词条 | Doob decomposition theorem |
释义 |
In the theory of stochastic processes in discrete time, a part of the mathematical theory of probability, the Doob decomposition theorem gives a unique decomposition of every adapted and integrable stochastic process as the sum of a martingale and a predictable process (or "drift") starting at zero. The theorem was proved by and is named for Joseph L. Doob.[1] The analogous theorem in the continuous-time case is the Doob–Meyer decomposition theorem. Statement of the theoremLet {{math|(Ω, {{mathcal|F}}, ℙ)}} be a probability space, {{math|I {{=}} {0, 1, 2, . . . , N}}} with {{math|N ∈ ℕ}} or {{math|I {{=}} ℕ0}} a finite or an infinite index set, {{math|({{mathcal|F}}n)n∈I}} a filtration of {{mathcal|F}}, and {{math|X {{=}} (Xn)n∈I}} an adapted stochastic process with {{math|E[{{!}}Xn{{!}}] < ∞}} for all {{math|n ∈ I}}. Then there exists a martingale {{math|M {{=}} (Mn)n∈I}} and an integrable predictable process {{math|A {{=}} (An)n∈I}} starting with {{math|A0 {{=}} 0}} such that {{math|Xn {{=}} Mn + An}} for every {{math|n ∈ I}}. Here predictable means that {{math|An}} is {{math|{{mathcal|F}}n−1}}-measurable for every {{math|n ∈ I \\ {0}}}. This decomposition is almost surely unique.[2][3][4] CorollaryA real-valued stochastic process {{math|X}} is a submartingale if and only if it has a Doob decomposition into a martingale {{math|M}} and an integrable predictable process {{math|A}} that is almost surely increasing.[5] It is a supermartingale, if and only if {{math|A}} is almost surely decreasing. RemarkThe theorem is valid word by word also for stochastic processes {{math|X}} taking values in the {{math|d}}-dimensional Euclidean space {{math|ℝd}} or the complex vector space {{math|ℂd}}. This follows from the one-dimensional version by considering the components individually. ProofsProof of the theoremExistenceUsing conditional expectations, define the processes {{math|A}} and {{math|M}}, for every {{math|n ∈ I}}, explicitly by {{NumBlk|:||{{EquationRef|1}}}}and {{NumBlk|:||{{EquationRef|2}}}}where the sums for {{math|n {{=}} 0}} are empty and defined as zero. Here {{math|A}} adds up the expected increments of {{math|X}}, and {{math|M}} adds up the surprises, i.e., the part of every {{math|Xk}} that is not known one time step before. Due to these definitions, {{math|An+1}} (if {{math|n + 1 ∈ I}}) and {{math|Mn}} are {{math|{{mathcal|F}}n}}-measurable because the process {{math|X}} is adapted, {{math|E[{{!}}An{{!}}] < ∞}} and {{math|E[{{!}}Mn{{!}}] < ∞}} because the process {{math|X}} is integrable, and the decomposition {{math|Xn {{=}} Mn + An}} is valid for every {{math|n ∈ I}}. The martingale property a.s. also follows from the above definition ({{EquationNote|2}}), for every {{math|n ∈ I \\ {0}}}. UniquenessTo prove uniqueness, let {{math|X {{=}} M{{'}} + A{{'}}}} be an additional decomposition. Then the process {{math|Y :{{=}} M − M{{'}} {{=}} A{{'}} − A}} is a martingale, implying that a.s., and also predictable, implying that a.s. for any {{math|n ∈ I \\ {0}}}. Since {{math|Y0 {{=}} A{{'}}0 − A0 {{=}} 0}} by the convention about the starting point of the predictable processes, this implies iteratively that {{math|Yn {{=}} 0}} almost surely for all {{math|n ∈ I}}, hence the decomposition is almost surely unique. Proof of the corollaryIf {{math|X}} is a submartingale, then a.s. for all {{math|k ∈ I \\ {0}}}, which is equivalent to saying that every term in definition ({{EquationNote|1}}) of {{math|A}} is almost surely positive, hence {{math|A}} is almost surely increasing. The equivalence for supermartingales is proved similarly. ExampleLet {{math|X {{=}} (Xn)n∈ℕ0}} be a sequence in independent, integrable, real-valued random variables. They are adapted to the filtration generated by the sequence, i.e. {{math|{{mathcal|F}}n {{=}} σ(X0, . . . , Xn)}} for all {{math|n ∈ ℕ0}}. By ({{EquationNote|1}}) and ({{EquationNote|2}}), the Doob decomposition is given by and If the random variables of the original sequence {{math|X}} have mean zero, this simplifies to and hence both processes are (possibly time-inhomogenious) random walks. If the sequence {{math|X {{=}} (Xn)n∈ℕ0}} consists of symmetric random variables taking the values {{math|+1}} and {{math|−1}}, then {{math|X}} is bounded, but the martingale {{math|M}} and the predictable process {{math|A}} are unbounded simple random walks (and not uniformly integrable), and Doob's optional stopping theorem might not be applicable to the martingale {{math|M}} unless the stopping time has a finite expectation. ApplicationIn mathematical finance, the Doob decomposition theorem can be used to determine the largest optimal exercise time of an American option.[6][7] Let {{math|X {{=}} (X0, X1, . . . , XN)}} denote the non-negative, discounted payoffs of an American option in a {{math|N}}-period financial market model, adapted to a filtration {{math| ({{mathcal|F}}0, {{mathcal|F}}1, . . . , {{mathcal|F}}N)}}, and let {{math|ℚ}} denote an equivalent martingale measure. Let {{math|U {{=}} (U0, U1, . . . , UN)}} denote the Snell envelope of {{math|X}} with respect to {{math|ℚ}}. The Snell envelope is the smallest {{math|ℚ}}-supermartingale dominating {{math|X}}[8] and in a complete financial market it represents the minimal amount of capital necessary to hedge the American option up to maturity.[9] Let {{math|U {{=}} M + A}} denote the Doob decomposition with respect to {{math|ℚ}} of the Snell envelope {{math|U}} into a martingale {{math|M {{=}} (M0, M1, . . . , MN)}} and a decreasing predictable process {{math|A {{=}} (A0, A1, . . . , AN)}} with {{math|A0 {{=}} 0}}. Then the largest stopping time to exercise the American option in an optimal way[10][11] is Since {{math|A}} is predictable, the event {{math|{τmax {{=}} n} {{=}} {An {{=}} 0, An+1 < 0}}} is in {{math|{{mathcal|F}}n}} for every {{math|n ∈ {0, 1, . . . , N − 1}}}, hence {{math|τmax}} is indeed a stopping time. It gives the last moment before the discounted value of the American option will drop in expectation; up to time {{math|τmax}} the discounted value process {{math|U}} is a martingale with respect to {{math|ℚ}}. GeneralizationThe Doob decomposition theorem can be generalized from probability spaces to σ-finite measure spaces.[12] Citations1. ^{{harvtxt|Doob|1953}}, see {{harv|Doob|1990|pp=296−298}} 2. ^{{harvtxt|Durrett|2005}} 3. ^{{harv|Föllmer|Schied|2011|loc=Proposition 6.1}} 4. ^{{harv|Williams|1991|loc=Section 12.11, part (a) of the Theorem}} 5. ^{{harv|Williams|1991|loc=Section 12.11, part (b) of the Theorem}} 6. ^{{harv|Lamberton|Lapeyre|2008|loc=Chapter 2: Optimal stopping problem and American options}} 7. ^{{harv|Föllmer|Schied|2011|loc=Chapter 6: American contingent claims}} 8. ^{{harv|Föllmer|Schied|2011|loc=Proposition 6.10}} 9. ^{{harv|Föllmer|Schied|2011|loc=Theorem 6.11}} 10. ^{{harv|Lamberton|Lapeyre|2008|loc=Proposition 2.3.2}} 11. ^{{harv|Föllmer|Schied|2011|loc=Theorem 6.21}} 12. ^{{harv|Schilling|2005|loc=Problem 23.11}} References
3 : Theorems regarding stochastic processes|Martingale theory|Articles containing proofs |
随便看 |
|
开放百科全书收录14589846条英语、德语、日语等多语种百科知识,基本涵盖了大多数领域的百科知识,是一部内容自由、开放的电子版国际百科全书。