activity, such as trade date, option type, order type, transaction price, trading volume, and trade direction. Second, the
dataset contains the information needed to determine the investor type on both ends of each transaction (i.e., domestic or foreign;
individual or investment firm) and identify who submits each order. Our investor type classification is based on the assumption that the
predictive ability of options orders for future spot volatility differs depending on the trading purpose and information processing ability
of the submitting investor. Third, and most importantly, our dataset provides full investor identification with investor accounts, enabling
us to track the positions of all options traders and classify each transaction as one that opens or closes an investor's position. Opening
(closing) trades are used when investors conduct options transactions to open (close) new (existing) positions.
Observations for the first three-month period (January to March 2010) are used to precisely trace investors’ positions on each
options contract. After April 2010, we begin measuring the information content of options volatility demand classified by options
market characteristics. Exploiting order and trade details in our dataset, we classify each options trade by initiating investor type, option
leverage, and trade type, including open-interest changes, to associate volatility information content with options market character-
istics. The degree of option leverage is measured by option moneyness, where the moneyness of a call (put) is the ratio of the underlying
(strike) to strike (underlying) price. An options contract is classified as deep out-of-the-money (DOTM) if its moneyness value is lower
than 0.955, out-of-the-money (OTM) if the value is between 0.955 and 0.985, at-the-money (ATM) when the value is between 0.985 and
1.015, in-the-money (ITM) between 1.015 and 1.045, and deep in-the-money (DITM) if its moneyness value exceeds 1.045.
Table 2 presents the summary statistics of options trades by investor type and option moneyness. The investor-moneyness categories
include the transaction prices, order sizes, number of transactions, number of contracts, trading values, and bid-ask spreads. Transaction
prices and order sizes are shown as per-order average values, while both daily average values and full-period total values are presented
for the number of transactions, the number of contracts, and trading values. The transaction price, order size measured in values, and
daily bid-ask spread monotonically increase as an option goes into the money. The average order sizes are 4.52, 8.6, 13.58, 20.37, and
52.23 points for DOTM, OTM, ATM, ITM, and DITM options, respectively, indicating that ITM options trades are generally larger than
ATM and OTM options trades. By contrast, trading volumes measured by the number of transactions, number of contracts, or trading
value tend to decrease with option moneyness. The bigger (smaller) trading volume and smaller (bigger) spread for OTM (ITM) options
indicate that the OTM options market is more liquid than the ITM options market.
In terms of the number of contracts (# of contracts (total)), 22.2% of total DOTM options trades are initiated by domestic individuals,
while only 8.79% of DITM options trades are initiated by that group. By contrast, trades initiated by foreign investment firms comprise
7.36% of DOTM options transactions, but 20.61% of DITM options trading. The figures in square brackets (parentheses) show the
relative trading proportions within each investor type (to total options transactions). For example, in terms of the number of contracts,
“[39.11%]” denotes the ratio of OTM options trading volume of foreign investment firms (173,920,135) to their total trading volume
(444,721,074), and “(2.96%)” is the ratio of OTM options trading volume of foreign investment firms (173,920,135) to all options
volume (5,869,884,122). These percentage values indicate that there are different participation rates for domestic and foreign investors
across different option leverage and moneyness categories. Because the information advantages, trading motives, and experiences of
each investor group may differ, the different investor compositions across option moneyness categories might imply differences in
information content and the effect of options trades across moneyness categories (Barber, Odean, & Zhu, 2009; Dorn, Huberman, &
Sengmueller, 2008; Erenburg, Kurov, & Lasser, 2006).

3.2. Variables and methodology

We construct the vega-weighted net options demand for volatility following Ni et al. (2008), to measure the volatility information
embedded in options trades. If an options investor has superior information about underlying market volatility and trades in the options
market on the basis of this information, the volatility demand change can be observed via the options trading volume. Since an option's
vega captures the sensitivity of the option price to the underlying volatility and is positive irrespective of the option type, her volatility
demand monotonically increases (deceases) if she buy (sells) calls or puts. Thus, for each option series, we calculate the net options
demand for volatility (Dσi ) in the ith intraday five-minute time interval for each trading day. Equation (1) shows that Dσi is computed
based on each tth option price observation in the ith interval:
X∂lnCt X∂lnPt
Dσi ¼ ðBuyCallt  SellCallt Þ þ ðBuyPutt  SellPutt Þ; (1)
t
∂σ t
∂σ

where Ct and Pt denote the tth option prices of call and put options with a specific strike price (K) and maturity (T), respectively. σ is the
volatility of the underlying asset.5 BuyCallt (SellCallt) represents the tth number of bought (sold) call contracts of a specific options series.
BuyPutt (SellPutt) denotes the number of bought (sold) put contracts. Through vega weighting, Equation (1) measures the aggregate
information on the underlying asset's volatility. Since the vegas of contracts differ depending on option prices, time-to-expiration, and
strike prices, the demand for each contract is weighted according to its return to the option's per unit change in volatility. We
approximate ∂lnCt/∂σ and ∂lnPt/∂σ using the respective Black–Scholes call and put vega values, as follows:

5
The volatility parameter (σ ) is estimated based on the historical volatility of the past 20 trading days using the one-minute intraday frequency
data of the underlying index.

417
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Table 2
Summary statistics.
Moneyness

All DOTM OTM ATM ITM DITM

Panel A. Overall
Price 1.74 0.66 1.69 3.70 7.89 18.49
Order size (in contracts) 6.39 10.4 6.04 3.72 2.53 2.82
Order size (in value) 8.11 4.52 8.60 13.58 20.37 52.23
# of transactions (daily) 689,705 203,003 300,366 176,305 8,871 1,159
# of contracts (daily) 5,569,150 2,449,788 2,265,477 821,646 28,105 4,134
Trading value (daily) 6,702,949 1,591,561 2,767,756 2,052,602 214,336 76,695
Bid–ask spread (daily) 0.02 0.01 0.02 0.04 0.10 0.35
# of transactions (total) 726,948,614 213,964,862 316,585,461 185,826,366 9,350,054 1,221,871
# of contracts (total) 5,869,884,122 2,582,076,032 2,387,813,107 866,015,371 29,622,755 4,356,857
Trading value (total) 7,064,908,525 1,677,504,923 2,917,214,930 2,163,442,175 225,909,926 80,836,571

Panel B. Foreign investment firms

Price 2.11 0.79 1.76 3.78 7.83 17.86
Order size (in contracts) 7.33 11.80 7.34 4.86 3.25 2.84
Order size (in value) 12.02 6.59 11.5 18.83 26.23 50.39
# of transactions (daily) 53,719 13,947 23,313 15,079 1187 195
# of contracts (daily) 421,937 180,269 165,010 70,780 5026 861
Trading value (daily) 662,628 149,457 240,020 216,144 41,224 15,948
Bid–ask spread (daily) 0.02 0.01 0.02 0.04 0.09 0.35
# of transactions (total) 56,619,802 14,699,752 24,572,141 15,893,001 1,251,198 203,710
[100%] [25.96%] [43.4%] [28.07%] [2.21%] [0.36%]
(7.79%) (2.02%) (3.38%) (2.19%) (0.17%) (0.03%)
# of contracts (total) 444,721,074 190,003,602 173,920,135 74,601,872 5,297,355 898,110
[100%] [42.72%] [39.11%] [16.77%] [1.19%] [0.2%]
(7.58%) (3.24%) (2.96%) (1.27%) (0.09%) (0.02%)
Trading value (total) 698,409,629 157,527,932 252,981,190 227,815,981 43,450,584 16,633,942
[100%] [22.56%] [36.22%] [32.62%] [6.22%] [2.38%]
(9.89%) (2.23%) (3.58%) (3.22%) (0.62%) (0.24%)

