We work on a filtered probability space (Ω,F,P), where Ω⊂R is the sample space, F a σ-algebra consisting of events and P a probability measure.
Definition 3.1 (Filtration): Let T be an index set with a total order. The non-decreasing family of σ-algebras (Ft)t∈T lying in F, i.e. ∀t∈T:Ft⊂F, is called a filtration.
For us, T will denote time and therefore be a subset of R+.
Definition 3.2 (Stochastic process): A real-valued stochastic process is a collection of R-valued random variables usually written as {Xt:Ω→R∣t∈T}.
Definition 3.3 (Sample path): A sample path or trajectory describes a single outcome of a stochastic process, i.e. it is a map X(⋅)(ω):T→R for some fixed ω∈Ω.
Definition 3.4 (Increment): An increment is defined as the difference between two random variables of a certain stochastic process, i.e. Xti−Xtj where ti,tj∈T.
3.1.1 Special Stochastic Processes
Definition 3.5 (Markov process): A stochastic process satisfies the Markov property if its value in the future depends on the current value but is conditionally independent of its previous behaviour, i.e. for every n and t1<t2<…<tn, it holds
P(Xtn≤xn∣Xt1=x1,…,Xtn−1=xn−1)=P(Xtn≤xn∣Xtn−1=xn−1)
Note: We can write the Markov property for continuous-time stochastic processes as P(Xt+s∈A∣Ft)=P(Xt+s∈A∣Xt) for all s>0.
Example (Markov process): We will later see that a simple Random Walk is a Markov process.
Definition 3.6 (Martingale): A martingale is a stochastic process where the conditional expectation of a future value Xt given the current value Xs and previous values Fs equals the current value, i.e. EP[Xt∣Fs]=Xs for all s<t.
In other words, our expected gain in the process is zero at all times. We can also view this definition as a mathematical formalization of a game of chance being fair.
Note: For the martingale definition to be valid, the first moment of the random variables of the martingale must exist, i.e. EP[∣Xt∣]<∞ for all t∈T.
A process can be a martingale without being a Markov process and vice versa. This is because a Markov process's definition is about the conditional distribution of future values depending only on the current value, while a martingale's definition is about the conditional expectation of future values being equal to the current value.
Definition 3.7 (Stationary stochastic process): A stochastic process is stationary if all its random variables are identically distributed, i.e. for all n∈N the random variables Xt1,…,Xtn have the same probability distribution.
Example (Stationary stochastic process): The process {Xt} with Xt∼N(0,σ2) for all t∈T is stationary.
Definition 3.8 (Lévy process): A Lévy process is a stochastic process with stationary and independent increments, i.e. for every n∈N and t1<t2<…<tn, the n−1 increments Xt2−Xt1,…,Xtn−Xtn−1 are independent of each other and the distribution of an increment is fully determined by its time difference Δt.
Note:
All Lévy processes are Markov processes.
A Lévy process is a martingale if and only if its mean is zero, i.e. E[X0]=0.
Example (Lévy process): We will later see that Brownian motion is a Lévy process.
3.1.2 Convergence of Random Variables
We recall the convergence of random variables.
Definition 3.9 (Convergence of random variables): Let X1,X2,… be a sequence of real-valued random variables and X be a random variable as well. Then the sequence converges
almost surely if P[limn→∞Xn=X]=1
in the Lr-norm if E[∣Xn∣r] and E[∣X∣r] exist and limn→∞E[∣Xn−X∣r]=0
in probability if for all ε>0limn→∞P[∣Xn−X∣>ε]=0
weakly/in distribution if limn→∞Fn(x)=F(x) for all x∈R
Note: 1. and 2. imply 3., which in turn implies 4. Moreover, Lp convergence implies Lq convergence for p>q≥1.
3.1.3 Quadratic variation and covariation
Let P be a partition of [0,T] and its norm ∥P∥ given by its mesh size max1≤k≤n∣tk−tk−1∣.
Definition 3.10 (Quadratic variation): Let Xt be a process. The quadratic variation is
[Xt]T=P∥P∥→0limk=1∑n(Xtk−Xtk−1)2
Note: The quadratic variation vanishes for differentiable functions. For stochastic processes this is not necessarily the case as we will see later.
Definition 3.11 (Covariation): Let Xt and Yt be processes. The covariation is
[Xt,Yt]T=P∥P∥→0limk=1∑n(Xtk−Xtk−1)(Ytk−Ytk−1)
Note: Both the quadratic variation and the covariation are defined as convergence in probability.
