betaFeedback

WorksMathematics

Two approaches of estimation for a tempered fractional Brownian motion with drift

Abstract

We investigate the problem of estimation for a tempered fractional Brownian motion with drift of the form: Xt=μt+σBH,λ(t)X_t=μt + σB_{H,λ}(t), for all t≥0t\geq 0. Based on discrete observations of XX and by using the method of moments with the ergodic theorem, we construct estimators of the drift parameter μ∈Rμ\in\R and noise parameters σ2,H,λ∈(0,∞) σ^2,H,λ\in(0,\infty). We establish the asymptotic properties of the proposed estimators, including convergence in probability, almost sure convergence, and asymptotic normality. Finally, by using Gladyshev's results \cite{gladyshev1961new} we provide some consistent alternative estimators of σ2σ^2, HH and λλ, when μ=0μ=0 and H∈(0,1)H\in (0,1).

Keywords: Asymptotic normality, Consistency, Ergodic theorem, Method of moments, Tempered fractional Brownian motion

MSC Classification: 60G15, 60G18, 60G22, 62H12, 62F12

Introduction

Tempered fractional Brownian motions (tfBms) of first or second kind have been proposed by [17, 26] as an extension of the fractional Brownian motion (fBm). They are constructed by exponentially tempering the power law kernel in the Mandelbrot and Van Ness representation of the fBm. A tfBm BH,λB_{H,\lambda} shares the Gaussianity and stationarity of increments with the fBm, but is characterized by a mean squared displacement which decays rapidly to a constant, i.e., Var⁡(BH,λ(t))=C+O(tH−1/2e−λt)\operatorname{Var}(B_{H,\lambda}(t)) = C + O(t^{H-1/2}e^{-\lambda t}), as t→∞t \to\infty, for some constant C>0C > 0. The autocorrelation of its increments exhibits the semi-long range dependence, that is, it behaves like a power over moderate scales, but quasi-exponentially over large scales. TfBms are widely used to model wind speed data summarized by the so-called Davenport spectrum (e.g., [7]). The latter is a modification of the classical Kolmogorov spectrum for the inertial range of turbulence at low frequencies. TfBms have also been used to model geophysical flow data [6] and in finance [30].

The theoretical aspects of tfBm have attracted considerable research interest (see [1, 16, 17, 18, 27] and references therein). In contrast, literature on their statistical aspect remains limited to a few papers [5, 6], in which wavelet estimation procedures were performed. Recently, least-squares estimators for the unknown drift parameters within Vasicek models driven by tfBms have been proposed and examined in [19, 20]. This work addresses the parameter estimation of a tfBm with drift of the form.

Xt=μt+σBH,λ(t)(1)X_t = \mu t + \sigma B_{H,\lambda}(t) \tag*{(1)}

for all t≥0t \ge0, where BH,λB_{H,\lambda} is a tfBm as defined below. Using the ergodic theorem and the method of moments (e.g., [2]) we derive suitable estimators for model (1). We investigate both the weak and almost sure behaviour of the proposed estimators. Consequently, our results provide alternatives to those given in [5, 6]. When H∈(0,1)H \in(0,1), we apply Gladyshev’s results to construct some alternative estimators of the noise parameters HH, σ2\sigma^2, λ\lambda and prove their consistency. The evaluation of their limiting distributions requires more advanced tools and is beyond the scope of this work.

The rest of the paper is organized as follows. In Section 2 we introduce some basic and preliminary results. Then we discuss the procedures of estimation to be used. In Section 3 we construct μ^\widehat{\mu} the estimator of the drift μ∈R\mu\in\mathbb{R} and investigate its asymptotic behaviour (independently of the other parameters). Section 4 is devoted to the joint estimation of the noise parameters (σ2,H,λ)∈(0,∞)3(\sigma^2,H,\lambda) \in(0,\infty)^3 by the method of moments (e.g., [2]). In Section 5 we use Gladyshev’s results [12] to propose a multistage estimation for the noise parameters HH, σ2\sigma^2, λ\lambda under the condition 0<H<10 < H < 1. In this section, the consistency of the estimators of HH and σ2\sigma^2 holds almost surely, while the estimator of λ\lambda is consistent in probability. Section 6 concludes the paper. Some auxiliary and technical lemmas on the modified Bessel function KH(⋅)K_H(\cdot) with the Birkhoff’s ergodic theorem are gathered in the Appendix.

In what follows we work on a probability space (Ω,F,P)(\Omega,\mathcal{F},\mathbb{P}) and each process should be understood as being defined with respect to its own natural filtration. We are concerned with four modes of convergence: almost surely P-a.s.⟶\mathbb{P}\text{-a.s.}\longrightarrow, in the mean square sense L2(Ω)⟶L^2(\Omega)\longrightarrow, in probability P⟶\mathbb{P}\longrightarrow and in law L⟶\mathcal{L}\longrightarrow. The following symbols O(⋅)O(\cdot) and o(⋅)o(\cdot) denote the usual “big-oh” and “small-oh” notations, respectively.

Preliminary results

Let (B(t):t∈R)(B(t):t \in\mathbb{R}) be a two-sided Brownian motion, that is, a process with stationary independent increments such that B(t)B(t) has a Gaussian distribution with mean zero and variance ∣t∣|t| for all t∈Rt \in\mathbb{R}. Define an independently scattered Gaussian random measure B(dx)B(dx) with control measure m(dx)=dxm(dx)=dx by setting B([a,b])=B(b)−B(a)B([a,b])=B(b)-B(a) for any a,b∈Ra,b \in\mathbb{R} with a<ba < b and then extending to all Borel sets with finite Lebesgue measure. It is well known that the Wiener integral ∫f(x)B(dx)\int f(x)B(dx) exists whenever ∫f(x)2dx<∞\int f(x)^2 dx < \infty. Many famous fractional processes were constructed by choosing a suitable integrand f(⋅)f(\cdot). For instance, the fractional Brownian motion BHB_H was introduced by setting f(x)=CH((t−x)+H−1/2−(−x)+H−1/2)f(x)=C_H\left((t-x)_+^{H-1/2}-(-x)_+^{H-1/2}\right), where CH>0C_H>0 is some normalizing constant. In [17], the integrand in the form of BHB_H is exponentially tempered to get the tfBm

BH,λ(t)=∫[e−λ(t−x)+(t−x)+H−1/2−e−λ(−x)+(−x)+H−1/2]B(dx)B_{H,\lambda}(t)=\int\left[e^{-\lambda(t-x)_+}(t-x)_+^{H-1/2}-e^{-\lambda(-x)_+}(-x)_+^{H-1/2}\right]B(dx)

for all t∈Rt \in\mathbb{R}. BH,λB_{H,\lambda} exhibits many interesting properties such as stationarity of increments and semi-long range dependence. It is worth mentioning that this process is not self-similar, but instead the following scaling property: BH,λ(ct)=dcHBH,cλ(t)B_{H,\lambda}(ct)\overset{\mathrm{d}}{=}c^H B_{H,c\lambda}(t) holds for any c>0c>0. This property implies

E[BH,λ(∣t∣)2]=t2HE[BH,λt(1)2]=:t2HCt2,⟶2Γ(2H)/(2λ)2H(2)\mathbb{E}\left[B_{H,\lambda}(|t|)^2\right]=t^{2H}\mathbb{E}\left[B_{H,\lambda t}(1)^2\right]=:t^{2H}C_t^2, \qquad \longrightarrow2\Gamma(2H)/(2\lambda)^{2H} \tag*{(2)}

as t→∞t \to\infty, where

Ct2=2Γ(2H)(2λ∣t∣)2H−2Γ(H+1/2)π(2λ∣t∣)HKH(λ∣t∣)C_t^2 = \frac{2\Gamma(2H)}{(2\lambda|t|)^{2H}} - \frac{2\Gamma(H+1/2)}{\sqrt{\pi}(2\lambda|t|)^H}K_H(\lambda|t|)

and C02=0C_0^2=0,

where KHK_H denotes the modified Bessel function of second kind. The covariance of BH,λB_{H,\lambda} can be expressed as

Cov⁡(BH,λ(t),BH,λ(s))=12(∣t∣2HCt2+∣s∣2HCs2−∣t−s∣2HCt−s2)(3)\operatorname{Cov}(B_{H,\lambda}(t),B_{H,\lambda}(s)) = \frac{1}{2}\left(|t|^{2H}C_t^2+|s|^{2H}C_s^2-|t-s|^{2H}C_{t-s}^2\right) \tag*{(3)}

for all t,s≥0t,s\geq0. When H∈(0,1)H\in(0,1), the tfBm BH,λB_{H,\lambda} becomes a quasi-hé lix process in the sense of [15], and the Hölder continuity follows with HH as critical exponent. For H>1H>1, the sample paths of BH,λB_{H,\lambda} are continuously differentiable. For more details on properties of the tfBm we refer to [17, 1].

Given a function g:R→Rg:\mathbb{R}\to\mathbb{R}, we define its kkth order increments with step h∈Rh\in\mathbb{R} as

Δh(k)g(t):=∑j=0k(−1)k−j(kj)g(t+hj)\Delta_h^{(k)}g(t):=\sum_{j=0}^{k}(-1)^{k-j}\binom{k}{j}g(t+hj)

for all t∈Rt\in\mathbb{R} and k∈Z+k\in\mathbb{Z}_+ with the convention Δh(0)g(t)=g(t)\Delta_h^{(0)}g(t)=g(t). We say that a process Y=(Y(t):t∈R)Y=(Y(t):t\in\mathbb{R}) is kk-stationary if for every h∈Rh\in\mathbb{R} the processes (Δh(r)Y(t):t∈R)(\Delta_h^{(r)}Y(t):t\in\mathbb{R}), r=0,…,(k−1)r=0,\ldots,(k-1) are non-stationary, but (Δh(k)Y(t):t∈R)(\Delta_h^{(k)}Y(t):t\in\mathbb{R}) exhibits stationarity. The process XX defined in (1) is 1-stationary and stationarity holds also for any increments of higher order k≥2k\geq2. Consider the following statistics to be used later

X‾Np,q(h):=1N∑k=0N−1(Δh(p)Xt∣t=kh)q,N=1,2,…,(4)\overline{X}_N^{p,q}(h):=\frac{1}{N}\sum_{k=0}^{N-1}\left(\left.\Delta_h^{(p)}X_t\right|_{t=kh}\right)^q,\qquad N=1,2,\ldots, \tag*{(4)}

for fixed h>0h>0 and p,q∈Z+∗p,q\in\mathbb{Z}_+^*. Then the following results are obtained.

Lemma 1. We have

X‾Np,q(h)→P-a.s.E(Δh(p)Xt∣t=0)q=:X‾∞p,q(h)\overline{X}_N^{p,q}(h)\xrightarrow{\mathbb{P}\text{-a.s.}}\mathbb{E}\left(\left.\Delta_h^{(p)}X_t\right|_{t=0}\right)^q=:\overline{X}_\infty^{p,q}(h)

as N→∞N\to\infty. In particular,

X‾∞1,1(h)=μh,X‾∞1,2(h)=μ2h2+σ2h2HCh2,X‾∞2,2(h)=σ2h2H(4Ch2−22HC2h2).(5)\begin{aligned} \overline{X}_\infty^{1,1}(h)&=\mu h,\qquad\overline{X}_\infty^{1,2}(h)=\mu^2h^2+\sigma^2h^{2H}C_h^2,\\ \overline{X}_\infty^{2,2}(h)&=\sigma^2h^{2H}\left(4C_h^2-2^{2H}C_{2h}^2\right). \tag*{(5)} \end{aligned}

