Usually, in most textbooks and research papers, the evolution of the stock price in the Black-Scholes-Merton (BSM) model is given by a so-called stochastic differential equation:
(see also equation (17.1.8) in Chapter 17). This last equation has an equivalent stochastic integral form:
In other words, in the BSM framework, the dynamic of the stock price S is the sum of two random integrals: the integral of the process {μSt, t ⩾ 0} with respect to the time variable and an integral of the process {σSt, t ⩾ 0} with respect to Brownian motion.
Before making sense of these integrals, recall that we mentioned in Chapter 14 that the stock price in the BSM framework is represented by a geometric Brownian motion. Therefore, the latter two representations, in terms of a stochastic differential equation and of the sum of random integrals, should be two equivalent representations of the same stochastic process, namely a GBM. To better understand these concepts, we will provide below a heuristic introduction to stochastic calculus.
Stochastic calculus is a set of tools, in the field of probability theory, to work with continuous-time stochastic processes, just like we do with functions in classical differential and integral calculus. One of the main mathematical objects is the so-called stochastic integral. Stochastic calculus is widely used in mathematical finance, but also in mathematical biology and physics. Stochastic calculus arises naturally in continuous-time actuarial finance whenever, for example, we need to consider dynamic (replicating) portfolios requiring continuous trading or if we need to determine the dynamic of a stochastic interest rate.
The overall objective of this chapter is to provide a heuristic introduction to stochastic calculus based on Brownian motion by defining Ito’s stochastic integral and stochastic differential equations. This is a rather complex topic so the presentation focuses on providing a working knowledge of the material. We aim at an understanding suitable for the pricing and hedging of options in the Black-Scholes-Merton model. This chapter is intended for readers with a stronger mathematical background and is not mandatory to understand the upcoming chapters.
More specifically, the learning objectives are to:
understand the definition of a stochastic integral and its basic properties;
compute the mean and variance of a given stochastic integral;
apply Ito’s lemma to simple situations;
understand how a stochastic process can be the solution to a stochastic differential equation;
recognize the SDEs for linear and geometric Brownian motions, the Ornstein-Uhlenbeck process and the square-root process, and understand the role played by their coefficients;
solve simple SDEs such as those corresponding to linear and geometric Brownian motions, and the Ornstein-Uhlenbeck process.
Before we begin, we emphasize there are two types of stochastic integrals that are of interest in line with the specific objectives:
stochastic Riemann integrals, i.e. classical Riemann integrals of the path of a stochastic process with respect to the time variable (e.g. ∫T0μStdt);
stochastic integrals with respect to Brownian motion, i.e. integrals of a stochastic process with respect to a Brownian motion (e.g. ∫T0σStdWt).
In both cases, the resulting integral will be a random variable, thus the name stochastic integral.
15.1 Stochastic Riemann integrals
Let us consider a stochastic process H = {Ht, 0 ⩽ t ⩽ T} with continuous trajectories, such as for example a linear Brownian motion or a geometric Brownian motion. If we draw a trajectory of the process H, i.e. if we look at one scenario ω of the underlying experiment, then the corresponding path t↦Ht(ω) is a function of the time variable. For the function f(t) = Ht(ω), we could compute the following Riemann integral:
Since a Riemann integral is obtained as the limit of Riemann sums, i.e.
(15.1.1)
where the time points given by ti = it/n partition the interval [0, t], then we can write
Note that even if the notation does not make it explicit, the tis depend on the number of time points n.
For each n, the realization ∑n − 1i = 0Hti(ω) × (ti + 1 − ti) is the (stochastic) total area of n rectangles, where the (random) height of the i-th rectangle is given by the value and the length of its base by the value ti + 1 − ti. This is illustrated in Figure 15.1.
As we can repeat this for every scenario ω ∈ Ω, we have defined a random variable which we will denote simply by
In some sense, it is the (random) area under the curves given by the trajectories of H.
It is possible to show that1 the average random area is such that:
i.e. we can interchange the order of the expectation and the Riemann integral (as long as each quantity is well defined). The integral on the right-hand-side is a genuine Riemann integral of the deterministic function .
Example 15.1.1Riemann integral of standard Brownian motion
Let us compute the average random area below a Brownian trajectory over a unit time interval, i.e. the expectation of the random variable ∫10Wtdt.
