In many cases statistical ratios are used to group data by the preliminary sorting by groups and pro quantification their importance. When the data to be examined are located along the interval scale, they form the basis of variable data normally screened by their frequency distribution. By definition, the frequency distribution is a statistical tool to help the maximum amount of information is presented with minimal details. Variable data can be presented as a whole and with fractions. For ease of analysis is often applied rounding of fractions to whole.
A basic requirement for the application of frequency
distribution as a statistical method for financial analysis, analyzed data to
be of homogeneous type. That is, the data must be comparable and compatible
with each other. In this respect, the analysis of the frequency distribution of
group shares on the stock exchange requires to choose stocks from a sector
rather than several sectors. Typically in this case is to be noted that:
 |
| Statistical analysis method |
First. The grouping of data into classes is the first phase
of implementation of the frequency distribution as a statistical method forfinancial analysis. The main idea of the frequency distribution is to
organize a large database by presenting it in an understandable and accessible
form. So if you develop example published daily stock quotes to a group of
selected stocks can trace the logic of the application of the method of
frequency distribution. In this respect (Video 1) represents the pooled data
for quotations of the shares of 170 companies at the close of the stock
exchange in New York. Graphical data are expressed using a histogram by (Video2).
Presented by
(Video 1) frequency distribution shows how the initial data for quotations of
170 selected stocks closing of the exchange are grouped into classes. Each
class has the same range (in this case the scope equals $ 10). There is a
correlation between the number of selected classes and robustness of the
analysis. The fewer classes are formed, the possibility of leakage of the
important features of the test set increases and vice versa. The introduction
of a plurality of classes, of course, also should be moderate. In other words,
it is advisable classes frequency distribution to be tailored to the specifics
of the population surveyed by financial data. In view of these considerations
database Stocks of selected stocks form 10 class data with an interval of $ 10.
Classes must be defined precisely in order to avoid double counting. This means
that for example the second class is defined by $ 10 and under $ 20 instead of
$ 10 respectively to $ 20. The second case would allow, exchange quotation of $
20 to be included in two classes, which is methodologically inadmissible.
 |
| Statistical analysis method |
Second. Frequency
distribution can be represented in relative expression. This is because the
share of each class within the population presents additional information about
its structuring. The last column (Video 1) demonstrates the kind of relatively
similar frequency distribution. The result can easily compare the distribution
of individual classes within the overall 100%.
Third. The
cumulative frequency distribution is an alternative way to implement this
modification of the statistical method of financial analysis. An example of a
cumulative frequency distribution is presented in (Video 2). The main advantage
of the cumulative frequency distribution is that it provides information on the
distribution of the meanings of the population surveyed to and above a certain
value.
Fourth. The profile of the frequency distribution allows to display graphically depending affecting the population surveyed by financial data. For example histogram of (Video 3) to support a finding that the profile of the frequency distribution of stock quotes of the 170 studied shares is distorted.
This is due to the fact that in the second grade groups formed the largest number of researched stocks, and therefore the profile of the frequency distribution layer reaches maximum.
 |
| Statistical analysis method |
This is due to the fact that in the second grade groups formed the largest number of researched stocks, and therefore the profile of the frequency distribution layer reaches maximum.
(Video 3)
Normal profile is
when both sides are symmetrical. Symmetry may be concentrated in a small space
(strongly convex profile) or be located within all derived classes (slightly
convex profile of the frequency distribution).
