2252 CHAPTER 65. STOCHASTIC INTEGRATION

The elementary functions {Φn} have values in L (U1,H)0 meaning they are restrictions offunctions in L (U1,H) to JQ1/2U and converge to Φ◦ J−1 in

L2([a,T ]×Ω;L2

(JQ1/2U,H

))where Φ ∈ L2

([a,T ]×Ω;L2

(Q1/2U,H

))is given. Let

Φ(t)≡n−1

∑k=0

Φ(tk)X(tk,tk+1] (t)

be an elementary function. In particular, let Φ(tk) be Ftk measurable as a map intoL (U1,H), and has finitely many values. As just mentioned, the topic of interest is theelementary functions Φn in the above diagram. Thus Φ will be one of these elementaryfunctions.

Let τ be a stopping time having values from the set of mesh points {tk} for the elemen-tary function. Then from the definition of the integral for elementary functions,

∫ t∧τ

aΦdW ≡

n−1

∑k=0

Φ(tk)(W (t ∧ τ ∧ tk+1)−W (t ∧ τ ∧ tk))

If ω is such that τ (ω) = t j, then to get something nonzero, you must have t j > tk sok ≤ j−1. Thus the above on the right reduces to

j−1

∑k=0

Φ(tk)(W (t ∧ tk+1)−W (t ∧ tk))

It clearly is 0 if j = 0. Define ∑−1k=0 ≡ 0. Thus the integral equals

n

∑j=0

X[τ=t j]

j−1

∑k=0

Φ(tk)(W (t ∧ tk+1)−W (t ∧ tk))

Interchanging the order of summation, k ≤ j−1 so j ≥ k+1 and this equals

n−1

∑k=0

n

∑j=k+1

X[τ=t j]Φ(tk)(W (t ∧ tk+1)−W (t ∧ tk))

=n−1

∑k=0

Ftk measurable︷ ︸︸ ︷X[τ>tk]Φ(tk)(W (t ∧ tk+1)−W (t ∧ tk))

Therefore ∫ t∧τ

aΦdW =

∫ t

a

n−1

∑k=0

X[τ>tk]Φ(tk)X(tk,tk+1]dW (65.8.15)

2252 CHAPTER 65. STOCHASTIC INTEGRATIONThe elementary functions {®, } have values in (U;,H) meaning they are restrictions offunctions in Y (U;,H) to JQ!/U and converge to ®oJ~! inVL ( T] xX QPL, (yo'u,H))where ® € L? ([a,T] x Q;.Y (Q"/U,H)) is given. Letn—1P(t)= u° (ti) Kt thr] (t)be an elementary function. In particular, let ®(t,) be -;, measurable as a map into-£ (U,,H), and has finitely many values. As just mentioned, the topic of interest is theelementary functions ®, in the above diagram. Thus ® will be one of these elementaryfunctions.Let T be a stopping time having values from the set of mesh points {t,} for the elemen-tary function. Then from the definition of the integral for elementary functions,tAT n—|| bdW = YD (q)(W (tA TAtey1) —W (tA TAR)a k=0If @ is such that tT(@) = tj, then to get something nonzero, you must have t; > t% sok < j—1. Thus the above on the right reduces toj—1¥ &(4) W(t Aten) —W An)k=0It clearly is 0 if j = 0. Define Yio = 0. Thus the integral equalsj-ly Berar] LP (te) (W (A teg1) — W (tAte))j=0Interchanging the order of summation, k < j—1 so j >k-+1 and this equalsn—l1 onLY Zire P (te) (W (Ate) —W (tte)k=0 j=k+1Fry, measurablen—-1i= LY ZesnjO (te) (W (tAters) — W (tt)k=0ThereforetT tn—l[caw [YE FeO) Msi (65.8.15)a 4 (=0