
Citation: Xin Liang, Xingfa Zhang, Yuan Li, Chunliang Deng. Daily nonparametric ARCH(1) model estimation using intraday high frequency data[J]. AIMS Mathematics, 2021, 6(4): 3455-3464. doi: 10.3934/math.2021206
[1] | Erlin Guo, Cuixia Li, Patrick Ling, Fengqin Tang . Convergence rate for integrated self-weighted volatility by using intraday high-frequency data with noise. AIMS Mathematics, 2023, 8(12): 31070-31091. doi: 10.3934/math.20231590 |
[2] | Wenhui Feng, Xingfa Zhang, Yanshan Chen, Zefang Song . Linear regression estimation using intraday high frequency data. AIMS Mathematics, 2023, 8(6): 13123-13133. doi: 10.3934/math.2023662 |
[3] | Yue Li, Yunyan Wang . Strong consistency of the nonparametric kernel estimator of the transition density for the second-order diffusion process. AIMS Mathematics, 2024, 9(7): 19015-19030. doi: 10.3934/math.2024925 |
[4] | Oussama Bouanani, Salim Bouzebda . Limit theorems for local polynomial estimation of regression for functional dependent data. AIMS Mathematics, 2024, 9(9): 23651-23691. doi: 10.3934/math.20241150 |
[5] | Ahmed Ghezal, Mohamed balegh, Imane Zemmouri . Markov-switching threshold stochastic volatility models with regime changes. AIMS Mathematics, 2024, 9(2): 3895-3910. doi: 10.3934/math.2024192 |
[6] | Gaosheng Liu, Yang Bai . Statistical inference in functional semiparametric spatial autoregressive model. AIMS Mathematics, 2021, 6(10): 10890-10906. doi: 10.3934/math.2021633 |
[7] | Junke Kou, Hao Zhang . Wavelet estimations of the derivatives of variance function in heteroscedastic model. AIMS Mathematics, 2023, 8(6): 14340-14361. doi: 10.3934/math.2023734 |
[8] | Omar Alzeley, Ahmed Ghezal . On an asymmetric multivariate stochastic difference volatility: structure and estimation. AIMS Mathematics, 2024, 9(7): 18528-18552. doi: 10.3934/math.2024902 |
[9] | Jiajia Zhao, Zuoliang Xu . Calibration of time-dependent volatility for European options under the fractional Vasicek model. AIMS Mathematics, 2022, 7(6): 11053-11069. doi: 10.3934/math.2022617 |
[10] | Fatimah Alshahrani, Wahiba Bouabsa, Ibrahim M. Almanjahie, Mohammed Kadi Attouch . $ k $NN local linear estimation of the conditional density and mode for functional spatial high dimensional data. AIMS Mathematics, 2023, 8(7): 15844-15875. doi: 10.3934/math.2023809 |
Since the autoregressive conditional heteroscedasticity (ARCH) model in Engle [1] and its generalized version, the generalized autoregressive conditional heteroscedasticity (GARCH) model in Bollerslev [2] were proposed, scholars have extended them and applied them to the financial market. See, for example, Nelson [3], Hentschel [4], Kl¨uppelberg et al. [5], Pan et al. [6], Zou et al. [7], Tetsuya [8], Davide [9], Samuel [10], Zhang et al. [11] and Linton et al. [12]. To avoid the misspecification problem for volatility function, nonparametric ARCH or GARCH models have been widely studied since the work of Engle and Ng [13]. See, for example, H¨ardle and Tsybakov [14], B¨uhlmann and McNeil [15], Yang [16], Giordano and Parrella [17], Chen et al. [18] and the references therein. With the popularity of electronic trading systems, intraday high frequency data become easily available. Such data are valuable in modeling and parameter estimation. Visser [19] utilized the intraday high frequency data to improve the estimation of daily GARCH model. The framework was further extended to the estimation of GJR-GARCH model and robust estimation of GARCH model (Huang et al. [20], Wang et al. [21]). Deng [22] utilized the intraday high frequency data to study the parameter test of daily GARCH model. Existent results following Visser [19] mainly focus on parametric GARCH type models, which may cause misspecification problem for practical data. An alternative perspective to study the volatility is from a nonparametric way.
