Semimartingales: Theory, Stochastic
Integration, and Financial Applications
KESSEL WILSON MBOUCHE NZALI
December 4, 2024
Lecturer : SIMON HARDEN
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Table of contents
1 Introduction
2 Semimartingales
3 Stochastic Integration versus Semimartingales
Integration with respect to martingales
Integration with respect to semimartingales
4 Application to finance : Semimartingale Market model
5 Conclusion
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Introduction
1 Extend the concept of
martingales to a broader
class of processes.
2 The need for
semimartingales arose from
the limitations of
martingales.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Introduction
1 Extend the concept of
martingales to a broader
class of processes.
2 The need for
semimartingales arose from
the limitations of
martingales.
• Modeling processes where the future value is expected to be
the same as the present
• Many real-world phenomena, or other forms of deterministic
behavior alongside randomness.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Martingale : Gambler’s Ruin
• Bet 1$ (Loss: 1 $).
• Bet 2$ (Loss: 3$).
• Bet 4$ (Loss: 7$)
• Bet 8$ (Win 8$ : Recover
7$ and make a profit of 1$)
The expected wealth after the next round is equal to their
current wealth.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Martingale: Formal definition
Martingales
• The concept of martingale in probability theory was
introduced by Paul Lévy in 1934.
• Main idea : The expected future value of a process, given
all current and past information, is equal to its present value.
• Mathematically can be written like :
E(X(t)|Fs ) = X(s) almost surely for 0 ≤ s < t,
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Local martingale and càdàlag function
Localmartingales
• While a martingale satisfies the martingale property globally
(over the entire time horizon), a local martingale satisfies
this property only locally, within certain restricted intervals of
time.
,
càdlàg function
• A function f : E −→ M is called càdlàg function if for every
t ∈ E.
1 The left limit f (t − ) = f (x ) exist.
x −→t −
2 The right limit f (t + ) = f (x ) exist and is equivalent to f (t).
x −→t +
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
SEMIMARTINGALES
Semmimartingales
A right-continuous with left limits (càdlàg) adapted process S(t) is
a semimartingale if it can be represented as a sum of two
processes: a local martingale M(t) and a process of finite variation
A(t), with M(0) = A(0) = 0, and :
S(t) = S(0) + M(t) + A(t) (1)
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
SEMIMARTINGALES : EXAMPLES
Examples
1 Let M be a martingale, then M is a semi-martingale.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
SEMIMARTINGALES : EXAMPLES
Examples
1 Let M be a martingale, then M is a semi-martingale.
2 Let B a Brownian motion, Then S(t) = B 2 (t) is a semi
martingale since :
S(t) = S(0) + M(t) + A(t) (2)
with M(t) = B 2 (t) − t and A(t) = t.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Definition : (Predictable processes)
(Predictable processes)
A process X is predictable if it’s one of the following:
1 A left-continuous adapted process, in particular a continuous
adapted process.
2 A limit of a left-continuous adapted process.
3 A Borel-measurable function of a predictable process.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Definition : (Predictable processes)
(Predictable processes)
A process X is predictable if it’s one of the following:
1 A left-continuous adapted process, in particular a continuous
adapted process.
2 A limit of a left-continuous adapted process.
3 A Borel-measurable function of a predictable process.
Importance
Z T
H(t)dS(t) (3)
0
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Stochastic integral of Simple processes
1 For a simple predictable process H(t), given by
n−1
X
H(t) = Hi 11[ti ,ti+1 ] (t) (4)
i=0
with 0 = t0 ≤ t1 ≤ t2 · · · ≤ tn = T and Hi are Fti -measurable
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Stochastic integral of Simple processes
1 For a simple predictable process H(t), given by
n−1
X
H(t) = Hi 11[ti ,ti+1 ] (t) (4)
i=0
with 0 = t0 ≤ t1 ≤ t2 · · · ≤ tn = T and Hi are Fti -measurable
2 The stochastic integral is :
Z T n−1
X
H(t)dM(t) = Hi (M(ti+1 − M(ti ))) (5)
0 i=0
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Broad class of processes
1 M(t) is locally square integrable martingale and H a
predictable process.
