Conditioned Brownian motion, 1st Laplacian eigenfunction, etc…

Brownian Motion

Brownian Motion

A few days ago, a student asked the following question:

Consider a bounded domain {D \subset \mathbb{R}^d} and a Brownian motion conditioned to stay inside {D}. How does it look like ? What is the invariant distribution ?

This is very simple, and surprisingly interesting. This involves the first eigenfunction of the Laplacian on {D}. To make everything simple, and because this does not change anything, it suffices to study the situation where {D=[0,1] \subset \mathbb{R}}. In other words, what does a real Brownian motion conditioned to stay inside the segment {[0,1]} look like.

Discretisation

One could directly do the computations in a continuous setting but, as it is often the case, it is simpler to consider the usual random walk discretisation of a Brownian motion. For this purpose, consider a time discretisation {\delta>0} and a standard random walk {X_k} with increment {\pm \sqrt{\delta}}: with probability {\frac12} the random walk goes up {+\sqrt{\delta}} and with probability {\frac12} it goes down {-\sqrt{\delta}}. For clarity, assume that there exists an integer {m_{\delta} \in \mathbb{N}} such that {m_{\delta} \sqrt{\delta} = 1}. Suppose as well that it is conditioned on the event {X_k \in [0,1]} for {k=1, \ldots, N}. Later, we will consider the limiting case {N \rightarrow \infty} and {\delta \rightarrow 0}, in this order, to recover the Brownian motion case.

Conditioning

First, let us compute the probability transitions of the random walk {\{X_k\}_{k=0}^N} conditioned on the event {X_k \in [0,1]} for {k=1, \ldots, N}. For clarity, let us denote the conditioned random walk by {\hat{X} = \big\{ \hat{X}_k \big\}_{k \geq 0}}. This is a Doob h-transform, and the resulting process is a non-homogenous Markov chain. In this simple case the computations are straightforward. The conditioned Markov chain has transition probabilities given by {\hat{P}(k,x,y) = \mathbb{P}[\hat{X}_{k+1}=y \,|\hat{X}_k=x]} with {x,y \in D_{\delta}} where {D_{\delta} = \sqrt{\delta} \mathbb{N} \; \cap \; [0,1]}. One can compute the probability that the conditioned Markov chain follows a given trajectory {(x_0, \ldots, x_N)} of {D_{\delta}},

\displaystyle  \begin{array}{rcl}  \mathbb{P}[\hat{X}_0 = x_0, \ldots, \hat{X}_N=x_N] = \frac{1}{Z(N,\delta)} P(x_0,x_1) \times \ldots \times P(x_{N-1}, x_N) \end{array}

where {P(x,y)} is the transition kernel of the unconditioned Markov chain and {Z(N,\delta) = \mathbb{P}[X_k \in D_{\delta} \; \text{for} \; k=0, \ldots, N]} is a normalisation constant. The Doob h-transform simply consists in noticing that this also reads

\displaystyle  \begin{array}{rcl}  \mathbb{P}[\hat{X}_0 = x_0, \ldots, \hat{X}_N=x_N] = \hat{P}(0,x_0,x_1) \times \ldots \times \hat{P}(N-1,x_{N-1}, x_N) \end{array}

where the conditioned Markov kernel is {\hat{P}(k,x_k,x_{k+1}) = \frac{P(x,y) \, h(k+1,y)}{h(k,x)}} and the function {h(\cdot, \cdot)} is defined by

\displaystyle  \begin{array}{rcl}  h(k,x) = \mathbb{P}[X_{j} \in D_{\delta} \; \text{for} \; k+1 \leq j \leq N \, |X_k=x]. \end{array}

Of course we have {h(N,x)=1} for all {x \in D_{\delta}}. The quantity {h(x,k)} is the probability that a random walk starting at {x \in D_{\delta}} at time {k} remains inside {D_{\delta}} for time {k,k+1, \ldots, N}. Consequently, in order to find the transition probabilities of the conditioned kernel, it suffices to compute the quantities {h(k, x)} for all {x \in D_{\delta}}. Since we are interested in the limiting case {N \rightarrow \infty}, it actually suffices to consider the case {k=0}. It can be computed recursively since {h(k,x) = \frac12 \, h(k+1, x+\sqrt{\delta}) + \frac12 \, h(k+1, x-\sqrt{\delta})} with the appropriate boundary conditions. In other words, adopting the obvious matrix notations, the vector {h(k,\cdot)} satisfies

