由连续半鞅驱动的倒向随机微分方程解的比较定理

彭实戈[1]首先建立了一维倒向随机微分方程的比较定理,本文在Lipschitz条件下研究由连续半鞅驱动的倒向随机微分方程,我们将比较定理推广到此类倒向随机微分方程,并且证明方法比彭实戈[1]的更加直接和简单.     

倒向随机微分方程解的比较定理

毛学荣新近将彭实戈和Ｐａｒｄｏｕｘ关于倒向随机策分方程解的存在性定理推广到非Ｌｉｐｓｃｈｉｔｚ系数情景,此文将彭实戈的比较定理推广到这一情形,主要工具是Ｔａｎａｋａ－Ｍｅｙｅｒ公式,Ｄａｖｉｓ不等式和Ｂｉｈａｒｉ不等式。     

随机游走和离散的倒向随机微分方程

本文研究了随机游走和离散的倒向随机微分方程。把随机游走到布朗运动的收敛推广到L^2情形；而且根据倒向随机微分方程的理论框架研究了离散的倒向随机微分方程，得到了离散的倒向随机微分方程解的存在唯一性和比较定理，这实际上给出了倒向随机微分方程的一种离散方法，为理论和实际研究提供了方便。     

倒向随机微分方程的一个一般的反比较定理(英文)

在由彭实戈引入的倒向随机微分方程的最基本的条件下,提出并证明了一个一般的反比较定理.     

一类非时齐非Lipschitz条件下带跳的倒向随机微分方程的适应解及比较定理

对系数f(t,y,z,k)满足非常一般的非时齐非Lipschitz条件,本文给出一类带跳的倒向随机微分方程局部和整体解的存在唯一性的证明,同时本文也研究了带跳的倒向随机微分方程的比较定理,从而把前人的相应结果推广到更一般情形．     

多维带跳倒向随机微分方程比较定理

研究了高维及矩阵值带跳倒向随机微分方程解的比较定理问题.利用倒向随机生存性质的相关理论,将比较定理转化为一个特定闭凸集上的生存性质问题,并得到了比较定理成立的一个充分必要条件.     

关于反射倒向随机微分方程的解的一些性质

研究了关于反射倒向随机微分方程的解的一些性质．同时在适当的条件下建立了关于反射倒向随机微分方程生成元的一个唯一性定理和一个逆比较定理．     

生成元连续且线性增长的反射倒向随机微分方程生成元的表示定理（英文）

本文建立了一个生成元满足连续且线性增长条件的反射倒向随机微分方程生成元的局部表示定理,此定理推广了一些已有的倒向随机微分方程生成元的表示定理.应用此表示定理,本文获得了一个一般的反射倒向随机微分方程的逆比较定理,同时讨论了此类方程的一些性质.     

正倒向随机微分方程的适应解及其比较定理

研究了一类正倒向随机微分方程的适应解，其中正向方程不需要满足非退化条件，我们证明了在某些单调条件下，正倒向随机微分方程存在唯一的适应解，并给出了该正倒向随机微分方程的比较定理。     

正倒向随机微分方程理论基础及相关应用

本文从随机微分方程和倒向随机微分方程基本理论和应用背景谈起,结合随机最优控制理论和金融市场中的期权定价理论导出完全耦合的正倒向随机微分方程的形式.进而从该类方程的可解性这一角度出发,对已有的理论方法进行分析和探讨,引入一种非马尔科夫框架下保证解的存在唯一性的“统一框架”方法,给出比较定理、解的高维估计等重要性质,并联系相关偏微分方程系统给出其概率解释.对实际中应用广泛的线性正倒向随机微分方程引入了一种线性变换的方法作为“统一框架”方法的重要补充和完善,使得正倒向随机微分方程的应用更加广泛.     