Z T
H 2 (t)d[M, M](t) < ∞ almost surely (6)
0
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Broad class of processes
1 M(t) is locally square integrable martingale and H a
predictable process.
Z T
H 2 (t)d[M, M](t) < ∞ almost surely (6)
0
RT
2
0 H 2 (t)d[M, M](t) is locally integrable, i.e, There is a
sequence of stopping time τn −→ ∞ such that
Z T ∧τn
E H 2 (t)d[M, M](t) < ∞ (7)
0
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Stochastic integral of a broad class of predictable process
Stochastic integral
There exist a sequence of simple predictable processes (Hn )n that
approximates H and te stochastic integral is defined by :
Z T Z T
H(t)dM(t) = lim Hn (t)dM(t) (8)
0 n−→∞ 0
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Integration with respect to semimartingales
•Let S be a semi martingale with the representation :
S(t) = S(0) + M(t) + A(t) (9)
• Let H be a predictable process such that condition (6) and
(10)hold.
Z T
| H(t) | dVA (t) < ∞ (10)
0
where VA (t) is the variation of the process of A.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Integration with respect to martingales
Stochastic Integration versus Semimartingales
Integration with respect to semimartingales
Application to finance : Semimartingale Market model
Conclusion
Integration with respect to semimartingales
Then the stochastic integral is defined as the sum of integrals,
Z T Z T Z T
H(s)dS(t) = H(t)dM(t) + H(t)dA(t). (11)
0 0 0
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Application to finance : Semimartingale Market model
Framework: No-arbitrage
• The main premise is the same as in discrete time, it should be
impossible to make “something” out of nothing without
taking risk.
• Let a(t) and b(t) denote the number of shares and bond units
respectively held at time t. The market value of the portfolio
at time t is given by
V (t) = a(t)S(t) + b(t)β(t) (12)
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Self-financing portfolio
Definition
A portfolio (a(t), b(t)), 0 ≤ t ≤ T , is called self-financing if the
change in value comes only from the change in prices of the assets,
dV (t) = a(t)dS(t) + b(t)dβ(t) (13)
Z t Z t
V (t) = V (0) + a(u)dS(u) + b(u)dβ(u). (14)
0 0
• It is assumed that S(t) and β(t) are semimartingales.
• The processes a(t) and b(t) must be predictable processes
satisfying (6)-(10).
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Self-financing portfolio : Crucial result
Theorem
(a(t), b(t)) is self-financing if, and only if, the discounted value
process Vβ(t)
(t)
is a stochastic integral with respect to the discounted
price process
s
Z t
V (t)
= V (0) + a(u)dZ (u) (15)
β(t) 0
where Z (t) = S(t)/β(t).
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Self-financing portfolio : Importance
Importance
1 There are no external cash injections or withdrawals after the
initial investment.
2 Self-financing strategies is crucial because it sets the
boundaries for what is considered possible or reasonable in the
financial market.
3 Track and manage the value of the portfolio in other to make
the maximum benefit.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Conclusion
What did we do?
1 Discussed about martingales and semimartingales
2 Talked about the Stochastic integration with respect to
semi-martingales.
3 Discussed their role in portfolio-management.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
Bibliography
Rogers, L.C.G. and Williams, D. (1990). Difusions, Markov
Processes, and Mar- tingales, Vol. 2, Ito Calculus, reprint,
Wiley, New York.
Fima C Klebaner Monash University, Australia Introduction To
Stochastic Calculus With Applications.
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
Introduction
Semimartingales
Stochastic Integration versus Semimartingales
Application to finance : Semimartingale Market model
Conclusion
THANK
YOU FOR
YOUR KIND
ATTENTION
Questions are welcome!
KESSEL WILSON MBOUCHE NZALI Semimartingales: Theory, Stochastic Integration, and Financial