References

  1. [1]Ehsan Azmoodeh, Yuliya Mishura, and Farzad Sabzikar. How does tempering affect the local and global properties of fractional Brownian motion? Journal of Theoretical Probability, pages 1–44, 2022.
  2. [2]Luis A Barboza and Frederi G Viens. Parameter estimation of Gaussian stationary processes using the generalized method of moments. 2017.arxiv.org/abs/1604.06511
  3. [3]Árpád Baricz. Bounds for modified bessel functions of the first and second kinds. Proceedings of the Edinburgh Mathematical Society, 53(3):575–599, 2010.DOI
  4. [4]Árpád Baricz, Dragana Jankov Maširević, Saminathan Ponnusamy, and Sanjeev Singh. Bounds for the product of modified bessel functions. Aequationes mathematicae, 90(4):859–870, 2016.DOI
  5. [5]B Cooper Boniece, Gustavo Didier, and Farzad Sabzikar. Tempered fractional Brownian motion: wavelet estimation, modeling and testing. Applied and Computational Harmonic Analysis, 51:461–509, 2021.
  6. [6]B Cooper Boniece, Farzad Sabzikar, and Gustavo Didier. Tempered fractional Brownian motion: Wavelet estimation and modeling of turbulence in geophysical flows. In 2018 IEEE Statistical Signal Processing Workshop (SSP), pages 174–178. IEEE, 2018.DOI
  7. [7]Alan G Davenport. The spectrum of horizontal gustiness near the ground in high winds. Quarterly Journal of the Royal Meteorological Society, 87(372):194–211, 1961.
  8. [8]Mohamed El Omari. Statistical inference for nth-order mixed fractional Brownian motion with polynomial drift. Modern Stochastics: Theory and Applications, pages 1–19, 2024.DOI
  9. [9]David Gale and Hukunane Nikaido. The Jacobian matrix and global univalence of mappings. Mathematische Annalen, 159(2):81–93, 1965.DOI
  10. [10]Robert E Gaunt. Inequalities for modified Bessel functions and their integrals. Journal of Mathematical Analysis and Applications, 420(1):373–386, 2014.DOI
  11. [11]Iosif I Gikhman and Anatoli Vladimirovich Skorokhod. The theory of stochastic processes I. Springer Science & Business Media, 2004.
  12. [12]EG Gladyshev. A new limit theorem for stochastic processes with Gaussian increments. Theory of Probability & Its Applications, 6(1):52–61, 1961.DOI
  13. [13]JL González-Santander. Closed-form expressions for derivatives of Bessel functions with respect to the order. Journal of Mathematical Analysis and Applications, 466(1):1060–1081, 2018.
  14. [14]Ken-ichi Inada. The production coefficient matrix and the Stolper-Samuelson condition. Econometrica: Journal of the Econometric Society, pages 219–239, 1971.DOI
  15. [15]Jean-Pierre Kahane. Hélices et quasi-hélices. Université de Paris-Sud. Département de Mathématique, 1980.
  16. [16]Xing Liu and Weihua Deng. Numerical methods for the two-dimensional Fokker-Planck equation governing the probability density function of the tempered fractional Brownian motion. Numerical Algorithms, 85(1):23–38, 2020.DOI
  17. [17]Mark M Meerschaert and Farzad Sabzikar. Tempered fractional Brownian motion. Statistics & Probability Letters, 83(10):2269–2275, 2013.DOI
  18. [18]Mark M Meerschaert and Farzad Sabzikar. Stochastic integration for tempered fractional Brownian motion. Stochastic processes and their applications, 124(7):2363–2387, 2014.pubmed.ncbi.nlm.nih.gov/24872598
  19. [19]Yuliya Mishura and Kostiantyn Ralchenko. Asymptotic growth of sample paths of tempered fractional Brownian motions, with statistical applications to Vasicek-type models. Fractal and Fractional, 8(2):79, 2024.DOI
  20. [20]Yuliya Mishura, Kostiantyn Ralchenko, and Olena Dehtiar. Asymptotic properties of parameter estimators in Vasicek model driven by tempered fractional Brownian motion. Austrian Journal of Statistics, 54(1):61–81, 2024.arxiv.org/abs/2406.02800
  21. [21]Czeslaw Olech, T Parthasarathy, and G Ravindran. Almost N-matrices and linear complementarity. Linear Algebra and its Applications, 145:107–125, 1991.DOI
  22. [22]Frank WJ Olver, Daniel W Lozier, Ronald F Boisvert, and Charles W Clark. NIST handbook of mathematical functions hardback and CD-ROM. Cambridge university press, 2010.
  23. [23]Thiruvenkatachari Parthasarathy. On global univalence theorems. Springer, 1983.DOI
  24. [24]Kostiantyn Ralchenko and Mykyta Yakovliev. Asymptotic normality of parameter estimators for mixed fractional Brownian motion with trend. Austrian Journal of Statistics, 52(SI):127–148, 2023.DOI
  25. [25]Kostiantyn Ralchenko and Mykyta Yakovliev. Parameter estimation for fractional mixed fractional Brownian motion based on discrete observations. Modern Stochastics: Theory and Applications, pages 1–29, 2023.DOI
  26. [26]Farzad Sabzikar and Donatas Surgailis. Tempered fractional Brownian and stable motions of second kind. Statistics & Probability Letters, 132:17–27, 2018.arxiv.org/abs/1702.07258
  27. [27]Guangjun Shen, Liangwen Xia, and Dongjin Zhu. A strong convergence to the tempered fractional Brownian motion. Communications in Statistics-Theory and Methods, 46(8):4103–4118, 2017.DOI
  28. [28]Albert N Shiryaev. Probability-1, volume 95. Springer, 1996.
  29. [29]Akira Moiseevich Yaglom. Correlation theory of stationary and related random functions. Springer Science & Business Media, 1987.
  30. [30]Xili Zhang and Weilin Xiao. Arbitrage with fractional Gaussian processes. Physica A: Statistical Mechanics and its Applications, 471:620–628, 2017.DOI

Paper details

vlnc:W_qtsr3w75
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Mathematics

Not recorded for this paper

View statement
Looking for related papers…
Contents