Panel C. Foreign non-investment firms

Price 1.85 0.69 1.72 3.79 7.94 18.59
Order size (in contracts) 6.29 8.52 6.25 3.86 2.46 2.72
Order size (in value) 8.93 4.79 9.36 14.17 19.81 50.76
# of transactions (daily) 347,850 104,892 144,908 91,010 6,243 798
# of contracts (daily) 2,719,434 1,139,760 1,126,042 431,446 19,374 2,811
Trading value (daily) 3,681,508 879,098 1,460,455 1,143,118 146,724 52,113
Bid–ask spread (daily) 0.02 0.01 0.02 0.04 0.10 0.35
# of transactions (total) 366,634,165 110,556,364 152,732,909 95,924,361 6,579,774 840,757
[100%] [30.15%] [41.66%] [26.16%] [1.79%] [0.23%]
(50.43%) (15.21%) (21.01%) (13.2%) (0.91%) (0.12%)
# of contracts (total) 2,866,283,361 1,201,307,348 1,186,848,527 454,744,166 20,420,623 2,962,697
[100%] [41.91%] [41.41%] [15.87%] [0.71%] [0.1%]
(48.83%) (20.47%) (20.22%) (7.75%) (0.35%) (0.05%)
Trading value (total) 3,880,309,523 926,568,914 1,539,319,835 1,204,846,453 154,646,908 54,927,413
[100%] [23.88%] [39.67%] [31.05%] [3.99%] [1.42%]
(54.92%) (13.12%) (21.79%) (17.05%) (2.19%) (0.78%)

Panel D. Domestic investment firms

Price 1.37 0.55 1.59 3.38 7.84 17.88
Order size (in contracts) 10.47 28.82 8.12 4.33 2.81 3.64
Order size (in value) 8.61 6.41 9.55 13.86 22.31 64.06
# of transactions (daily) 76,010 20,986 38,269 16,626 114 29
# of contracts (daily) 1,098,856 547,090 436,025 115,304 386 100
Trading value (daily) 873,264 224,665 425,506 219,134 2,738 1,891
Bid–ask spread (daily) 0.01 0.01 0.01 0.03 0.09 0.29
# of transactions (total) 80,114,123 22,119,026 40,335,816 17,523,706 113,519 22,056
[100%] [27.61%] [50.35%] [21.87%] [0.14%] [0.03%]
(11.02%) (3.04%) (5.55%) (2.41%) (0.02%) (0%)
# of contracts (total) 1,158,194,023 576,632,609 459,570,158 121,529,991 385,094 76,171
[100%] [49.79%] [39.68%] [10.49%] [0.03%] [0.01%]
(19.73%) (9.82%) (7.83%) (2.07%) (0.01%) (0%)
Trading value (total) 920,420,446 236,797,194 448,483,551 230,966,886 2,732,164 1,440,650
[100%] [25.73%] [48.73%] [25.09%] [0.3%] [0.16%]
(13.03%) (3.35%) (6.35%) (3.27%) (0.04%) (0.02%)

Panel E. Domestic individuals

Price 1.55 0.62 1.66 3.58 7.73 18.14
Order size (in contracts) 4.75 7.78 4.35 2.81 1.98 2.57
(continued on next page)

418
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Table 2 (continued )
Moneyness

All DOTM OTM ATM ITM DITM

Order size (in value) 5.50 3.07 5.96 9.62 15.43 46.79
# of transactions (daily) 204,597 60,847 90,627 51,696 1,287 140
# of contracts (daily) 1,256,981 543,690 513,406 196,358 3,163 364
Trading value (daily) 1,424,373 322,862 617,278 4,550,392 22,448 6,759
Bid–ask spread (daily) 0.02 0.01 0.02 0.03 0.10 0.36
# of transactions (total) 215,645,489 64,132,240 95,520,854 54,487,912 1,356,852 147,631
[100%] [29.74%] [44.3%] [25.27%] [0.63%] [0.07%]
(29.66%) (8.82%) (13.14%) (7.5%) (0.19%) (0.02%)
# of contracts (total) 1,324,857,866 573,049,227 541,130,241 206,961,097 3,334,258 383,043
[100%] [43.25%] [40.84%] [15.62%] [0.25%] [0.03%]
(22.57%) (9.76%) (9.22%) (3.53%) (0.06%) (0.01%)
Trading value (total) 1,501,288,884 340,296,814 650,611,002 479,611,212 23,659,749 7,110,107
[100%] [22.67%] [43.34%] [31.95%] [1.58%] [0.47%]
(21.25%) (4.82%) (9.21%) (6.79%) (0.33%) (0.1%)

Note: Based on five-minute time intervals, this table presents the average mid-quote price (Price), average order size in contracts (Order size (in
contracts)), and average order size in value (Order size (in value)) for each investor type and moneyness category. It also shows the daily average number
of transactions (# of transactions (daily)), daily average number of contracts (# of contracts (daily)), daily average trading value (Trading value (daily)),
daily average bid–ask spread (Bid-ask spread (daily)), total number of transactions (# of transactions (total)), total number of contracts (# of contracts
(total)), and total trading value (Trading value (total)) for each category. There are four types of investors: foreign investment firms (Panel B), foreign
non-investment firms (Panel C), domestic investment firms (Panel D), and domestic individuals (Panel E). DOTM, OTM, ATM, ITM, and DITM refer to
deep out-of-the-money, out-of-the-money, at-the-money, in-the-money, and deep in-the-money, respectively. The figures in square brackets denote the
percentage values within each investor type. The figures in parentheses beneath the bracketed figures denote the percentage values of that item to the
overall sample. The sample period is April 2010 to June 2014.

∂lnCt 1 ∂Ct ∂lnPt 1 ∂Pt

¼ and ¼ : (2)
∂σ Ct ∂σ ∂σ Pt ∂σ
The five-minute intraday regression equation (Equation (3)) is estimated to test whether the net volatility demand in the KOSPI200
options market predicts the future implied volatility of the underlying asset market after controlling for various potential determinants
of the volatility dynamics:

IVi ¼ αþβDσi-jþγIVi-jþθ1OVi-jþθ2SRi-jþθ3SVi-jþθ4jDri-jjþλ1Beginþλ2Endþεi, for j ¼ 1, 2, …, 5, (3)

6
where the dependent variable IVi is the ith return of the KOSPI200 model-free implied volatility (i.e., the VKOSPI). Considering that the
VKOSPI is highly persistent and clustered (Han, Kutan, & Ryu, 2015), we use the volatility returns and incorporate the five lagged
variables IVi-j (j ¼ 1, 2, …, 5) to ensure stationarity of the volatility series. The independent variable Dσ i-j captures the net options de-
mand for volatility. The regression equation is estimated separately for different j values, and the coefficient β measures the j-step-ahead
prediction of net options demand. A significantly positive value of β would suggest that sophisticated investors with volatility infor-
mation constitute the majority of options traders.
We examine the net volatility forecasting ability of the options demand and whether the volatility demand conveys information
beyond directional movements, market liquidity, and other information conveyed by other options trading activities as well. To this end,
we control for several variables related to the options and underlying markets. First, we consider the relationship between the volatility
dynamics and trading activities in related markets (Girma & Mougoue, 2002; Herbert, 1995; Moosa & Silvapulle, 2000). OVi denotes the
total number of options contracts traded in each ith five-minute trading interval. SRi and SVi are the log return and log trading volume in
the underlying KOSPI200 spot market, respectively. Second, if a substantial share of options traders uses their directional information,
the ability of the net options demand to predict volatility can be deemed spurious rather than reflective of accurate volatility information
contained in options trades. For example, options traders with positive (negative) directional information choose to buy call (put)
options. If their information is superior, the underlying spot price increases (decreases) after they buy calls (puts), and its volatility
increases. Meanwhile, their long positions in calls (puts) increase the net demand for options, resulting in a seemingly positive asso-
ciation between the net demand for options and future volatility of the underlying spot. To control the directional trading behavior in
the options market, we incorporate the “return” demand, Dri, calculated as the difference between the sums of delta-weighted long and
short options volumes. Finally, intraday dummy variables Begin and End are included to capture intraday volatility patterns. Begin (End)
equals one if the options transaction occurs during the beginning (ending) period of the day from 9:00 to 10:00 a.m. (2:00 to 3:05 p.m.)
and zero otherwise.
Table 3 presents the summary statistics of the variables used for our empirical analysis. IV and SR are multiplied by 100 for scale