\displaystyle  \begin{array}{rcl}  h(k,\cdot) = A h(k+1, \cdot) \end{array}

where {\big( A_{i,j} \big)_{0 \leq i,j \leq m_{\delta}}} is the usual tridiagonal matrix given by {A_{i,j} = 1} if, and only if, {|i-j|=1} and {A_{i,j} = 0} otherwise. It is related to the discrete Laplacian operator. Indeed, because all the eigenvalues of {A} are real with modulus strictly inferior to {1}, it follows that {h(0,\cdot) = A^{N} \textbf{1} \approx \lambda^N \, \varphi_{\delta}(\cdot)} where {\lambda_{\delta}} is the highest eigenvalue of {A} and {\varphi_{\delta}(\cdot)} the associated eigenfunction. The eigenvalues of {A} are well-known, and as {\delta \rightarrow 0}, the highest eigenvalue converges to {1} and the associated eigenfunction converges to the first eigenfunction of the Laplacian on the domain {D=[0,1]} with Dirichlet boundary. In our case it is {\varphi(u) = \sin(\pi \, u)} and

\displaystyle  \begin{array}{rcl}  \lim_{N \rightarrow \infty} \frac{ h(1,x \pm \sqrt{\delta})}{h(0,x)} = \lambda_{\delta}^{-1} \frac{ \varphi_{\delta}(x \pm \sqrt{\delta})}{\varphi_{\delta}(x)}. \end{array}

In other words, the random walk with increments {\pm \sqrt{\delta}} conditioned to stay inside {D_{\delta}} has probability transitions given by

\displaystyle  \begin{array}{rcl}  \hat{P}(x, x \pm \sqrt{\delta}) = \frac{{\lambda_{\delta}}^{-1}}{2} \frac{ \varphi_{\delta}(x \pm \sqrt{\delta})}{\varphi_{\delta}(x)}. \end{array}

Next section investigates the limiting case {\delta \rightarrow 0}.

Conclusion

We have computed the dynamics of the conditioned random walk with space-increments {\sqrt{\delta}}. To obtain the dynamics of the conditioned Brownian motion it suffices to consider the limiting case {\delta \rightarrow 0}. The drift of the resulting diffusion is given by

\displaystyle  \begin{array}{rcl}  \text{(drift)} &=& \lim_{\delta \rightarrow 0} \frac{1}{\delta} \Big\{ \hat{P}(0,x,x+\sqrt{\delta})\sqrt{\delta} - \hat{P}(0,x,x-\sqrt{\delta})\sqrt{\delta} \Big\} \\ &=& \lim_{\delta \rightarrow 0} \frac{{\lambda_{\delta}}^{-1}}{2 \sqrt{\delta}} \Big\{ \frac{ \varphi_{\delta}(x + \sqrt{\delta})}{\varphi_{\delta}(x)} - \frac{ \varphi_{\delta}(x - \sqrt{\delta})}{\varphi_{\delta}(x)} \Big\} = \frac{1}{2} \frac{\varphi'(x)}{\varphi(x)}. \end{array}

The same computation gives the volatility of the resulting diffusion. It is given by

\displaystyle  \begin{array}{rcl}  \text{(volatility)} &=& \lim_{\delta \rightarrow 0} \frac{1}{\delta} \Big\{ \hat{P}(0,x,x+\sqrt{\delta})\delta + \hat{P}(0,x,x-\sqrt{\delta})\delta \Big\} \\ &=& \lim_{\delta \rightarrow 0} \frac{{\lambda_{\delta}}^{-1}}{2} \Big\{ \frac{ \varphi_{\delta}(x + \sqrt{\delta})}{\varphi_{\delta}(x)} + \frac{ \varphi_{\delta}(x - \sqrt{\delta})}{\varphi_{\delta}(x)} \Big\} = 1. \end{array}

