Step Sum study of Nifty : 2019

Step Sum is in the form of a new technical tool to evaluate the psychology of the market. It has been developed by Mark Lundeen and it offers a unique insight into the psychology of the market. The step sum is simply a running total of UP vs. DOWN  days and it reflects trader psychology.

The step sum is a single item Advance – Decline Line. In other words; if Nifty is down from the previous day’s close, we give it a value of (-1), if they end the day up from the previous day’s close, we give it a value of (+1). So a step sum plot is merely the net sum of up and down market days over time.

Initially, step sum fluctuated in negative zone as Nifty dipped below 1000 (Arbitrarily set as OHLC available from 1^st November 1995) for considerable period & then remained in positive since PE August 2002; a graph is shared here in below:

There were seven occasions when Nifty showed Zero change over the previous day’s closing & listed below : 

A Chart of Nifty Close-StepSum (on ‘end of month’ basis) is shared below: 

Next is a Chart (on ‘end of Year’ basis) is shared below:

 “15 Count” is  running sum of the daily step used in constructing their step sum over consecutive 15 days. As for example,
 “Count +/-  13”  means 14 up (down) & 1 Down(up),
“Count +/-  11”  means 13 up (down) & 2 Down(up),
“Count +/-  9”  means 12 up (down) & 3 Down(up),
and so on
“Count +/-  1”  means 8 up (down) & 7 Down(up)
Now I am trying to measure the "Quality aspect" of Step sum with a simple logic & mathematics ; simply stated here is a plan to quantify the sum of all 'Count of 15' over a period of calendar year. First refer to the Table :

Table is based on OHLC data sourced from NSE site. Step sum were calculated on daily basis, thereafter, 'Count of 15' were generated by excel (by adding step up or step down days over 15 days). Thereafter, total number of occasions for diverse Cnt_15 are tabulated in the Table.

Note there had been 7 cases of 'no change' in closing Nifty , resulting in 0 change in Step Sum; consequently 'Count of 15' were rendered even numbers ; such cases were added to nearest odd number (e.g., 2 were counted with 1, 4 with 3, etc)

Z (Sigma) = Net weighted Count_15 for the Year =

 Summation over the year of  [ (Number of count) X (Value of Cnt) ]  for full range of  available Cnt  { from 15 to minus 15 } with the help of Template shared below :

A Chart showing correlation of the derived value for Z (Sigma) with year-end incremental value for Step Sum is shared below:

Ratio is found to be Median value of 1.80 & Average value of 2.50 even though broad range is wide from peak of 4.20 to low of 0.62.