6
IVi is calculated as the average value of five one-minute VKOSPI returns for the ith interval. The VKOSPI represents options-implied information
and, because options market trading activity affects the dynamics of the VKOSPI, possibly contains the information embedded in options volumes.
However, the VKOSPI measures the underlying asset's volatility (i.e., it is one of the best proxies for the return volatility of the underlying KOSPI200
spot market), not the options market volatility.

419
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Table 3
Descriptive statistics.
Panel A. Descriptive statistics of key variables
Mean StdDev Min Median Max
IV 0.0018 0.0874 1.7306 0.0000 2.2736
Dσ 0.0022 0.0478 1.2950 0.0004 2.3516
SR 0.0004 0.1399 5.4955 0.0000 3.9157
SV 13.8531 0.5068 12.4387 13.8062 16.8444
r
jD j 0.2349 0.6386 0.0000 0.0558 32.5024
Panel B. Correlation coefficients
IV Dσ SR SV jDrj
IV 1
Dσ 0.1323*** 1
SR 0.4145*** 0.2265*** 1
SV 0.0346*** 0.0340*** 0.0088** 1
r
jD j 0.0261*** 0.0720*** 0.0074* 0.2264*** 1
Panel C. Descriptive statistics of net demand for volatility (Dσ) by subsample
Mean StdDev Min Median Max
Investor type
Foreign Invest. 0.0000 0.0020 0.0742 0.0000 0.1523
Foreign Non-invest. 0.0005 0.0152 0.3494 0.0001 0.6539
Domestic Invest. 0.0003 0.0297 0.9565 0.0000 1.2136
Individuals 0.0013 0.0149 0.4435 0.0002 0.7366
Moneyness
DOTM 0.0002 0.0402 1.2882 0.0000 2.3578
OTM 0.0021 0.0185 0.5868 0.0003 0.8034
ATM 0.0004 0.0017 0.0800 0.0000 0.0411
ITM 0.0000 0.0001 0.0020 0.0000 0.0028
DITM 0.0000 0.0000 0.0005 0.0000 0.0009
Open-interest changes
Open 0.0025 0.0307 0.7109 0.0004 1.3540
Close 0.0047 0.0297 0.9350 0.0011 1.0813

Note: Based on five-minute time intervals, this table presents the summary statistics (mean, standard deviation, and minimum, median, and maximum
values) and correlation coefficients of the main variables. IV is the average return of the model-free implied volatility index of the KOSPI200 derived
from KOSPI200 options. Dσ is the net demand for volatility; SR and SV are the log return and log trading volume of the KOSPI200 index, respectively.
jDrj is the absolute difference between the sum of the delta-weighted long and short volumes. There are four investor types: foreign investment firms
(Foreign Invest.), foreign non-investment firms (Foreign Non-invest.), domestic investment firms (Domestic Invest.), and domestic individuals (Individuals).
Moneyness is classified as deep out-of-the-money (DOTM), out-of-the-money (OTM), at-the-money (ATM), in-the-money (ITM), and deep in-the-money
(DITM). Open-interest changes are categorized as open volume (Open) or close volume (Close). ***, **, and * indicate statistical significance at the level
1%, 5%, and 10% levels, respectively. The sample period is April 2010 to June 2014.

adjustments. The net demand measures, Dσ and jDrj, are divided by 10,000,000. As shown in Panel A, the net volatility demand of
options, Dσ , has a negative mean value of 0.0022, indicating that, on average, investors take short volatility options positions. The
average absolute net return demand for options (jDrj) is 0.2349, reflecting the existence of directional trading in the KOSPI200 options
market. Panel B presents the correlation coefficients between the volatility returns (IV) and other independent variables (Dσ , SR, SV, and
jDrj). IV and Dσ are negatively related, implying that the overall net demand for options may not predict volatility. IV and SV and jDrj,
which measure trading activities, are all positively related to each other. The negative correlations between spot returns (SR) and
volatilities reflect asymmetric volatility (Chun, Cho, & Ryu, 2019). Panel C presents the descriptive statistics of net volatility demand
(Dσ ) by investor type, option moneyness, and open-interest changes. The average value of Dσ is positive for the trades of foreign in-
vestment firms but negative for domestic individuals, indicating that foreign investment firms are net volatility demanders while do-
mestic individuals generally take short volatility positions. Panel C also shows that options traders who initiate opening (closing) trades
generally take long (short) volatility positions.

4. Empirical results

Table 4 reports the results of Equation (3) without the classification of options trades. When the options demands are motivated by
volatility information trading, the coefficient of Dσ should be significantly positive. Our basic intraday time-unit period is a five-minute
interval. We measure the ability of net options demand to predict the underlying volatility one period (i.e., 5 min) through five periods
(i.e., 25 min) ahead. Panel A of Table 4 shows that none of the coefficients of volatility demand (Dσ ) are significant, indicating insuf-
ficient evidence of volatility information in the options market.7 Panels B and C show whether the ability to predict volatility differs
before and after the recent market reform. As presented in the panels, the difference is not significant despite the change in the
KOSPI200 options market trading environment. This finding suggests that the determinants of predictive ability are unchanged after the

7
Although Bae and Dixon (2018) report that options trades have some predictive power for future volatility, they fail to control for directional
information trading. Furthermore, they examine only arbitrary straddle positions using near-the-money options.