As the consequence, this shows that a Brownian motion conditioned to stay inside {D=[0,1]} follows the stochastic differential equation {dX = \frac{1}{2} \frac{\varphi'(X)}{\varphi(X)} \, dt + dW_t} where {\varphi(\cdot)} is the first eigenfunction of the Laplacian on {D} with Dirichlet boundary conditions. More generaly, the same argument would show that a Brownian motion in {\mathbb{R}^d} conditioned to stay inside a nice bounded domain {D \subset \mathbb{R}^d} evolves according to the stochastic differential equation

\displaystyle  \begin{array}{rcl}  dX = \frac{1}{2} \frac{\nabla \varphi(X)}{\varphi(X)} \, dt + dW_t \end{array}

where {\varphi:D \rightarrow \mathbb{R}} is the first eigenfunction of the Laplacian on {D}. This is a Langevin diffusion and one can immediately see that the invariant distribution of this diffusion is given by

\displaystyle  \begin{array}{rcl}  \pi_{\infty}(x) \, \propto \varphi(x). \end{array}

For example, the following plot depicts the first eigenfunction of the Laplacian on the domain {D=[0,1]^2 \subset \mathbb{R}^2}.

First Eigenfunction

Advertisements

Curvature for Markov Chains

Recently, Yann Ollivier developed a nice theory of Ricci curvature for Markov chains. In many ways, this can be seen as a geometric language giving another view on the notion of path coupling, developed at the end of the {90}‘s by Martin Dyer and co-workers. It has to be noted that this new notion of curvature is very general and does not need the state space where the Markov chain evolves to have any differential structure, as can be expected at first sight. Any state space endowed with a metric suffices.

Let {P} be a Markov kernel on a metric state space {(S,d)}. We would like to quantify how long it takes for two different particles evolving according to the Markovian dynamic given by {P} to meet. If the first particle starts at {x \in S} and the second at {y \in S}, the initial distance between them is {d(x,y)}. At time {t>0}, what is the average distance between these two particles. For example, if {W^x} and {W^y} are two Brownian motions in {{\mathbb R}^n} started from {x} and {y} respectively, there is no reason why {W^x_t} and {W^y_t} should be closer from each other than {x=W^x_0} and {y=W^y_0}. Indeed, one can even show that whatever the coupling of these two Brownian motions we have {\mathop{\mathbb E}[d(W^x_t, W^y_t)] \geq d(x,y)}: this is roughly speaking because the Euclidean space {{\mathbb R}^n} has no curvature. The situation is quite different if we were instead considering Brownian motions on a sphere: in this case, trajectories tend to coalesce.

1. Wasserstein distance

In the sequel, we will need to use a notion of distance between probability distributions on the metric space {(S,d)}. The usual total variation distance {d(\mu,\nu)} defined by

\displaystyle  d(\mu,\nu) \;=\; \sup_{A \subset S} \; |\mu(A)-\nu(A)| \ \ \ \ \ (1)

is not adapted to our purpose since the metric structure of the space is not exploited. Instead, in order to take into account the distance {d(\cdot,\cdot)} of the space {E} and develop a notion of curvature, we use the Wasserstein distance {W(\mu,\nu)} between probability measures. It is defined as

\displaystyle  W(\mu,\nu) \;=\; \sup\Big\{ \mu(f) - \nu(f) \;:\; \text{Lip}(f) \leq 1\Big\}. \ \ \ \ \ (2)

The distance {d(\cdot,\cdot)} is crucial to this definition: a change of distance implies a change of the class of {1}-Lipschitz functions. Since {\mu(f) - \nu(f) = \mathop{\mathbb E}[f(X) - f(Y)]} for any coupling {(X,Y)} of {\mu} and {\nu}, and since the function {f} is {1}-Lipschitz, it follows that {\mathop{\mathbb E}[f(X) - f(Y)] \leq \mathop{\mathbb E}[d(X,Y)]}. Consequently, for any coupling {(X,Y)} we have {W(\mu,\nu) \leq \mathop{\mathbb E}[d(X,Y)]}. Taking the infimum over all the couplings {(X,Y)} leads to the inequality

