Multi-dimensional anticipated backward stochastic differential equations with quadratic growth
Ying Hu, Feng Li, Jiaqiang Wen · 2025
Read Multi-dimensional anticipated backward stochastic differential equations with quadratic growth by Ying Hu, Feng Li, Jiaqiang Wen (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-20255.txt.
Introduction
Ying Hu, Feng Li and Jiaqiang Wen widen the theory of anticipated backward stochastic differential equations, equations whose present value depends on future states. Past work solved them in one dimension with bounded end values. The new results cover many dimensions and allow unbounded end values as well, under quadratic growth conditions on the driver.
Three results are proved on when solutions exist and are unique. Local solutions hold for bounded end values under a general growth form. Global solutions need the growth to be skew sub-quadratic yet strictly and diagonally quadratic in one variable. For unbounded end values, two global cases are settled, one with a diagonal quadratic driver and one mixing quadratic and linear growth.
Catalog record
| CREATOR | Ying Hu, Feng Li, Jiaqiang Wen |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 144 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-20255
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Multi-dimensional anticipated backward stochastic differential equations with quadratic growth Thanks: YH is partially supported by Lebesgue Center of Mathematics “Investissements d’avenir” program-ANR-11-LABX-0020-01, by CAESARS-ANR-15-CE05-0024 and by MFG-ANR16-CE40-0015-01. Jiaqiang Wen is supported by National Natural Science Foundation of China (Grant No. 12101291), Guangdong Basic and Applied Basic Research Foundation (Grant No. 2025B151502009), and Shenzhen Fundamental Research General Program grant (Grant No. JCYJ20230807093309021). Ying Hu, Feng Li, Jiaqiang Wen Abstract Abstract. This paper is devoted to the general solvability of anticipated backward stochastic differential equations with quadratic growth by relaxing the assumptions made by Hu, Li, and Wen [ 21 , Journal of Differential Equations, 270 (2021), 1298–1311] from the one-dimensional case with bounded terminal values to the multi-dimensional situation with bounded/unbounded terminal values. Three new results regarding the existence and uniqueness of local and global solutions are established. More precisely, for the local solution with bounded terminal values, the generator f(t,Y_{t},Z_{t},Y_{t+\delta_{t}},Z_{t+\zeta_{t}}) is of general growth with respect to Y_{t} and Y_{t+\delta_{t}} . For the global solution with bounded terminal values, the generator f(t,Y_{t},Z_{t},Y_{t+\delta_{t}},Z_{t+\zeta_{t}}) is of skew sub-quadratic but also “strictly and diagonally” quadratic growth in Z_{t} . For the global solution with unbounded terminal values, the generator f(t,Y_{t},Z_{t},Y_{t+\delta_{t}}) is of diagonal quadratic growth in Z_{t} in the first case; and in the second case, the generator f(t,Z_{t})+\mathbb{E}[g(t,Y_{t},Z_{t},Y_{t+\delta_{t}},Z_{t+\zeta_{t}})] is of diagonal quadratic growth in Z_{t} and linear growth in Z_{t+\zeta_{t}} . Key words: Anticipated backward stochastic differential equation; quadratic generator; multi-dimension; unbounded terminal value AMS subject classifications. 60H10, 60H30. 1 Introduction Let \{W_{t};t\geqslant 0\} be a d -dimensional standard Brownian motion defined on some complete probability space (\Omega,\mathscr{F},\mathbb{P}) and \mathbb{F}\triangleq\{\mathscr{F}_{t};t\geqslant 0\} be the natural filtration of {W} augmented by all the \mathbb{P} -null sets in \mathscr{F} . Fix a terminal time T\in(0,+\infty) , a constant K\in[0,+\infty) , and two positive integers n and d . Consider the following anticipated backward stochastic differential equations (ABSDEs, for short) over a finite time horizon [0,T+K] : \left\{\begin{aligned} &Y_{t}=\xi_{T}+\int_{t}^{T}f(s,Y_{s},Z_{s},Y_{s+\delta_{s}},Z_{s+\zeta_{s}})ds-\int_{t}^{T}Z_{s}dW_{s},\quad t\in[0,T];\\ &Y_{t}=\xi_{t},\quad Z_{t}=\eta_{t},\quad t\in[T,T+K],\end{aligned}\right. (1.1) where \delta and \zeta are two deterministic \mathbb{R}^{+} -valued continuous functions defined on [0,T] , and \xi and \eta are some given processes defined on [T,T+K] . The process \xi is called the terminal value and the function f is called the generator of ABSDE ( 1.1 ). The unknown processes (Y,Z) , taking values in \mathbb{R}^{n}\times\mathbb{R}^{n\times d} , are called a pair of \mathbb{F} -progressively measurable solutions if they make Eq. ( 1.1 ) hold almost surely. In 2009, to deal with the delayed stochastic optimal control problem, Peng