Factors to consider in the construction of stock market index:
1. Sample size
The sample should be a statistically significant fraction of the population studied because larger samples tend to produce more accurate indications about the underlying population. If a sample is too large, it can be costly to compile.
The sample should contain heterogeneous elements representing all segments of the population.
The various elements in the sample should be assigned weights that correspond to investment opportunities in the population under study:
(i) A security‟s weight in some in some index might be proportional to the total market value of all the firm‟s shares that are outstanding stated as a fraction on the total market value of all the securities being traded in its market. Such value-weighing of an index is done to reflect investment opportunities in existence at any movement.
(ii) Equal weights could be used to represent the probability of selecting any given security with random sampling (or equivalently, selecting stocks throwing un anaimed dart.). An equally weighed index represents a “no skill” or naïve buy and hold investment strategy
(4) Convenient units
An index should be sated in units that are easy to understand and which facilitate answering questions.
(5) Computation of the mean
Most security market indicators are calculated as some sort of arithmetic average. The geometric average is an alternative computational procedure, it results in a smaller but similar value that is less volatile than an arithmetic mean calculated from the same sample of securities.
(6) Timing interval on which calculations of index are based.
(7) The prices to use for computation we close, mean, low, high.