Processing math: 76%
Research article

Parameter estimation for discretely observed Cox–Ingersoll–Ross model driven by fractional Lévy processes

  • Received: 29 January 2023 Revised: 03 March 2023 Accepted: 17 March 2023 Published: 22 March 2023
  • MSC : 62E20, 62F03, 62F12, 62P05, 62P20

  • This paper deals with least squares estimation for the Cox–Ingersoll–Ross model with fractional Lévy noise from discrete observations. The contrast function is given to obtain the least squares estimators. The consistency and asymptotic distribution of estimators are derived when a small dispersion coefficient ε0, n, εn12d0, and nε simultaneously.

    Citation: Jiangrui Ding, Chao Wei. Parameter estimation for discretely observed Cox–Ingersoll–Ross model driven by fractional Lévy processes[J]. AIMS Mathematics, 2023, 8(5): 12168-12184. doi: 10.3934/math.2023613

    Related Papers:

    [1] Jafar Biazar, Fereshteh Goldoust . Multi-dimensional Legendre wavelets approach on the Black-Scholes and Heston Cox Ingersoll Ross equations. AIMS Mathematics, 2019, 4(4): 1046-1064. doi: 10.3934/math.2019.4.1046
    [2] Xiuxian Chen, Zhongyang Sun, Dan Zhu . Mean-variance investment and risk control strategies for a dynamic contagion process with diffusion. AIMS Mathematics, 2024, 9(11): 33062-33086. doi: 10.3934/math.20241580
    [3] Huiping Jiao, Xiao Zhang, Chao Wei . $ L_{\infty} $-norm minimum distance estimation for stochastic differential equations driven by small fractional Lévy noise. AIMS Mathematics, 2023, 8(1): 2083-2092. doi: 10.3934/math.2023107
    [4] Guiwen Lv, Ping Xu, Yanxue Zhang . Pricing of vulnerable options based on an uncertain CIR interest rate model. AIMS Mathematics, 2023, 8(5): 11113-11130. doi: 10.3934/math.2023563
    [5] Abdulaziz Alsenafi, Mishari Al-Foraih, Khalifa Es-Sebaiy . Least squares estimation for non-ergodic weighted fractional Ornstein-Uhlenbeck process of general parameters. AIMS Mathematics, 2021, 6(11): 12780-12794. doi: 10.3934/math.2021738
    [6] Riza Andrian Ibrahim, Sukono, Herlina Napitupulu, Rose Irnawaty Ibrahim . Modeling earthquake bond prices with correlated dual trigger indices and the approximate solution using the Monte Carlo algorithm. AIMS Mathematics, 2025, 10(2): 2223-2253. doi: 10.3934/math.2025103
    [7] Tian Xu, Ailong Wu . Stabilization of nonlinear hybrid stochastic time-delay neural networks with Lévy noise using discrete-time feedback control. AIMS Mathematics, 2024, 9(10): 27080-27101. doi: 10.3934/math.20241317
    [8] Danhua He, Liguang Xu . Boundedness analysis of non-autonomous stochastic differential systems with Lévy noise and mixed delays. AIMS Mathematics, 2020, 5(6): 6169-6182. doi: 10.3934/math.2020396
    [9] Teh Raihana Nazirah Roslan, Sharmila Karim, Siti Zulaiha Ibrahim, Ali Fareed Jameel, Zainor Ridzuan Yahya . Stochastic pricing formulation for hybrid equity warrants. AIMS Mathematics, 2022, 7(1): 398-424. doi: 10.3934/math.2022027
    [10] Jingjing Yang, Jianqiu Lu . Stabilization in distribution of hybrid stochastic differential delay equations with Lévy noise by discrete-time state feedback controls. AIMS Mathematics, 2025, 10(2): 3457-3483. doi: 10.3934/math.2025160
  • This paper deals with least squares estimation for the Cox–Ingersoll–Ross model with fractional Lévy noise from discrete observations. The contrast function is given to obtain the least squares estimators. The consistency and asymptotic distribution of estimators are derived when a small dispersion coefficient ε0, n, εn12d0, and nε simultaneously.