\displaystyle  W(\mu,\nu) \;=\; \sup_{\text{Lip}(f) \leq 1} \; |\mu(f) - \nu(f)| \;\leq\; \inf_{(X,Y)} \; \mathop{\mathbb E}[d(X,Y)]. \ \ \ \ \ (3)

This is a deep result that on any reasonable space {(S,d)} the inequality is in fact an equality. Indeed, Kantorovich duality states that on any Radon space {(S,d)} we have

\displaystyle   W(\mu,\nu) \;=\; \sup_{\text{Lip}(f) \leq 1} \; |\mu(f) - \nu(f)| \;=\; \inf_{(X,Y)} \; \mathop{\mathbb E}[d(X,Y)]. \ \ \ \ \ (4)

It is interesting to note that under mild conditions on the state space {(S,d)} one can always find a coupling that achieves the infimum of (4): this is an easy compactness argument.

2. Notion of Curvature

Denoting by {m_x = \delta_x P} the one step distribution of the Markov chain started from {x} in the sense that {m_x(A) = \mathop{\mathbb P}[X_1 \in A \;| X_0 = x ]}, we define the local (Ricci) curvature {\kappa(x,y) \in {\mathbb R}} between {x} and {y} as

\displaystyle   W(m_x, m_y) = d(x,y) \cdot (1-\kappa(x,y)). \ \ \ \ \ (5)

The closer to {1} is {\kappa(x,y)}, the more the trajectories started at {x} tend to meet the trajectories started at {y}.

Trajectories tend to coalesce

The interesting case is when the infimum {\inf_{x,y} \, \kappa(x,y)} is strictly positive,

\displaystyle   \inf_{x,y \in E} \kappa(x,y) \;=\; \kappa > 0. \ \ \ \ \ (6)

In this case we say that the Markov kernel {P} is positively curved on {(S,d)}. It should be noted that in many natural spaces it suffices to ensure that {\kappa(x,y) \;\geq\; \kappa} for all neighbouring states {x} and {y} to ensure that {\kappa(x,y) \;\geq\; \kappa} for any pair {x,y \in S}. This can be proved thanks to the so called Gluing Lemma. A space without curvature correspond to the case {\kappa=0}: for example, a symmetric random walk on {\mathbb{Z}^d} and a Brownian motion on {{\mathbb R}^d} have both zero curvature. The curvature {\kappa} is a property of both the metric space {(S,d)} and the Markov kernel {P}: indeed, different Markov chain on the same metric space {(S,d)} have generally different associated curvature. Given a metric space {(S,d)} carrying a probability distribution {\pi}, this is an interesting problem to construct a {\pi}-invariant Markov chain with the highest possible curvature {\kappa}.

Indeed, the notion of curvature readily generalizes to continuous time Markov processes by taking a limiting case of (5). For example, one can define the curvature of the continuous time Markov process {\{X_t\}_{t \geq 0}} as the largest real number {\kappa} such that for any {x,y \in (S,d)} and {\kappa' < \kappa} we have

\displaystyle  W(m_x^{\delta}, m_y^{\delta}) \;\leq\; (1-\delta \kappa') \; d(x,y) \ \ \ \ \ (7)

for every {\delta} small enough. The quantity {m_x^{\delta}} is the distribution of {X_{\delta}} when started from {x} in the sense that {m_x^{\delta}(A) = \mathop{\mathbb P}[X_{\delta} \in A \; |X_0=x]}.

3. Contraction property

We now show that a positive curvature implies a contraction property. Equation (5) shows that {W(\delta_x P, \delta_y P) \leq W(\delta_x,\delta_y) \cdot (1-\kappa)} for any {x,y \in S}. A simple argument shows that one can indeed generalize the situation to any two distributions {\mu,\nu} in the sense that

\displaystyle   W(\mu P, \nu P) \leq W(\mu,\nu) \cdot (1-\kappa). \ \ \ \ \ (8)

Proof: For any pair {x,y \in S} consider a coupling {(U_{x,y}, V_{x,y})} of {m_x} and {m_y} such that {W(m_x,m_y)=\mathop{\mathbb E}[d(U_{x,y}, V_{x,y})]}. Now, choose an optimal coupling {(X,Y)} of {\mu} and {\nu}. This is straightforward to check that {(U_{X,Y}, V_{X,Y})} is a coupling (in general not optimal) of {\mu P} and {\nu P} so that