As Brownian motion starts from zero and is highly symmetrical, we should expect this expectation to be equal to zero. Indeed, we have
as the (deterministic) function is identically equal to zero, i.e. for all values of t.
◼
Example 15.1.2Riemann integral of a stochastic process
What is the expectation of the random variable given by ∫1.51W2tdt?
We have
where we used the fact that for all values of t.
◼
Stochastic force of mortality
Stochastic Riemann integrals appear in various areas of actuarial science, actuarial finance and mathematical finance. To quantify longevity risk, there exists a string of literature that models the (stochastic) force of mortality by a stochastic process H = {Ht, t ⩾ 0}. In such models, the remaining lifetime τ of an individual subject to a stochastic mortality intensity H = {Ht, t ⩾ 0} is such that
which involves a stochastic Riemann integral of the stochastic force of mortality H.
In credit risk modeling, the force of mortality is interpreted as the default intensity to represent the time until default (in the class of reduced-form models) and failure/hazard rate in engineering to characterize the time until failure of an object.
15.2 Ito’s stochastic integrals
Now, we want to define the stochastic integral of a stochastic process H = {Ht, 0 ⩽ t ⩽ T} with respect to Brownian motion, i.e. give a meaning to a random variable denoted by
As in a stochastic Riemann integral, the integrand is a stochastic process H, but now the integrator is a standard Brownian motion. Symbolically, dt needs to be replaced by dWt. Intuitively, we want to compute the area under the trajectories of H using a Brownian motion W, meaning that both the height (Ht) and the width (dWt) of the rectangles will be random.
As for other integrals, a stochastic integral over the time interval [0, t] is obtained as the limit of Riemann sums as the partition of the interval [0, t] becomes finer and finer. More precisely, the random variable called Ito’s stochastic integral ofHwith respect to Brownian motion is defined by
(15.2.1)
where the time points are given by ti = it/n, for each i = 0, 1, …, n. Again, let us keep in mind that the tis depend on n, so they change when n goes to infinity. The limiting random variable is denoted by ∫t0HsdWs because, as we will see below, it behaves like other types of integrals in many ways; see Section 15.2.4.
Example 15.2.1
Fix a real number c. If Ht = c for all t, i.e. if it is not affected by the scenario ω or by the time index t, then
Since tn = nt/n = t, we have and, by the definition of the stochastic integral given above, we have
In particular, this means that a linear Brownian motion Xt = X0 + μt + σWt can also be written as
◼
Limit of random variables
The limit of a sequence of random variables is a topic often overlooked in a first probability course. But in fact, both the Law of Large Numbers and the Central Limit Theorem rely on limits of random variables. The results are: given a sequence of independent and identically distributed random variables (Xn, n ⩾ 1) with common mean m and common variance σ2, there exists a random variable and a random variable such that
and
where is a constant (not affected by the scenario ω) and is a normally distributed random variable with mean 0 and variance 1.
Similar conclusions will be obtained for stochastic integrals: the properties of the random variable ∫t0HsdWs, given by the limit of the random variables , will be derived from the properties of those Riemann sums.
15.2.1 Riemann sums
Intuitively, if we fix a scenario ω ∈ Ω, then the corresponding paths Ht(ω) and Wt(ω) are both functions of t and then we can write down the non-random Riemann sum:
For each n, this is the total area of n rectangles, where the height of the i-th rectangle is given by and the width of its base is given by .
Example 15.2.2Realization of a Riemann sum
Let us fix n = 4 and , yielding ti = i/12, and let us choose Ht = |Wt|. Consider a scenario ω in which we have:
Note that W0(ω) = 0 by definition, no matter which scenario ω occurs. Since Ht = |Wt| at any time t, we have
One should remark that in this scenario, the area of the first (non-null) rectangle is negative: |0.382927566| × ( − 0.141739491 − 0.382927566) = −0.2009095.
◼
As we can repeat this for every scenario ω ∈ Ω, then for each n we have a doubly random Riemann sum that we denote by
(15.2.2)
It is a random variable that we will call the n-th (random) Riemann sum (of H).