420
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

market reform. Thus, we focus on the results obtained using the entire sample.
As shown in Table 4, an analysis of the aggregate options trading volume indicates a low predictive power of the demand for
volatility (Dσ ). This unexpected result can be explained by the dominance of noisy, unsophisticated, and behaviorally biased investors
with ambiguous volatility information in the KOSPI200 options market (Ahn et al., 2008; Yang, Choi, & Ryu, 2017; Yang, Lee, & Ryu,
2018). Even if a substantial number of sophisticated, professional investors successfully conduct volatility trading in the options market
based on their superior volatility information, the effect of such trades can be diluted by that of noisy trades.
Therefore, the type of investors who submit each order and initiate each trade must be considered to examine the volatility in-
formation content of options demand. Furthermore, the trading motives of options market participants, their trading vehicles, and other
traits of their options transactions differ substantially depending on their option moneyness choices and position changes. The literature
on the KOSPI200 options market shows stark differences in the ITM and OTM options trading (Kim & Ryu, 2015; Park, Kutan, & Ryu,
2019; Sim, Ryu, & Yang, 2016; Yang, Kutan, & Ryu, 2018). Additionally, Pan and Poteshman (2006) and Ryu and Yang (2018) analyze
the information quality of options volumes and claim that opening trades (i.e., buy or sell orders for brand new options positions) are
generally more informative than closing trades (i.e., buy or sell orders to close existing positions). Consequently, we examine whether
the new volatility demand from opening trades conveys better information. Our microstructure dataset enables us to identify investor
categories for each options trade and to trace the real-time positions of each options market participant. An analysis of this informative
dataset allows us to identify the initiator of each trade and determine whether each order is submitted to open or close the investor's
position, thus tracking open-interest changes.
To determine the type of investors that possess superior volatility information and how they use their information in the index
options market, which is the central theme of our research, we consider the investor type, option moneyness, and open-interest changes.
Table 5 reports the estimation results of the regression for volatility demand by category.8 We first classify options market participants as
foreign investment firms (Foreign Invest.), foreign non-investment firms (Foreign Non-invest.), domestic investment firms (Domestic
Invest.), and domestic individuals (Individuals), because these categories of investors have different information-processing abilities,
trading motives and experiences, investor psychology, performance, and wealth.9 Panel A of Table 5 shows the estimation results when
volatility demand (Dσ) is categorized by investor type. We find that foreign non-investment firms, domestic investment firms, and
domestic individuals do not initiate options trades with volatility information, whereas foreign investment firms do. The volatility
demand of foreign investment firms predicts the short-term, one-period-ahead underlying volatility; the estimated coefficient is 0.541
and t-statistic is 2.29 (i.e., the estimated β is significantly positive).
Panel A of Table 5 shows that foreign investors have superior volatility information. Most foreign investors in the Korean derivatives
market are sophisticated and have better information-processing skills, market timing, trading knowledge, and experience than do-
mestic traders do; as such, they can profit from public and/or market-wide information in addition to their sophistication and trading
advantages. Our results support prior findings (Chang, Hsieh, & Lai, 2009; Chang et al., 2010; Kuo, Chung, & Chang, 2015; Lee & Wang,
2016) about the sophistication, trading experience, and trading skills of foreign investors in emerging options markets, as well as
findings (Froot, O'Connell, & Seasholes, 2001; Richards, 2005) about the high influence of their capital flows in such markets.10
Chang et al. (2010) find that the volatility demand in the Taiwanese options market created by a specific volatility trading strategy of
foreign institutional investors contains vital information and can predict one-day-ahead volatility, which is partially consistent with our
findings. According to the authors, foreign investors profit from their volatility information by combining options and futures trades;
other volatility transactions, such as straddle, strangle, and calendar spread trades (typical volatility trading strategies), convey little
information. Their vague and contrasting results may be attributable to an inaccurate process of identifying volatility–sensitive options
strategies and/or the small share contributed by such volatility trading to the total volatility trading. The Korean and Taiwanese options
markets have similar characteristics and structures, and most of the speculators and hedgers who dominate options markets are
directional traders rather than volatility traders (Lakonishok, Lee, Pearson, & Poteshman, 2006). Accordingly, we examine the infor-
mation content of volatility demand based on key option market characteristics—including initiating investor type, option moneyness,
and open-interest changes—rather than specific volatility–sensitive options strategies.
Panel B of Table 5 examines whether option moneyness explains the effects of volatility information trading. The option moneyness
category reflects the leverage effect and sensitivity of options prices to underlying asset price changes, suggesting option moneyness may
explain the predictive ability of volatility demand. The characteristics of each KOSPI200 options contract differ significantly in terms of
spread size, depth, investor participation rate, and degree of informed trading depending on the contract's moneyness. We find that the
net volatility demand constructed from OTM options trading predicts 10-min-ahead volatility, but the demand from other moneyness

8
To conserve space, we report only the estimated coefficients of the volatility demand (Dσ) and not those of the control variables.
9
Existing studies report that the information superiorities, performances, and traits of the investors are quite different and heterogeneous (Bae,
Stulz, & Tan, 2008; Brandt, Brav, Graham, & Kumar, 2009; Chan, Menkveld, & Yang, 2007; Chang et al., 2009; Chung, Kim, & Ryu, 2017; Dvorak,
2005; Griffin, Harris, & Topaloglu, 2003; Grinblatt & Keloharju, 2000; Hau, 2001; Hsieh & He, 2014; Huang & Shiu, 2009; Kaniel, Liu, Saar, &
Titman, 2012; Kaniel, Saar, & Titman, 2008; Kim, Ryu, & Seo, 2015; Kuo et al., 2015; Lee, 2015; Ng & Wu, 2007; Seok, Cho, & Ryu, 2019).
10
For all subsamples, we estimate our models and replicate all tables before and after the market reform. We find the same patterns of foreign
investment firms' volatility-informed trading in the pre-reform sample as in the full sample, and OTM options trades have predictive power for future
volatility. Foreign investment firms have significantly better volatility information than domestic traders before the reform. After the reform, their
relative information advantage and dominance somewhat decrease, reflecting the reform's exclusion of noisy and novice individuals from the options
market. We also find that the ATM options trades of domestic investors show some predictive power for future volatility in both pre-reform and post-
reform periods, indicating that domestic investors with volatility information primarily trade the near-the-money options for their high sensitivity to
volatility movements.

421
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Table 4
Volatility predictions by options volume.
Panel A. Entire sample period (April 1, 2010 to June 30, 2014)
j¼1 j¼2 j¼3 j¼4 j¼5
Intercept 0.041** (-2.57) 0.031** (-2.01) 0.018 (-1.18) 0.022 (-1.45) 0.019 (-1.30)
Dσ t-j 0.012 (0.81) 0.011 (0.89) 0.014 (1.50) 0.015 (1.58) 0.003 (0.29)
IVt-j 0.074*** (-5.14) 0.010 (-0.87) 0.005 (-0.46) 0.006 (-0.58) 0.013 (1.17)
SRt-j 0.134*** (-13.00) 0.015 (-1.51) 0.011 (-1.34) 0.010 (-0.90) 0.016 (1.55)
SVt-j 0.003** (2.44) 0.002* (1.90) 0.001 (1.09) 0.002 (1.37) 0.001 (1.22)
r
jD t-jj 0.002** (2.32) 0.001 (1.43) 0.000 (0.76) 0.000 (-0.40) 0.000 (-0.11)
Begin 0.004*** (-2.77) 0.003** (-2.15) 0.003* (-1.83) 0.003* (-1.89) 0.001 (-0.92)
End 0.004*** (-3.45) 0.003*** (-3.24) 0.003*** (-2.86) 0.003*** (-3.00) 0.003*** (-3.00)
adj. R2 0.0167 0.0004 0.0002 0.0002 0.0003
Panel B. Pre-regulation period (April 1, 2010 to June 14, 2012)
Intercept 0.070*** (-2.60) 0.050* (-1.93) 0.028 (-1.01) 0.032 (-1.24) 0.030 (-1.23)
Dσ t-j 0.023 (1.53) 0.013 (0.98) 0.018* (1.73) 0.016 (1.59) 0.001 (0.13)
IVt-j 0.079*** (-4.05) 0.016 (-1.05) 0.006 (-0.45) 0.005 (0.39) 0.020 (1.32)
SRt-j 0.16*** (-11.93) 0.022 (-1.6) 0.017 (-1.51) 0.010 (-0.70) 0.020 (1.46)
SVt-j 0.005** (2.54) 0.004* (1.89) 0.002 (1.00) 0.002 (1.23) 0.002 (1.22)
jDrt-jj 0.002** (2.10) 0.001 (1.38) 0.000 (0.50) 0.000 (-0.60) 0.000 (-0.40)
Begin 0.006*** (-2.84) 0.005** (-2.25) 0.004* (-1.85) 0.004* (-1.82) 0.002 (-0.68)
End 0.006*** (-3.63) 0.006*** (-3.49) 0.005*** (-3.05) 0.005*** (-3.23) 0.005*** (-3.28)
2
adj. R 0.023 0.001 0.000 0.000 0.001
Panel C. Post-regulation period (June 15, 2012, 2012 to June 30, 2014)
Intercept 0.010 (-0.64) 0.002 (-0.15) 0.010 (0.68) 0.008 (0.58) 0.017 (1.12)
Dσ t-j 0.015 (-0.27) 0.007 (-0.12) 0.025 (-0.53) 0.034 (0.73) 0.051 (1.11)
IVt-j 0.067*** (-5.06) 0.004 (0.37) 0.002 (-0.21) 0.032*** (-3.48) 0.003 (-0.28)
SRt-j 0.068*** (-7.86) 0.002 (0.21) 0.006 (0.76) 0.007 (-0.82) 0.003 (0.41)
SVt-j 0.001 (0.46) 0.000 (0.02) 0.001 (-0.81) 0.001 (-0.72) 0.001 (-1.25)
r
jD t-jj 0.009** (2.05) 0.003 (-0.85) 0.001 (-0.28) 0.003 (-0.82) 0.000 (-0.05)
Begin 0.001 (-0.77) 0.001 (-0.32) 0.000 (-0.12) 0.000 (-0.19) 0.000 (-0.15)
End 0.001 (-0.51) 0.000 (-0.22) 0.000 (0.09) 0.000 (0.04) 0.000 (0.17)
adj. R2 0.006 0.000 0.000 0.001 0.000

