Large volatility-stabilized markets (Q1761492)

Motivated by modeling of the evolution of market capitalizations in equity markets, the author studies a system of coupled stochastic differential equations \[ d X_i(t) = {\eta \over 2} S(t) dt + \sqrt{X_i (t)S(t)} d W_i(t),\quad 1 \leq i \leq N, \] where \(\eta > 1\) is a constant, \(S(t) = \sum_{i=1}^N X_i(t)\), and \(W_1,\dots,W_N\) are independent standard Brownian motions, with a random initial value \((X_1(0),\dots,X_N(0))\) in \([0,\infty)^N\). The aim of the paper is to study the asymptotic behavior of this system when \(N\) is large. More specifically, the author considers the empirical distribution \[ \rho^N(t) = {1 \over N} \sum_{i=1}^N \delta_{X_i(t/N)}, \] where \(\delta_x\) stands for the Dirac delta at \(x \in \mathbb R\), with \(0 \leq t \leq T\) for some fixed time horizon \(T>0\). The process \(\rho^N(t)\), \(0 \leq t \leq T\), is continuous and assumes values in the space \(M_1(\mathbb R)\) of probability measures on \(\mathbb R\). The main result of the paper establishes that, under a mild condition on the asymptotic behavior of the initial value \((X_1(0),\dots,X_N(0))\), the law of the process \(\rho^N(t)\), \(0 \leq t \leq T\), converges weakly, as \(N \rightarrow \infty\), to a Dirac delta \(\delta_\rho\), where the measure-valued function \(\rho \in C([0,T],M_1([0,\infty)))\) can be described as a distributional solution to a degenerate parabolic partial differential equation. Furthermore, it is shown that \(\rho\) admits a stochastic representation in terms of a time-changed squared Bessel process and that the density of \(\rho(t)\) with respect to the Lebesgue measure, for any \(0 \leq t \leq T\), can be described as an integral involving a Bessel function.