an empirical analysis of the price volatility

Research ArticleAn Empirical Analysis of the Price Volatility Characteristics ofChinarsquos Soybean Futures Market Based on ARIMA-GJR-GARCH Model

Yang Xu 1 Zhihao Xia 2 Chuanhui Wang 2 Weifeng Gong 23 Xia Liu 2

and Xiaodi Su 2

1Management College Ocean University of China Qingdao 266100 China2School of Economics Qufu Normal University Rizhao 276826 China3School of Economics and Management Nanjing University of Aeronautics and Astronautics Nanjing 211006 China

Correspondence should be addressed to Chuanhui Wang chhwang001163com

Received 15 June 2021 Accepted 6 October 2021 Published 5 November 2021

Academic Editor Niansheng Tang

Copyright copy 2021 YangXu et al+is is an open access article distributed under the Creative CommonsAttribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

As the main force in the futures market agricultural product futures occupy an important position in the Chinarsquos market Takingthe representative soybean futures in Dalian Commodity Futures Market of China as the research object the relationship betweenprice fluctuation characteristics and trading volume and open position was studied +e empirical results show that the pricevolatility of Chinarsquos soybean futures market has a ldquoleverage effectrdquo+e trading volume and open interest are divided into expectedparts and unexpected parts which are added to the conditional variance equation +e expected trading volume coefficient isestimated Also the estimated value of the expected open interest coefficient is respectively smaller than the estimated value of theunexpected trading volume coefficient and the estimated value of the unexpected open interest coefficient +erefore the impactof expected trading volume on the price fluctuation of Chinarsquos soybean futures market is less than that of unexpected tradingvolume on the price of soybean futures market +is paper adds transaction volume as an information flow to the variance of theconditional equation innovatively and also observes transaction volume as the relationship between conditional variance andprice fluctuations

1 Introduction

In bulk commodity trade soybean is one of the agriculturalproducts with large demand in China and its price fluc-tuation is more prominent Chinarsquos soybean import is at adisadvantage and China has become the worldrsquos largestimporter Due to the further opening of the soybean marketthe price fluctuation of domestic soybean futures is affectedby many factors at home and abroad which has arousedwide attention in China Since 1996 China has become theworldrsquos major soybean importer In order to optimize thesoybean supply structure the Chinese government hasimplemented the ldquoreduction of corn and the beansrdquo since2016 With the Sino-US trade frictions intensified in 2018the United States increased import tariffs by 25 and import

costs rose Russia and Canada have thrown out olivebranches to increase exports to China In 2019 the Ministryof Agriculture and Rural Affairs decided to implement theSoybean Revitalization Plan and put forward six subsidypolicies to support the development of soybean In 2020 theNo 1 document of the Chinese government central com-mittee pointed out increasing support for high-yieldingsoybean varieties and the promotion of new agronomictechniques for intercropping corn and soybean +e reasonthat China imports more soybeans is mainly reflected in thelarge domestic supply and demand gap the inadequategenetic modification technology and low tariffs China has ahigh degree of dependence on international imports ofsoybeans +erefore there is a large price fluctuation ofChinarsquos soybeans

HindawiJournal of MathematicsVolume 2021 Article ID 7765325 9 pageshttpsdoiorg10115520217765325

+e classical theory often assumes that the return issubject to conditional or unconditional normal distributionso its volatility is stable However the return of financialassets has more complex volatility in the actual markettransactions FAMA [1] studies show that although thevolatility of financial asset prices is similar in the similar timeperiods there is volatility aggregation effect+e change of thereturn rate appears a form of peak and thick tail which doesnot conform to the simple normal distribution +e returnrate is more likely to fall in the limit value region which hashigh risk Foster [2] studied the international crude oil pickupmarket and found that trading volume has a significantpositive impact on futures price volatility Liu and Wei [3]also found that there was a significant positive correlationbetween trading volume and futures price volatility Wanget al [4] used the ARCH (autoregressive conditional heter-oscedasticity) model to analyze the price fluctuation ofsoybean futures daily trading price data from 2006 to 2011+e research shows that the yield of soybean futures has asecond-order arch process and its price fluctuation hasvolatility clustering in which trading volume has a positiverole in promoting price fluctuation Tang et al [5] tooksoybean futures and wheat futures in agricultural futures asexamples to analyze the long-term correlation of futures pricefluctuations and found that the agricultural futures markethas state continuity and volatility aggregation Hua andZhong [6] Zhou and Qi [7] and Cai [8] respectively useddaily trading data of high-frequency data to analyze thecharacteristics of futures price fluctuations from differentperspectives when studying different futures Based on theEGARCH (exponential generalized autoregressive condi-tional heteroscedasticity) model Li et al [9] studied the re-lationship among the depth liquidity and volatility of CSI300 Index futures market which showed that stock indexfutures can improve the structure of capital market anddeepen the reform of capital market In the research on theprice fluctuation of China and American soybean futuresmarket experts and scholars believe that the price ofAmerican soybean futures plays a major role in the trans-mission of information and there is fluctuation spillovereffect between the two markets [10ndash16] Macroeconomicregulation and control may affect commodity price volatilityand some studies show that there is volatility spillover effectfrom energy market to agricultural product market [17ndash19]Jiang and Zhang [20] established a stochastic volatility modelby using the non-parametric Bayesian method and studiedthe data characteristics of the SSE 50 Index so as to providesuggestions for preventing financial risks Chen et al [21]predicted the volatility of Brent crude oil futures market byintroducing the hiddenMarkov model+ey believed that thepseudo-structure mutation might occur in this state and theycorrected it by using the ICSS (intra-cranial self-stimulation)model to make the results more accurate Liu et al [22]conducted modeling and prediction based on three types ofhigh-frequency extreme volatility data demonstrating thesignificance and effectiveness of high-frequency data Danet al [23] constructed a copper futures price fluctuationprediction model based on symbolic high-frequency timeseries which shows that K-NN (k-nearest neighbors)

algorithm is more accurate Cai and Liao [24] predicted thevolatility of the GEM market and measured the risk byconstructing the dynamic higher moment realized EGARCHmodel and improved the prediction accuracy and riskmeasurement accuracy

+e futures price fluctuates frequently which can reflectthe market information sensitively Domestic and foreignscholars mostly use daily high-frequency data to analyze therelationship between trading volume and price fluctuationIt is of great significance to change the perspective to analyzethe characteristics of futures price fluctuation In the soy-bean futures market most scholars study the impact ofvolatility agglomeration effect spillover effect and infor-mation symmetry on volatility while there are few studies onprice volatility using yield to express the degree of pricevolatility It is more to take the trading volume as an ex-planatory variable and add it into the model but it is less tointroduce the conditional variance equation and use therelated models to divide the forecast

In this paper the soybean futures of Dalian CommodityExchange will be researched +e trading volume and po-sition as information flow will be innovatively added to thevariance of conditional equation +e relationship betweentrading volume as conditional variance and price fluctuationwill be analyzed +e ARIMA-GJR-GARCH combinationmodel will be constructed +e relationship between tradingvolume and price fluctuation will be studied and the re-lationship between open positions and price fluctuation willalso be studied +e relationship between trading volumeand open positions is mainly divided into expected tradingvolume and open positions and unexpected trading volumeand open positions +e relationship between trading vol-ume and price fluctuations can be analyzed

2 Methods

21 (e Established Model In Chinarsquos soybean futuresmarket trading volume and open position are two impor-tant variables Trading volume is mainly used to describe theamount of market information and open position mainlyreflects the depth of the market +ey can reflect the role ofasymmetric information and the activity of market tradingTrading volume and open position can also describe thedifferent responses of speculators to price changes under thecondition of information asymmetry In the futures marketthe time series of trading volume and open position aresignificantly correlated +erefore by entering the predic-tion model the trading volume and open position are di-vided into expected and unexpected to forecast

In the financial market AR MA ARMA and ARIMAmodels are usually used to predict the transaction price orvolatility of the in-sample and out-of-sample marketsARMA and ARIMA models are established on the basis ofAR and MA and can be more comprehensively predicted+e paper will choose both expected and unexpected volumeand open interest on the fluctuation of prices which canonly be applicable to the stationary time series ARMAmodel +e model for the seasonal time series has no directmodel so the paper chooses the ARIMA model

2 Journal of Mathematics

In order to deeply and comprehensively describe thevolatility characteristics of Chinarsquos soybean futures price anARIMA-GJR-GARCH model will be constructed Takinginto account the trading volume and open interest therelationship between expected and unexpected tradingvolume and open interest and the price volatility of Chinarsquossoybean futures will be comprehensively analyzed

A ldquoleverage effectrdquo of the financial time series often exists+e characteristic of ldquoleverage effectrdquo is that the price fluc-tuations caused by bad news or negative shocks are greaterthan the price fluctuations caused by good news or positiveshocks +erefore an asymmetric ARCH model needs to beestablished+e trading volume and open interest are dividedinto expected and unexpected parts and the ARIMA-GJR-GARCHmodel is constructed to analyze the characteristics ofprice fluctuations For such non-stationary and seasonal se-quences a model can be established as follows

ϕp(L)Φp(L)(1 minus L)d 1 minus L

S1113872 1113873

Dyt θq(L)Θ Q(L)εt (1)

where p represents the order of the seasonal autoregressiveprocess SAR Q represents the order of the seasonal movingaverage process SMA p q respectively represent the orderof the non-seasonal autoregressive process AR and the orderof the non-seasonal moving average process MA d Q re-spectively represent the order of the non-seasonal andseasonal difference of the sequence yt ϕp(L)Φp(L) re-spectively represent the lag operator polynomial of the non-seasonal autoregressive process AR and the seasonalautoregressive process SAR (1 minus Ld) (1 minus LS)D respec-tively represent the order of the non-seasonal difference andthe seasonal difference lag operator of the sequence S is thestep size of the seasonal difference and θq(L)Θq(L) re-spectively represent the lag operator polynomial of the non-seasonal moving average process MA and the seasonalautoregressive process SMA

+e constructed model form is as follows

Rt α + 1113944n

j1cjRtminusj + 1113944

n

j1πj1113954σtminusj + Ut

1113954σt δ + 1113944

n

j1ωj

1113954Utminusj + 1113944

m

k1μkAk + 1113944

n

j1βj1113954σtminusj + et

1113954σt α + μ1ETVt + μ2UTVt + ρ1EOIt

+ ρ2UOIt + 1113944n


(2)

where t stands for time Rt is the price return variable on theday t ETVt represents the expected trading volume on the dayt UTVt represents the unexpected trading volume on the dayt EOIt is the expected position on day t UOIt is the un-expected position on day t and α cj πj δ ωj μk and βj arethe coefficients of the relevant variables in the above models

22 Data Collection and Selection In recent years Chinarsquossoybean futures contracts have been active on the DalianCommodity Exchange so the paper chooses the soybean

futures contract of Dalian Commodity Exchange for re-search +e volatility of soybean futures market in China ismainly caused by the fluctuation of price and yield rate sothis paper chooses soybean futures market price as the re-search object Two problems should be noted in the selectionof data Firstly unlike the stock market futures contractshave a continuous time series of stock prices Futurescontracts are discontinuous which means that they will beliquidated and stopped trading after the future trading dateSecondly because there are multiple contracts of differentdelivery months on the same trading day there will bedifferent trading prices for the same futures product on thesame trading day Because a number of contracts withdifferent delivery months will participate in the trading onthe same trading day the same futures product has differenttrading prices on the same trading day +erefore in theselection of data both the discontinuity of futures marketprices and the price of futures contracts should be taken intoaccount +is paper selects data of Dalian CommodityExchange which are from the database of China StockMarket amp Accounting Research Database (CSMAR) Inorder to make the transaction data more representative andmore convincing the contracts with large settlement pricevolume and position in the transaction contract are selectedIf two contracts are included in a day the contract with largevolume of trading volume and large position is selected +emain contract of Chinarsquos soybean futures market will beconstituted according to the above selection +e settlementprice of the contract is selected as the transaction data toform a continuous time series Soybean futures that can bedelivered in Dalian Commodity Exchange include soybeanNo 1 and soybean No 2 As a non-genetically modifiedsoybean soybean No 1 is a good representative in the studyof price fluctuation in Chinese soybean futures market+erefore the main contract of soybean No 1 is selected asthe research object A total of 1219 sample data from De-cember 31 2014 to December 31 2019 were selected

+rough logarithmic processing of the settlement pricedata of Chinarsquos soybean futures the return rate series datawere obtained and the statistical characteristics were analyzed+e formula for calculating the return rate was as follows

Rt lnPt

Ptminus1 (3)

where Rt represents the return rate of the day Pt representsthe settlement price of the day and Ptminus1 represents thesettlement price of the previous day +us the time series ofreturn rate can be obtained and analyzed

3 Empirical Analysis

31 (e Statistical Characteristics of Chinarsquos SoybeanFutures Market

311 Statistical Characteristics of Return Rate +roughlogarithmic processing of sample data the basic statisticalresults are obtained A linear chart of daily yield fluctuationof Chinarsquos soybean futures main contract is obtained asshown in Figure 1

Journal of Mathematics 3

+e fluctuation curve of yield rate directly shows thefluctuation degree of yield in China soybean futures marketIt can be seen from Figure 1 that the yield sequence ofChinarsquos soybean futures market has a volatility agglomer-ation effect that is when the volatility is large it willcontinue to fluctuate significantly for a period of time andwhen the volatility is small for a period of time it willcontinue to fluctuate to a smaller extent and the appearanceof volatility agglomeration effect often means that there willbe an ARCH effect in the sequence of returns +e emer-gence of volatility aggregation effect usually represents theARCH effect of return series

+e statistical test of the return rate sequence of Chinarsquossoybean futures market is carried out Standard deviation ismainly used to reflect the degree of dispersion betweenindividuals of data within the group Skewness is used as theskewness direction and degree of statistical data Kurtosis is astatistic used to describe the steep degree of distribution ofall values in the overall data When the data are normallydistributed K equals 33 It has a high and thin shape and athick tail that is a sharp peak and a thick tail Most financialsequences have a sharp peak and a thick tail and areasymmetrically distributed J-B test statistics are used to testwhether the sample sequence obeys normal distribution andstudy the fluctuation information of return rate +esestatistics can best reflect it +e histogram and statisticaltable of the daily return rate sequence of Chinarsquos soybeanfutures continuous contract are shown in Table 1

From the statistical results in Table 1 it can be seen thatthe skewness of the daily yield series of Chinarsquos soybeanfutures is 1436 It means that the distribution is skewed tothe right relative to the normal distribution and the rate ofreturn series has a right trailing phenomenon Kurtosis Kequals 1474 which is greater than the kurtosis value 3 in thenormal distribution It means that the yield series have thecharacteristics of sharp peaks +erefore the yield series ofChinarsquos soybean futures show the characteristics of sharp

peaks and thick tails +e J-B statistic is 7417 and thecorresponding p value is not more than 0001 so the nullhypothesis should be rejected It means that the yield seriesdo not obey the standard normal distribution Consideringthat financial time series generally have heteroscedasticityan ARCH model is considered to solve the problem+erefore stationarity test and autocorrelation test are alsoneeded for sample time return series

312 Stability Test and Partial Autocorrelation Test of ReturnRate Before establishing a model for the rate of return datait needs to be tested for stationarity autocorrelation andpartial autocorrelation and ARCH effect It has been de-termined that the model to be established subsequently haspractical significance

+ere is no unit root in the daily yield series of Chinarsquosoybean futures which is a stationary series +e autocor-relation and partial autocorrelation tests on the yield data ofChinarsquos soybean futures market are shown in Table 2

As can be seen from Table 2 various order sequencecorrelation values are two times the standard deviation of theset range It means that Chinarsquos soybean futures have noobvious yield sequence truncation and trailing phenome-non Sample sequence lag order autocorrelation and partialautocorrelation coefficient is close to zero +e sample se-quence rejects the non-autocorrelation hypothesis under thecondition of 5 confidence level +erefore there are au-tocorrelation and partial autocorrelation in the yield series ofChinarsquos soybean futures market

313 ARCH Effect Test +is paper chooses the Lagrangemultiplier test to verify whether the residual series has theARCH effect Engle and Granger [25] proposed a Lagrangemultiplier test to test whether the residual sequence has theARCH effect +e ARCH LM test results of the residualsequence are shown in Table 3

-006

-004

-002

0

002

004

006

008

01

2014

-12-

3120

15-0

2-25

2015

-04-

1320

15-0

5-28

2015

-07-

1420

15-0

8-27

2015

-10-

2120

15-1

2-04

2016

-01-

2020

16-0

3-11

2016

-04-

2720

16-0

6-15

2016

-07-

2920

16-0

9-13

2016

-11-

0720

16-1

2-21

2017

-02-

1320

17-0

3-29

2017

-05-

1720

17-0

7-04

2017

-08-

1720

17-1

0-09

2017

-11-

2220

18-0

1-08

2018

-02-

2820

18-0

4-17

2018

-06-

0420

18-0

7-19

2018

-09-

0320

18-1

0-25

2018

-12-

1020

19-0

1-25

2019

-03-

1920

19-0

5-08

2019

-06-

2420

19-0

8-07

2019

-09-

2320

19-1

1-13

2019

-12-

27

r

Figure 1 Line chart of daily yield fluctuation in Chinese soybean futures market

Table 1 Statistics of the daily return rate sequence of the main continuous contracts of Chinarsquos soybean futures

Statistics Standard deviation Skewness Kurtosis J-B statistics p valueSoybean futures daily return sequence 00086258 1436 1474 7417 le0001


According to Table 3 it can be obtained that the p valuecorresponding to each statistic of the residual squared lag oforder from 1 to 5 is no more than 0001 It means that all thelagging residual squares are jointly significant and the pvalue of the ARCH effect test is also not more than 0001 sothe null hypothesis is rejected +e residual sequence hasconditional heteroscedasticity and has an ARCH effect

32VolatilityCharacteristics ofChinarsquosSoybeanFuturesPriceA volatility curve between trading volume and open interestand yield is drawn as shown in Figure 2

According to Figure 2 it can be seen that the volatility oftrading volume is relatively large +e range of change be-tween open interest and trading volume is basically the same+e range of change in the rate of return is relatively small

Firstly the ARIMA model is used to predict the volumeand open interest in the sample +e predictable part is calledthe expected volume and the open interest which are recordedas ETVt and EOVt respectively +e relative error of thedifference between the actual value and the predicted value iscalled the unexpected transaction +e volume and open in-terest are recorded as UTVt and UOVt respectively +en theabove two variables are added to the conditional varianceequations of the GJR-GARCH (2 1) model to explore theempirical analysis of the impact of soybean futures pricefluctuations

+e regression analysis results are shown in Tables 4 and 5As can be seen from Tables 4 and 5 c represents the

coefficient of the constant term which is greater than zeroμ1 represents the estimated value of the expected tradingvolume coefficient μ2 represents the estimated value of theunexpected trading volume coefficient ρ1 represents theexpected open position coefficient and ρ2 represents theunexpected open position coefficient

According to Tables 4 and 5 the following conclusionscan be drawn

+e estimated value μ1 of expected trading volume co-efficient and the estimated value μ2 of unexpected tradingvolume coefficient are more than 0 respectively +e resultsare significant which indicates that there is a positive cor-relation between expected trading volume and unexpectedtrading volume and price fluctuation in Chinarsquos soybeanfutures market +e estimated values of the expected positioncoefficient ρ1 and the unexpected position coefficient ρ2 arenot more than 0001 and the results are significant indicatingthat there is a negative correlation between the expectedposition and the unexpected position and the price fluctu-ation in Chinarsquos soybean futures market Increasing a certainnumber of positions will reduce the impact of price fluctu-ation caused by the increase of trading volume

From the perspective of the relationship between currenttrading volume and open interest soybean futures pricefluctuations have a deeper level +e newly opened trades arefar less than the impact of closing trades and handovertransactions on price fluctuations from the perspective oftrader behavior Set the marginal impact of current tradingvolume on price fluctuations asMμ and themarginal impact ofcurrent holdings on price fluctuations as Mρ At the beginningof the new position trading it has an impact on the currenttrading volume +e current trading volume will increase butthe change of the current position is uncertain If the newtransaction is a new open position futures contract then thecurrent trading volume and position will have the sameamount of increase At this time the impact on the pricefluctuation in the soybean futures market is the sum of themarginal impact of the current trading volume on the pricefluctuation and the marginal impact of the current position onthe price fluctuation which is recorded asMμ + Mρ If the newtransaction is a soybean futures contract with closed tradingunits the impact on the current trading volume and position isthat when the trading volume increases by a certain amountthe position will decrease by a certain amount At this time theimpact on price fluctuation in soybean futures market is themarginal impact of current trading volume on price fluctuationminus the marginal impact of current position on pricefluctuation which is recorded as Mμ minus Mρ If the newtransaction is to change the number of trading units of soybeanfutures contract the impact on the current trading volume andposition is that the trading volume will have a certain amountof increase and the position will remain unchanged At thistime the impact on the price fluctuation in the soybean futuresmarket is themarginal impact of the current trading volume onthe price fluctuation which is recorded as Mμ In the soybeanfutures market there is such a relationship Mμ gt 0 Mρ lt 0then Mμ minus Mρ gtMμ gtMμ + Mρ When the soybean futuresmarket rises unilaterally the trading parties establish newtransactions +e market participants will increase their in-terests in the new trading contract so that a large number offunds will enter the new trading futures contract and themarket depth will continue to increase +us it can reduce theimpact of current trading volume changes on pricefluctuations

+e estimated value of the expected trading volumecoefficient μ1 is less than the estimated value of the unex-pected trading volume coefficient μ2 and the estimated value

Table 3 +e ARCH LM test of residual sequence

Lags(p) Chi2 Df Probgt chi21 24996 1 le00012 25293 2 le00013 26459 3 le00014 26457 4 le00015 28198 5 le0001

Table 2 +e autocorrelation and partial autocorrelation tests onthe yield data of Chinarsquos soybean futures market

Lag AC PAC Q ProbgtQ1 01433 01433 25077 le000002 00361 00159 26673 le000003 00384 00317 28476 le000004 00078 minus00026 2855 le000005 00401 00387 30525 le000006 00163 00042 30852 le000007 00086 00041 30942 000018 00112 00066 31096 000019 minus0011 minus00147 31245 0000310 00127 00145 31444 00005


of the expected position coefficient ρ1 is less than the esti-mated value of the unexpected trading volume coefficient ρ2+erefore the impact of the expected trading volume on theprice fluctuation of Chinarsquos soybean futures market is lessthan the impact of the unexpected trading volume on theprice fluctuation of Chinarsquos soybean futures market +isphenomenon can be explained as follows in Chinarsquos soybeanfutures market when new information appears there will betransactions caused by information asymmetry and liquiditydemand difference +e emergence of new information willlead to the emergence of transactions to a certain extent InChinarsquos soybean futures market the expected trading vol-ume and position are not generated by new information butmainly by market participants through changing liquiditydemand or adjusting positions +e unexpected tradingvolume and position are mainly generated by the arrival ofnew information in the soybean futures market whichcontains more information +erefore the impact of un-expected trading volume and position on the price fluctu-ation of Chinarsquos soybean futures market is stronger than thatof the expected trading volume and position on the pricefluctuation of Chinarsquos soybean futures market

33 (e Influence of Trading Volume and Open Position onARIMA-GJR-GARCH Model +e influence of trading vol-ume and open position on ARIMA-GJR-GARCH model isfurther analyzed +e trading volume is introduced into theconditional variance equation based on the ARIMA-GJR-GARCHmodel as information flow+e results are shown inTable 6

+e coefficient of current trading volume is greater thanzero and significant at the significance level of 5 in Table 6It means that there is a positive correlation between tradingvolume and price volatility After the current trading volumeis added into the conditional variance equation the GARCHterm coefficient changes from the previous 09615618 to01830656 which decreases significantly +e result is still

significant at the significance level of 5 indicating that theGARCH effect in the model is obviously weakened but theGARCH effect still exists

However the coefficient of ARIMA-GJR-GARCH termchanged from minus00453458 to 02090938 and the result wassignificant at the significance level of 5 +e coefficientchanged from negative to positive indicating that the modelno longer had asymmetry after adding volume It means thatvolume absorbs part of the persistence and asymmetry ofprice fluctuations indicating that volume has a strong abilityto explain price fluctuations In the mixed distributionhypothesis (MDH) there is a positive correlation betweenthe volatility variance of asset prices and information var-iables after the introduction of information variables +econditional expected value of trading volume mainly de-pends on information variables so there is a positive cor-relation between trading volume and price volatility in thefutures market +rough the empirical analysis of Chinarsquossoybean futures market the mixed distribution hypothesis isproved Adding the current soybean futures trading volumeinto the model as a substitute index of mixed variables has astrong ability to explain price fluctuations

Open positions are introduced into the conditionalvariance equation based on ARIMA-GJR-GARCH model asinformation flow +e results are shown in Table 7

In Table 7 the coefficient of open position in the currentperiod is less than zero and significant at the significancelevel of 5 indicating that there is a negative correlationbetween open position and price volatility and open po-sition has a strong explanatory power on the variance ofprice volatility After adding the current open position intothe conditional variance equation the GARCH term coef-ficient decreases slightly and the GARCH effect exists Itmeans that the current open position has little influence onthe persistence of soybean futures price volatility and thevolatility variance of soybean futures price still has a strongpersistence +e coefficient of ARIMA-GJR-GARCH termchanges from minus00453458 to 00158608 and the result is not

-8

-6

-4

-2

0

2

4

6

2014

-12-

3120

15-0

2-17

2015

-04-

0920

15-0

5-25

2015

-07-

0820

15-0

8-20

2015

-10-

1320

15-1

1-25

2016

-01-

0820

16-0

2-29

2016

-04-

1320

16-0

5-27

2016

-07-

1320

16-0

8-25

2016

-10-

1820

16-1

1-30

2017

-01-

1320

17-0

3-06

2017

-04-

2020

17-0

6-07

2017

-07-

2020

17-0

9-01

2017

-10-

2320

17-1

2-05

2018

-01-

1820

18-0

3-09

2018

-04-

2520

18-0

6-11

2018

-07-

2520

18-0

9-06

2018

-10-

2920

18-1

2-11

2019

-01-

2520

19-0

3-18

2019

-05-

0620

19-0

6-19

2019

-08-

0120

19-0

9-16

2019

-11-

0520

19-1

2-18

Trading volumeOpen interestYield

Figure 2 +e volatility curve of trading volume open interest and yield


significant at the significance level of 5 +e coefficientchanges from negative to positive indicating that the modelis no longer asymmetric after the addition of position+erefore open position has a strong explanatory power onprice fluctuations

4 Conclusion

+is paper analyzes the price fluctuation characteristics ofChinarsquos soybean futures market by constructing an ARIMA-GJR-GARCH model and draws the following conclusions

(1) +e price volatility of Chinarsquos soybean futures isstable +e impact of the previous shock on thevariance of the subsequent conditions is long lastingand will act on future volatility for a long time+erefore the volatility and market risk are relativelyhigh +e main reason is that Chinarsquos soybeans are

mainly derived from imports and changes in theinternational political situation will have a hugeimpact on Chinarsquos soybean futures market SecondlyChinarsquos soybean futures price fluctuations have aleverage effect+e impact of negative news is greaterthan the impact of positive news Exogenous inter-ference will affect Chinarsquos soybeans Fluctuations infutures prices have an impact

(2) In this paper the current trading volume and po-sition of Chinarsquos soybean futures contract are addedto the conditional variance equation based onARIMA-GJR-GARCH (2 1) model respectivelyand the relationship between expected and unex-pected trading volume and position and pricefluctuation is studied +e results show that theestimated value of the expected trading volumecoefficient and the estimated value of the expected

Table 6 +e relationship between trading volume and Chinarsquos soybean futures price volatility

α1 α2 β1 c1 μ

Untraded modelCoefficient estimation 04193364 minus03562949 09615618 minus00453458

Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000

Traded modelCoefficient estimation 02300784 minus0026109 01830656 02090938 1317224

Z 708 minus11 687 342 1565P 0000 0272 0000 0001 0000

Table 7 +e relationship between open position holding and price fluctuation of Chinarsquos soybean futures market

α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000

Traded modelCoefficient estimation 02467478 minus01721645 08575355 00158608 minus3950086

Z 703 minus542 5140 081 minus2016P 0000 0000 0000 0420 0000

Table 5 +e relationship between unexpected trading volume and open interest and price fluctuations in Chinarsquos soybean futures market

Coefficient Standard error z value p value 25 quantile 975 quantilec 03687889 00014382 25643 le0001 03659701 03716077μ2 1681456 00767077 2192 le0001 1531108 1831797ρ2 minus1179658 01932675 minus610 le0001 minus1558455 minus08008604α1 01919649 00284119 676 le0001 01362786 02476511α2 minus0004453 00235912 minus019 0850 minus00506908 00417848c 01492003 0057995 257 0010 00355234 02628773β1 02068837 00231923 892 le0001 01614275 02523398

Table 4 +e relationship between expected trading volume and open interest and price fluctuations in Chinarsquos soybean futures market

Coefficient Standard error z value p value 25 quantile 975 quantileC 03687882 00017501 21439 le0001 03654168 03721596μ1 09586294 01421858 674 le0001 06799503 12373090ρ1 minus1559571 07651638 minus2038 le0001 minus1709541 minus14096020α1 01640964 00397881 412 le0001 00861133 02420796α2 minus00977562 00355365 minus275 0006 minus01674065 00281059c 00745078 0029619 252 0012 00164555 01325600β1 07738417 0025848 2994 le0001 07231805 08245029


position coefficient are less than the estimated valueof the unexpected trading volume coefficient and theestimated value of the unexpected trading volumecoefficient respectively When the trading volume ofthe current soybean futures contract increases andthe open interest decreases the impact on the var-iance of price fluctuations is greater than that of thecurrent increase in trading volume and the openinterest increases or remains unchanged +ereforethe impact of the expected trading volume on theprice fluctuation of Chinarsquos soybean futures marketis less than that of the unexpected trading volume onthe price fluctuation of Chinarsquos soybean futuresmarket +e main reason is that the expression ofnew information is mainly realized by unexpectedtrading volume and new information is an impor-tant factor affecting price volatility

Data Availability

Previously reported data were used to support this study andare available at httpswwwgtarsccom

Conflicts of Interest

+e authors declare that they have no conflicts of interest

Acknowledgments

+is paper benefited from years of thinking about theseissues and discussion with many colleagues related toeconomics at that time +is research was supported by theHumanities and Social Sciences Youth Foundation ofMinistry of Education of China under grant no19YJC790128

References

[1] FAMA ldquoEfficient capital markets a review of theory andempirical workrdquo (e Journal of Finance vol 25 no 2pp 383ndash471 1970

[2] A J Foster ldquoVolume-volatility relations for crude oil futuresmarketsrdquo Journal of Futures Markets vol 15 no 8pp 929ndash951 1995

[3] X M Liu and Z X Wei ldquoDiscussion on the relationshipbetween trading volume and income volatility in wheat fu-tures marketrdquo Financial (eory amp Practice vol 11 pp 87ndash902011

[4] X D Wang B Liu and Y Yan ldquoFluctuation analysis ofChinarsquos soybean futures price based on arch modelrdquo Journalof Agrotechnical Economics vol 12 pp 73ndash79 2013

[5] YW Tang G Chen and C H Zhang ldquoAn empirical researchon the long-term correlation of the price volatility of theagricultural products futures marketsrdquo Systems Engineeringvol 12 pp 79ndash84 2005

[6] R H Hua and W J Zhong ldquoAn empirical analysis on thedynamic relationship between futures price fluctuation andtrading volume and short offer volume in Chinese futuresmarketrdquo (e Journal of Quantitative amp Technical Economicsvol 7 pp 123ndash132 2004

[7] B Zhou and Z Y Qi ldquoEmpirical research on the volatility intwo different stages of Chinese futures marketsrdquo Journal ofApplied Sport Management vol 3 pp 518ndash527 2007

[8] C Cai ldquoResearch on the price volatility of major commodityfutures in this economic crisisrdquo Financial (eory amp Practicevol 2 pp 64ndash69 2010

[9] J L Li Y Lei and S J Li ldquoMarket depth liquidity and volatilitythe impact of CSI 300 stock index futures on spot marketrdquoJournal of Financial Research vol 6 pp 124ndash138 2012

[10] T Xia and X Y Cheng ldquoResearch on the dynamic rela-tionship between domestic and foreign futures prices anddomestic spot pricesmdashmdashbased on the empirical analysis ofDCE and CBOT soybean futures markets and domesticsoybean marketsrdquo Journal of Financial Research vol 2pp 110ndash117 2006

[11] Y H Zhou and L G Zou ldquoResearch on the price relationshipbetween Chinarsquos soybean futures market and internationalsoybean futures marketmdashmdashan empirical analysis based onVAR modelrdquo Journal of Agrotechnical Economics vol 1pp 55ndash62 2007

[12] X G Li and Y H Zhou ldquoResearch on volatility spillover effectamong Chinese and international soybean future marketsrdquoJournal of Technical Economics amp Management vol 6pp 103ndash107 2014

[13] L Sun K K Ni and X G Li ldquo+e dynamic correlation offood price between domestic and abroad based on DCC-MGARCHmodelrdquo Journal of Nanjing Agricultural University(Social Sciences Edition) vol 14 no 2 pp 65ndash72 2014

[14] J Y Zheng ldquoEffects of international genetically modifiedsoybean on Chinarsquos soybean industry and its futures marketrdquoAsia-pacific Economic Review vol 5 pp 39ndash46 2015

[15] H L Wang and Y F Zhao ldquoAn analysis on price relationshipbetween China and US soybean futures markets based onstructural breaks viewpointrdquo Journal of China AgriculturalUniversity vol 21 no 9 pp 156ndash165 2016

[16] J H Liu J H Tian Y B Wang and H Z Wu ldquoVolatilityspillover effect between Chinese and American soybean fu-tures markets based on variable structure copula functionrdquoSoybean Science vol 38 no 3 pp 469ndash476 2019

[17] X Du C L Yu and D J Hayes ldquoSpeculation and volatilityspillover in the crude oil and agricultural commodity marketsa bayesian analysisrdquo Energy Economics vol 33 no 3pp 497ndash503 2011

[18] S Nazlioglu and U Soytas ldquoOil price agricultural commodityprices and the dollar a panel co-integration and causalityanalysisrdquo Energy Economics vol 34 no 4 pp 1098ndash11042012

[19] Q Ji and Y Fan ldquoHow does oil price volatility affect non-energy commodity marketsrdquo Applied Energy vol 89 no 1pp 273ndash280 2012

[20] Y Y Jiang and B Zhang ldquoNon-parameter bayesian stochasticvolatility model and its application in the financial marketsrdquoJournal of Applied Sport Management vol 38 no 1 pp 49ndash612019

[21] Z Chen Y Lin D S Huang and Y X Chen ldquoForecasting ofstructure breakthrough points in brent crude oil futuresmarketrdquo Journal of Systems amp Management vol 28 no 6pp 1095ndash1105 2019

[22] W Y Liu H Y Jiang T W Zhang and W Chen ldquoVolatilitymodeling and forecasting based on high frequency extremevalue datardquo Systems Engineering-(eory amp Practice vol 40no 12 pp 3095ndash3111 2020

[23] W Dan B Jbha and B Mrza ldquoPrediction of metal futuresprice volatility and empirical analysis based on symbolic time


series of high-frequencyrdquo Transactions of Nonferrous MetalsSociety of China vol 30 no 6 pp 1707ndash1716 2020

[24] G H Cai and Y Q Liao ldquoDynamic higher moment realizedEGARCH model and application based on structural breaksrdquo(e Journal of Quantitative amp Technical Economics vol 38no 1 pp 157ndash173 2021

[25] R F Engle and C Granger ldquoCointegration and error-cor-rection representation estimation and testingrdquo Econo-metrica vol 55 no 2 pp 251ndash276 1987


+e classical theory often assumes that the return issubject to conditional or unconditional normal distributionso its volatility is stable However the return of financialassets has more complex volatility in the actual markettransactions FAMA [1] studies show that although thevolatility of financial asset prices is similar in the similar timeperiods there is volatility aggregation effect+e change of thereturn rate appears a form of peak and thick tail which doesnot conform to the simple normal distribution +e returnrate is more likely to fall in the limit value region which hashigh risk Foster [2] studied the international crude oil pickupmarket and found that trading volume has a significantpositive impact on futures price volatility Liu and Wei [3]also found that there was a significant positive correlationbetween trading volume and futures price volatility Wanget al [4] used the ARCH (autoregressive conditional heter-oscedasticity) model to analyze the price fluctuation ofsoybean futures daily trading price data from 2006 to 2011+e research shows that the yield of soybean futures has asecond-order arch process and its price fluctuation hasvolatility clustering in which trading volume has a positiverole in promoting price fluctuation Tang et al [5] tooksoybean futures and wheat futures in agricultural futures asexamples to analyze the long-term correlation of futures pricefluctuations and found that the agricultural futures markethas state continuity and volatility aggregation Hua andZhong [6] Zhou and Qi [7] and Cai [8] respectively useddaily trading data of high-frequency data to analyze thecharacteristics of futures price fluctuations from differentperspectives when studying different futures Based on theEGARCH (exponential generalized autoregressive condi-tional heteroscedasticity) model Li et al [9] studied the re-lationship among the depth liquidity and volatility of CSI300 Index futures market which showed that stock indexfutures can improve the structure of capital market anddeepen the reform of capital market In the research on theprice fluctuation of China and American soybean futuresmarket experts and scholars believe that the price ofAmerican soybean futures plays a major role in the trans-mission of information and there is fluctuation spillovereffect between the two markets [10ndash16] Macroeconomicregulation and control may affect commodity price volatilityand some studies show that there is volatility spillover effectfrom energy market to agricultural product market [17ndash19]Jiang and Zhang [20] established a stochastic volatility modelby using the non-parametric Bayesian method and studiedthe data characteristics of the SSE 50 Index so as to providesuggestions for preventing financial risks Chen et al [21]predicted the volatility of Brent crude oil futures market byintroducing the hiddenMarkov model+ey believed that thepseudo-structure mutation might occur in this state and theycorrected it by using the ICSS (intra-cranial self-stimulation)model to make the results more accurate Liu et al [22]conducted modeling and prediction based on three types ofhigh-frequency extreme volatility data demonstrating thesignificance and effectiveness of high-frequency data Danet al [23] constructed a copper futures price fluctuationprediction model based on symbolic high-frequency timeseries which shows that K-NN (k-nearest neighbors)

algorithm is more accurate Cai and Liao [24] predicted thevolatility of the GEM market and measured the risk byconstructing the dynamic higher moment realized EGARCHmodel and improved the prediction accuracy and riskmeasurement accuracy

+e futures price fluctuates frequently which can reflectthe market information sensitively Domestic and foreignscholars mostly use daily high-frequency data to analyze therelationship between trading volume and price fluctuationIt is of great significance to change the perspective to analyzethe characteristics of futures price fluctuation In the soy-bean futures market most scholars study the impact ofvolatility agglomeration effect spillover effect and infor-mation symmetry on volatility while there are few studies onprice volatility using yield to express the degree of pricevolatility It is more to take the trading volume as an ex-planatory variable and add it into the model but it is less tointroduce the conditional variance equation and use therelated models to divide the forecast

In this paper the soybean futures of Dalian CommodityExchange will be researched +e trading volume and po-sition as information flow will be innovatively added to thevariance of conditional equation +e relationship betweentrading volume as conditional variance and price fluctuationwill be analyzed +e ARIMA-GJR-GARCH combinationmodel will be constructed +e relationship between tradingvolume and price fluctuation will be studied and the re-lationship between open positions and price fluctuation willalso be studied +e relationship between trading volumeand open positions is mainly divided into expected tradingvolume and open positions and unexpected trading volumeand open positions +e relationship between trading vol-ume and price fluctuations can be analyzed

2 Methods

21 (e Established Model In Chinarsquos soybean futuresmarket trading volume and open position are two impor-tant variables Trading volume is mainly used to describe theamount of market information and open position mainlyreflects the depth of the market +ey can reflect the role ofasymmetric information and the activity of market tradingTrading volume and open position can also describe thedifferent responses of speculators to price changes under thecondition of information asymmetry In the futures marketthe time series of trading volume and open position aresignificantly correlated +erefore by entering the predic-tion model the trading volume and open position are di-vided into expected and unexpected to forecast

In the financial market AR MA ARMA and ARIMAmodels are usually used to predict the transaction price orvolatility of the in-sample and out-of-sample marketsARMA and ARIMA models are established on the basis ofAR and MA and can be more comprehensively predicted+e paper will choose both expected and unexpected volumeand open interest on the fluctuation of prices which canonly be applicable to the stationary time series ARMAmodel +e model for the seasonal time series has no directmodel so the paper chooses the ARIMA model





S1113872 1113873




Rt α + 1113944n


n


1113954σt δ + 1113944

n

j1ωj

1113954Utminusj + 1113944

m

k1μkAk + 1113944

n



+ ρ2UOIt + 1113944n


(2)





Rt lnPt

Ptminus1 (3)














-006

-004

-002

0

002

004

006

008

01

2014

-12-

3120

15-0

2-25

2015

-04-

1320

15-0

5-28

2015

-07-

1420

15-0

8-27

2015

-10-

2120

15-1

2-04

2016

-01-

2020

16-0

3-11

2016

-04-

2720

16-0

6-15

2016

-07-

2920

16-0

9-13

2016

-11-

0720

16-1

2-21

2017

-02-

1320

17-0

3-29

2017

-05-

1720

17-0

7-04

2017

-08-

1720

17-1

0-09

2017

-11-

2220

18-0

1-08

2018

-02-

2820

18-0

4-17

2018

-06-

0420

18-0

7-19

2018

-09-

0320

18-1

0-25

2018

-12-

1020

19-0

1-25

2019

-03-

1920

19-0

5-08

2019

-06-

2420

19-0

8-07

2019

-09-

2320

19-1

1-13

2019

-12-

27

r



























-8

-6

-4

-2

0

2

4

6

2014

-12-

3120

15-0

2-17

2015

-04-

0920

15-0

5-25

2015

-07-

0820

15-0

8-20

2015

-10-

1320

15-1

1-25

2016

-01-

0820

16-0

2-29

2016

-04-

1320

16-0

5-27

2016

-07-

1320

16-0

8-25

2016

-10-

1820

16-1

1-30

2017

-01-

1320

17-0

3-06

2017

-04-

2020

17-0

6-07

2017

-07-

2020

17-0

9-01

2017

-10-

2320

17-1

2-05

2018

-01-

1820

18-0

3-09

2018

-04-

2520

18-0

6-11

2018

-07-

2520

18-0

9-06

2018

-10-

2920

18-1

2-11

2019

-01-

2520

19-0

3-18

2019

-05-

0620

19-0

6-19

2019

-08-

0120

19-0

9-16

2019

-11-

0520

19-1

2-18





4 Conclusion






α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 708 minus11 687 342 1565P 0000 0272 0000 0001 0000


α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 703 minus542 5140 081 minus2016P 0000 0000 0000 0420 0000







Data Availability




Acknowledgments


References
































S1113872 1113873




Rt α + 1113944n


n


1113954σt δ + 1113944

n

j1ωj

1113954Utminusj + 1113944

m

k1μkAk + 1113944

n



+ ρ2UOIt + 1113944n


(2)





Rt lnPt

Ptminus1 (3)














-006

-004

-002

0

002

004

006

008

01

2014

-12-

3120

15-0

2-25

2015

-04-

1320

15-0

5-28

2015

-07-

1420

15-0

8-27

2015

-10-

2120

15-1

2-04

2016

-01-

2020

16-0

3-11

2016

-04-

2720

16-0

6-15

2016

-07-

2920

16-0

9-13

2016

-11-

0720

16-1

2-21

2017

-02-

1320

17-0

3-29

2017

-05-

1720

17-0

7-04

2017

-08-

1720

17-1

0-09

2017

-11-

2220

18-0

1-08

2018

-02-

2820

18-0

4-17

2018

-06-

0420

18-0

7-19

2018

-09-

0320

18-1

0-25

2018

-12-

1020

19-0

1-25

2019

-03-

1920

19-0

5-08

2019

-06-

2420

19-0

8-07

2019

-09-

2320

19-1

1-13

2019

-12-

27

r



























-8

-6

-4

-2

0

2

4

6

2014

-12-

3120

15-0

2-17

2015

-04-

0920

15-0

5-25

2015

-07-

0820

15-0

8-20

2015

-10-

1320

15-1

1-25

2016

-01-

0820

16-0

2-29

2016

-04-

1320

16-0

5-27

2016

-07-

1320

16-0

8-25

2016

-10-

1820

16-1

1-30

2017

-01-

1320

17-0

3-06

2017

-04-

2020

17-0

6-07

2017

-07-

2020

17-0

9-01

2017

-10-

2320

17-1

2-05

2018

-01-

1820

18-0

3-09

2018

-04-

2520

18-0

6-11

2018

-07-

2520

18-0

9-06

2018

-10-

2920

18-1

2-11

2019

-01-

2520

19-0

3-18

2019

-05-

0620

19-0

6-19

2019

-08-

0120

19-0

9-16

2019

-11-

0520

19-1

2-18





4 Conclusion






α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 708 minus11 687 342 1565P 0000 0272 0000 0001 0000


α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 703 minus542 5140 081 minus2016P 0000 0000 0000 0420 0000







Data Availability




Acknowledgments


References





































-006

-004

-002

0

002

004

006

008

01

2014

-12-

3120

15-0

2-25

2015

-04-

1320

15-0

5-28

2015

-07-

1420

15-0

8-27

2015

-10-

2120

15-1

2-04

2016

-01-

2020

16-0

3-11

2016

-04-

2720

16-0

6-15

2016

-07-

2920

16-0

9-13

2016

-11-

0720

16-1

2-21

2017

-02-

1320

17-0

3-29

2017

-05-

1720

17-0

7-04

2017

-08-

1720

17-1

0-09

2017

-11-

2220

18-0

1-08

2018

-02-

2820

18-0

4-17

2018

-06-

0420

18-0

7-19

2018

-09-

0320

18-1

0-25

2018

-12-

1020

19-0

1-25

2019

-03-

1920

19-0

5-08

2019

-06-

2420

19-0

8-07

2019

-09-

2320

19-1

1-13

2019

-12-

27

r



























-8

-6

-4

-2

0

2

4

6

2014

-12-

3120

15-0

2-17

2015

-04-

0920

15-0

5-25

2015

-07-

0820

15-0

8-20

2015

-10-

1320

15-1

1-25

2016

-01-

0820

16-0

2-29

2016

-04-

1320

16-0

5-27

2016

-07-

1320

16-0

8-25

2016

-10-

1820

16-1

1-30

2017

-01-

1320

17-0

3-06

2017

-04-

2020

17-0

6-07

2017

-07-

2020

17-0

9-01

2017

-10-

2320

17-1

2-05

2018

-01-

1820

18-0

3-09

2018

-04-

2520

18-0

6-11

2018

-07-

2520

18-0

9-06

2018

-10-

2920

18-1

2-11

2019

-01-

2520

19-0

3-18

2019

-05-

0620

19-0

6-19

2019

-08-

0120

19-0

9-16

2019

-11-

0520

19-1

2-18





4 Conclusion






α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 708 minus11 687 342 1565P 0000 0272 0000 0001 0000


α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 703 minus542 5140 081 minus2016P 0000 0000 0000 0420 0000







Data Availability




Acknowledgments


References



















































-8

-6

-4

-2

0

2

4

6

2014

-12-

3120

15-0

2-17

2015

-04-

0920

15-0

5-25

2015

-07-

0820

15-0

8-20

2015

-10-

1320

15-1

1-25

2016

-01-

0820

16-0

2-29

2016

-04-

1320

16-0

5-27

2016

-07-

1320

16-0

8-25

2016

-10-

1820

16-1

1-30

2017

-01-

1320

17-0

3-06

2017

-04-

2020

17-0

6-07

2017

-07-

2020

17-0

9-01

2017

-10-

2320

17-1

2-05

2018

-01-

1820

18-0

3-09

2018

-04-

2520

18-0

6-11

2018

-07-

2520

18-0

9-06

2018

-10-

2920

18-1

2-11

2019

-01-

2520

19-0

3-18

2019

-05-

0620

19-0

6-19

2019

-08-

0120

19-0

9-16

2019

-11-

0520

19-1

2-18





4 Conclusion






α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 708 minus11 687 342 1565P 0000 0272 0000 0001 0000


α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 703 minus542 5140 081 minus2016P 0000 0000 0000 0420 0000







Data Availability




Acknowledgments


References






























4 Conclusion






α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 708 minus11 687 342 1565P 0000 0272 0000 0001 0000


α1 α2 β1 c1 μ


Z 1324 minus1082 15391 minus538P 0000 0000 0000 0000


Z 703 minus542 5140 081 minus2016P 0000 0000 0000 0420 0000







Data Availability




Acknowledgments


References






























Data Availability




Acknowledgments


References





























an empirical analysis of the price volatility

Documents