发帖
雷达卡
书名:Statistical Arbitrage___ Algorithmic Trading Insights and Techniques
作者:ANDREW POLE
页数:257
出版商:wiley
278258.rar
(1.31 MB, 需要: 19 个论坛币)
本附件包括：
Mean reversion in prices, as in much of human activity, is a
powerful and fundamental force, driving systems and markets
to homeostatic relationships. Starting in the early 1980s, statistical
arbitrage was a formal and successful attempt to model this behavior
in the pursuit of profit. Understanding the arithmetic of statistical
arbitrage (sometimes abbreviated as stat. arb.) is a cornerstone to
understanding the development of what has come to be known as
complex financial engineering and risk modeling.
The trading strategy referred to as statistical arbitrage is generally
regarded as an opaque investment discipline. The view is that it is
being driven by two complementary forces, both deriving from the
core nature of the discipline: the vagueness of practitioners and the
lack of quantitative knowledge on the part of investors. Statistical
arbitrage exploits mathematical models to generate returns from
systematic movements in securities prices. Granted, no investment
manager is inclined to divulge the intricate ‘‘how-tos’’ of his business.
While stock pickers can tell a good story without revealing the heart
of their decision making, that is not the case with model-based
strategies developed by ‘‘quants.’’ A description with any meaningful
detail at all quickly points to a series of experiments from which an
alert listener can try to reverse-engineer the strategy. That is why
quant practitioners talk in generalities that are only understandable
by the mathematically trained.
Opacity has also increased the need for mathematical maturity
on the part of investors seeking to assess managers. To comprehend
what a statistical arbitrageur is saying beyond a glib level, one needs
to understand advanced mathematics beyond the college level. This,
naturally, limits the audience. The limitation is perpetuated by the
lack of reference material from which to learn. Statistical Arbitrage
now fills that void.
Statistical arbitrage has been in existence for approximately 25
years. During that time, the general concepts have been widely
disseminated via the storytelling of early implementers to interested
investment bank analysts and academics. Nevertheless, opacity
remains because practitioners have steadily increased the sophistica-
tion of their modeling—and for good commercial reasons remained
obscure about their innovations. In the wide dissemination of basic
stat. arb. concepts, the term mean reversion as well as its variant,
reversion to the mean, looms very large. Reversion to the mean is a
simple concept to illustrate: Children of unusually tall parents are typ-
ically shorter than their parents; children of unusually short parents
are typically taller than their parents. This is a concept that is easy for
most people to grasp. Translating this idea to the motions of security
prices means that securities prices return to an average value. So far,
so good. But then we hit a problem. Height reversion is an intergen-
erational phenomenon, while price reversion is an entity dynamic.
Prices returning from where? And to what average value? The
average height of adults is a familiar concept, even if the precise
quantification requires a little work. Even children as young as
grade-school age can give a reasonable estimate of the average height
of the adults they know, and by extension, of the average height
of local adult populations. There is no such common grounding of
observation or experience to apply to securities prices. They are all
over the map. Scaling is arbitrary. They can grow many fold. And
they can collapse to zero. People do not grow to the sky and then
revert back to some average, but security prices can.
Even if we suppose that the questions have been reasonably
answered, other technicalities immediately pose themselves: How
does one identify when a price is away from the mean and by how
much? How long will the return to the mean take?
Here is where the opacity enters the discussion and makes its
permanent home. The language of mathematical models compounds
the unfamiliarity of the notions, generating a sense of disquiet, a fear
of lack of understanding.
In Statistical Arbitrage, Pole has given his audience a didactic tour
of the basic principles of statistical arbitrage, eliminating opacity at
the Statistical Arbitrage 101 level. In the 1980s and early 1990s,
Stat. Arb. 101 was, for the most part, all there was (exceptions such
as D.E. Shaw and Renaissance aside). Today, more than a decade
later, there is a much more extensive and complex world of statistical
arbitrage.
This is not unlike the natural world, which is now populated
by incredibly complex biological organisms after four billion years
of evolution. Yet the simplest organisms thrive everywhere and still
make up by far the largest part of the planet’s biomass. So is it true in
statistical arbitrage,where the basics underpinmuch of contemporary
practice.
Statistical Arbitrage describes the phenomena, the driving forces
generating those phenomena, the patterns of dynamic development
of exploitable opportunities, and models for exploitation of the basic
reversion to the mean in securities prices. It also offers a good deal
more, from hints at more sophisticated models to valuable practi-
cal advice on model building and performance monitoring—advice
applicable far beyond statistical arbitrage.
Chapters 1 and 2 speak to the genesis of statistical arbitrage, the
venerable pairs trading schemes of the 1980s, with startling illustra-
tion of the enormous extent and productivity of the opportunities.
This demonstration sets the scene for theoretical development, pro-
viding the first step to critical understanding of practical exploitation
with rules for calibrating trade placement.More penetration of opac-
ity follows in Chapter 5 where the relationship between (a) reversion
in securities prices watched day-by-day and (b) statistical descriptions
(distributions) of collections of such daily prices viewed as a glob
devoid of the day-by-day context, is clearly spelled out.
Chapters 8 and 9 tell of the midlife crisis of statistical arbitrage.
The roiling of United States financial markets for many months,
beginning with the Enron debacle in 2000 and running through
the terrorist attacks of 2001 and what Pole calls ‘‘an appalling
litany’’ of corporate misconduct, is dissected for anticipated impact
on statistical arbitrage performance. Adding to that mix have been
technical changes in the markets, including decimalization and the
decline of independent specialists on the floor of the NYSE. Pole
draws a clear picture of why statistical arbitrage performance was
disrupted. Very clearly the impression is made that the disruption
was not terminal.
Chapters 10 and 11 speak to the arriving future of statistical
arbitrage. Trading algorithms, at first destroyers of classical stat. arb.
are now, Pole argues, progenitors of new, systematically exploitable
opportunities. He labels one of the new motions the ‘‘catastrophe
move’’; a detailed exposition of modeling the dynamics follows a
catastrophe-theory explication of a possible rationale for the behav-
ioral pattern. The unmistakable impression is that statistical arbitrage
is rising once again.
Thetoneof Statistical Arbitrage is direct and thorough. Obfus-
cation is in short supply. Occasionally, the tone is relieved with a bit
of lightheartedness—the tadpole-naming story in a note to Chapter
11 is a gem—and throughout, refreshing prose is to be found.
In describing mathematical models, authors readily produce
unmemorable, formulaic wording offering nothing by way of inter-
pretation or explanation beyond what is provided by the algebra
itself. Statistical Arbitrage is an exception—a break in the cloud of
opacity—a mean that Pole has avoided reverting to!
[此贴子已经被作者于2008-12-19 9:06:15编辑过]
提升卡
置顶卡
沉默卡
变色卡
抢沙发
千斤顶
显身卡
京公网安备 11010802022788号
