Published by EH.NET (February 2003)

Bertrand M. Roehner, Patterns of Speculation: A Study in Observational

Econophysics. Cambridge: Cambridge University Press, 2002. xvii + 230 pp.

$50 (hardcover), ISBN: 0-521-80263-6.

Reviewed for EH.NET by Graeme D. Snooks, Institute of Advanced Studies,

Australian National University.

Physicists are taking a renewed interest in economics. As is well known, many

of the economic pioneers were trained in either science or engineering. Pareto,

Walras, Jevons, and Marshall, for example, brought a training in science and

mathematics to the study of economics. And general equilibrium theory derives

from classical mechanics. Today a growing number of physicists are again

investigating economic phenomena and publishing their results in physics

journals. Since 1997 this activity has been known as “econophysics.”

As in orthodox economics, the practitioners of econophysics fall into either

the deductive or empirical camps. Bertrand Roehner, a theoretical physicist

based at the University of Paris, is a highly effective pioneer and advocate of

the empirical approach that he calls “observational econophysics.” He has been

publishing on economic issues in physics journals since the mid 1980s, and has

even published (jointly with Tony Syme) a fascinating book on comparative

history (Pattern and Repertoire in History, Harvard University Press,


Roehner’s objective in Patterns of Speculation is to develop

mathematical models based on the “regularities” that can be identified from a

systematic observation of the boom-and-bust cycles of speculative activities in

both financial and non-financial markets. Unhappy with the large number of

“adjustable,” or unmeasured, parameters in econometric models, his aim is to

“sharpen the picture” by reducing these to a minimum through more accurate

statistical tests arising from close observation of reality. This objective has

shaped the structure and tone of a book that is highly systematic and precise

— in a word, scientific.

Patterns of Speculation consists of nine chapters organized into four

parts. Part I briefly describes the nature and history of econophysics, which

will be unknown to most readers. Part II outlines the organization of

speculative markets of both a financial and non-financial kind, together with

Roehner’s view of the collective behavior of decision makers — characterized

as social rather than economic man — which is highly influenced by sociology.

Part III, the main section of the book, attempts to identify “statistical

regularities” concerning “speculative peaks” through comparative analysis. This

involves the presentation of a series of charts and basic statistical

measurements of speculative variables — such as the prices and volumes of

stocks and commodities — that characterize similar events through space and

time. And Part IV presents a “theoretical framework” based on these

observations. Here he adopts mathematical models that are capable of

replicating the observed regularities. While this is similar to econometric

simulation, Roehner persuasively claims that econophysics has the advantage of

being able to reduce the number of “adjustable” parameters in the theoretical

model through systematic observation. It should be noted that the theoretical

models adopted by Roehner are based on pre-existing mathematical forms rather

than being derived directly from empirical relationships.

Roehner could have profitably compared the methodology of “observational

econophysics” with that of a discipline pursuing similar objectives from within

economics. The historical school of economics — which flourished particularly

in Germany but also in Britain, North America, and Australia during the last

quarter of the nineteenth century — also attempted to develop economic theory

from the observation of historical patterns. The resulting confrontation

between the deductivists and the historicists became known as the

methodenstreit — or the battle of the methods — which was ultimately

won by the newly emerging neoclassical economists trained (ironically) in

science and mathematics, largely because the historicists focused myopically on

the patterns in history and not the underlying economic processes. In the early

twentieth century, the historicists joined the ranks of economic historians and

institutionalists, and historical economics collapsed (Snooks, Economics

without Time, 1993). Only at the century’s end did a more robust form of

historical economics emerge, which was able to develop a viable general dynamic

theory to explain these historical patterns — see the reviewer’s Laws of

History (1998), Longrun Dynamics (1998), and Global

Transition (1999). This was something the deductivists were unable to

achieve despite their institutional dominance during the intervening century.

In omitting the history of historical economics, Roehner appears to overstate

the contribution of econophysics when he concludes that “the search for

regularities is probably one of its most important innovations” (p.20), and

that observational econophysics is unique in constructing theory only once the

patterns in reality have been identified (p.xii).

Econophysics is an interesting new development that has been clearly explained

and expertly employed by Bertrand Roehner, who is to be congratulated on his

pioneering contribution. This new discipline is to be welcomed not only because

it provides new insights into economic phenomena, but also because it endorses

the attempt by economic historians to systematically examine the patterns of

the past. In particular it supports the longstanding attempt by historical

economists to employ the inductive approach to build a realist body of economic

theory. Over the past decade the reviewer has employed a new historicist

technique to develop a general dynamic theory — the dynamic-strategy theory —

to examine the issues of economic growth and economic development, and, more

recently, to explore the dynamics of all living systems (see my The Collapse

of Darwinism, Lexington Books, July 2003). For the first time in a century,

therefore, deductive economics is under concerted and effective attack by

empiricists from within and without the old empire. No doubt it will go the way

of all seemingly impregnable empires of the past.