\displaystyle  \begin{array}{rcl}  W(\mu P, \nu P) &\leq& \mathop{\mathbb E}[d(U_{X,Y}, V_{X,Y})] = \mathop{\mathbb E}[\; \mathop{\mathbb E}[d(U_{x,y}, V_{x,y}) \;|X=x, Y=y] ] \\ &=& \mathop{\mathbb E}[ W(m_X, m_Y) ] = \mathop{\mathbb E}[ d(X,Y) \cdot (1-\kappa(X,Y)) ]\\ &\leq& (1-\kappa) \; \mathop{\mathbb E}[ d(X,Y) ] = (1-\kappa) \; W(\mu,\nu). \end{array}

\Box

Equation (8) is extremely powerful since it immediately shows that

\displaystyle  W(\mu P^t, \pi) \leq (1-\kappa)^{t} \; W(\mu,\pi). \ \ \ \ \ (9)

In other words, there is exponential convergence (in the Wasserstein metric) to the invariance distribution {\pi} at rate {(1-\kappa)^t}. In continuous time, this reads

\displaystyle  W(\mu P^t, \pi) \;\leq\; e^{-\kappa t} \; W(\mu,\pi). \ \ \ \ \ (10)

In other words, the higher the curvature, the faster the convergence to equilibrium.

4. Examples

Let us give examples of positively curved Markov chains.

  1. Langevin diffusion with convex potential: consider a convex potential {\Psi:{\mathbb R} \rightarrow {\mathbb R}} that is uniformly elliptic in the sense {\Psi^{''}(x) \geq \lambda > 0}. The Langevin diffusion {dz = -\frac{1}{2} \Psi'(z) \, dt + dW} has invariant distribution {\pi} with density proportional to {e^{-\Psi(x)}}. Given a time step {\delta}, the Euler discretization of this diffusion reads
    \displaystyle  x^{k+1} = x^k - \frac{1}{2} \Psi'(x^k) \, \delta + \sqrt{\delta} \; \xi \ \ \ \ \ (11) 

    where {\xi \sim {\mathcal N}(0,1)}. Given two starting points {x^0=x} and {y^0=y}, using the same noise {\xi} to define {x^1} and {y^1} it immediately follows that

    \displaystyle  \begin{array}{rcl}  W(x^1, y^1) &\leq& (x-y) \; \Big(1 - \frac{\delta}{2} \frac{\Psi'(x)-\Psi'(y)}{x-y} \Big)\\ &\leq& (x-y) \; (1-\frac{\lambda}{2} \delta). \end{array}

    In other words, the Langevin diffusion {\{z_t\}_{t \geq 0}} is positively curved with curvature (at least) equal to {\kappa = \frac{\lambda}{2}}.

     

  2. Brownian motion on a sphere: consider a Brownian motion on the unit sphere of {{\mathbb R}^n}. Consider two points {X,Y} on this unit sphere: by symmetry, one can always rotate the coordinates so that that {X=(\sqrt{1-h^2},0,h)} and {X=(\sqrt{1-h^2},0,-h)} for some {h \in [0,1]}. For {h \ll 1} the (geodesic) distance {d(X,Y)} is approximated by {d(X,Y) \approx 2h}. One can couple two Brownian motions {W^X} and {W^Y}, one started at {X} and the other one started at {Y}, by the usual symmetry with respect to the plane {\mathcal{P} = \{(x,y,z): z=0\}}: in other words, {W^Y} is the reflexion of {W^X} with respect to {\mathcal{P}}. One can check (good exercise!) that the diffusion followed by the {z}-coordinate of a Brownian motion on the unit sphere of {{\mathbb R}^n} is simply given by
    \displaystyle  dz = -\frac{1}{2}(n-1)z \, dt + \sqrt{1-z^2} \, dW. \ \ \ \ \ (12) 