    Stochastic differential equations are an important tool to describe random phenomena, and parameter estimation is a non-negligible problem in the study of stochastic differential equations. In the past few decades, some scholars have studied the parameter estimation problem of stochastic models. For example, Bishwa [1] presented the estimation of the unknown parameters in stochastic differential models based on continuous and discrete observations and examined asymptotic properties of several estimators under maximum likelihood, minimum contrast and Bayesian methods. Genon-Catalot [2] studied the asymptotic distribution of the maximum contrast estimator of a one-dimensional diffusion process based on the sample path and gave the condition that the maximum contrast estimator is asymptotically equivalent to the maximum likelihood estimator. Pasemann and Stannat [3] used the maximum likelihood method to study the parameter estimation problem of a class of semi-linear stochastic evolution equations and gave the conditions of consistency and asymptotic normality according to the growth and continuity of the nonlinear part. Cialenco et al. [4] studied the asymptotic properties of maximum likelihood estimators for stochastic partial differential equations. Neuenkirch and Tindel [5] derived the least squares estimator for stochastic differential equations with additive fractional noise and verified its strong consistency. Fei [6] studied the consistency of least squares estimation of parameters of stochastic differential equations under distributional uncertainty. Karimi and McAuley [7] developed a Bayesian algorithm for estimating measurement noise variance, disturbance strength, and model parameters in nonlinear stochastic differential equation models. Kan et al. [8] studied the parameter estimation of linear stochastic systems using Bayesian methods. When the system is observed discretely, Wei et al. [9] studied the problem of parameter estimation for square radial Ornstein-Uhlenbeck processes driven by α-stable noise from discrete observations. Long [10] performed parameter estimation for stochastic differential equations driven by small stable noise under discrete observations. Wei [11] focused on the parameter estimation problem of the stochastic Lotka-Volterra model driven by small Lévy noise, verifying the consistency of the estimator and the asymptotic distribution of the estimation error.

    Among them, the Lévy process is a natural noise model in the random process, but it has long-term dependence. It is inappropriate to describe random fluctuations in many aspects only by the Lévy process or even the general Markov process without aftereffect. Fractional Lévy noise, as an important non-Gaussian noise, can more accurately reflect actual fluctuations. Because of this, more and more scholars have devoted themselves to the qualitative analysis of stochastic differential equations driven by fractional Lévy processes. Laskin et al. [12] explored fractional Lévy process probability density functions and applied them in network traffic modeling. Lin et al. [13] studied the existence and joint continuity of local time for multiparameter fractional Lévy processes. Bender et al. [14,15] discussed a certain class of stochastic integrals of real-valued fractional Lévy processes and the finite variation of fractional Lévy processes. Lu and Dai [16] explored the stochastic integration of pure jump 0-mean square-integrable fractional Lévy processes and their driven stochastic differential equations. Bender et al. [17] listed maximum inequalities for fractional Lévy and related processes. Shen et al. [18] studied the problem of parameter estimation for Ornstein-Uhlenbeck processes driven by fractional Lévy processes. Bishwal [19] incorporated jumps and long memory into the volatility process driven by fractional Lévy processes to obtain refined inference results like confidence interval. Boniece et al. [20] studied sample path properties and stochastic integration with respect to fractional Lévy processes. Rao [21] studied the problem of nonparametric estimation of linear multipliers of stochastic differential equations driven by small noise fractional Lévy processes. Prakasa [22] described and predicted trends in nonparametric estimation of stochastic differential equations driven by fractional Lévy processes.

    The Cox–Ingersoll–Ross (CIR) model, proposed in 1985 [23,24], is an extension of the Vasicek model and solves the problem of possible negative interest rates. However, due to the observational discontinuity and heavy tails of financial samples, the CIR model cannot capture these characteristics [25], and it is necessary to replace the Brownian motion in the CIR model with other processes. There have been some studies, such as Ma et al. [26], who studied the central limit theorem of the least squares estimator of the CIR model driven by α-stable noise. Li and Ma [27] derived the consistency of weighted least squares estimators in the CIR model. Yang [28] studied the maximum likelihood estimation of discrete-observation CIR models with small α-stable noise. There are few studies on building CIR models driven by fractional Lévy processes and making statistical inferences, and the equations in the existing literature contain only one unknown parameter. Inspired by the above results, in this paper, we consider the model with two unknown parameters. Parameter estimation and statistical inference are performed on stochastic financial models driven by fractional Lévy processes. Research on the consistency of estimators, the asymptotic distribution of estimation errors, hypothesis testing and other asymptotic properties can reflect the effectiveness of estimators and estimation methods and can more accurately grasp the dynamic changes of assets. Therefore, this study has certain practical value and significance.