Example 15.2.3Riemann sums of geometric Brownian motion
Let us compute the expectation of the n-th Riemann sum of the process Ht = exp (Wt), i.e.
Recall that since increments of Brownian motion are independent, we have that is independent of , for each i. Consequently,
as increments of Brownian motion are zero-mean random variables.
◼
15.2.2 Elementary stochastic processes
Depending on the integrand, the limiting procedure involved in defining the stochastic integral as given in (15.2.1) is not always necessary. We have seen such a trivial illustration in example 15.2.1. In this section, we introduce a class of stochastic processes called elementary stochastic processes whose stochastic integrals do not require the limiting procedure. Hence, each such stochastic integral is equal to a Riemann sum.
The simplest stochastic process one can imagine2 is a one-step elementary stochastic process, i.e. a process of the form
(15.2.3)
where φ is a random variable3 and where 0 ⩽ t1 < t2 ⩽ T. Here, φ, t1 and t2 specify the elementary process: they are the parameters of H.
Figure 15.2 shows four different trajectories of this process H associated to four different scenarios.
As we can see in Figure 15.2, the trajectories of the process H given in (15.2.3) consist of a single step of height φ(ω) when the scenario is ω. Consequently, for such a process, its stochastic integral does not require the limiting procedure: it is directly equal to the Riemann sum . In other words, the stochastic integral of the process H given in (15.2.3), over the time interval [0, T], is given by the following random variable:
(15.2.4)
Example 15.2.4
If , with 0 ⩽ t1 < t2 ⩽ T, then it is a one-step elementary process as given in (15.2.3) with a constant. Consequently, the stochastic integral of H is given by
In this case, the stochastic integral is normally distributed with mean 0 and variance 3 × (t2 − t1).
◼
Example 15.2.5
If , with 0 ⩽ t1 < t2 ⩽ T, then it is a one-step elementary process as given in (15.2.3) with . Consequently, the stochastic integral of H is given by
In this case, the stochastic integral is the product of two independent zero-mean normal random variables. In particular, we have
and consequently
◼
More generally, a stochastic process H = {Ht, 0 ⩽ t ⩽ T} is said to be an elementary stochastic process if it is such that
(15.2.5)
where 0 = t0 < t1 < … < tn − 1 < tn = T generate non-overlapping time intervals and where, for each i, the random variable φi is completely determined by the Brownian motion up to time ti; this is a mathematical technicality we have to deal with. Note that we have H0 = 0 for any scenario ω.
For a given ω ∈ Ω, the trajectory Ht(ω) is made of steps with various heights determined by the draws of φ1(ω), φ2(ω), …, φn(ω). This is illustrated in Figure 15.3 showing trajectories of Ht(ω) for two scenarios.
Again, as we can see in Figure 15.3, since the trajectories of the process H given in (15.2.5) consist of steps of height φi(ω) when the scenario is ω, then its stochastic integral does not require the limiting procedure. The stochastic integral of H given in (15.2.5) over the time interval [0, T] is the extension of equation (15.2.4): it is given by the following random variable:
(15.2.6)
Example 15.2.6
For the elementary process , the corresponding stochastic integral is simply
Since increments of Brownian motion over non-overlapping intervals are independent, this is a sum of independent normally distributed random variables. Hence, the stochastic integral is normally distributed with mean
and variance
◼
As observed in the last example, if H is an elementary stochastic process as in (15.2.5) with non-random φi, then its stochastic integral is normally distributed with mean zero and variance equal to
Note that in other/most cases, we cannot say much about the distribution of a stochastic integral.
Example 15.2.7
For the process , its stochastic integral is the random variable given by
Using the properties of Brownian motion, we can compute its expectation, which is equal to zero. However, the variance is more challenging to compute because W4 − W2 and W6 are not independent. First, since the mean of the integral is 0, we have
We then need to expand the square within the expectation, which yields three expectations. The first expectation is
The third expectation is
since we are taking the expectation of independent random variables (Brownian increments over non-overlapping intervals). The second expectation is more complicated. We have
which is the expectation of the product of three random variables, some of them being dependent. However, since we can write
then
and its expectation is equal to zero. Combining everything, we obtain
◼
15.2.3 Ito-integrable stochastic processes
We would like to analyze the stochastic integral of a more general class of stochastic processes, which we will call Ito-integrable stochastic processes. Such processes need to have the following properties.