3.1.4 Mathematical Formulation of Brownian Motion
Let Wt denote the position at time t∈T in one dimension of a physical particle. The collection of these random variables indexed by the continuous-time parameter t is a stochastic process with the properties mentioned below.
Theorem 3.12 (Brownian motion): The Brownian motion {Wt} is the real-valued stochastic process that satisfies the following:
starting point:W0=0
Lévy process: non-overlapping increments Wt−Ws for all t,s∈T such that s≤t are independent and follow the normal distribution Wt−Ws∼N(0,t−s)
continuous paths:W has continuous paths, i.e. t→Wt(ω) are continuous almost surely
From the definition of Brownian motion, we derive the probability density and probability distribution.
Proposition 3.13 (Probability density of Brownian motion): The probability density of the position x of a Brownian motion at the end of the time period [0,t] is
p(x)=2πt1exp(−2tx2)
Proposition 3.14 (Probability distribution of Brownian motion): The probability distribution of the increment Wt+u−Wt is
P[Wt+u−Wt≤a]=∫−∞a2πu1exp(−2ux2)dx
Note: As the probability distribution of Wt is normal with standard deviation t, it is the same as tZ,Z∼N(0,1). When calculating expectations of Brownian motions, it is convenient to use tZ.
3.2 Stock Price Dynamics
Our goal is to formulate a mathematical object S that can describe the movement of stock prices through time. We define {St} as a real-valued continuous-time process on a filtered probability space (Ω,{Ft}t∈[0,T],P). We would like it to satisfy the following:
If we know the present, then the future is independent of the past, i.e. {St} should be a Markov process.
The paths should be continuous.
The paths should be almost surely “erratic”, i.e. they should not be differentiable, and hence predictable, with respect to time.
With this wishlist we can define the concept of a diffusion process as follows.
Definition 3.15 (Diffusion process): A diffusion process {St} is a real-valued stochastic process that satisfies the following:
Markov process:P(St+h∈A∣Ft)=P(St+h∈A∣St)
continuous paths:P[∣St+h−St∣≤ε∣St]=1−o(h)
local expectation:EP[St+h−St∣St]=μh+o(h), i.e. the local expectation of the changes are approximately proportional to the length h
local variance:EP[(St+h−St)2∣St]=σ2h+o(h), i.e the local variance of the changes is approximately proportional to the length h
The return of a stock is the percentile difference between the prices at consecutive times, i.e. StSt+Δt−St. Written infinitesimally, this is
StdSt=μdt+σdWt
where we model the returns with a percentage drift μ and percentage volatility σ. We chose the Brownian motion for the erratic volatility term to fulfill the requirements imposed on {St}. The next chapter will cover how to solve this stochastic differential equation.
3.3 Random Walk
To get a first intuition, we look at a discrete-time model, the Random Walk, which will also show us why the variance of Brownian motions is proportional to time. We assume that the time t is discrete, i.e. T={t0,t1,…,tn} with layer spacing Δt=tk+1−tk=nT.
Definition 3.16 (Random Walk): The Random Walk {Wk} is the discrete-time real-valued process that satisfies the following:
starting point:W0=0
evolution: up and down moves have size Δt with probability p=21
Hence, if {X1,X2,…} is a sequence of independent binomial random variables taking values ±Δt with equal probability, then the position at time T is Wn=∑i=1nXi.
Proposition 3.17 (Properties of the Random Walk): The Random Walk Wn has the following properties:
E[Wn]=0
Var[Wn]=T
{Wk} is a Markov process
Wn can reach n+1 values
limn→∞Wn∼CLTN(0,T)
We will show 5. of Proposition 3.17. For that, recall the definition of the moment generating function.
Recap (Moment Generating Function): The moment generating function or MGF of a random variable X is defined as
mX(θ)=E[exp(θX)]
for all values θ for which the expectation exists.
Proof: The probability distribution of Wn is determined uniquely by its MGF, i.e.
mWn(θ)=E[exp(θWn)]=k=1∏nE[exp(θXk)]=E[exp(θX)]n
Since E[exp(θX)]=21exp(θΔt)+21exp(−θΔt)≈1+21θ2Δt for small Δt using Taylor series expansion, we have
mWn(θ)≈(1+21θ2Δt)n
We write the MGF in logarithmic form
ln(mWn(θ))≈nln(1+21θ2Δt)≈n21θ2Δt=21θ2T
where we have used ln(1+y)≈y for small y. Hence mWn(θ)n→∞exp(21θ2T) which is the MGF of the normal distribution N(0,T).
