Noise traders, mispricing, and price adjustments in derivatives markets

ABSTRACT

This study examines disagreements between actual and options-implied futures prices and the corresponding adjustments in a sophisticated setting. We identify the market that triggers each type of price disagreement and find that the market in which the disagreement is initiated adjusts more to eliminate the mispricing. Futures prices adjust less for options-initiated price disagreements with out-of-the-money (OTM) options-implied prices than they do for disagreements with at-the-money (ATM) prices. Options markets adjust more when disagreements are initiated by OTM options than they do when disagreements are initiated by ATM options. Adjustments in both the futures and options markets consistently suggest that OTM options trading provides inferior information. Price disagreements are positively correlated with the participation of domestic investors, especially when they trade OTM options, implying that domestic investors are noisier and less informed than foreign investors are.

Highlights

• We examine disagreements between actual and options-implied futures prices and the corresponding adjustments.

• The market in which the disagreement is initiated adjusts more to eliminate the mispricing.

• Out-of-the-money options-initiated price disagreements are positively correlated with domestic investors’ trades.

KEYWORDS:

• Cost-of-carry
• domestic investor
• market efficiency
• noise trading
• price adjustment
• price disagreement
• put-call parity

JEL CLASSIFICATIONS:

• G13
• G14
• G15

Acknowledgment

The authors appreciate helpful comments and suggestions from Robert I. Webb (University of Virginia). This work was supported by the LG Yonam Foundation (of Korea).

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Doojin Ryu http://orcid.org/0000-0002-0059-4887

Heejin Yang http://orcid.org/0000-0003-2479-459X

Notes

1 Our disagreement framework is somewhat similar to the lead-lag price relationship. After a price disagreement between two related assets occurs, we regard the asset whose price moves to resolve this disagreement as a mispriced asset and its price as uninformative.

2 In recent finance research, the quantile approach is adopted to examine price and trading dynamics in derivatives markets (Lee, Lee, and Lien 2019; Lee, Lee, and Ryu 2019).

3 Domestic individual and domestic institutional investors are known to be relatively uninformed and to perform poorly in index options trading. In contrast, foreign traders are mostly institutional investors, who are known to be professional investors and possess trading skills, knowledge, and better information-processing abilities than other traders in the KOSPI200 options market (Ahn, Kang, and Ryu 2008; Ryu and Yang 2018, 2019). For these reasons, we compare the trades of domestic and foreign investors in this study.

4 The quoting unit for the underlying KOSPI200 spot prices is the ‘point.’ This unit is also used for quotes in the KOSPI200 futures and options markets, but it indicates different dollar amounts in each market. The minimum tick size for futures (options) contracts is 0.05 (0.01) points.

5 In Table 1, we consider all maturity contracts and all option moneyness types. However, in Table 2, we only consider the nearest maturity contracts, selected option moneyness types, and transactions during continuous trading sessions. More importantly, we do not differentiate between initiating and initiated trades in Table 1 when we show total trading volumes for each investor type. In contrast, in Table 2, we trace whether each trade is initiated by a domestic or foreign investor. For example, if a foreign investor submits a market buy order and the order is transacted using standing orders submitted by domestic investors, this transaction is initiated by the foreign investor and is recorded as part of the foreign investor trading volume. We confirm that the trading proportions of domestic and foreign investors are quite similar to those shown in Table 1 by computing the combined volume of initiating and initiated trades. This finding indicates that foreign investors in the KOSPI200 options market are more likely to initiate trades and submit aggressive orders than their domestic counterparts are.

6 We also analyze the median values of the implied prices as a robustness check. Using median instead of mean prices does not qualitatively affect our results. Thus, our results are robust to changes in the measurement of the implied prices.

7 In Equations (2(2) $TypeI\text{ disagreement}:P_{f,t}-P_{o,t}\ge Threshold$(2) ) and (3(3) $TypeII\text{ disagreement}:P_{o,t}-P_{f,t}\ge Threshold$(3) ), we can set the thresholds as 0.01, 0.02, or 0.05 points based on the minimum tick sizes in the options (i.e., 0.01 points) and futures markets (i.e., 0.05 points). Among these candidate thresholds, we set the greatest threshold level (i.e., 0.05 points) as our basic criterion to define price disagreements strictly. In the case of KOSPI200 futures and options trading, the explicit transaction costs are very low, and the markets are highly liquid, meaning that the implicit transaction costs are also very minimal. Furthermore, the bid-ask spreads usually equal the minimum tick size, which also supports the conclusion that the barriers and transaction costs of derivatives trading are trivial. For these reasons, we are confident that our disagreements involve sufficiently large price differences to overcome any transaction costs that may discourage investors from trading and adjusting prices.

8 Although we consider using returns instead of price changes as the dependent variable, our main variables of interest are given as absolute values. For example, we use an absolute price difference as the threshold when identifying price disagreements. Furthermore, if transaction costs are the main reasons for disagreements and their resolutions, price changes are more important than relative returns.

9 We do not include data points for which a price disagreement occurs multiple times within a thirty-second evaluation period. Any lags exceeding thirty seconds prior to an event are not included to eliminate any biases that may arise from other disagreements with other effects.

10 Unreported analyses also confirm that domestic investors have an information disadvantage, whereas the option trades of foreign investors predict future spot returns and volatilities, indicating that foreign institutions have information regarding both the directional and volatility information about the KOSPI200 options market. These results support the adequacy of our classification of domestic and foreign investors in that these investor types are clearly distinct in terms of information superiority.

11 For more details on the characteristics, dynamics, and history of the VKOSPI, refer to Han, Kutan, and Ryu (2015), Song, Ryu, and Webb (2016, 2018), Park, Ryu, and Song (2017), Song, Park, and Ryu (2018), and Chun, Cho, and Ryu (2019).

12 Because the results in Table 7 suggest the possibility of a nonlinear or U-shaped relationship, we use a quadratic specification. Consistent with our previous results, we find that OTM options-initiated price disagreements occur more frequently when the trading proportion of domestic investors is greater. For example, for Type I-O (Type II-O) events, the estimated coefficients on Domestic² for a thirty-second evaluation period are 0.5819 (0.7977) for the OTM Call group and 0.5938 (0.7105) for the OTM Put group.

13 Very short-term maturity options and deep OTM options are actively traded in the KOSPI200 options market. The possibility that a deep OTM option goes in the money in the short remaining period before maturity is quite low. Considering their low probability of being exercised, deep OTM KOSPI200 options are typically overpriced. The excess speculation on OTM options contracts may reflect the lottery-buying behavior and sentiment of noisy individual traders in the KOSPI200 options market (Kim and Ryu 2015; Ryu, Kang, and Suh 2015).

Additional information

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT; Ministry of Science and ICT) [grant number 2019R1G1A1100196].