Chaos theory is an exciting new development. The modelling of non-linear deterministic mathematical system can help to explain events in several disciplines, e.g., biology. Partly because of this general excitement, partly because the very name, 'chaos', has a certain resonance, there have been many commentators, who have, somewhat prematurely, assumed that chaos theory should be applicable to financial markets. For example, chaos theory was described in one of last year's (1991) Amex Prize Essays as "one of the most exciting new scientific ideas to hit the world of theoretical finance".
Yet some aspects of this model seem distinctly implausible in the context of financial markets. Whereas most of us have no great difficulty with the concept of the existence of non-linear relationships, the idea that the time-path of financial markets is endogenously deterministic, rather than buffeted by the stochastic shocks of unforeseeable 'news', seems hard to accept. Nevertheless one must test hypotheses, rather than just accept or reject on the basis of prior plausibility.
These tests are mathematically quite complex, involving techniques and terminology that will prove opaque to many. Such tests also require an extremely large data set. Since I had created one such set, I encouraged the authors of the attached paper to use it as a test bed to search for the existence of 'chaos' in these financial markets.
As the title indicates they found some indications of non-linear relationships, but none of chaos.
Download is not available
