It is important to know what is the probability of a trend to be in the market before it changes direction. This question can be answered in 2 parts. The first is what is the probability of having a continuos trend without the market changing direction. The second would be to consider anchor points from the most significant turning points and determine the number of days that the trend is in effect.
To answer the first question, we have computed the number of days form the closing in which the market does not change direction. This for the main indexes DJIA, SP500, NASDAQ and RUT. The results are presented in Table 2.1 that contains the frequency of occurrence as a function of the number of days. It is important to say that a trend of 4 days contains a trend of 3 days and a trend of 2 days and a trend of 1 day. It would be interesting to find the true frequency trends of the trends eliminating this effect and it is found on the columns on the right. As can be observed the values are very similar and because the former is much easier to calculate we will use those in the future.
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
||
|
|
|
|
|
|
|
|
|
|
The data is plotted in Graph 2.2 in a logarithmic scale and it can be observed that data fits the equation
It is important to notice that about half the time it is most probable to find a reversal, then
To answer the second question we need to establish certain criteria to decide which are anchor points. Therefore we define an anchor point like a point that changes a market direction. either an oversold or overbought market.