First, the process H = {Ht, 0 ⩽ t ⩽ T} should be an adapted stochastic process, that is a stochastic process such that for each time t, the random variable Ht is completely determined by the Brownian trajectory up to time t, i.e. by the random variables Ws, for all 0 ⩽ s ⩽ t. In some sense, an adapted process follows the flow of information provided by the Brownian motion W: it does not look into the future. Second, the process H also needs to be sufficiently integrable for the variance of its stochastic integral to be finite; more on this below. These are mathematical technicalities we have to deal with, but we will not say more than we already have on this matter. Elementary stochastic processes defined in (15.2.5) are Ito-integrable processes by definition.
The stochastic integral of an Ito-integrable process H, as defined in (15.2.1), requires the limiting procedure, meaning that
is obtained as the limit of the Riemann sums where ti = it/n, for each i = 0, 1, …, n. We can also say that each n-th Riemann sum is the stochastic integral of an elementary stochastic process given by
implying that
Said differently, the stochastic integral of a process H is obtained by approximating H with elementary stochastic processes having an increasing number of time steps n. This is illustrated in Figure 15.4.
Example 15.2.8Stochastic integrals of functions of Brownian motion
First, let us have a look at the following stochastic integral: ∫t0WsdWs. In this case, we have H = W and then, by definition,
Similarly, the stochastic integral ∫t0exp (Ws)dWs is obtained by the following limiting procedure:
Recall that these last Riemann sums have been studied in example 15.2.3.
◼
Convergence in quadratic mean
The type of convergence involved in the definition of the stochastic integral given in (15.2.1) is the convergence in quadratic mean. Mathematically, it means that
In other words, as n increases, the mean-square error between the Riemann sums and the stochastic integral gets closer to zero.
15.2.4 Properties
Let us now analyze some of the properties of Ito’s stochastic integral. It must be noted that since a stochastic integral is a random variable, there will be two types of properties:
properties related to its random nature, i.e. its mean and variance, and whenever possible its probability distribution;
properties related to its integral nature, i.e. its approximation by Riemann sums and its linearity.
As mentioned before, the properties of stochastic integrals are mostly inherited from those of the approximating Riemann sums.
Let H = {Ht, 0 ⩽ t ⩽ T} and G = {Gt, 0 ⩽ t ⩽ T} be Ito-integrable stochastic processes and let α1, α2 be real numbers. Then, for any time T > 0, we have:
Stochastic integrals are zero-mean random variables:
Stochastic integrals have a variance given by:
(15.2.7)
which is also known as Ito’s isometry.
If H is a non-random function of time, i.e. if it is not affected by the scenario ω, then
Stochastic integrals are linear:
Let us show all these properties and how they are inherited from the Riemann sums. For any Ito-integrable process H = {Ht, 0 ⩽ t ⩽ T}, its n-th Riemann sum
where ti = iT/n for each i, is a zero-mean random variable. Indeed, we have
where in the second last equality we used the fact that, for each i, the random variable is independent of the random variable ; this is because is completely determined by the Brownian trajectory up to time ti.
Given that the mean of the Riemann sum is equal to zero, its variance is such that
Distributing the square and using the linearity property of the expectation, we get
When i = j, again by independence, we have
When i < j, using the formula of iterated expectations, we get
where, once more, we used the independence properties of Brownian increments. Of course, changing the role played by i and j, we get a symmetric result when j < i. In conclusion,
Therefore, using the above discussion, we can conclude that
where the last equality comes from the definition of the (deterministic) Riemann integral of the function , as defined in (15.1.1).
Finally, if H = {Ht, 0 ⩽ t ⩽ T} is a non-random function of time, then the s are no longer genuine random variables: they are simple real numbers. Therefore, the n-th Riemann sum
is a linear combination of independent normal random variables given by the increments . This means that the Riemann sum itself is normally distributed. We have already checked that the mean is equal to zero and now the variance becomes
because the expectation of a constant is equal to the constant itself. If we summarize, we have: if H is deterministic, then
The normality of the Riemann sums is not affected by the limiting procedure, so if H is deterministic, then ∫T0HtdWt is indeed normally distributed with mean 0 and variance given by:
Finally, since Riemann sums are clearly linear (they are sums) and since limits are themselves linear, it should be clear that stochastic integrals are linear with respect to their integrands. The details are left to the reader.