    Definition 1. (Marquardt [29]) Let L={L(t),tR} be a zero-mean two-sided Lévy process with E[L(1)2]< and without a Brownian component. For fractional integration parameter d(0,12), a stochastic process

    Ldt:=1Γ(d+1)[(ts)d+(s)d]L(ds),tR

    is called a fractional Lévy process, where x+=x0.

    In this paper, we study the parametric estimation problem for the Cox–Ingersoll–Ross model driven by fractional Lévy processes described by the following stochastic differential equation:

    {dXt=(αβXt)dt+εXtdLdt,t[0,1],d(0,12)X0=x0 (1)

    where α and β are unknown parameters, and Ldt is a fractional Lévy process. Assume that this process is observed at n regularly spaced time points {ti=in,i=1,2,...,n}, ε[0,1].

    To get the least squares estimators, we introduce the following contrast function:

    ρn,ε(α,β)=ni=1|XtiXti1(αβXti1)Δti1|2 (2)

    where Δti1=titi1=1n. Then, the least squares estimators ˆαn,ε and ˆβn,ε are defined as follows:

    ρn,ε(ˆαn,ε,ˆβn,ε)=minρn,ε(α,β).

    It is easy to obtain the least square estimators:

    {ˆαn,ε=nni=1(XtiXti1)Xti1ni=1Xti1(ni=1Xti1)2nni=1X2ti1nni=1(XtiXti1)ni=1X2ti1(ni=1Xti1)2nni=1X2ti1ˆβn,ε=n2ni=1(XtiXti1)Xti1nni=1(XtiXti1)ni=1Xti1(ni=1Xti1)2nni=1X2ti1. (3)

    Let X0=(X0t,t0) be the solution to the underlying ordinary differential equation under the true values of the parameters:

    dX0t=(αβX0t)dt,X00=x0. (4)

    Note that

    XtiXti1=1nαβtiti1Xsds+εtiti1XsdLds. (5)

    Bring the formula into (3), and a more explicit decomposition for ˆαn,ε and ˆβn,ε can be obtained:

    ˆαn,ε=α+nβni=1titi1Xsdsni=1X2ti1(ni=1Xti1)2nni=1X2ti1nβni=1Xti1titi1Xsdsni=1Xti1(ni=1Xti1)2nni=1X2ti1
    +nεni=1Xti1titi1XsdLdsni=1Xti1(ni=1Xti1)2nni=1X2ti1nεni=1titi1XsdLdsni=1X2ti1(ni=1Xti1)2nni=1X2ti1
    =α+βni=1titi1Xsds1nni=1X2ti1(1nni=1Xti1)21nni=1X2ti1βni=1Xti1titi1Xsds1nni=1Xti1(1nni=1Xti1)21nni=1X2ti1
    +εni=1Xti1titi1XsdLds1nni=1Xti1(1nni=1Xti1)21nni=1X2ti1εni=1titi1XsdLds1nni=1X2ti1(1nni=1Xti1)21nni=1X2ti1,
    ˆβn,ε=nβni=1titi1Xsdsni=1Xti1n2ni=1Xti1ni=11Xti1n2βni=1Xti1titi1Xsdsn2ni=1Xti1ni=11Xti1
    +n2εni=1Xti1titi1XsdLdsn2ni=1Xti1ni=11Xti1nεni=1titi1XsdLdsni=1Xti1n2ni=1Xti1ni=11Xti1
    =βni=1titi1Xsds1nni=1Xti1(1nni=1Xti1)21nni=1X2ti1βni=1Xti1titi1Xsds(1nni=1Xti1)21nni=1X2ti1
    +εni=1Xti1titi1XsdLds(1nni=1Xti1)21nni=1X2ti1εni=1titi1XsdLds1nni=1Xti1(1nni=1Xti1)21nni=1X2ti1.

    Before giving the theorems, we need to establish some preliminary results.