Continuous dependence properties on solutions of backward stochastic differential equation

The existence theorem and continuous dependence property in ”L²” sense for solutions of backward stochastic differential equation (shortly BSDE) with Lipschitz coefficients were respectively established by Pardoux-Peng and Peng in [1,2], Mao and Cao generalized the Pardoux-Peng’s existence and uniqueness theorem to BSDE with non-Lipschitz coefficients in [3,4]. The present paper generalizes the Peng’s continuous dependence property in ”L²” sense to BSDE with Mao and Cao’s conditions. Furthermore, this paper investigates the continuous dependence property in “almost surely” sense for BSDE with Mao and Cao’s conditions, based on the comparison with the classical mathematical expectation.     

The adapted solution and comparison theorem for backward stochastic differential equations with Poisson jumps and applications

This paper deals with a class of backward stochastic differential equations with Poisson jumps and with random terminal times. We prove the existence and uniqueness result of adapted solution for such a BSDE under the assumption of non-Lipschitzian coefficient. We also derive two comparison theorems by applying a general Girsanov theorem and the linearized technique on the coefficient. By these we first show the existence and uniqueness of minimal solution for one-dimensional BSDE with jumps when its coefficient is continuous and has a linear growth. Then we give a general Feynman-Kac formula for a class of parabolic types of second-order partial differential and integral equations (PDIEs) by using the solution of corresponding BSDE with jumps. Finally, we exploit above Feynman-Kac formula and related comparison theorem to provide a probabilistic formula for the viscosity solution of a quasi-linear PDIE of parabolic type.     

连续系数倒向随机微分方程最小解的Levi定理

本文研究了倒向随机微分方程解的连续依赖性问题.利用文献[4]中使用的方法,提出并证明了连续系数的一维倒向随机微分方程最小解的Levi定理,推广了文献[10]中的相应结果.     

g-上鞅的Riesz分解定理

本文给出了当终端时间趋于无穷时一类有限时间区间上的倒向随机微分方程的解的收敛性,并且证明了这类解平方收敛到特定的无穷时间区间上的倒向随机微分方程的解.本文主要研究了由倒向随机微分方程生成的非线性期望及其鞅的性质,证明了当生成元g是超线性时的g-上鞅Riesz分解定理.并且指出经典鞅论中的Riesz分解定理和下期望(又称最小期望)对应的上鞅Riesz分解定理是g-上鞅Riesz分解定理的两种特殊情况.     

A tropical analogue of the Pauli problem and a splitting of quasithermodynamics

We consider a tropical analogue of the quantum mechanical Pauli problem (the Fourier transform is replaced by the Legendre transform). The statement of the problem is motivated by a generalized theory of thermodynamic fluctuations. The paper contains a theorem about a splitting of solutions in one-dimensional quasithermodynamic case.     

Solvability of some quadratic BSDEs without exponential moments

We establish the existence and uniqueness of square integrable solutions for a class of one-dimensional quadratic backward stochastic differential equations (QBSDEs). This is done with a merely square integrable terminal condition, and in some cases with a measurable generator. This shows, in particular, that neither the existence of exponential moments for the terminal condition nor the continuity of the generator are needed for the existence and/or uniqueness of solutions for quadratic BSDEs. These conditions are used in the previous papers on QBSDEs. To do this, we show that Itô?s formula remains valid for functions having a merely locally integrable second (generalized) derivative. A comparison theorem is also established.     

A financial market with interacting investors: does an equilibrium exist?

While trading on a financial market, the agents we consider take the performance of their peers into account. By maximizing individual utility subject to investment constraints, the agents may ruin each other even unintentionally so that no equilibrium can exist. However, when the agents are willing to waive little expected utility, an approximated equilibrium can be established. The study of the associated backward stochastic differential equation (BSDE) reveals the mathematical reason for the absence of an equilibrium. Presenting an illustrative counterexample, we explain why such multidimensional quadratic BSDEs may not have solutions despite bounded terminal conditions and in contrast to the one-dimensional case.