Example 15.2.9Stochastic integral of a Brownian motion
In Example 15.2.8, we considered the stochastic integral of a standard Brownian motion, namely the random variable ∫T0WtdWt.
By the above properties, we have that . Note that, in Example 15.2.3, we had already verified that the corresponding n-th Riemann sums were also zero-mean random variables. By Ito’s isometry given in (15.2.7), the variance is given by
Note that since the integrand is W, i.e. a stochastic process (as opposed to a function), we do not know the distribution of ∫T0WtdWt.
◼
Example 15.2.10.Linearity of stochastic integrals
In example 15.2.8, we also considered the stochastic integral ∫T0exp (Wt)dWt.
Let us now consider the following stochastic integral:
We know that it has a zero mean and that its variance can be computed using Ito’s isometry.
Moreover, by the linearity property of stochastic integrals, we can write
where the two stochastic integrals on the right-hand-side have been defined in example 15.2.8.
◼
15.3 Ito’s lemma for Brownian motion
As we now know, a stochastic integral is a random variable with an expectation equal to 0 and an easy-to-compute variance (using Ito’s isometry). We also discovered that in most cases its probability distribution is unknown.
On the other hand, because stochastic integrals are also integrals, we would like to be able to simplify their expressions, in the spirit of the fundamental theorem of calculus in classical calculus.
Before going any further, let us emphasize some differences between stochastic calculus and classical calculus. In classical calculus, we have that
and, more generally,
if f is differentiable and such that f(0) = 0. We used the fact that for a differentiable function, we have .
The analog stochastic integral is given by
Unfortunately, we cannot write something like because the trajectories of Brownian motion are not differentiable functions. Even if we know that this stochastic integral has a mean equal to zero and a variance equal to T2/2 (see Example 15.2.9), we can wonder whether a simpler expression can be obtained or not.
As alluded to above, in classical calculus, one of the main results is the fundamental theorem of calculus: if f is a differentiable function, then
(15.3.1)
The proof of this result is based on a first-order Taylor expansion of f.
There is a similar result in stochastic calculus whose simplest version is as follows: if f is a twice differentiable function, then
(15.3.2)
for all 0 ⩽ t ⩽ T. This result is known as Ito’s lemma and its proof is based on a second-order Taylor expansion of f.
Note that there are two types of integrals in Ito’s lemma of equation (15.3.2):
the stochastic integral of f′(Ws);
the stochastic Riemann integral of f′′(Ws).
Both types have been studied above.
Example 15.3.1
Let us go back to the following stochastic integral:
In order to use Ito’s lemma, we must find a function f( · ) such that Wt = f′(Wt). If we choose f(x) = x2/2, then f′(x) = x and f′′(x) = 1. By Ito’s lemma in (15.3.2), we have
where we used the linearity property of stochastic integrals.
In conclusion, if we reorganize the above equation and since W0 = 0, then we can write
As , we can obtain the probability density function of W2T (by using the change-of-variable formula as seen in an elementary probability course) and then the distribution of this stochastic integral.
◼
Example 15.3.2
Let us use Ito’s lemma to simplify .
For f(x) = ex, we have f′(x) = f′′(x) = ex. Using Ito’s lemma in (15.3.2), we can write
Since , we can write
(15.3.3)
Somehow, the expression on the right-hand-side is simpler:
there is no stochastic integral left, only a stochastic Riemman integral;
the distribution of is known: it is a lognormal distribution.
◼
A glimpse at the proof of Ito’s lemma
Note the similarities between equation (15.3.1) and equation (15.3.2). To gain more intuition as to why they are similar but yet different, consider the following second-order Taylor expansion of f around x:
where the quality of this approximation is linked directly to the size of h. For standard Riemann integrals, the second-order term is negligible. However, this term does play an important role for stochastic integrals. The reason is that standard Brownian motion has a non-zero quadratic variation: for each T > 0, we have
where ti = iT/n, for each i = 0, 1, …, n.
Heuristically speaking, we could replace x + h by Wt + h and replace x by Wt, while keeping in mind that h = (x + h) − x. Then, we can write symbolically
With h = T/n and n large, if we sum up and take the limit:
15.4 Diffusion processes
An important class of processes, applied in various areas including mathematical and actuarial finance, is known as diffusion processes. A diffusion process X = {Xt, 0 ⩽ t ⩽ T} is a continuous-time stochastic process that can be written in the following form:
(15.4.1)
where a( · ) and b( · ) are deterministic functions5 and where is the initial value of the process X. The function a(x) is usually known as the drift coefficient whereas b(x) is typically called the diffusion or volatility coefficient.
In some sense, diffusion processes are generalizations of linear Brownian motions. Indeed, recall from example 15.2.1 that a linear Brownian motion Xt = X0 + μt + σWt can also be written as
In other words, such a linear Brownian motion is a diffusion process as defined in (15.4.1), with a constant drift coefficient a(x) = μ and a constant volatility coefficient b(x) = σ. Note also that the terminology is the same as the one used in Chapter 14. Diffusion processes also extend the commonly used GBM to create richer and more realistic dynamics for asset prices and interest rates.
15.4.1 Stochastic differential equations
In many actuarial and financial models, the dynamics of a financial variable over time (e.g. the stock price) is expressed with a stochastic differential equation. Such an equation specifies the desirable behavior of this quantity over the next infinitesimal period of time.
Symbolically, a stochastic differential equation (SDE) is a stochastic equation: for given functions a( · ) and b( · ), we are looking for a process X such that
(15.4.2)
with initial condition X0 = x0. More rigorously, we are looking for a diffusion process X such that
as already defined in (15.4.1). The stochastic differential equation in (15.4.2) is also called the differential form whereas the expression in (15.4.1) is known as the integral form of a diffusion process.
Intuitively, a SDE can be interpreted as follows: given that we know the value of Xt at time t, the future value of the process one millisecond after is explained by two components:
a component a(Xt)dt: a change of magnitude a(Xt) in the direction of the infinitesimal increase in time dt;
a component b(Xt)dWt: a change of magnitude b(Xt) in the direction of the infinitesimal increment of the Brownian motion dWt.
As a SDE describes the evolution of a quantity Xt as time passes, we are interested in finding an explicit expression for Xt: this is what we will mean by solving the SDE. But recall that a SDE of the form given in (15.4.2) has no meaning by itself: it is rather a symbolic representation for the integral equation given in (15.4.1). Hence, solving a SDE and studying the underlying stochastic process means analyzing its corresponding integral form and its existence. Most of the time, this is a difficult task.
Example 15.4.1Examples of diffusion processes
Let us consider the following SDEs: for given numbers μ and σ,
, with initial value X0 = −1;
, with initial value X0 = 2.
For the first SDE, according to the general form given in (15.4.2), we have a constant drift coefficient a(x) = μ and a constant volatility coefficient b(x) = σ. Therefore, the integral form is given by
and we have recovered a linear Brownian motion issued from − 1. Note also that for this SDE, there is clearly a solution (we found it).
For the second SDE, according to the general form given in (15.4.2), we have a drift coefficient given by a(x) = μx and a volatility coefficient given by b(x) = σx. Therefore, the integral form is given by
For this SDE, it is not clear whether there exists a solution X that verifies this equality.
◼
15.4.2 Ito’s lemma for diffusion processes
In the last example, we were not able to obtain an explicit expression for the diffusion process solving
with initial value X0 = 2. In order to solve such SDEs, we will need a generalized version of Ito’s lemma for diffusion processes. Indeed, in Section 15.3, we saw the simplest form of Ito’s lemma; the most general form, still known as Ito’s lemma, applies to deterministic transformations f(t, Xt) of a diffusion process X.
Let f(t, x) be a function of two variables, i.e. of time and space respectively, such that ∂f/∂t and ∂2f/∂x2, i.e. the first partial derivative in time and the second partial derivative in space,6 are both continuous functions. If X = {Xt, 0 ⩽ t ⩽ T} is a diffusion process solving the general SDE given in (15.4.2), then the process Yt = f(t, Xt) solves the following SDE:
(15.4.3)
Before going any further, we need to make two clarifications. First, the terms and must be understood as follows:
Compute the partial derivatives , and of the function f(t, x).
Evaluate each partial derivative at (t, Xt).
Second, the expression (dXt)2 must be understood as follows: it can be shown that
This is an application of the product rule; see the box below for more details. Hence, substituting dXt by its expression given in (15.4.2) and reorganizing equation (15.4.3), we get
(15.4.4)
This last SDE is equivalent to the one in (15.4.3). They are both referred to as Ito’s lemma for diffusion processes.
While the last expression in (15.4.4) can be tedious, the one in (15.4.3) is more intuitive: it is more or less a Taylor expansion, of order one in time and order two in space. Then, one only needs to be familiar with the product rule to further simplify the expressions.
Product rule
Whenever we need to multiply two SDEs, there exists a convenient and symbolic tool (backed by rigorous mathematics) known as the product rule. It is described in the following table:
Using Ito’s lemma, let us verify that the following simple geometric Brownian motion, i.e. the process
is a diffusion process. We must show that S satisfies an SDE as given in (15.4.2), that is we must find the SDE coefficients.
Clearly, we have St = f(t, Wt), with f(t, x) = ex − t/2, and S0 = 1. First, we compute the partial derivatives:
Second, we replace x by Wt and we use Ito’s lemma (15.4.3) to f(t, Wt) = exp (Wt − t/2), noticing that . We get
since . Substituting f(t, Wt) by St, we see that exp (Wt − t/2) is a solution to the SDE given by
with S0 = 1. In conclusion, we have indeed verified that the simple geometric Brownian motion
is a diffusion process with drift a(St) = 0 and diffusion b(St) = St.
◼
Note that if we consider a transformation f(Xt) in space only, all partial derivatives simplify to ordinary derivatives and then Ito’s lemma of (15.4.3) and (15.4.4) simplify respectively to:
(15.4.5)
and, equivalently,
(15.4.6)
15.4.3 Geometric Brownian motion
We saw in example 15.4.1 that an SDE of the form
is that of a linear Brownian motion with drift μ and diffusion/volatility σ. In other words, Xt = x0 + μt + σWt is a diffusion process.
In this section, we are interested in obtaining a similar characterization, as a diffusion process, for geometric Brownian motions. We want to extend what we did in example 15.4.2 and thus solve the second SDE in example 15.4.1.
We know from Chapter 14 that a geometric Brownian motion S is obtained as the exponential of a linear Brownian motion X. As this is a simple deterministic transformation of a diffusion process, then, using Ito’s lemma, we should be able to find its corresponding SDE.
Let us apply Ito’s lemma to St = exp (Xt), where Xt = x0 + μt + σWt. Here, as f(x) = ex, we will use the version of Ito’s lemma in (15.4.5). First, we have
Consequently, as St = f′(Xt) = f′′(St), using Ito’s lemma we can write
where . Since, by the product rule,
we further have
with . By analogy with Example 15.4.1, this is an SDE with drift coefficient given by a(x) = (μ + σ2/2)x and volatility coefficient given by b(x) = σx. In other words, a geometric Brownian motion is indeed a diffusion process.
It is important to realize that if we consider instead the GBM S given by
then it will be the solution to the following SDE:
This representation corresponds to the SDE in Equation (17.1.8) of Chapter 17.
In summary, we have obtained Table 15.1, for converting a geometric Brownian motion from its SDE representation to its exponential representation, and vice versa.
Table 15.1Equivalence of representations for geometric Brownian motions (with S0 = 1)
SDE representation
Exponential representation
⇔
St = exp ((μ − σ2/2)t + σWt)
⇔
St = exp (μt + σWt)
15.4.4 Ornstein-Uhlenbeck process
In this section, we analyze the Ornstein-Uhlenbeck (OU) process, which is widely used in actuarial finance, especially to model mortality/longevity risk and interest rates in continuous-time short rate models.
First, let us look at a simple version of the Ornstein-Uhlenbeck SDE, namely
(15.4.7)
with an arbitrary initial condition X0 = x0. We know this SDE means that
for all t ⩾ 0. Unfortunately, this cannot be considered as an explicit expression: the process X itself appears on both sides of the equation.
To find the solution to the SDE in (15.4.7), we can use Ito’s lemma along with the function f(t, x) = xet. The partial derivatives of this function are
Thus, using the version of Ito’s lemma in (15.4.3), we can write
with f(0, x0) = x0, which is equivalent to
If we move things around, we get
(15.4.8)
In other words, is a solution to . This is now an explicit expression for the OU process X.
For a fixed time t, since the integrand is deterministic, the stochastic integral in (15.4.8) is normally distributed (see Section 15.2.4) with
and
In conclusion, for each fixed time t, we have
More generally, the simple OU SDE in (15.4.7) can be extended to get the following general OU SDE:
for parameters α, β, σ > 0, and initial value X0 = x0.
This OU dynamic is said to be mean reverting:
if Xt < β, then the drift is positive, pushing the process upward (toward β);
if Xt > β, then the drift is negative, pushing the process downward (toward β).
Note that this is a desirable feature for interest rate modeling. We interpret β as an equilibrium level whereas α is the speed of mean reversion, thus amplifying or mitigating the effect of the gap between Xt and β. This version of the mean-reverting OU process is used in the Vasicek model for the short rate.
It is also possible to find an explicit expression for the general OU process. Again, using Ito’s lemma with f(t, x) = xeαt, we obtain
As a consequence, we have, for each fixed time t > 0, that
15.4.5 Square-root diffusion process
A general OU process can take both positive and negative values. Indeed, since the random variable Xt is normally distributed, it has a non-zero probability of being negative. This can be a major impediment for some actuarial and financial modeling.
This is not the case for the square-root diffusion process. A simple version of this other process can be obtained as the square of a specific OU process. Let X be given by the SDE in (15.4.7), i.e.
Set Yt = X2t. If we use Ito’s lemma with the function f(x) = x2, for which the first two derivatives are
then, after a few manipulations, we get
The general mean-reverting square-root process is defined as the solution of
for parameters α > 0, β > 0, σ > 0, and initial condition X0 = x0. In general, there is no closed-form expression for the solution of this SDE.
This version of the square-root dynamic is used for interest rate modeling, e.g. in the Cox-Ingersoll-Ross (CIR) model, and for volatility modeling, e.g. in the Heston model.
15.5 Summary
Notation
Stochastic processes:
– W = {Wt, 0 ⩽ t ⩽ T} is a standard Brownian motion;
– H = {Ht, 0 ⩽ t ⩽ T} is a continuous-time stochastic process.
Partition of the time interval [0, t]: for each i = 0, 1, …, n, set ti = it/n.
Stochastic Riemann integrals
Definition:
Important property:
Ito’s stochastic integrals
Definition:
where is the n-th Riemann sum.
Elementary stochastic process:
where the φis are random variables.
Stochastic integral of an elementary process:
Important properties:
Mean:
Variance (Iso’s isometry):
If H is a non-random function of time, then
Linearity:
Diffusion processes
Diffusion process: a continuous-time stochastic process X verifying
Ornstein-Uhlenbeck process: solves
and is such that
Square-root diffusion process:
Ito’s lemma
Ito’s lemma for Brownian motion: if f is a twice differentiable function, then
Ito’s lemma for diffusion processes (simple version): if X is a diffusion process and if f(x) is twice differentiable, then
Ito’s lemma for diffusion processes (general version): if X is a diffusion process and if f(t, x) is twice differentiable in x and differentiable in t, then
From the product rule: .
15.6 Exercises
Compute the expected value of the following stochastic Riemann integrals:
∫T0(Wt)2dt;
∫T0(Wt + t)dt;
.
Find the distribution of the following random variables:
;
∫T0Wtdt (hint: use the definition of the Riemann integral as a limit).
Let S = {St, t ⩾ 0} be a geometric Brownian motion given by . Let Yt = f(t, St) be a two-variable transformation of S. Derive the SDE of the following transformations and determine the SDE of which Y = {Yt, t ⩾ 0} is a solution:
Yt = e− rtSt;
Yt = (St)α.
Use Ito’s lemma to show that is a stochastic integral.
Compute for a mean-reverting square-root diffusion process X given by
Hint: consider the process Yt = eαtXt.
Consider the stochastic process
Verify that Xt is the solution to the following SDE: