variational estimates for martingale transforms
play

Variational estimates for martingale transforms Pavel Zorin-Kranich - PowerPoint PPT Presentation

Variational estimates for martingale transforms Pavel Zorin-Kranich University of Bonn 2020-06-09 Joint work with P. Friz. Pavel Zorin-Kranich (U Bonn) Variational estimates for martingale transforms 1/ 9 Rough paths Defjnition Pavel


  1. Variational estimates for martingale transforms Pavel Zorin-Kranich University of Bonn 2020-06-09 Joint work with P. Friz. Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms 1/ 9

  2. Rough paths Defjnition Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms . 1 / p |𝕐 u l βˆ’ 1 , u l | p ) l = 1 βˆ‘ l max ( l max , u 0 <β‹―< u lmax sup V p 𝕐 = , 1 / p l = 1 βˆ‘ l max A p -rough path, 2 < p < 3, is a pair X ∢ [ 0 , ∞) β†’ H , 𝕐 ∢ Ξ” = {( s , t ) | 0 ≀ s < t < ∞} β†’ H βŠ— H such that X ∈ V p loc , 𝕐 ∈ V p / 2 2/ 9 (Chen’s relation) p-variation: V p X = sup l max , u 0 <β‹―< u lmax ( loc , and for s < t < u 𝕐 s , u = 𝕐 s , t + 𝕐 t , u + ( X u βˆ’ X t ) βŠ— ( X t βˆ’ X s ). | X u l βˆ’ X u l βˆ’ 1 | p ) How to check the conditions X ∈ V p and 𝕐 ∈ V p / 2 ?

  3. Rough path lifts of martingales Let M = ( M t ) be a (Hilbert space valued) cΓ dlΓ g martingale. Let Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms How to incorporate e.g. fractional Brownian motion? Q: what is the appropriate generality for these lifting results? There exist rough path lifts of over processes, e.g. LΓ©vy processes. 3/ 9 ( s , t ] Then, a.s., the pair ( M , 𝕅) is a p -rough path for any p > 2. 𝕅 s , t ∢= ∫ ( M u βˆ’ βˆ’ M s ) βŠ— dM u . β–Ά Chen’s relation – from ItΓ΄ integration β–Ά Bound for V p M : LΓ©pingle 1976. β–Ά Bounds for V p / 2 𝕅 : β–Ά M Brownian motion: Lyons 1998 β–Ά M has continuous paths: Friz+Victoir 2006 β–Ά M dyadic: Do+Muscalu+Thiele 2010, β–Ά M has cΓ dlΓ g paths: Chevyrev+Friz 2017, Kovač+ZK 2018.

  4. Joint rough path lifts All martingales and processes are adapted, cΓ dlΓ g, Hilbert space valued. Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms The proof also recovers existence of ItΓ΄ integrals and estimates for New in this result: Motivation: dY = a ( Y ) d X + b ( Y ) dM is a p-rough path. ) 4/ 9 𝕐 ( X p-rough path ( 2 < p < 3 ). Then, a.s., the pair of processes Let M = ( M t ) be a cΓ dlΓ g martingale and ( X , 𝕐) a deterministic cΓ dlΓ g Theorem (Friz+ZK 2020+) ∫ X u βˆ’ βŠ— dM u M ) , ( ∫ M u βˆ’ βŠ— dX u ∫ M u βˆ’ βŠ— dM u β–Ά Variational estimates for ItΓ΄ integrals ∫ X dM , β–Ά existence of and estimates for integrals ∫ M dX . 𝕅 = ∫ M dM from previous slide.

  5. Martingale transforms 𝜐 Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms RHS is ≲ β€– Sf β€– L q 0 β€– Sg β€– L q 0 . In this case, any r > 1 works. The supremum is taken over increasing sequences of stopping times 𝜐 = (𝜐 k ) . β€– L q 1 β€– Sg β€– L q 0 . 1 / p β€– 𝜐 k βˆ’ 1 < j β‰€πœ k sup ( k β€–(βˆ‘ β€– 5/ 9 β€– Theorem (Main estimate) Martingale in t variable, discrete version of area integral. β€– V r Ξ ( f , g )β€– s < j ≀ t Let ( f n ) n βˆˆβ„• be a discrete time adapted process and ( g n ) n βˆˆβ„• a discrete time martingale. Defjne paraproduct Ξ  s , t ( f , g ) ∢= βˆ‘ ( f j βˆ’ 1 βˆ’ f s ) dg j , dg j = g j βˆ’ g j βˆ’ 1 . Let 1 ≀ p ≀ ∞ , 0 < q 1 ≀ ∞ , 1 ≀ q 0 < ∞ . Defjne q by 1 / q = 1 / q 0 + 1 / q 1 and suppose 1 / r < 1 / 2 + 1 / p. Then, with β€– g β€– L q = (𝔽| g | q ) 1 / q , Sg = [ g ] 1 / 2 , β€– L q ≲ sup | f j βˆ’ 1 βˆ’ f 𝜐 k βˆ’ 1 |) p ) β–Ά If f is a martingale, p = 2, 1 ≀ q 1 < ∞ , then by BDG inequality the β–Ά For general f , RHS is ≀ β€– V p f β€– L q 0 β€– Sg β€– L q 0 and r = p / 2 works.

  6. Discrete approximation of adapted processes lim Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms Given πœ— > 0, consider the adapted partition Proof. p ∈ ( p , ∞) βˆͺ {∞} . for any Μƒ p ) in Defjnition 6/ 9 If f ∈ L q ( V p ) for some q > 0 and p > 1 , then Lemma ∢= f ⌊ t ,πœŒβŒ‹ . t f (𝜌) ⌊ t , πœŒβŒ‹ ∢= max { s ∈ 𝜌 | s ≀ t }, For an adapted partition 𝜌 , let An adapted partition 𝜌 = (𝜌 j ) j is an increasing sequence of stopping times. Adapted partitions are ordered by a.s. inclusion of the sets {𝜌 j | j ∈ β„•} . The set of adapted partitions is directed, so lim 𝜌 makes sense. 𝜌 f (𝜌) = f L q ( V Μƒ 𝜌 j + 1 (πœ•) ∢= inf { t > 𝜌 j (πœ•) | 𝜌 0 ∢= 0 , | | f t βˆ’ f 𝜌 j (πœ•) |(πœ•) > πœ—}.

  7. Stopping time reduction ∞ Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms where the supremum is taken over all adapted partitions 𝜐 . (1) β€– L q , 1 /𝜍 β€– 𝜍 ) |Ξ  t , t β€² |) 𝜐 j βˆ’ 1 ≀ t < t β€² β‰€πœ j sup ( j = 1 βˆ‘ β€–( f adapted process, g martingale β€– 𝜐 Then, for every 0 < 𝜍 < r < ∞ and q ∈ ( 0 , ∞] , we have Lemma Suppose 1 / r < 1 / p + 1 / 2 . Then Theorem (Main estimate) Square function: Sg = [ g ] 1 / 2 , 7/ 9 Martingale transform: Ξ  s , t ( f , g ) = βˆ‘ s < j ≀ t ( f j βˆ’ 1 βˆ’ f s ) dg j HΓΆlder exponents: 1 / q = 1 / q 0 + 1 / q 1 . β€– V r Ξ β€– L q (Ω) ≲ β€– V p f β€– L q 1 (Ω) β€– Sg β€– L q 0 (Ω) . The V r norm is estimated as follows. Let (Ξ  s , t ) s ≀ t be a cΓ dlΓ g adapted sequence with Ξ  t , t = 0 for all t. β€– V r Ξ β€– L q ≲ sup

  8. LΓ©pingle’s inequality Above stopping time argument fjrst used in the following result. Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms inequalities that follow from weighted inequalities by OsΘ©kowski. For dealing with martingale transforms, we use vector-valued BDG Weighted inequalities imply vector-valued inequalities. Classical LΓ©pingle’s inequality is the case w ≑ 1, A p ( w ) = 1. p βˆ’ 1 β€– L ∞ ( w ) 𝜐 ) ‖𝔽( w | β„± 𝜐 stopping time sup A p ( w ) ∢= For 1 < p < ∞ and 2 < r, we have Let M be a martingale and w a positive random variable. Theorem (ZK 2019) 8/ 9 β€– V r M β€– L p ( w ) ≀ C p , r A p ( w ) max ( 1 , 1 /( p βˆ’ 1 )) β€– M β€– L p ( w ) , where the A p charactersitic is given by 𝜐 )𝔽( w βˆ’ 1 /( p βˆ’ 1 ) | β„±

  9. Integration by parts β€– L q β€–( ∞ βˆ‘ j = 1 ( βˆ‘ 𝜐 j βˆ’ 1 < u β‰€πœ j |Ξ” X u Ξ” u M | 2 ) 𝜍/ 2 ) 1 /𝜍 β€– HΓΆlder ≲ ≀ β€–( ∞ βˆ‘ j = 1 βˆ‘ 𝜐 j βˆ’ 1 < u β‰€πœ j |Ξ” u M | 2 ) 1 / 2 β€– β€– L q Pavel Zorin-Kranich (U Bonn) – Variational estimates for martingale transforms β€– vector BDG ( X , 𝕐) rough path, M martingale β€– We do this by partial integration: Ξ ( M , X ) ∢= πœ€ M πœ€ X βˆ’ Ξ ( X , M ) βˆ’ πœ€[ X , M ]. The bracket is given by u ≀ T Ξ” X u Ξ” M u , Variation norm estimate for the bracket: β€– V r [ X , M ]β€– L q β€– L q ≲ stopping β€–( ∞ βˆ‘ j = 1 ( sup 𝜐 j βˆ’ 1 < t < t β€² β‰€πœ j |πœ€[ X , M ] t , t β€² |) 𝜍 ) 1 /𝜍 β€– 9/ 9 So far we have estimated ∫ X dM and ∫ M dM . Next, we want to construct and estimate Ξ ( M , X ) = ∫ M dX . [ X , M ] T = βˆ‘ Ξ” M u = M u βˆ’ M u βˆ’ . V p X β‹… β€–

Recommend


More recommend