The main goal of this paper is to use the intraday high frequency data to improve the estimation of daily nonparametric ARCH model. For simplicity, we focus on the nonparametric ARCH (1) model. Giordano and Parrella [17] showed GARCH (1, 1) model can be well approximated by a nonparametric ARCH (1) model, implying this model could be adequate in many situations. Moreover, the idea of this article can be easily extended to other nonparametric or semiparametric ARCH/GARCH models.
The rest of the paper is organized as follows. Section 2 introduces the models and estimators. Section 3 derives the asymptotic results. Section 4 investigates the estimation performance based on simulation studies. Empirical study is provided in Section 5. A concluding remark is given in Section 6. The appendix shows the proofs of theorems.
Let Yt be the daily returns of a certain asset. The nonparametric ARCH (1) model, denoted as NARCH(1), has the form
Yt=g1/2(Yt−1)ξt, | (2.1) |
where g(Yt−1) is the unknown volatility function depending on Yt−1. The errors ξt are independent and identically distributed random noises such that E(ξt)=0 and E(ξ2t)=1, and they are independent of Ys for s≤t. Model (2.1) can be rewritten as
Y2t=g(Yt−1)+g(Yt−1)(ξ2t−1). | (2.2) |
It follows that
E(Y2t|Yt−1=y)=g(y), Var(Y2t|Yt−1=y)=g2(y)(m4−1) and m4=Eξ4t. | (2.3) |
According to the equation E(Y2t|Yt−1=y)=g(y) in (2.3), the unknown function g(y) can be estimated by many nonparametric regression techniques such as kernel regression, local polynomial regression and spline method.
To introduce intraday high frequency data, following Visser [19], we provide the following nonparametric scaling model by normalizing the trading day into the unit time interval:
Yt(u)=g1/2(Yt−1)ξt(u), 0≤u≤1, | (2.4) |
where Yt(u) denotes the continuous-time intraday log-return process. On different days, the standard noise processes ξk(u) and ξl(u), k≠l, are assumed to be independent and to follow the same probability distribution. Through standardization, when u=1, we have
Eξ2t(1)=1, Yt=Yt(1), ξt=ξt(1). | (2.5) |
Let Ht≡H(Yt(u)) be a function of Yt(u) with the property of positive homogeneity, given by
Ht=H(αYt(u))=αH(Yt(u))>0, for α>0. | (2.6) |
In this paper a proxy is a positive statistic that satisfies the conditions in (2.6), which can be easily satisfied. An example is the daily realized volatility of the form
Ht=RV=√∑k(rt,k−rt,k−1)2, | (2.7) |
where rt,k denotes the return over the k-th intraday interval in day t. According to (2.4), homogeneity implies H(Yt(u))=g1/2(Yt−1)H(ξt(u)). Denote
Ht≡H(Yt(u)), ZHt≡H(ξt(u)), μ2H=EZ2Ht,ξHt=ZHt/√μ2H, gH(y)=g(y)μ2H, | (2.8) |
then we have the following nonparametric volatility proxy model:
Ht=g1/2H(Yt−1)ξHt, | (2.9) |
where E(H2t|Yt−1)=gH(Yt−1) and Eξ2Ht=1. Note that the above equation can not only use the information of high frequency data, but also retain the volatility function of NARCH(1) model except for a certain constant, gH(y)=g(y)μ2H. When Ht=|Yt(1)|=|Yt|, E(H2t|Yt−1)=gH(Yt−1) reduces to E(Y2t|Yt−1)=g(Yt−1), which means that only daily information Yt is adopted for estimation. Consequently, (2.9) includes the traditional daily model (2.1) as a special case.
In practice, {Yt}nt=1 is observable and {Ht}nt=1 can be calculated based on discrete intraday high frequency sequence. Let K(.) be a given kernel function, and h be the bandwidth. We firstly define certain symbols as follows:
V=(V2V3⋯Vn)=(H22H23⋯H2n), Z=(1V1−y1V2−y⋯⋯1Vn−1−y), |
W=diag{1nKh(V1−y),⋯,1nKh(Vn−1−y)}, Kh(u)=1hK(uh) and E1=(1,0). Recall gH(y)=E(H2t|Yt−1=y) and g(y)=E(Y2t|Yt−1=y). Then, according to Yang [16], the local linear estimator of gH(y) is given by
ˆgH(y)=E1(ZτWZ)−1ZτWV, | (2.10) |
and the local linear estimator of g(y), denoted as ˆg(y), can be obtained by setting Ht=|Yt(1)|=|Yt| in (2.10). Note μZH=gH(y)/g(y) is an unknown parameter depending on Ht. Based on (2.10), an estimator for μZH is
ˆμZH=1n−1n∑t=2ˆgH(Yt−1)ˆg(Yt−1). | (2.11) |
Then the final estimator for volatility function g(y) that takes the intraday high frequency information into account is given by
˜g(y)=ˆgH(y)ˆμZH. | (2.12) |
Before the statements of the limiting theory, we need to define some symbols and list certain assumptions. Let ||K||22=∫K2(u)du, μr(K)=∫μrK(u)du,
S=(μ0(K)μ1(K)μ1(K)μ2(K)), S−1=(S00S01S10S11), |
K∗λ(u)=1∑λ′=0Sλλ′μλ′K(u), Λλ,2=∫K∗λ(u)u2du, λ=0,1. | (3.1) |
Throughout this section, we assume (Y1,V2),⋯,(Yn−1,Vn) is a stationary α mixing sequence with mixing coefficient α(k). The following assumptions are needed for our theoretical development, which have been used in Fan and Yao [23] and Yang [16].
Assumptions:
A1 The kernel function K is bounded with a bounded support.
A2 The conditional density fY0,Yl|V1,Vl+1(y0,yl|v1,vl+1)≤A1<∞,∀l≥1.
A3 For α mixing sequences, we assume that for some δ>2 and a>1−2/δ, ∑lla[α(l)]1−2/δ<∞, E|V0|δ<∞, fY0|V1(y0|v1)≤A2<∞.
A4 For α mixing sequences, there exists a sequence of positive integers satisfying sn→∞ and sn=o{(nhn)1/2} such that (n/hn)1/2α(sn)→0 as n→∞.
A5 The random variable ξHt has a continuous density function, which is positive everywhere and mH4=Eξ4Ht<∞.
A6 The random variable Yt has a stationary density φ(y). gH(y) and φ(y) have Lipschitz continuous 2th derivatives. Further, infy∈Aφ(y)>0, where A is a compact subset of R with nonempty interior.
Theorem 1. Under assumptions A1–A6, for any fixed y∈A, as nh→∞,nh5=O(1),
√nh(ˆgH(y)−gH(y)−h2bH(y))D⟶N(0,VH(y)) |
where bH(y)=Λ0,2g(2)H(y)/2! and VH(y)=||K∗0||22(mH4−1)g2H(y)φ−1(y).
Theorem 2. Under assumptions A1–A6, if h∼n−r for some r∈(1/4,1), then as n→∞, (ˆμZH−μZH)=op(n−12) and
√nh(˜g(y)−g(y)−h2˜b(y))D⟶N(0,˜V(x)), |
where ˜b(y)=Λ0,2g2(y)/2! and ˜V(y)=(mH4−1)||K∗0||22g2(y)φ−1(y).
Remark. In Theorem 1, when Ht=|Yt|, ˆgH(y) and gH(y) become ˆg(y) and g(y) respectively. Based on Theorem 2, the revised estimator ˜g(y) retains the same bias term with ˆg(y) while the asymptotic variance terms are different. The impact of Ht lies in the quantity (mH4−1), and smaller mH4 will cause lower asymptotic variance of ˜g(y) and hence will lead to a more precise estimator.
In practice, mH4=c⋅EH4t/(EH2t)2 and c=[E(g(Yt−1))]2/E[g2(Yt−1)]. Let
mH∗4=EH4t/(EH2t)2. | (3.2) |
Then we can choose the optimal proxy Ht according to the value of mH∗4.
In this section, we assess the finite-sample performance of the proposed estimator ˜g(y). To simulate Yt and Yt(u), we first need to simulate the intraday noise process ξt(u). Following Visser [19], ξt(u) was simulated by the following two processes:
dγt(u)=−δ(γt(u)−μ)du+σγdB(2)t(u), | (4.1) |
dξt(u)=eγt(u)dB(1)t(u), u∈[0,1]. | (4.2) |
The Brownian motions B(1)t and B(2)t are uncorrelated, ξt(0)=0 and γ(0) is sampled from N(μ,σ2r). We divide the unit time interval [0,1] into 240 small intervals and set δ=1/2, σγ=1/4, μ=−1/16, g(y)=0.1+0.5y2. For each day, we consider three settings for the proxy Ht in (2.9): 5-minute realized volatility (denoted as H5t), 30-minute realized volatility (denoted as H30t), and |Yt|. Here, the realized volatility is computed according to (2.7).
To get the estimator ˜g(y) based on (2.12), the bandwidth is simply set as 1.06∗std(Yt)∗n−1/5, the kernel function is K(x)=0.75(1−x2)+. The sample sizes of n=200, 400, 800 are considered, and the replication time is set to be 100. For each proxy H5t, H30t and |Yt|, according to the 20% and 80% percentiles of the simulated Yt−1, the subset A in Theorem 1 is given as [−0.45,0.45], and the grid point vector is U=[−0.45:0.025:0.45]. For each sample size and each proxy, the 100 replicated estimated curves ˜g(Ui) are plotted in subplot of Figure 1 (green curve), together with the true function curve in bold black line for comparisons. From left to right, the three columns correspond to the cases with n=200, 400 and 800, respectively. In the first column, subplots of (ai)(i=1,2,3) are the estimation results corresponding the proxy H5t, H30t and |Yt| respectively. Similarly, subplots of (bi) and (ci)(i=1,2,3) are the estimation results for n=400 and 800 respectively. Subplots (a4),(b4) and (c4) are the box plots of mH∗4 in (3.2) for H5t, H30t and |Yt| under different sample sizes.
It is found that the estimator under the proxy H5t performs best among the three proxies considered for each sample size, especially for the case with a small sample size. This is consistent with the box plots of mH∗4 in subplots (a4),(b4) and (c4), where the estimated values under H5t are generally smaller than the others. The estimator under the proxy H30t shows more precise estimation than that of proxy |Yt|. When the sample size increases, each proxy shows better fitting performance, justifying the asymptotic normality in Theorem 2. According to the simulation results, it is shown that introducing the intraday high frequency data can significantly improve the estimation accuracy of NARCH(1) model.
In this section, based on NARCH(1) model, the proposed method is applied to estimate the volatility function of CSI (China Shanghai-Shenzhen) 300 index, which consist of the 300 largest and most liquid Chinese A-share stocks. The data span the period from 01 Sep 2017 to 12 July 2019, which consist of 466 daily observations. There are 241 price observations in each day based on the intrady sampling frequency of 1 minute. Denote Pt(u) as the t-th intraday price sequence. We can calculate the intraday log-return as
Yt(u)=100[logPt(u)−logPt−1(1)], u∈[0,1]. | (5.1) |
For each day, we first consider 11 different volatility proxies: 1-minute realized volatility H1 up to 10-minute realized volatility H10, and daily absolute return |Yt|. Here, the 1-minute proxy is computed as
H1t=√Σ240k=1[Yt(uk)−Yt(uk−1)]2, | (5.2) |
where Yt(uk) denotes the k-th observation of intraday sequence Yt(u). The formulas of other proxies are similar. According to (3.2), the estimated values of mH∗4 for proxies H1-H10 and |Yt| are: (3.9324, 1.6049, 1.6864, 1.7488, 1.8208, 1.8264, 1.7181, 1.7436, 1.7882, 1.7999, 5.5620), with H2 the smallest. To compare the impact of frequency, we consider H1, H2, H10 and |Yt| for our studies.
To estimate the volatility function for the considered data, according to the 20% and 80% percentiles of Yt−1, we set the subset A in Theorem 1 as [−0.9,0.9] and the grid point vector as U=[−0.9:0.05:0.9]. In terms of (2.8), the ratio gH(Yt−1)/g(Yt−1) equals to a constant μ2H and this property is justified by plotted {ˆgH(Ui)/ˆg(Ui)} in Figure 2, where all the ratios nearly remain constant. This implies that the proposed method is reasonable and adequate for the considered data set. From Figure 3, it is seen the volatility function estimators for g(Yt−1) show significant differences among considered proxies. According to estimated values of mH∗4, the proxy H2 is supposed to obtain the most precise estimation. From Figure 3, the volatility function under H2 shows an asymmetric behavior: the function values for Yt−1<0 are larger than those when Yt−1>0. The function shape is also analogous to that of Giordano and Parrella [17] where the volatility of Dow Jones index is studied. Based on the popular view that negative returns usually cause larger volatility, it can be found that the curves estimated by proxy H2 and H10 are similar and more reasonable than other two curves based on H1 and |Yt|. The above results could be explained as follows. H1 may contain much noise causing inefficiency of the proxy, while |Yt−1| makes no use of intraday information and could be inadequate compared to proxies H2 and H10.
In this article, an approach is given to utilize the intraday high-frequency data for the estimation of daily nonparametric ARCH(1) model, which has been widely used to forecast the volatility of financial market. The method has potential applications in estimating volatility function of financial asset where the mixed-frequency data are available. Both the theoretical and simulation results show that introducing the intraday high frequency data can significantly improve the estimation precision of daily nonparametric ARCH(1) model, compared to the cases where only daily data is used. The idea of this article is of certain novelty and can provide insights motivating future research on daily nonparametric or semiparametric ARCH/GARCH model estimation by taking the intraday high-frequency data into account.
The proof of Theorem 1 is routine and hence omitted. Detailed proof can be found in Yang [16] and Fan and Yao [23]. Next we give simple deduction for Theorem 2. Based on (2.11),
ˆμZH−μZH=1n−1n∑t=21ˆg(Yt−1){ˆgH(Yt−1)−gH(Yt−1)}−1n−1n∑t=2gH(Yt−1)ˆg(Yt−1)g(Yt−1){ˆg(Yt−1)−g(Yt−1)}≡I1+I2. |
The above two terms are analogous to the term I3 in page 383 of Yang [16]. Based on Assumptions A1–A6, when h∼n−r for some r∈(1/4,1), following the steps showing I3=op(n−1/2) in Yang [16], we can prove that both I1 and I2 are op(n−1/2) and hence ˆμZH−μZH=op(n−1/2). According to (2.10),
√nh{˜g(y)−g(y)}=√nh{1ˆμZH[ˆgH(y)−gH(y)]}−√nh{gH(y)ˆμZHμZH[ˆμZH−μZH]}≡I3+I4. |
Further, using the conclusion ˆμZH−μZH=op(n−1/2), one can show I3=(1/μZH)√nh{ˆgH(y)−gH(y)}+op(1) and I4=op(1). Consequently,
√nh{˜g(y)−g(y)}=1μZH√nh{ˆgH(y)−gH(y)}+op(1), |
and Theorem 2 holds from the asymptotic normality of ˆgH(y) in Theorem 1.
The work is partially supported by National Natural Science Foundation of China 11731015, 11571148, 11701116, Natural Science Foundation of Guangdong Province 192 2018A030310068, Young Innovative Talents Program for Colleges and Universities 191 in Guangdong Province of China 2018KQNCX241 and Guangzhou 193 University Research Funding 69-6209254, 220030401.
The authors declare no conflict of interest.
[1] | R. F. Engle, Autoregressive conditional heteroskedasticity with estimates of the variance of UK inflation, Economics, 50 (1982), 987–1008. |
[2] |
T. Bollerslev, Generalized autoregressive conditional heteroskedasticity, J. Econometrics, 31 (1986), 307–327. doi: 10.1016/0304-4076(86)90063-1
![]() |
[3] |
D. B. Nelson, Conditional heteroskedasticity in asset returns: A new approach, Econometrica: J. Econometric Soc., 59 (1991), 347–370. doi: 10.2307/2938260
![]() |
[4] |
L. Hentschel, All in the family Nesting symmetric and asymmetric GARCH models, J. Financ. Econ., 39 (1995), 71–104. doi: 10.1016/0304-405X(94)00821-H
![]() |
[5] |
C. Kl¨uppelberg, A. Lindner, R. Maller, A continuous-time GARCH process driven by a Lˊevy process: Stationarity and second-order behaviour, J. Appl. Probab., 41 (2004), 601–622. doi: 10.1239/jap/1091543413
![]() |
[6] |
J. Pan, H. Wang, H. Tong, Estimation and tests for power-transformed and threshold GARCH models, J. Econometrics, 142 (2008), 352–378. doi: 10.1016/j.jeconom.2007.06.004
![]() |
[7] |
Y. Zou, L. Yu, K. He, Estimating portfolio value at risk in the electricity markets using an entropy optimized BEMD approach, Entropy, 17 (2015), 4519–4532. doi: 10.3390/e17074519
![]() |
[8] |
T. Tetsuya, Volatility estimation using a rational GARCH model, Quant. Finance Econ., 2 (2018), 127–136. doi: 10.3934/QFE.2018.1.127
![]() |
[9] |
D. G. Davide, Forecasting volatility using combination across estimation windows: An application to S&P500 stock market index, Math. Biosci. Eng., 16 (2019), 7195–7216. doi: 10.3934/mbe.2019361
![]() |
[10] |
A. G. Samuel, Modelling the volatility of Bitcoin returns using GARCH models, Quant. Finance Econ., 3 (2019), 739–753. doi: 10.3934/QFE.2019.4.739
![]() |
[11] | X. Zhang, R. Zhang, Y. Li, S. Q. Ling, LADE-based inferences for autoregressive models with heavy-tailed G-GARCH(1, 1) noise, J. Econometrics, 2020. Available from: https://doi.org/10.1016/j.jeconom.2020.06.011. |
[12] |
O. Linton, J. Wu, A coupled component DCS-EGARCH model for intraday and overnight volatility, J. Econometrics, 217 (2020), 176–201. doi: 10.1016/j.jeconom.2019.12.015
![]() |
[13] |
R. F. Engle, V. K. Ng, Measuring and testing the impact of news on volatility, J. Finance., 48 (1993), 1749–1778. doi: 10.1111/j.1540-6261.1993.tb05127.x
![]() |
[14] |
W. Härdle, A. Tsybakov, Local polynomial estimation of the volatility function in nonparametric autoregression, J. Econometrics, 81 (1997), 223–242. doi: 10.1016/S0304-4076(97)00044-4
![]() |
[15] |
P. Bühlmann, A. J. McNeil, An algorithm for nonparametric GARCH modelling, Comput. Stat. Data Anal., 40 (2002), 665–683. doi: 10.1016/S0167-9473(02)00080-4
![]() |
[16] |
L. J. Yang, A semiparametric GARCH model for foreign exchange volatility, J. Econometrics, 130 (2006), 365–384. doi: 10.1016/j.jeconom.2005.03.006
![]() |
[17] |
F. Giordano, L. M. Parrella, Efficient nonparametric estimation and inference for the volatility function, Statistics, 53 (2019), 770–791. doi: 10.1080/02331888.2019.1615066
![]() |
[18] | X. Chen, Z. Huang, Y. Yi, Efficient estimation of multivariate semi-nonparametric GARCH filtered copula models, J. Econometrics, 2020. Available from: https://doi.org/10.1016/j.jeconom.2020.07.012. |
[19] |
M. P. Visser, Garch parameter estimation using high-frequencydata, J. Financ. Econometrics, 9 (2011), 162–197. doi: 10.1093/jjfinec/nbq017
![]() |
[20] |
J. S. Huang, W. Q. Wu, Z. Chen, J. J. Zhou, Robust M-estimate of GJR model with high frequency data, Acta Math. Appl. Sin., 31 (2015), 591–606. doi: 10.1007/s10255-015-0488-y
![]() |
[21] |
M. Wang, Z. Chen, C. D. Wang, Composite quantile regression for GARCH models using high-frequency data, Econometrics Stat., 7 (2018), 115–133. doi: 10.1016/j.ecosta.2016.11.004
![]() |
[22] |
C. Deng, X. Zhang, Y. Li, Q. Xiong, Garch model test using high-frequency data, Mathematics, 8 (2020), 1922. doi: 10.3390/math8111922
![]() |
[23] | J. Q. Fan, Q. W. Yao, Nonlinear Time Series: Nonparametric and Parametric Methods, New York: Springer, 2005. |
1. | Xiaoling Chen, Xingfa Zhang, Yuan Li, Qiang Xiong, Daily LGARCH model estimation using high frequency data, 2021, 1, 2769-2140, 165, 10.3934/DSFE.2021009 | |
2. | Fangrou Chai, Xingfa Zhang, Yuan Li, Yanshan Chen, The nonparametric GARCH model estimation using intraday high-frequency data, 2024, 0361-0918, 1, 10.1080/03610918.2024.2374900 | |
3. | Fangrou Chai, Yuan Li, Xingfa Zhang, Zhongxiu Chen, Daily Semiparametric GARCH Model Estimation Using Intraday High-Frequency Data, 2023, 15, 2073-8994, 908, 10.3390/sym15040908 | |
4. | Yanshan Chen, Xingfa Zhang, Chunliang Deng, Yujiao Liu, Portmanteau Test for ARCH-Type Models by Using High-Frequency Data, 2024, 13, 2075-1680, 141, 10.3390/axioms13030141 |