Pitman type theorem for one-dimensional diffusion processes (Q756290)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Pitman type theorem for one-dimensional diffusion processes
scientific article

    Statements

    Pitman type theorem for one-dimensional diffusion processes (English)
    0 references
    0 references
    0 references
    1990
    0 references
    For a regular diffusion X in \({\mathbb{R}}\) starting from 0 with generator \({\mathcal L}=d/dm\cdot d/ds\), \(s(0)=0\), and for any fixed \(r>0\), define \[ \tau_ 1=\inf \{t>0:\;X(t)=r\},\;\tau_ 2=\sup \{t>0:\;X(t)=0,\;t<\tau_ 1\}. \] Then Williams' theorem states that \(\{X(\tau_ 2+t):\) \(0\leq t\leq \tau_ 1-\tau_ 2\}\) is identical in distribution to \(\{\tilde X(t):\;0\leq t\leq \tau \},\) where \(\tilde X\) is a diffusion process starting from 0 with generator \(\tilde {\mathcal L}f=(1/s){\mathcal L}(sf)\) and \(\tau =\inf \{t>0:\;\tilde X(t)=r\}.\) In the case where X is a one-dimensional Brownian motion B with \(B(0)=0\), \(\tilde X\) is a Bessel process with index 3 (the radial part of a three-dimensional Brownian motion) and Pitman proved that \[ (*)\;\{\tilde X(t),\;t\geq 0\}=^{d}\{B(t)=2L(t),\;t\geq 0\},\;L(t)=-\min_{0\leq u\leq t}B(u), \] where \(=^{d}\) means the equality in distribution. In this paper we consider the case where X is the one-dimensional diffusion process defined by the stochastic differential equation \[ X(t)=\int^{t}_{0}\sigma (X(u))dB(u)+\int^{t}_{0}b(X(u))du, \] and will prove that \(\tilde X\) admits a representation similar to (*).
    0 references
    0 references
    regular diffusion
    0 references
    Bessel process
    0 references
    stochastic differential equation
    0 references
    0 references