Note: This table presents the estimated coefficients of the prediction regression based on the aggregate options volume level. Panel A shows the results
for the entire sample period of April 1, 2010 to June 30, 2014. To check for possible effects of the regulatory change in June 2012, we divide our sample
period into two subperiods. Panel B presents the results for the pre-regulation period (April 1, 2010 to June 14, 2012) and Panel C shows the results for
the post-regulation period (June 15, 2012 to June 30, 2014). The VKOSPI average return is the dependent variable and the independent variables are
the demand for volatility (Dσ) and various control variables with lagged periods from one to five. IVi ¼ αþβDσi-jþγIVi-jþθ1OVi-jþθ2SRi-jþθ3SVi-jþθ4jDri-
jjþλ1Beginþλ2Endþεi, for j ¼ 1, 2, …, 5, where IV is the return of the model-free implied volatility index of the KOSPI200 derived from KOSPI200
options; Dσ is the net demand for volatility; OV is the number of options contracts; SR and SV are the log return and log trading volume of the KOSPI200
index, respectively. jDrj is the absolute difference between the sum of the delta-weighted long and short volumes, and Begin (End) equals one if a given
option corresponds to the time period from 9:00 to 10:00 a.m. (from 2:00 to 3:05 p.m.), and zero otherwise. Numbers in parentheses are the New-
ey–West t-statistics. ***, **, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

options trading does not. This finding contrasts with that of Ni, Pan, and Poteshman's (2008) study, which attributes the significant
volatility information content of ATM options to the high vega levels of ATM options (i.e., ATM options are sensitive to volatility
changes) in the US market. Our results indicate greater importance of factors such as liquidity, speculative opportunities, and leverage
than of sensitivity to volatility changes.
Panel C of Table 5 estimates the regression results separately for options trades that open new positions (open volumes) and close
existing positions (close volumes). Studies consistently report that opening trades convey higher-quality directional information than
closing trades (Hsieh & He, 2014; Pan & Poteshman, 2006). We test whether there is a higher concentration of informed trading in the
volatility demand created by open volumes than that of close volumes. In contrast to the findings of existing studies on directional
information, the estimated coefficients for both opening and closing trades are negative or insignificant, indicating insufficient evidence
of volatility information trading. This puzzling result on an aggregate level motivates us to further classify options trades across multiple
dimensions, considering various options market characteristics.
We estimate the regression using a two-dimensional classification: investor–moneyness categories (see Table 6) and investor–trade
type categories (see Table 7). This would provide greater insight into the type of investors that use their superior volatility information
and how they profit from this information through options trades. Table 6 shows that, regardless of which option moneyness category
they use as their trading vehicle, the volatility trading of foreign non-investment firms is not informative. Panel A of Table 6 shows that
the DOTM options trades of foreign investment firms predict future volatility, and foreign investment firms trade on their volatility
information by exploiting the high leverage of DOTM options. The estimated coefficient for foreign investment firms using DOTM
options is 0.753, with a t-statistic of 2.69. DOTM options may provide a cheaper or more effective means of trading on the information,
given their higher degree of leverage and liquidity (Lee & Wang, 2016). Because uninformed individual investors provide ample
liquidity and trading opportunities in the KOSPI200 (deep) out-of-the-money options market (Ryu, 2011; Ryu & Yang, forthcoming),
foreign investment firms that possess sophisticated volatility information can find profitable opportunities in highly levered options
trading. The volatility information embedded in the DOTM options trades of foreign investment firms could also be associated with the

422
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Table 5
Volatility predictions of options volumes by investor type, option moneyness, and open-interest changes.
j¼1 j¼2
Coeff. t-stat. Coeff. t-stat.
Panel A. Investor type
Foreign Invest. 0.541** (2.29) 0.237 (-1.19)
Foreign Non-invest. 0.037 (1.09) 0.015 (-0.47)
Domestic Invest. 0.015 (0.75) 0.024 (1.15)
Individuals 0.017 (0.33) 0.026 (0.79)
Panel B. Moneyness
DOTM 0.010 (0.52) 0.008 (0.50)
OTM 0.030 (1.42) 0.032** (2.01)
ATM 0.219 (0.99) 0.034 (0.17)
ITM 0.827 (0.09) 2.462 (-0.3)
DITM 260.264 (-0.89) 174.749 (-0.94)
Panel C. Open-interest changes
Open 0.032 (1.49) 0.020 (1.19)
Close 0.002 (-0.12) 0.007 (0.40)

Note: This table presents the estimation coefficients of the prediction regression by investor type, option moneyness, and trading strategy. In Panel A,
there are four investor types: foreign investment firms (Foreign Invest.), foreign non-investment firms (Foreign Non-invest.), domestic investment firms
(Domestic Invest.), and domestic individuals (Individuals). In Panel B, option moneyness is categorized as deep out-of-the money (DOTM), out-of-the-
money (OTM), at-the-money (ATM), in-the-money (ITM) or deep in-the-money (DITM). In Panel C, open-interest changes are categorized as open
volume (Open) and close volume (Close). The VKOSPI is the dependent variable and the independent variables are the demand for volatility (Dσ) and
various control variables with lagged periods from one to five. IVi ¼ αþβDσ i-jþγIVi-jþθ1OVi-jþθ2SRi-jþθ3SVi-jþθ4jDri-jjþλ1Beginþλ2Endþεi, for j ¼ 1, 2,
where IV is the return of the model-free implied volatility index of the KOSPI200 derived from KOSPI200 options; Dσ is the net demand for volatility;
OV is the number of options contracts; SR and SV are the log return and log trading volume of the KOSPI200 index, respectively. jDrj is the absolute
difference between the sum of the delta-weighted long and short volumes, and Begin (End) equals one if each option corresponds to the time period
from 9:00 to 10:00 a.m. (2:00 to 3:05 p.m.), and zero otherwise. The sample period is April 2010 to June 2014. Numbers in parentheses are New-
ey–West t-statistics. ** indicates statistical significance at the 5% level.

Table 6
Volatility predictions of options volumes by investor type and option moneyness.
j¼1 j¼2

Coeff. t-stat. Coeff. t-stat.

