The concept of the nature of market memory in the study of market dynamics
Financial University under the government of the Russian Federation, Department of AD, PR and FT, Russia, 125993, Moscow, Leningrad Ave., 49
Filtering random perturbations in market observations allows to identify the trend, provided the presence of internal determination. The existence of an objective regularity is usually assumed, but it is not always possible to prove this existence. Moreover, if the regularity thus revealed is confirmed empirically, then this is not a proof of the existence of the desired regularity. It is possible to disprove the existence of determination by increasing the sample size.
We introduce the concept of a memory field, in the sense of the relationship between current and previous observations. The weak field of memory is comparable to the lack of regularity. We define a strong field as a situation of random" inclusion "of a regularity and its random "shutdown". The presence of determination at a particular time interval may be obvious, but long-term observation of the market will lead to the appearance of inconsistency of the detected determination, which can be interpreted as a random shutdown of the pattern.
The econometric approach to market analysis treats the deterministic and random component as mutually exclusive components, thereby limiting the set of methods that can be used for analysis.
Following the concept of the field of memory, formulated with respect to market processes, it is possible to assume both the existence in different historical periods of different qualities (randomness and determinism) of the market, and their simultaneous presence in the market nature. The most important tool useful in identifying the current field is the financial series smoothing procedure. Traditional smoothing schemes have a common drawback-the loss of information at the ends of the series, which can be eliminated by an alternative method of one-sided smoothing, which includes the previous levels of the series during the sliding averaging.
An illustration of the approach is given by analyzing the dynamics of major currency pairs in the FOREX market. The paper proposes a systematic approach to the problem of analyzing the dynamics of market trends, and considers the principle of complementation of different methods of data processing in the problem of system analysis.