Thus in the continuous-time limit of the discrete-time framework, the probability density of the terminal position of the Random Walk Wn is the same as that of a Brownian motion that has run an amount of time T. In other words, if the mesh size tends to zero, the Random Walk converges to the Brownian motion.
Finally, we prove that Var[Wn]=T, i.e. property 2. of Proposition 3.17, arises only if we choose ±Δt as the size of the Random Walk increment.
Proof: Let y be the increment over time step Δt such that Xk=±y. Then
Var(Wn)=nVar(Xk)=ny2=ΔtTy2=TΔty2
In order to ensure that Var(Wn) stays finite as Δt→0 and y→0 we choose y2=cΔT where c>0 is a constant. Then Var(Wn)=Tc and since time units are arbitrary there is no sense in choosing c differently from 1. Hence, y=±Δt.
This legitimizes our choice of ±Δt as the increment.
3.4 Three Key Properties
We will prove three key properties of the Brownian motion.
3.4.1 Non-Differentiability
Recall the definition of differentiability.
Recap (Differentiable function): A continuous function f is differentiable at x if the below and above limits exist and they coincide, i.e.
h↑0limhf(x+h)−f(x)=h↓0limhf(x+h)−f(x)
For the Random Walk, the slope of the path is Δt∣Wk+1−Wk∣=Δt1 which becomes infinite as Δt→0. This gives us a hint that Brownian motion paths are nowhere differentiable. We prove this in the following.
Proof: For the Brownian motion path, we consider the rate of change over [t,t+Δt], i.e. XΔt=ΔtWt+Δt−Wt∼N(0,Δt1). Hence, we can write XΔt∼ΔtZ with Z standard normal. Take any interval A=(a,b], then P[XΔt∈A]=P[aΔt<Z≤bΔt]. As Δt→0 we get P[XΔt∈A]=0. As A can be chosen arbitrarily, we conclude that the rate of change is not finite. And since t is arbitrary, we also conclude that the Brownian motion path is nowhere differentiable.
3.4.2 Continuity
We prove that Brownian motion paths are continuous.
Proof: We calculate
P[∣Wt+Δt−Wt∣<ε]=2πΔt1∫−εεexp(−2Δtx2)dx=2π1∫−ΔtεΔtεexp(−2u2)du
For Δt→0 we have
2π1∫−∞∞exp(−2u2)du=1
Hence, the Brownian motion path is continuous in probability.
3.4.3 Variability
Proposition 3.18 (Quadratic variation of Brownian motion): The quadratic variation of a Brownian motion Wt after time T is [Wt]T=PT.
We prove this property by showing that [Wt]T converges in L2-norm to T, which implies convergence in probability by Definition 3.9.
Proof: We consider [0,T] with partition Δ=nT and tk=kΔt. Then
=E(k=0∑n−1(Wtk+1−Wtk)2−T)2E(k=0∑n−1((Wtk+1−Wtk)2−Δt))2
Expanding the square results in full square terms ((Wtk+1−Wtk)2−Δt)2 and cross-terms
((Wtk+1−Wtk)2−Δt)((Wtj+1−Wtj)2−Δt)=Δt2(Zk2−1)(Zm2−1)
where Zk,Zm are iid N(0,1). Taking the expectation of the cross-terms results in
Δt2(E[Zk2]−1)(E[Zm2]−1)=0
as Zk2,Zm2 are iid χ12 and have mean E[Zk2]=E[Zm2]=1. Hence
==E(k=0∑n−1(Wtk+1−Wtk)2−T)2E[k=0∑n−1((Wtk+1−Wtk)2−Δt)2]k=0∑n−1E[((Wtk+1−Wtk)2−Δt)2]
Expanding the square, results in
===E[(Wtk+1−Wtk)4]−2E[(Wtk+1−Wtk)2]Δt+Δt2E[(ΔtZ)4]−2E[(ΔtZ)2]Δt+Δt2Δt2(3−2+1)2n2T2
which tends to 0 as n→∞. Hence [Wt]T=L2T and thus [Wt]T=PT.
Proposition 3.19 (First variation of Brownian motion): A Brownian motion has unbounded first variation, i.e. given a partition P of [0,T] the limit lim∥P∥→0∑k=1n∣Xtk−Xtk−1∣ is unbounded.
Note: The first variation is also called total variation. In general, any continuous process with non-zero quadratic variation has unbounded first variation.