Panel A. Foreign investment firms

DOTM 0.753*** (2.69) 0.248 (-0.98)
OTM 0.018 (-0.06) 0.078 (-0.30)
ATM 3.272** (-2.24) 2.225 (-1.58)
ITM 23.906 (0.76) 19.697 (0.72)
DITM 447.457 (1.13) 215.046 (-0.67)
Panel B. Foreign non-investment firms
DOTM 0.045 (1.18) 0.017 (-0.46)
OTM 0.024 (0.52) 0.009 (0.21)
ATM 1.418*** (-3.50) 0.774* (-1.89)
ITM 0.990 (-0.11) 6.618 (-0.75)
DITM 358.738 (-1.18) 158.51 (-0.82)
Panel C. Domestic investment firms
DOTM 0.009 (0.36) 0.020 (0.83)
OTM 0.039 (1.40) 0.053** (2.09)
ATM 1.164*** (2.77) 0.509 (1.48)
ITM 84.628 (-0.79) 5.075 (-0.05)
DITM 547.462 (0.41) 189.289 (0.16)
Panel D. Domestic individuals
DOTM 0.008 (0.13) 0.018 (0.42)
OTM 0.032 (0.56) 0.047 (1.05)
ATM 1.189*** (2.96) 0.414 (1.22)
ITM 1.507 (0.05) 10.418 (0.40)
DITM 116.887 (-0.09) 62.717 (0.11)

Note: This table presents the estimated coefficients of the prediction regression by investor type, considering option moneyness. There are four investor
types: foreign investment firms, foreign non-investment firms, domestic investment firms, and domestic individuals. DOTM, OTM, ATM, ITM, and
DITM refer to deep out-of-the-money, out-of-the-money, at-the-money, in-the-money, and deep in-the-money options, respectively. The VKOSPI return
is the dependent variable and the independent variables are the demand for volatility (Dσ) and various control variables with lagged periods from one
to five. IVi ¼ αþβDσi-jþγIVi-jþθ1OVi-jþθ2SRi-jþθ3SVi-jþθ4jDri-jjþλ1Beginþλ2Endþεi, for j ¼ 1, 2, where IV is the return of the model-free implied volatility
index of the KOSPI200 derived from KOSPI200 options, Dσ is the net demand for volatility; OV is the number of options contracts; SR and SV are the log
return and log trading volume of the KOSPI200 index, respectively. jDrj is the absolute difference between the sum of the delta-weighted long and short
volumes, and Begin (End) equals one if each option corresponds to the time period from 9:00 to 10:00 a.m. (from 2:00 to 3:05 p.m.), and zero otherwise.
The sample period is April 2010 to June 2014. Numbers in parentheses are Newey–West t-statistics. ***, **, and * indicate statistical significance at the
1%, 5%, and 10% levels, respectively.

423
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Table 7
Volatility prediction by investor type and open-interest changes.
j¼1 j¼2

Coeff. t-stat. Coeff. t-stat.

Panel A. Foreign investment firms

Open 0.153 (0.47) 0.292 (-1.15)
Close 0.711*** (2.78) 0.035 (-0.16)
Panel B. Foreign non-investment firms
Open 0.000 (0.01) 0.029 (-0.72)
Close 0.061* (1.70) 0.004 (0.10)
Panel C. Domestic investment firms
Open 0.035 (1.31) 0.029 (1.22)
Close 0.001 (-0.04) 0.025 (0.91)
Panel D. Domestic individuals
Open 0.073 (1.54) 0.057 (1.64)
Close 0.041 (-1.11) 0.019 (-0.57)

Note: This table presents the estimated coefficients of the prediction regression by detailed investor type, considering trading strategies. There are
four investor types: foreign investment firms, foreign non-investment firms, domestic investment firms, and domestic individuals. Trading strategies
are categorized as open volume (Open) or close volume (Close). The VKOSPI return is the dependent variable and the independent variables are the
demand for volatility (Dσ) and various control variables with lagged periods from one to five. IVi ¼ αþβDσ i-jþγIVi-jþθ1OVi-jþθ2SRi-jþθ3SVi-jþθ4jDri-
jjþλ1Beginþλ2Endþεi, for j ¼ 1, 2, where IV is the return of the model-free implied volatility index of the KOSPI200 derived from KOSPI200 options;
Dσ is the net demand for volatility; OV is the number of options contracts; SR and SV are the log return and log trading volume of the KOSPI200
index, respectively. jDrj is the absolute difference between the sum of the delta-weighted long and short volumes, and Begin (End) equals one if the
option corresponds to the time period from 9:00 to 10:00 a.m. (from 3:00 to 3:05 p.m.), and zero otherwise. The sample period is April 2010 to June
2014. Numbers in parentheses are Newey–West t-statistics. *** and * indicate statistical significance at the 1% and 10% levels, respectively.

previous finding that they are net volatility demanders (see Panel C of Table 3). When investors open their positions, they are generally
net volatility demanders (see Table 3), indicating that creating new options positions is related to volatility demand. Volatility informed
investors might prefer smaller investments and avail the leverage effect of investing in DOTM options by taking long positions in the
options market as a strategy for generating profits.
Although the aggregate volatility demand of domestic individuals and that of investment firms are not informative, Panels C and D of
Table 6 show that the ATM options trades of domestic investors have some predictive power for future volatility. The estimated co-
efficient of volatility trades initiated by domestic investment firms (domestic individuals) using ATM options is 1.164 (1.189), with a t-
statistic of 2.77 (2.96), showing that the ATM options trades initiated by them convey some volatility information. Following rule-of-
thumb volatility trading strategies, domestic investors with volatility information primarily trade the near-the-money options because
they have the highest volatility sensitivity. The volatility trading of domestic investors is somewhat conservative compared to that of
foreign investment firms, which are involved in speculative options trading and use highly leveraged out-of-the-money options, possibly
reflecting their relative confidence and experience.
Table 7 reports the estimation results of the regression for volatility demand based on which type of investor initiates each options
trade and the change in investors’ positions (i.e., open-interest) following the transaction. Both domestic institutional and individual
investors conduct options trades without volatility information, irrespective of whether they open or close positions. By contrast,
although the demand for opening and closing trades does not convey volatility information when it is not considered separately by
investor type, Panel A of Table 7 shows that the closing trades of informed foreign investment firms significantly predict underlying
volatility. It is noteworthy that closing trades also contain significant volatility information considering that most studies examining
directional information report that only opening trades predict future asset returns, whereas closing trades do not.
Consider this example (and interpretation) of informative closing trades. Informed traders who expect a price increase (decrease) in
an underlying asset can write puts (calls) and open a new short position, that is, an open-sell trade using put (call) options at time t. The
trader profits from an increase (decrease) in the underlying spot price at time tþ1. The information flows are fast, volatility levels are
relatively high, and speculative trading prevails in the KOSPI200 spot and options markets; this means that the spot price often decreases
(increases) after a price increase (decrease). Specifically, in this example, the spot price can decrease (increase) at time tþ2. The
informed trader, having an open-sell put (call) position, should close the existing position by submitting a close-buy order in the put
(call) options market right before time tþ2 (i.e., at time tþ1). Such closing trades successfully predict volatility changes, and our analysis
captures the volatility information contained in close-buy orders. Furthermore, the rapid information flow in the index options market
induces informed traders to conduct “cheap lunch” trading. Timing is an essential factor when informed investors close their positions.
These closing orders should be carefully considered and submitted based on the existing positions and portfolio dynamics. While
opening trades are conducted by uninformed, noisy investors as well, we can infer that informed investors with open positions in the
options market tend to discreetly close their positions based on prudent predictions and expectations. However, other explanations for
the higher information quality of closing trades are also plausible. In the case of opening trades, hedging motives are included because
hedgers open their options positions and maintain their positions without closing them. By contrast, the closing trades do not include
such hedging needs. The net demand of volatility-informed investors may be better reflected in closing trades.

424
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

5. Conclusions

We empirically investigate whether intraday volatility information exists in the index options market using a unique, high-quality
microstructure dataset of KOSPI200 options. Based on the vega-weighted net demand for volatility, we classify investors into four
groups, to determine the types of investors that are informed about future stock market volatility and that conduct volatility information
trading in a highly liquid options market. Our results have significant implications for forecasting market volatility and examining the
effect of foreign firms versus domestic traders on market volatility. We find that the overall options market demand for volatility does
not predict market volatility. This result suggests that the information quality of options trades is relatively low because uninformed and
noisy domestic investors dominate the KOSPI200 options market. Second, the vega-weighted net demand of foreign investment firms
conveys significant information about future volatility dynamics. By tracking the positions of all options market participants for all
option moneyness categories and maturities, we find that foreign investment firms conduct volatility trading with highly levered options
and that their intraday volatility information superiority is more prominent when they close rather than open positions. Finally, we find
that the volatility demand of domestic traders who trade near-the-money options partially predicts market volatility, implying that their
trades also convey some volatility information.

Acknowledgment

This work was supported by the IREC, The Institute of Finance and Banking, Seoul National University.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.iref.2019.07.006.

References

Ahn, H.-. J., Kang, J., & Ryu, D. (2008). Informed trading in the index option market: The case of KOSPI 200 options. Journal of Futures Markets, 28(12), 1118–1146.
Ahn, H.-. J., Kang, J., & Ryu, D. (2010). Information effects of trade size and trade direction: Evidence from the KOSPI 200 index options market. Asia-Pacific Journal of
Financial Studies, 39(3), 301–339.
Bae, K. H., & Dixon, P. (2018). Do investors use options and futures to trade on different types of information? Evidence from an aggregate stock index. Journal of
Futures Markets, 38(2), 175–198.
Bae, K. H., Stulz, R. M., & Tan, H. (2008). Do local analysts know more? A cross-country study of the performance of local analysts and foreign analysts. Journal of
Financial Economics, 88(3), 581–606.
Banerjee, P. S., Doran, J. S., & Peterson, D. R. (2007). Implied volatility and future portfolio returns. Journal of Banking & Finance, 31(10), 3183–3199.
Barber, B. M., Odean, T., & Zhu, N. (2009). Systematic noise. Journal of Financial Markets, 12(4), 547–569.
Bollen, N. P., & Whaley, R. E. (2004). Does net buying pressure affect the shape of implied volatility functions? Journal of Finance, 59(2), 711–753.
Brandt, M. W., Brav, A., Graham, J. R., & Kumar, A. (2009). The idiosyncratic volatility puzzle: Time trend or speculative episodes? Review of Financial Studies, 23(2),
863–899.
Busch, T., Christensen, B. J., & Nielsen, M. (2011). The role of implied volatility in forecasting future realized volatility and jumps in foreign exchange, stock and bond
markets. Journal of Econometrics, 160(1), 48–57.
Cao, C., Chen, Z., & Griffin, J. M. (2005). Informational content of option volume prior to takeovers. Journal of Business, 78(3), 1073–1109.
Chan, K., Chung, Y. P., & Fong, W. M. (2002). The informational role of stock and option volume. Review of Financial Studies, 15(4), 1049–1075.
Chang, C. C., Hsieh, P. F., & Lai, H. N. (2009). Do informed option investors predict stock returns? Evidence from the Taiwan stock exchange. Journal of Banking &
Finance, 33(4), 757–764.
Chang, C. C., Hsieh, P. F., & Lai, H. N. (2013). The price impact of options and futures volume in after-hours stock market trading. Pacific-Basin Finance Journal, 21(1),
984–1007.
Chang, C. C., Hsieh, P. F., & Wang, Y. H. (2010). Information content of options trading volume for future volatility: Evidence from the Taiwan options market. Journal
of Banking & Finance, 34(1), 174–183.
Chan, K., Menkveld, A. J., & Yang, Z. (2007). The informativeness of domestic and foreign investors' stock trades: Evidence from the perfectly segmented Chinese
market. Journal of Financial Markets, 10(4), 391–415.
Chen, C. H., Chung, H., & Yuan, S. F. (2014). Deviations from put–call parity and volatility prediction: Evidence from the Taiwan index option market. Journal of Futures
Markets, 34(12), 1122–1145.
Chiang, C. H., Chung, S. G., & Louis, H. (2017). Insider trading, stock return volatility, and the option market's pricing of the information content of insider trading.
Journal of Banking & Finance, 76, 65–73.
Chun, D., Cho, H., & Ryu, D. (2019). Forecasting the KOSPI200 spot volatility using various volatility measures. Physica A: Statistical Mechanics and its Applications, 514,
156–166 (Corresponding Author).
Chung, C. Y., Kim, H., & Ryu, D. (2017). Foreign investor trading and information asymmetry: Evidence from a leading emerging market. Applied Economics Letters,
24(8), 540–544.
Chung, K. H., Park, S. G., & Ryu, D. (2016). Trade duration, informed trading, and option moneyness. International Review of Economics and Finance, 44, 395–411.
Dorn, D., Huberman, G., & Sengmueller, P. (2008). Correlated trading and returns. Journal of Finance, 63(2), 885–920.
Dvor
ak, T. (2005). Do domestic investors have an information advantage? Evidence from Indonesia. Journal of Finance, 60(2), 817–839.
Easley, D., O'hara, M., & Srinivas, P. S. (1998). Option volume and stock prices: Evidence on where informed traders trade. Journal of Finance, 53(2), 431–465.
Erenburg, G., Kurov, A., & Lasser, D. J. (2006). Trading around macroeconomic announcements: Are all traders created equal? Journal of Financial Intermediation, 15(4),
470–493.
Fahlenbrach, R., & Sandås, P. (2010). Does information drive trading in option strategies? Journal of Banking & Finance, 34(10), 2370–2385.
Froot, K. A., O'Connell, P. G., & Seasholes, M. S. (2001). The portfolio flows of international investors. Journal of Financial Economics, 59(2), 151–193.
Ge, L., Lin, T. C., & Pearson, N. D. (2016). Why does the option to stock volume ratio predict stock returns? Journal of Financial Economics, 120(3), 601–622.
Gharghori, P., Maberly, E. D., & Nguyen, A. (2017). Informed trading around stock split announcements: Evidence from the option market. Journal of Financial and
Quantitative Analysis, 52(2), 705–735.
Giot, P. (2005). Relationships between implied volatility indexes and stock index returns. Journal of Portfolio Management, 31(3), 92–100.

