xiv Preface
designing, developing, testing, and operating financial econometric applica
tions in asset management.
In this book we explain and illustrate the basic tools that are needed
to implement financial econometric models. While many books describe the
abstract mathematics of asset management, the unique feature of this book is
to address the question of how to construct asset management strategies using
financial econometric tools. We discuss all aspects of this process, including
model risk, limits to the applicability of models, and the economic intuition
behind models. We describe the critical issues using real life examples.
We start by discussing the process of applying financial econometrics
to asset management. The three basic steps of model selection, estimation,
and testing are discussed at length. We emphasize how in this phase eco
nomic intuition plays an important role. Before designing models we have to
decide what phenomena we want to exploit in managing assets.
We then discuss the most fundamental financial econometric technique:
regression analysis. Despite its apparent simplicity, regression analysis is a
powerful tool the application of which requires careful consideration. We
describe different types of regression analysis, including quantile regressions
and regressions with categorical variables, their applicability, and the condi
tions under which regression fails. We discuss the robustness of regression
analy sis, introducing the concept and technique of robust regression. All
concepts are illustrated with reallife examples.
Next, we analyze the dynamic behavior of time series, introducing vec tor
and scalar autoregressive models. We formalize meanreversion, intro ducing
the concept of cointegration, and describe the heteroscedastic behav ior of
financial time series. We discuss the economic intuition behind each model,
their estimation, and methods for parameter testing. We also analyze the
limits of the applicability of autoregressive techniques, the advantage of
exploiting mean reversion when feasible, and the model risk associated with
autoregressive models. We again use reallife examples to illustrate.
Subsequently, we move to consider large portfolios and discuss the tech
niques used to model large numbers of simultaneous time series, in particu
lar factor models and principal components analysis. The issues associated
with the estimation and testing of large models and techniques to separate
information from noise in large sets of mutually interacting time series are
discussed.
Finally, we discuss the specific process of implementing a financial
econometric model for asset management. We describe the various steps of
this process and the techniques involved in making modeling decisions.
One important characteristic of model development today is the avail
ability of good econometric software. Many building blocks of the pro
cess of implementing a financial econometric application are available