We are organizing a session that welcomes papers dealing with:
computationally efficient methods for the estimation of stochastic volatility models;
non parametric estimation of the conditional/unconditional densities of the stochastic volatility;
numerical procedures for the solution of the PDE associated with a particular SV model.
These items are closely related to the rapid development of GARCH but devote more attention to stochastic volatility models that are natural discretizations of continuous time diffusions. For this class of models, there is no explicit statement of the likelihood function, so numerically intensive procedures are required. Despite this, these models allow recovery of the continuous time parameters of the associated PDE from discrete time observations.
The continuous-time equivalent of a given SV model allows derivation of the marginal density of an asset price, given the drift and the diffusion. This allows one to formulate tests from which the correct diffusion for the short term interest rate can be chosen by matching it to the observed marginal density, the latter evaluated in a non parametric fashion.
The stochastic volatility model has increased appeal and ease of use because of recent development of simulations tools, such as importance sampling simulations applied to diffusions and Markov Chain Monte Carlo. These have all been used to filter the unobservable volatility of given price changes or to evaluate conditional expectations of particular expressions that define the intrinsic value of a contingent claim.
Deadline:
Abstracts (at least 300 words) or complete papers should be sent in duplicate by January 15, 1999. In any case, full papers are to be made available by May 1, 1999.
Submission details:
E-mail submission should be made to: alti.3441@interbusiness.it
Fax submission should be made to: +39 06 474-7820
Mail submission to:
Filippo Altissimo or Fabio Fornari
Bank of Italy - Research Department
Via Nazionale 91, 00814 Roma, Italy