    Lemma 1. ([29]) Let |f|,|g|H, where H is the completion of L1(R)L2(R) with respect to the norm g2H=E[L2(1)]R(Idg)2(u)du. Then,

    E[Rf(s)dLdsRg(s)dLds]=Γ(12d)E[L2(1)]Γ(d)Γ(1d)RRf(t)g(s)|ts|2d1dsdt.

    Lemma 2. ([29]) For any 0b2b1a1,0b2a2a1,b1b2=a1a2, there exists a constant C that only depends on r and d such that

    |b1b2a1a2er(u+v)|uv|2d1dudv|{C|er(a1+b1)er(a2+b2)||a1b2|2d,r0C|a1b2|2d,r=0

    where r denotes a constant, and d is the fractional integration parameter of the fractional Lévy process.

    Lemma 3. When ε0,n, we have

    sup0t1|XtX0t|P0.

    Proof. Observe that

    XtX0t=βt0(XsX0s)ds+εt0XsdLds. (6)

    By using the Cauchy-Schwarz inequality, we obtain

    |XtX0t|22β2|t0(XsX0s)ds|2+2ε2|t0XsdLds|2
    2β2tt0|XsX0s|ds+2ε2sup0t1|t0XsdLds|2.

    According to Gronwall's inequality, we get

    |XtX0t|22ε2e2β2t2sup0t1|t0XsdLds|2. (7)

    Then,

    sup0t1|XtX0t|2εeβ2sup0t1|t0XsdLds|. (8)

    By the Markov inequality and the results in Lemma 1 and Lemma 2, when f(s)=g(s)=1,r=0, for any given δ>0, when ε0, we have

    P(2εeβ2sup0t1|t0XsdLds|>δ)2δ2ε2e2β2E[sup0t1|t0XsdLds|]22δ2ε2e2β2E[|t0XsdLds|]22Cδ2ε2e2β21010|ts|2d1dsdt2Cδ2ε2e2β20

    where C is a constant.

    Therefore, it is easy to check that

    sup0t1|XtX0t|P0. (9)

    The proof is complete.

    Lemma 4. When ε0,n, we have

    1nni=1X2ti1p10(X0t)2dt.

    Proof. Since

    1nni=1X2ti1 = 1nni=1(X0ti1)2+1nni=1(X2ti1(X0ti1)2), (10)

    it is clear that

    1nni=1(X0ti1)2P10(X0t)2dt. (11)

    For 1nni=1(X2ti1(X0ti1)2), according to Lemma 3 and the fact that 1nni=1|X0ti1|P10|X0t|dt, when ε0,n, we have

    |1nni=1(X2ti1(X0ti1)2)|=|1nni=1(Xti1+X0ti1)(Xti1X0ti1)|1nni=1|Xti1X0ti1|(|Xti1|+|X0ti1|)1nni=1(|Xti1X0ti1|2+2|X0ti1||Xti1X0ti1|)=1nni=1|Xti1X0ti1|2+21nni=1|X0ti1||Xti1X0ti1|(sup0t1|XtX0t|)2+2sup0t1|XtX0t|1nni=1|X0ti1|P0

    and therefore obtain

    1nni=1X2ti1p10(X0t)2dt. (12)

    The proof is complete.

    Lemma 5. When ε0,n, we have

    1nni=11Xti1p101X0tdt.

    Proof.

    1nni=11Xti1=1nni=11X0ti1+1nni=1(1Xti11X0ti1),

    and

    1nni=11X0ti1P101X0tdt.

    Assume inf0t1{Xt}=XN>0. According to Lemma 1, when ε0,n, it can be checked that

    |1nni=1(1Xti11X0ti1)|=|1nni=1X0ti1Xti1Xti1X0ti1|1nni=1|X0ti1Xti1||Xti1X0ti1|sup0t1|XtX0t||XtX0t|sup0t1|XtX0t|X2NP0.

    Therefore, we obtain

    1nni=11Xti1p101X0tdt.

    The proof is complete.

    In the following theorem, the consistency of the least squares estimators is proved.

    Theorem 1. When ε0,n,εn12d0, the least squares estimators ˆα,ˆβ are consistent, namely,

    ˆαn,εPα,ˆβn,εPβ.

