A time series usually consists of two regular components and one irregular component.
The regular component is the trend, and in the present case, the season. The irregular
component represents the random and serves as noise or shock or innovation. Trend is
the backbone of development, which is described by a time series, because it indicates
the direction—upward, downward or sideward (Enders, 2004). The season, however,
represents regularity depending on the course of time, whose data accomplishes the
trend data to a vibrating process (Franses, 1998). Additionally, the irregular component
shapes the process with the result that the time series oscillates more or less. The study
explores in more detail the role of trend in time series with respect to market shares.
Methodology
In the context of analyzing a competitive situation, the development of market shares is
an important aspect because it brings to light the market dynamics that are driven by
the ruling power structure. Both the study of the development of different market shares
and the statistical analysis of the quantitative relations between the changing market
shares provide information on competitors’ interaction. In this regard, it is of great
importance to know whether the different developments show trends. If they show, it is
crucial to find out what type of trend is present. Typically, market shares do not often
show any deterministic trends, such as linear, quadratic and cubic, but rather very
often show stochastic trends which have a stochastic component that is additively linked
(multiplicatively linked terms, which are characteristic of non-linear trends, are omitted
from this discussion for reasons of space) with the intercept of the trend.
|