and Yang [ 38 ] introduced ABSDE ( 1.1 ) and established the existence and uniqueness under the condition that the generator is of uniformly Lipschitz and linear growth. From then on, this theory has attracted a lot of attentions. By regarding ABSDE as a new duality type of stochastic differential delay equations, Chen and Wu [ 6 ] derived the maximum principle for stochastic optimal control problem with delay. Along this line of research, Meng, Shi, Wang, and Zhang [ 32 ] obtained a general maximum principle for a stochastic optimal control problem with delay. Hamaguchi [ 18 ] studied a maximum principle for optimal control problems of with delay. Nie, Wang, and Wu [ 34 ] studied a type of linear-quadratic delayed mean-field social optimization, and de Feo and Świech [ 11 ] studied a verification theorem and constructed the optimal feedback controls for the optimal control problems with the stochastic delay differential equations. At the same time, the theory of ABSDE has also been well-developed. For example, Wu, Wang, and Ren [ 42 ] , along with Lu and Ren [ 30 ] , proved the existence and uniqueness of ABSDEs with non-Lipschitz generators and linear growth, as well as ABSDEs driven by Markov chains, respectively. Yang and Elliott [ 44 ] obtained some properties for generalized ABSDEs. Hu and Chen [ 20 ] obtained the L^{p} solutions of ABSDEs under monotonicity and general increasing conditions. Cheridito and Nam [ 10 ] studied a class of anticipated BSDEs when the generator is a path-functional of the solution under Lipschitz condition. Wen and Shi [ 41 ] and Douissi, Wen, and Shi [ 12 ] proved the solvability of ABSDEs driven by a fractional Brownian motion. Moreover, by using ABSDEs, Menoukeu-Pamen [ 33 ] studied a non-linear time-advanced backward stochastic partial differential equations with jumps. Chen and Yang [ 46 ] obtained the well-posedness for anticipated backward stochastic Schrödinger equations. Some other recent developments of ABSDEs can be found in Chen, Wu and Yu [ 7 ] , Guatteri and Masiero [ 17 ] , Klimsiak and Rzymowski [ 28 ] , Han and Li [ 19 ] , etc. However, the framework of Peng and Yang [ 38 ] failed to address some general delayed stochastic optimal control problems, such as the delayed stochastic linear-quadratic (LQ, for short) optimal control problems with random coefficients. The reason is that the related stochastic Riccati equations are essentially ABSDEs with quadratic growth (see Chen, Wu, and Yu [ 7 ] ). On one hand, the related problem of BSDEs with quadratic growth plays an intrinsic mathematical interest and Peng [ 37 , Section 5, page 270] listed it as an open problem. On the other hand, this topic has important applications in various fields, such as nonzero-sum risk-sensitive stochastic differential games, financial price-impact models, financial market equilibrium problems for several interacting agents, and stochastic equilibria problems in incomplete financial markets (see Hu, Liang, and Tang [ 23 , 24 ] ). Therefore, it is necessary and valuable to study ABSDEs with quadratic growth. In general, in multi-dimensional situation, ABSDE is called of quadratic growth if the generator f is of quadratic growth in Z for each i=1,...,n ; and it is called of diagonally quadratic growth if f^{i} is quadratic growth in the i -th row of Z and grows at a rate strictly slower than quadratic in the j -th component of Z when j\neq i . For ABSDE ( 1.1 ) with quadratic growth, Fujii and Takahashi [ 16 ] study one-dimensional case when the generator f is independent of the anticipated term Z_{t+\zeta_{t}} . Hu, Li, and Wen [ 21 ] studied the one-dimensional case with bounded terminal value. Based on the work of Hu, Li, and Wen [ 21 ] , Pak, Kim, and Kim [ 35 ] studied one-dimensional case with unbounded terminal value when the generator f is independent of the anticipated term Z_{t+\zeta_{t}} . Unfortunately, due to the fact that the comparison property does not hold for the general ABSDE ( 1.1 ) (see the counter Example 5.3 of Peng and Yang [ 38 ] ), the topic of multi-dimensional ABSDEs, the focus of the present paper, poses a great challenge. Next, in order to present our work in more detail, let’s recall some results about BSDEs. In ABSDE ( 1.1 ), if the generator f is independent of the anticipated term Y_{t+\delta_{t}} and Z_{t+\zeta_{t}} , then it becomes the following classical BSDE: Y_{t}=\xi_{T}+\int_{t}^{T}f(s,Y_{s},Z_{s})ds-\int_{t}^{T}Z_{s}dW_{s},\quad t\in[0,T]. For the classical BSDE, Pardoux and Peng [ 36 ] firstly solved the nonlinear situation when the generator has uniformly Lipschitz and linear growth in 1990. Since then, BSDEs have