    Proof. According to Lemma 4 and Lemma 5, it is clear that

    11nni=1Xti11nni=11Xti1p110X0tdt101X0tdt. (13)
    (1nni=1Xti1)21nni=1X2ti1p(10X0tdt)210(X0t)2dt. (14)

    When ε0,n, it can be verified that

    βni=1titi1XsXti1ds1nni=1Xti1Pβ10XtX0tdt10X0tdt (15)

    and

    βni=1titi1XsdsPβ10Xtdt. (16)

    At the same time, according to Lemma 3, the limits of items 1 and 2 of the detailed decomposition formula ˆα can be obtained.

    βni=1titi1XsXti1ds1nni=1Xti1βni=1titi1XsdsP0. (17)

    Since

    P(|εni=1titi1XsdLds|>δ)P(εni=1|titi1XsdLds|>δ),

    by the Markov inequality,

    P(εni=1|titi1XsdLds|>δ)δ2ε2E(ni=1|titi1XsdLds|)2δ2ε2nni=1E|titi1XsdLds|2Cδ2ε2n2sup0t1titi1titi1E[XsXt]|ts|2d1dsdtCδ2(εn12d)2E[XM]0.

    When ε0,n,εn1d0, we obtain

    εni=1titi1XsdLdsP0. (18)

    According to Lemma 5 and (18), it is obvious that

    εni=1titi1XsdLds11nni=1Xti11nni=11Xti1P0. (19)

    Then, we have

    |εni=1titi1XsXti1dLds|
    =|εni=11Xti1titi1XsdLds|εni=1|1Xti1||titi1XsdLds|
    εni=1(|1X0ti1|+|1Xti11X0ti1|)|titi1XsdLds|
    εni=1|1X0ti1||titi1XsdLds|+εsup0t1|1Xti11X0ti1||titi1XsdLds|.

    By the Markov inequality, we obtain

    P(|εni=1|1X0ti1||titi1XsdLds||>δ)δ2ε2(ni=1|1X0ti1|)2E|titi1XsdLds|2Cδ2ε2(ni=1|1X0ti1|)2titi1titi1XsXt|ts|2d1dsdtCδ2(εn1d)2(1nni=1|1X0ti1|)20,

    which implies that εni=1|1X0ti1||titi1XsdLds|P0 as ε0,n,εn12d0.

    According to Lemma 3, when ε0,n, it is obvious that

    εsup0t1|1Xti11X0ti1||titi1XsdLds|P0. (20)

    Then, we have

    εni=1titi1XsXti1dLds1nni=1Xti1P0. (21)

    Therefore, with the results of (17), (19) and (21), when ε0,n,εn12d0, we have

    ˆαn,εPα.

    Using the same methods, it can be easily checked that

    βni=1titi1XsXti1dsPβ. (22)
    βni=1titi1Xsds1nni=11Xti1Pβ10X0tdt101X0tdt. (23)
    βni=1titi1XsXti1ds11nni=1Xti11nni=11Xti1βni=1titi1Xsds1nni=11Xti111nni=1Xti11nni=11Xti1Pβ. (24)

    Moreover,

    εni=1titi1XsdLds1nni=11Xti111nni=1Xti11nni=11Xti1P0. (25)
    εni=1titi1XsXti1dLds11nni=1Xti11nni=11Xti1P0. (26)

    Therefore, when ε0,n,εn12d0, we have ˆβn,εPβ.

    The proof is complete.

    Theorem 2. When ε0,n,εn12d0,nε,

    ε1(ˆαn,εα)d10X0sdLds101X0sdLds10X0sds110X0sds101X0sdsε1(ˆβn,εβ)d10X0sdLds101X0sds101X0sdLds110X0sds101X0sds.

    Proof. According to the explicit decomposition for ˆαn,ε, it is obvious that

    ε1(ˆαn,εα)=ε1βni=1titi1XsXti1ds1nni=1Xti111nni=1Xti11nni=11Xti1ε1βni=1titi1Xsds11nni=1Xti11nni=11Xti1
    +ni=1titi1XsdLds11nni=1Xti11nni=11Xti1ni=1titi1XsXti1dLds1nni=1Xti111nni=1Xti11nni=11Xti1.

