|  DEC 1986

SIDEBAR: Markov analysis by Technical Analysis, Inc.

SIDEBAR: Markov analysis by Technical Analysis, Inc. A natural process which has a random element but also shows the effects of previous events influencing subsequent events is a Markov process. Markov models lie in the spectrum between fully deterministic models (such as y = 3 * x) and independent events models (where the probability of a system being in a particular state is not affected by previous states). They lie somewhere between the beliefs of random walk theorists and those who explain the markets using deterministic equations. Although they do not allow one to pin down exactly what is going to happen, they are particularly useful for when one wants to look for significant patterns in data sets. In trading, they can help assign risk factors to trades and identify recurring relationships. Markov analysis is really a family of techniques based on transition matrices. The terminology involved depends on data collection and processing techniques. The list below summarizes some of them.

by Technical Analysis, Inc.