    With this coupling, for small time {\delta \ll 1}, it follows that

    \displaystyle  \begin{array}{rcl}  z^X_{\delta} &\approx& h - \frac{1}{2} (n-1) h \, \delta + \sqrt{1-h^2} \sqrt{\delta} \; \xi\\ z^Y_{\delta} &\approx& -h + \frac{1}{2} (n-1) h \, \delta - \sqrt{1-h^2} \sqrt{\delta} \; \xi \end{array}

    where {\xi \sim {\mathcal N}(0,1)} is used as the same source of randomness for {z^X_{\delta}} and {z^Y_{\delta}} since {W^Y} is the reflexion of {W^X}. Since {d(W^X_{\delta}, W^X_{\delta}) \approx |z^X_{\delta} - z^Y_{\delta}|} it readily follows that

    \displaystyle  \begin{array}{rcl}  d(W^X_{\delta}, W^X_{\delta}) \; \leq \; \big(1- \frac{1}{2}(n-1)\delta \big)\; d(x,y). \end{array}

    In other words, the curvature of a Brownian motion on the unit sphere of {{\mathbb R}^n} is equal to {\frac{1}{2}(n-1)}. Maybe surprisingly, the higher the dimension, the faster the convergence to equilibrium. This is not so unreal if one notices that the Brownian increment satisfies {\mathop{\mathbb E} \|W_{t+\delta}-W_t\|^2 \approx n \delta}.

  3. Other examples: see the original text for many other examples.

Doob H-transforms

I read today about Doob h-transforms in the Rogers-Williams … It is done quite quickly in the book so that I decided to practice on some simple examples to see how this works.

So we have a Markov process {X_t} living in the state space {S}, and we want to see how this process looks like if we condition on the event {X_T \in A} where {A} is a subset of the state space. To fix the notations we define {p(t,t+s,x,y) = P(X_{t+s}=y|X_t=x)} and {h(t,x)=P(X_T \in A \, | X_t=x)}. The conditioned semi-group {\hat{p}(t,t+s,x,y)=P(X_{t+s}=y|X_t=x, X_T \in A)} is quite easily computed from {p} and {h}. Indeed, this also equals

\displaystyle \hat{p}(t,t+s,x,y) = \frac{P(X_{t+s}=y; X_T \in A\;|X_t=x)}{P(X_T \in A \,|X_t=x)} = p(t,t+s,x,y) \frac{h(t+s,y)}{h(t,x)}.

Notice also that {\hat{p}(t,t+s,x,y) = p(t,t+s,x,y) \frac{h(t+s,y)}{h(t,x)}} is indeed a Markov kernel in the sense that {\int_{y} \hat{p}(t,t+s,x,y) \, dy = 1}: the only property needed for that is

\displaystyle  h(t,x) = \int_{y} p(t,t+s,x,y)h(t+s,y)\,dy = E\left[ h(t+s,X_{t+s}) \, |X_t=x\right].

In fact, we could take any function {h} that satisfies this equality and define a new Markovian kernel {\hat{p}} and study the associated Markov process. That’s what people usually do by the way.

Remark 1 we almost never know explicitly the quantity {h(t,x)}, except in some extremely simple cases !

Before trying these ideas on some simple examples, let us see what this says on the generator of the process:

  1. continuous time Markov chains, finite state space:let us suppose that the intensity matrix is {Q} and that we want to know the dynamic on {[0,T]} of this Markov chain conditioned on the event {X_T=z}. Indeed {p(t,t+s,i,j) = [\exp(sQ)]_{i,j}} so that {\hat{p}(t,t+s,i,j) = [\exp(sQ)]_{i,j} \frac{p(t+s,T,j,z)}{p(t,T,i,z)}} so that in the limit we see that at time {t}, the intensity of the jump from {i} to {j} of the conditioned Markov chain is
    \displaystyle  Q(i,j) \frac{p(t+s,T,j,z)}{p(t,T,i,z)}.

    Notice how this behaves while {t \rightarrow T}: if at {t=T-\epsilon} the Markov chain is in state {i \neq z} then the intensity of jump from {i} to {z} is equivalent to {\approx \frac{1}{\epsilon}}.

  2. diffusion processes:this time consider a {1}-dimensional diffusion {dX_t = \mu(X_t) \, dt + \sigma(X_t) \, dW_t} on {[0,T]} conditioned on the event {X_T \in A} and define as before {h(t,x)=P(X_T \in A \,|X_t=x)}. The generator of the (non-homogeneous) conditioned diffusion is defined at time {t} by
    \displaystyle  \begin{array}{rcl}  \mathcal{G}^{(t)} f(x) &=& \lim_{s \rightarrow 0} \frac{1}{s} \Big( E\left[f(X_{t+s}) \,| X_t=x, X_T \in A\right]-f(x) \Big)\\ &=& \lim_{s \rightarrow 0} \frac{E\left[ f(X_{t+s}) h(t+s, X_{t+s}) \,| X_t=x\right]-f(x) }{s\,h(t,x)} \end{array}

    so that if {\mathcal{L} = \mu \partial_x + \frac{1}{2} \sigma^2 \partial^2_{xx}} is the generator of the original diffusion we get

    \displaystyle  \mathcal{G}^{(t)} f = \frac{1}{h} \Big(\partial_t + \mathcal{L})(hf).

    Because {(\partial_t + \mathcal{L})h=0}, this also reads

    \displaystyle  \mathcal{G}^{(t)} f = \mathcal{L}f + \sigma^2 \frac{\partial_x h}{h} \partial_x f.

    This means that the conditioned diffusion {Z} follows the SDE:

    \displaystyle  dZ_t = \Big( \mu(Z_T) + \sigma(Z_t)^2 \frac{\partial_x h(t,Z_t)}{h(t,Z_t)} \Big) \, dt+ \sigma(Z_t) dW_t.

    The volatility function remains the same while an additional drift shows up.

We will try these ideas on some examples where the probability densities are extremely simple. Notice that in the case of diffusions, if we take {A=\{x^+\}}, the function {(t,x) \mapsto P(X_T = x^+ \, |X_t=x)} is identically equal to {0} (except degenerate cases): to condition on the event {X_T=x} we need instead to take {h(t,x)} to be the transition probability {p(t,T,x,x^+)}. This follows from the approximation {P(X_T \in (x^+,x^+ + dx) \,|X_t=x ) \approx p(t,T,x,x^+) \, dx + o(dx)}. Let’s do it:

  • Brownian Bridge on {[0,T]}:in this case {p(t,T,x,0) \propto e^{-\frac{|x-y|^2}{2(T-t)}}} so that the additional drift reads {\frac{-x}{T-t}}: a Brownian bridge follows the SDE
    \displaystyle  dX_t = -\frac{X_t}{T-t} \, dt + dW_T.

    This might not be the best way to simulate a Brownian bridge though!

  • Poisson Bridge on {[0,T]}:we condition a Poisson process of rate {\lambda} on the event {X_T=N}. The intensity matrix is simply {Q(k,k+1)=\lambda=-Q(k,k)} and {0} everywhere else while the transition probabilities are given by {p(t,T,k,N) = e^{-\lambda (T-t)} \frac{(\lambda (T-t) )^{N-k}}{(N-k)!}}. This is why at time {t}, the intensity from {k} to {k+1} is given by
    \displaystyle  \lambda(t,k,k+1) = \frac{N-k}{T-t}.

    Again, that might not be the most efficient way to simulate a Poisson Bridge ! Notice how the intensity {\lambda} has disappeared …

  • Ornstein-Uhlenbeck Bridge:Let’s consider the usual OU process given by the dynamic {dX_t = -X_t + \sqrt{2}dW_t}: the invariant probability is the usual centred Gaussian distribution. Say that we want to know how does such an OU process behave if we condition on the event {X_T = z}. Because {p(t,T,x,z) \propto \exp(-\frac{|z-e^{-(T-t)}x|^2}{2(1-e^{-2(T-t)})^2})} we find that the conditioned O-U process follows the SDE
    \displaystyle dX_t = \Big(\frac{z-e^{-(T-t)x}}{e^{T-t}(1-e^{-2(T-t)})^2} - X_t \Big)\, dt+ \sqrt{2} \, dW_t.

    If we Taylor expand the additonal drift, it can be seen that this term behaves exactly as in the case of the Brownian bridge. Below is a plot of an O-U process conditioned on the event {X_{10} = 10}, starting from {X_0=0}.

    conditioned O-U process

    conditioned O-U process