    From Lemma 3, when ε0,n,nε,

    |ε1βni=11Xti1titi1Xsds|ε1βni=1|1Xti1||titi1Xsds|ε1n1βni=1(|1Xti11X0ti1|+|1X0ti1|)supti1tti|Xt|P0.

    Then, it is easy to check that

    ε1βni=1titi1XsdsP0..

    Combining with Lemma 4, we have

    ε1βni=1titi1XsXti1ds1nni=1Xti111nni=1Xti11nni=11Xti1P0, (27)

    and

    ε1βni=1titi1Xsds11nni=1Xti11nni=11Xti1P0.. (28)

    Since

    ni=1titi1XsdLds = ni=1titi1X0sdLds+ni=1titi1(XsX0s)dLds, (29)

    using Markov's inequality, for any given δ>0, we have

    P(|ni=1titi1(XsX0s)dLds|>δ)δ2E|ni=1titi1(XsX0s)dLds|2δ2ni=1E|titi1(XsX0s)dLds|2Cδ2nsup0t1Etiti1titi1(XsX0s)(XtX0t)|ts|2d1dsdt
    0.

    Moreover,

    ni=1titi1X0sdLds=10X0sdLds,ni=1titi1XsdLdsP10X0sdLds. (30)
    ni=1titi1XsXti1dLds1nni=1Xti1P101X0sdLds10X0sds. (31)

    It is obvious that

    (32)

    According to the detailed decomposition formula of ˆβn,ε, we get

    Then, we have

    (33)

    and

    (34)

    Then, we have

    (35)

    The proof is complete.

    In this experiment, we use an iterative approach to generate a discrete sample (Xti1)i=1,...,n and compute ˆαn,ε and ˆβn,ε from the sample. We let x0=0.01 and d=0.02. The first column of the table is the true value of the parameter (α,β). The size of the sample is represented as "Size n" and given in the table. In Table 1, ε=0.1, and the size is increasing from 1000 to 5000. In Table 2, ε=0.01, and the size is increasing from 10000 to 50000. The table lists the values of "ˆαn,ε", " {\widehat{\beta }}_{n, \varepsilon } " and the absolute errors (AE) of least squares estimators.

    Table 1.  Least squares estimator simulation results of \alpha and \beta .
    True Aver AE
    (\alpha , \beta ) size n {\hat \alpha _{n, \varepsilon }} {\hat \beta _{n, \varepsilon }} \left| {{{\hat \alpha }_{n, \varepsilon }} - \alpha } \right| \left| {{{\hat \beta }_{n, \varepsilon }} - \beta } \right|
    (1, 1) 1000 1.2162 1.2061 0.2162 0.2061
    2000 1.0823 1.1071 0.0823 0.1071
    5000 1.0421 1.0529 0.0421 0.0529
    (2, 3) 1000 2.2377 3.1907 0.2377 0.1907
    2000 2.1193 3.1249 0.1193 0.1249
    5000 2.0524 3.0693 0.0524 0.0693
    (4, 5) 1000 4.2556 5.2294 0.2556 0.2294
    2000 4.1372 5.1291 0.1372 0.1291
    5000 4.0583 5.0487 0.0583 0.0487

     | Show Table
    DownLoad: CSV
    Table 2.  Least squares estimator simulation results of \alpha and \beta .
    True Aver AE
    (\alpha , \beta ) size n {\hat \alpha _{n, \varepsilon }} {\hat \beta _{n, \varepsilon }} \left| {{{\hat \alpha }_{n, \varepsilon }} - \alpha } \right| \left| {{{\hat \beta }_{n, \varepsilon }} - \beta } \right|
    (1, 1) 10000 1.1265 1.1182 0.1265 0.1182
    20000 1.0372 1.0525 0.0372 0.0525
    50000 1.0017 1.0012 0.0017 0.0012
    (2, 3) 10000 2.1373 3.1264 0.1373 0.1264
    20000 1.9432 3.0473 0.0568 0.0473
    50000 2.0026 3.0037 0.0026 0.0037
    (4, 5) 10000 4.1775 5.1643 0.1775 0.1643
    20000 4.0413 5.0518 0.0413 0.0518
    50000 4.0041 5.0032 0.0041 0.0032

     | Show Table
    DownLoad: CSV

    The two tables indicate that the absolute error between the estimator and the true value depends on the size of the true value samples for any given parameter. According to the simulation results, when n is large enough, and \varepsilon is small enough, the estimator is very close to the true parameter value. If we let n go to infinity and \varepsilon converge to zero, the estimator will converge to the true value.

    Fractional Lévy noise, as an important non-Gaussian noise, can more accurately reflect actual fluctuations. Because of this, more and more scholars have devoted themselves to the qualitative analysis of stochastic differential equations driven by fractional Lévy processes. Due to the observational discontinuity and heavy tails of financial samples, the CIR model cannot capture these characteristics, and it is necessary to replace the Brownian motion in the CIR model with fractional Lévy noise.

    The purpose of this paper is to estimate the parameters of the CIR model driven by a fractional Lévy process with discrete observations. First, the comparison function is introduced to obtain the explicit expression of the least square estimator. Then, the consistency and asymptotic distribution of the estimator are derived according to the Markov inequality, Gronwall inequality and Cauchy-Schwarz inequality. The research topic can be extended to the parameter estimation problem for other stochastic models driven by fractional Lévy process.

    The authors declare that there are no conflicts of interest.



    [1] J. P. N. Bishwal, Parameter estimation in stochastic differential equations, Springer, 2007.
    [2] V. Genon-Catalot, Maximnm contrast estimation for diffusion processes from discrete observations, Statistics, 21 (1990), 99–116.
    [3] G. Pasemann, W. Stannat, Drift estimation for stochastic reaction-diffusion systems, Electron. J. Stat., 14 (2020), 547–579. https://doi.org/10.1214/19-EJS1665 doi: 10.1214/19-EJS1665
    [4] I. Cialenco, F. Delgado-Vences, H. J. Kim, Drift estimation for discretely sampled SPDEs, Stochastic Partial Differ. Equ.: Anal. Comput., 8 (2020), 895–920. https://doi.org/10.1007/s40072-019-00164-4 doi: 10.1007/s40072-019-00164-4
    [5] A. Neuenkirch, S. Tindel, A least square-type procedure for parameter estimation in stochastic differential equations with additive fractional noisem, Stat. Inference Stochastic Proc., 17 (2014), 99–120. https://doi.org/10.1007/s11203-013-9084-z doi: 10.1007/s11203-013-9084-z
    [6] C. Fei, W. Fei, Consistency of least squares estimation to the parameter for stochastic differential equations under distribution uncertainty, arXiv, 2019. https://doi.org/10.48550/arXiv.1904.12701 doi: 10.48550/arXiv.1904.12701
    [7] H. Karimi, K. B. McAuley, Bayesian estimation in stochastic differential equation models via laplace approximation, IFAC-PapersOnLine, 49 (2016), 1109–1114. https://doi.org/10.1016/j.ifacol.2016.07.351 doi: 10.1016/j.ifacol.2016.07.351
    [8] X. Kan, H. Shu, Y. Che, Asymptotic parameter estimation for a class of linear stochastic systems using Kalman-Bucy filtering, Math. Prob. Eng., 2012 (2012), 342705. https://doi.org/10.1155/2012/342705 doi: 10.1155/2012/342705
    [9] C. Wei, D. Li, H. Yao, Parameter estimation for squared radial Ornstein-Uhlenbeck process from discrete observation, Eng. Lett., 29 (2021).
    [10] H. Li, Parameter estimation for a class of stochastic differential equations driven by small stable noises from discrete observations, Acta Math. Sci., 30 (2010), 645–663. https://doi.org/10.1016/S0252-9602(10)60067-7 doi: 10.1016/S0252-9602(10)60067-7
    [11] C. Wei, Parameter estimation for stochastic Lotka-Volterra model driven by small Lévy noises from discrete observations, Commun. Stat.-Theory Methods, 50 (2021), 6014–6023. https://doi.org/10.1080/03610926.2020.1738489 doi: 10.1080/03610926.2020.1738489
    [12] N. Laskin, I. Lambadaris, F. C. Harmantzis, M. Devetsikiotis, Fractional Lévy motion and its application to network traffic modeling, Comput. Networks, 40 (2002), 363–375. https://doi.org/10.1016/S1389-1286(02)00300-6 doi: 10.1016/S1389-1286(02)00300-6