425
H. Yang et al. International Review of Economics and Finance 64 (2019) 412–426

Girma, P. B., & Mougoue, M. (2002). An empirical examination of the relation between futures spreads volatility, volume, and open interest. Journal of Futures Markets,
22(11), 1083–1102.
Griffin, J. M., Harris, J. H., & Topaloglu, S. (2003). The dynamics of institutional and individual trading. Journal of Finance, 58(6), 2285–2320.
Grinblatt, M., & Keloharju, M. (2000). The investment behavior and performance of various investor types: A study of Finland's unique data set. Journal of Financial
Economics, 55(1), 43–67.
Guo, H., & Whitelaw, R. F. (2006). Uncovering the risk-return relation in the stock market. Journal of Finance, 61(3), 1433–1463.
Han, H., Kutan, A. M., & Ryu, D. (2015). Effects of the US stock market return and volatility on the VKOSPI. Economics: The Open-Access. Open-Assessment E-Journal,
9(2015–35), 1–34.
Hau, H. (2001). Location matters: An examination of trading profits. Journal of Finance, 56(5), 1959–1983.
Herbert, J. H. (1995). Trading volume, maturity and natural gas futures price volatility. Energy Economics, 17(4), 293–299.
Holowczak, R., Hu, J., & Wu, L. (2014). Aggregating information in option transactions. Journal of Derivatives, 21(3), 9–23.
Hsieh, W. G., & He, H. R. (2014). Informed trading, trading strategies and the information content of trading volume: Evidence from the Taiwan index options market.
Journal of International Financial Markets, Institutions and Money, 31, 187–215.
Hu, J. (2014). Does option trading convey stock price information? Journal of Financial Economics, 111(3), 625–645.
Huang, R. D., & Shiu, C. Y. (2009). Local effects of foreign ownership in an emerging financial market: Evidence from qualified foreign institutional investors in Taiwan.
Financial Management, 38(3), 567–602.
Jiang, G. J., & Tian, Y. S. (2005). The model-free implied volatility and its information content. Review of Financial Studies, 18(4), 1305–1342.
Johnson, T. L., & So, E. C. (2012). The option to stock volume ratio and future returns. Journal of Financial Economics, 106(2), 262–286.
Kaniel, R., Liu, S., Saar, G., & Titman, S. (2012). Individual investor trading and return patterns around earnings announcements. Journal of Finance, 67(2), 639–680.
Kaniel, R., Saar, G., & Titman, S. (2008). Individual investor trading and stock returns. Journal of Finance, 63(1), 273–310.
Kim, J. S., & Ryu, D. (2015). Are the KOSPI 200 implied volatilities useful in value-at-risk models? Emerging Markets Review, 22, 43–64.
Kim, J. S., Ryu, D., & Seo, S. W. (2015). Corporate vulnerability index as a fear gauge? Exploring the contagion effect between U.S. and Korean markets. Journal of
Derivatives, 23(1), 73–88.
Kuo, W. H., Chung, S. L., & Chang, C. Y. (2015). The impacts of individual and institutional trading on futures returns and volatility: Evidence from emerging index
futures markets. Journal of Futures Markets, 35(3), 222–244.
Lai, Y. (2017). Macroeconomic factors and index option returns. International Review of Economics & Finance, 48, 452–477.
Lakonishok, J., Lee, I., Pearson, N. D., & Poteshman, A. M. (2006). Option market activity. Review of Financial Studies, 20(3), 813–857.
Lee, E. J. (2015). High frequency trading in the Korean index futures market. Journal of Futures Markets, 35(1), 31–51.
Lee, J., Kang, J., & Ryu, D. (2015). Common deviation and regime-dependent dynamics in the index derivatives markets. Pacific-Basin Finance Journal, 33, 1–22.
Lee, J., & Ryu, D. (2019). The impacts of public news announcements on intraday implied volatility dynamics. Journal of Futures Markets, 39(6), 656–685.
Lee, Y. H., & Wang, D. K. (2016). Information content of investor trading behavior: Evidence from Taiwan index options market. Pacific-Basin Finance Journal, 38,
149–160.
Le, V., & Zurbruegg, R. (2016). The impact of short sale restrictions on informed trading in the stock and options markets. International Review of Economics & Finance,
41, 262–273.
Lin, W. T., Tsai, S. C., Zheng, Z., & Qiao, S. (2018). Retrieving aggregate information from option volume. International Review of Economics & Finance, 55, 220–232.
Moosa, I. A., & Silvapulle, P. (2000). The price–volume relationship in the crude oil futures market Some results based on linear and nonlinear causality testing.
International Review of Economics & Finance, 9(1), 11–30.
Ng, L., & Wu, F. (2007). The trading behavior of institutions and individuals in Chinese equity markets. Journal of Banking & Finance, 31(9), 2695–2710.
Ni, S. X., Pan, J., & Poteshman, A. M. (2008). Volatility information trading in the option market. Journal of Finance, 63(3), 1059–1091.
Pan, J., & Poteshman, A. M. (2006). The information in option volume for future stock prices. Review of Financial Studies, 19(3), 871–908.
Park, Y. J., Kutan, A. M., & Ryu, D. (2019). The impacts of overseas market shocks on the CDS-option basis. North American Journal of Economics and Finance, 47,
622–636.
Park, S. G., & Ryu, D. (2019). Speed and trading behavior in an order-driven market. Pacific-Basin Finance Journal, 53, 145–164.
Richards, A. (2005). Big fish in small ponds: The trading behavior and price impact of foreign investors in Asian emerging equity markets. Journal of Financial and
Quantitative Analysis, 40(1), 1–27.
Roll, R., Schwartz, E., & Subrahmanyam, A. (2010). O/S: The relative trading activity in options and stock. Journal of Financial Economics, 96(1), 1–17.
Rourke, T. (2014). The delta-and vega-related information content of near-the-money option market trading activity. Journal of Financial Markets, 20, 175–193.
Ryu, D. (2011). Intraday price formation and bid-ask spread components: A new approach using a cross-market model. Journal of Futures Markets, 31(12), 1142–1169.
Ryu, D. (2015). The information content of trades: An analysis of KOSPI 200 index derivatives. Journal of Futures Markets, 35(3), 201–221.
Ryu, D. (2016). Considering all microstructure effects: The extension of a trade indicator model. Economics Letters, 146, 107–110.
Ryu, D., Ryu, D. and Yang, H., Vega-informed trading and options market reform, Applied Economics Letters, forthcoming.
Ryu, D., & Yang, H. (2017). Price disagreements and adjustments in index derivatives markets. Economics Letters, 151, 104–106.
Ryu, D., & Yang, H. (2018). The directional information content of options volumes. Journal of Futures Markets, 38(12), 1533–1548.
Ryu, D., Yang, H., Who has volatility information in the index options market? Finance Research Letters, forthcoming.
Schlag, C., & Stoll, H. (2005). Price impacts of options volume. Journal of Financial Markets, 8(1), 69–87.
Seok, S. I., Cho, H., & Ryu, D. (2019). Firm-specific investor sentiment and the stock market response to earnings news. North American Journal of Economics and Finance,
48, 221–240.
Sim, M., Ryu, D., & Yang, H. (2016). Tests on the monotonicity properties of KOSPI 200 options prices. Journal of Futures Markets, 36(7), 625–646.
Song, W., Ryu, D., & Webb, R. I. (2018). Volatility dynamics under an endogenous Markov-switching framework: A cross-market approach. Quantitative Finance, 18(9),
1559–1571.
Webb, R. I., Ryu, D., Ryu, D., & Han, J. (2016). The price impact of futures trades and their intraday seasonality. Emerging Markets Review, 26, 80–98.
Yang, H., Choi, H.-. S., & Ryu, D. (2017). Option market characteristics and price monotonicity violations. Journal of Futures Markets, 37(5), 473–498.
Yang, H., Kutan, A. M., & Ryu, D. (2018). Option moneyness and price disagreements. Applied Economics Letters, 25(3), 192–196.
Yang, H., Lee, J., & Ryu, D. (2018). Market depth, domestic investors and price monotonicity violations. Applied Economics Letters, 25(10), 688–692.
Yang, H., Ryu, D., & Ryu, R. (2018). Market reform and efficiency: The case of KOSPI200 options. Emerging Markets Finance and Trade, 54(12), 2687–2697.

426