See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/265189766

From the sde package to the Yuima Project

Article

CITATION READS

1 168

1 author:

Stefano M. Iacus
University of Milan
130 PUBLICATIONS   11,934 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Causal inference View project

simulation and inference for stochastic processes View project

All content following this page was uploaded by Stefano M. Iacus on 25 January 2015.

The user has requested enhancement of the downloaded file.

 From the sde package to the Yuima Project
Stefano Maria Iacus1,∗

1. Department of Economics, Business and Statistics * Contact author: stefano.iacus@unimi.it

Keywords: financial econometrics, stochastic differential equations, simulation, mathematical finance

We will briefly discuss the R package sde which contains generic functions for simulation and inference
for one dimensional stochastic differential equations driven by Wiener process and we introduce the new S4
framework for simulation and inference for general stochastic differential equations under development by
the Yuima Project1 Team, which is a joint venture between the Universities of Tokyo, Osaka, Kyushu and
Milan and the Institute of Statistical Mathematics in Tokyo.
Most of the theoretical results in modern finance rely on the assumption that the underlying dynamics of
asset prices, currencies exchange rates, interest rates, etc are continuous time stochastic processes driven by
stochastic differential equations. Continuous time models are also at the basis of option pricing and option
pricing often requires Monte Carlo methods. In turn, the Monte Carlo method requires a preliminary good
model to simulate (Kloden and Planten, 1999) whose parameters has to be estimated from the data. On the
other side, most applications in financial econometrics make use of pure time series modeling because many
statistical procedures are already available in many statistical packages.
The discrepancy between theoretical and applied mathematical finance is motivated by the fact that
while the model is continuous, the observations always come in discrete time. Inference for continuous time
data from stochastic differential equation dates back to Jacod and Shiryayev (1987) and today is considered
as a solved problem. On the contrary, the likelihood function for discretized stochastic differential equations
is available only for a very limited class of models and exact likelihood inference is usually not possible.
Recently, many authors have considered ways to establish approximate and/or quasi-likelihood inference
for stochastic differential equations (for a recent review see Iacus, 2008) and the sde package implements
those methods in the one dimensional framework but, although generic, still in a “classical way”.
The Yuima Project is an S4 framework for simulation and inference of stochastic differential equations in
the multidimensional case driven by either Wiener, Lévy, etc. noise. The building blocks are the definition
of the S4 class named sde.model in a very abstract way for which a series of S4 methods for simulation and
inference are defined. The definition of the sde.model does not include the description of the data which may
come in many forms (e.g., mts, tseries, zoo, xts, etc). The combination of the “Model” (Jump diffusion;
stoch. volatility; fractional BM; etc), the “Data” and the sampling scheme (high frequency, non small ∆,
tick data, etc.) and the “inference/financial problem” (parametric/nonparametric estimation; change point
analysis; covariance estimation; asymptotic expansion; estimation of functionals; etc) produces the optimal
(up to present knowledge) inference solution.
In this talk we present the basic infrastructure of the yuima package (not yet released) and discuss a
couple of toy examples to show generality and applicability of this approach.

References
Iacus. S.M. (2008). Simulation and Inference for Stochastic Differential Equations with R Example,
Springer, New York.

Jacod, J., Shiryayev, A.N. (1987) Limit Theorems for Stochastic Processes, Springer-Verlag, New York.
Kloden, P., Platen, E. (1999). Numerical Solution of Stochastic Differential Equations, Applied Mathemat-
ics, 23, Third corrected printing, Springer, New York.

1 This
project is partly supported by JST Basic Research Programs PRESTO, Grants-in-Aid for Scientific Research No.
19340021, the Global COE program “The research and training center for new development in mathematics” of Graduate
School of Mathematical Sciences, University of Tokyo, and by Cooperative Research Program of the Institute of Statistical
Mathematics.

View publication stats