    [13] Z. Lin, Z. Cheng, Existence and joint continuity of local time of multi-parameter fractional Lévy processes, Appl. Math. Mech., 30 (2009), 381–390. https://doi.org/10.1007/s10483-009-0312-y doi: 10.1007/s10483-009-0312-y
    [14] C. Bender, T. Marquardt, Stochastic calculus for convoluted Lévy processes, Bernoulli, 14 (2008), 499–518.
    [15] C. Bender, A. Lindner, M. Schicks, Finite variation of fractional Lévy processes, J. Theor. Probab., 25 (2012), 594–612.
    [16] X. Lu, W. Dai, Stochastic integration for fractional Lévy process and stochastic differential equation driven by fractional Lévy noise, arXiv, 2013. https://doi.org/10.48550/arXiv.1307.4173 doi: 10.48550/arXiv.1307.4173
    [17] C. Bender, R. Knobloch, P. Oberacker, Maximal inequalities for fractional Lévy and related processes, Stochastic Anal. Appl., 33 (2015), 701–714. https://doi.org/10.1080/07362994.2015.1036167 doi: 10.1080/07362994.2015.1036167
    [18] Shen G, Li Y, Gao Z. Parameter estimation for Ornstein–Uhlenbeck processes driven by fractional Lévy process, J. Inequal. Appl., 2018 (2018), 1–14. https://doi.org/10.1186/s13660-018-1951-0 doi: 10.1186/s13660-018-1951-0
    [19] J. P. N. Bishwal, Parameter estimation in stochastic volatility models, Springer Nature, 2022.
    [20] B. C. Boniece, G. Didier, F. Sabzikar, On fractional Lévy processes: tempering, sample path properties and stochastic integration, J. Stat. Phys., 178 (2020), 954–985. https://doi.org/10.1007/s10955-019-02475-1 doi: 10.1007/s10955-019-02475-1
    [21] B. L. S. P. Rao, Nonparametric estimation of linear multiplier for stochastic differential equations driven by fractional Lévy process with small noise, Bull. Inform. Cybern., 52 (2020), 1–13. https://doi.org/10.5109/4150376 doi: 10.5109/4150376
    [22] B. L. S. P. Rao, Nonparametric estimation of trend for stochastic differential equations driven by fractional Lévy process, J. Stat. Theory Pract., 15 (2021), 1–13. https://doi.org/10.1007/s42519-020-00138-z doi: 10.1007/s42519-020-00138-z
    [23] J. C. Cox, Jr J. E. Ingersoll, S. A. Ross, An intertemporal general equilibrium model of asset prices, Econometrica, 53 (1985), 363–384.
    [24] J. C. Cox, Jr J. E. Ingersoll, S. A. Ross, A theory of the term structure of interest rates, Econometrica, 53 (1985), 385–407.
    [25] C. Wei, Estimation for the discretely observed Cox–Ingersoll–Ross model driven by small symmetrical stable noises, Symmetry, 12 (2020), 327. https://doi.org/10.3390/sym12030327 doi: 10.3390/sym12030327
    [26] C. Ma, X. Yang, Small noise fluctuations of the CIR model driven by α-stable noises, Stat. Probab. Lett., 94 (2014), 1–11. https://doi.org/10.1016/j.spl.2014.07.001 doi: 10.1016/j.spl.2014.07.001
    [27] Z. Li, C. Ma, Asymptotic properties of estimators in a stable Cox–Ingersoll–Ross model, Stochastic Proc. Appl., 125 (2015), 3196–3233. https://doi.org/10.1016/j.spa.2015.03.002 doi: 10.1016/j.spa.2015.03.002
    [28] X. Yang, Maximum likelihood type estimation for discretely observed CIR model with small α-stable noises, Stat. Probab. Lett., 120 (2017), 18–27. https://doi.org/10.1016/j.spl.2016.09.014 doi: 10.1016/j.spl.2016.09.014
    [29] T. Marquardt, Fractional Lévy processes with an application to long memory moving average processes, Bernoulli, 12 (2006), 1099–1126.
  • Reader Comments
  • © 2023 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1663) PDF downloads(72) Cited by(0)

Figures and Tables

Tables(2)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog