Stochastic Calculus

Abstract
Would be including the notes and topics from Shreve’s book as well topics that would fall into this category. Will be only including theory , the practical and implementation part woulld be on a different page and that would be bifurcated further based on the programming language. C++ is currently the frontrunner but since visualization is easier in Python i would be writing code for that as well.

DISCLAIMER THERE IS AT TIMES LOT OF HAND WAVING ,THIS IS NOT AS MATHEMATICALLY RIGOROUS AS YOU WOULD EXPECT LIKE IN A PURE MATH TEXTBOOK.

Resources

These are list of books , videos and courses that I am currently referencing for content on this page , for more detailed list please check out the resource section above.

  • Mathematical modelling and computation in finance by Cornerlis Ooseterlee and Lech Grzelak (Main Text)

  • Stochastic CalcuIus for Finance II Continuous-Time Models by Steven Shreve (Main Reference text)

  • Stochastic Differential Equations By Bernt Oksendal

Preliminaries

Definition 1 (\sigma algebra) : Let \Omega be a nonempty set , and let \mathcal{F} be a set of collections of subsets of \Omega. We say that \mathcal{F} is a \sigma algebra provided that

  1. The empty set \emptyset belongs to \mathcal{F}
  2. A and A^{c} belong to \mathcal{F} whenever A \in \mathcal{F}
  3. whenever A_1 ,A_2 ,A_3 ..... \in \mathcal{F} then \bigcup_{i=1}^\infty A_{i} \in \mathcal{F}



Definition 2 (Probability measure) : Let \Omega be a nonempty set , and let \mathbb{F} be \sigma algebra of subsets of \Omega. A probability measure \mathbb{P} is a function that, to every set A \in \mathcal{F} , assigns a number in [0,1] called the probability of A written as \mathbb{P}(A) .We require

  1. \mathbb{P}(\Omega) = 1 and

  2. (Countable Additivity) Whenever A_1,A_2,A_3.... is a sequence of disjoint sets in \mathcal{F} , then \mathbb{P}(\bigcup_{n=1}^\infty A_n) = \sum_{n=1}^{\infty} \mathbb{P}(A_n)


The Triple (\Omega, \mathcal{F},\mathbb{}P) is called probability measure


From the above it follows that for finitely many disjoint sets A_1, A_2,A_3....A_n \mathbb{P}(\bigcup_{n=1}^N) = \sum_{n=1}^N \mathbb{P}(A_n)

and \mathbb{P}(A^c) = 1 - \mathbb{P}(A)

Example 1 (Uniform(Lebesgue) measure on [0,1])
\Omega = [0,1]
\mathbb{P}[a,b] = b-a ,0 \leq a \leq b \leq1
single points have zero probability and
(a,b) can be written as \bigcup_{n=1}^\infty[a+\frac{1}{n} , b -\frac{1}{n}]



The \sigma-algebra beginning with closed intervals and adding everything else necessary in order to have a \sigma- algebra is called a Borel \sigma-algebra and denoted by \mathcal{B}[0,1]. Borel \sigma-algebras are \sigma-algebra generated by open sets and equivalently closed sets over any topological space.

Definition 3 let (\Omega,\mathcal{F},\mathbb{P}) be a probability space. If a set A \in \mathcal{F} satisfies \mathbb{P}(A) = 1, we say the event A occurs almost surely

Example 2 Let \mathbb{P} be the uniform measure as defined in example 1. Define X(\omega) = \omega and Y(\omega) = 1 - \omega for all \omega \in [0,1] , then the distribution measure of X is uniform \mu_{X}[a,b] = \mathbb{P}\{\omega;a\leq X(w)\leq b\} = \mathbb{P}[a,b] = b-a, 0\leq a \leq b \leq 1 by definition of \mathbb{P}.
Its easy to see to see that X and Y have the same distribution. But under the probability measure\mathbb{\widetilde{P}} on [0,1] defined by \mathbb{\widetilde{P}[a,b]} = \int_a^b 2\omega \, d\omega = b^2-a^2 , 0 \leq a\leq b \leq1 X and Y have different distributions.

Definition 4 (cumulative distribution function and density function) F(x) = \mathbb{P}\{X \leq x\} , x \in \mathbb{R} \mu_X[a,b] = \mathbb{P}\{a \leq X \leq b \} = \int_a^b f(x) \, dx , -\infty < a \leq b < \infty f(x) is called the density funtion and F(x) is called the cummulative distribution function.

Example 3 (Standard normal random variable) Let \phi(x) = \frac {a}{\sqrt{2\pi}}e^{\frac{-x^2}{2}} be the standard normal density and definiing the cummulative normal distribution function as N(x) = \int_{-\infty}^x \phi(\xi)\,d\xi The function N(x) is strictly increasing function which is surjective onto (0,1) from \mathbb{R}, so it has a strictly increasing inverse function N^{-1}(y),y \in (0,1). Let Y be a uniformly distributed random variable defined on some probability space (\Omega,\mathcal{F},\mathbb{P}) and set X = N^{-1}(Y) then

\begin{align*} \mu_X[a,b] &= \mathbb{P}\{\omega \in \Omega ; a \leq X(\omega) \leq b\} \\ &= \mathbb{P}\{\omega \in \Omega ; a \leq N^{-1}(Y(\omega)) \leq b\}\\ &= \mathbb{P}\{\omega \in \Omega ; N(a) \leq Y(\omega) \leq N(b) \} \\ &= N(b) - N(a) \\ &= \int_a^b \phi(x) \, dx \end{align*}

Any random variable that has this distribution regardless of the probability space is called standard normal distribution.
Thing to note the use of uniformly distributed random variable for generating a standard random variable is called probability integral transform and this is commonly used in Monte Carlo simulation



Stochastic Processes



Black-Scholes model