I have just finished reading your book for the second time, and I found your work one of the best book I have ever read, if not the best one.
I really like very much the concepts you show and I have now a very clear view about how to look at the market: I want to thank you for disclosing your thoughts and methods.
I would like to code all the tools in my Java platform to be used for my own trades in the Italian stock market, and for this reason I have some questions just to be sure to code it properly because there are some things not really clear to me.
1) page 32 "What is effective volume flow?" is explaind as the simple sum of all the single effective volume: but it is then called "Total effective volume flow", then "Total effective volume".
I suppose all three statements means the same thing, the sum of all single effective volume, summed with its polarity ( + or -).
Am I correct?
2) page 32, about the PI number: this is not clear to me. Is PI the tick value for the stock to be analized? For tick I mean the minimum distance (step) the stock can move. Tick value change depending on the value of the stock: in example, a stock trading @ 20 € 1 tick = 0.01 € (20.00 then 20.01 and so on) while stocks trading @ 8 € has 1 tick = 0.005 € ( 8.000 then 8.005 etc.)
So my question is: should I always use a fixed PI = 0.01 or change the PI value according with the stock tick value?
3) pag 123, in order to verify my code, I have compared the EV result as from table 3.1 and it is all fine, but the EV value at the second row where you have a -5,057 value where it should be positive 5,057 since there is a price increase.
Is this a print error or I am missing something?
4) LEV SEV separation: I am using the equi power method as describe at page 54. The method is clear to me but still I have some doubt about how use the data to be anylized.
Should I use an incremental data method or a moving window data?
In order to be more clear, let's suppose I would like to perform an analysis on 1 day (1 minute data). Italian market open @ 0900 and close @ 1725 with 5 min extension for close fixing @ 1730. Live data ranges from 0900 up to 1725, that's 505 minutes per day since market stop any trading activities @ 17.25:00
Having set = 0 the first trading minute (0900), do I have to take all these 505 1 min data and compute the LEV SEV separation or do I have to use a moving window (in example, x data such as 100 or 200 1 min data) in which add the next data and remove the first one?
Page 214 describe the use of a moving window for the positive and negative EV, so I am a little confused for the approach to use.
What if we would analyze more data, in example, 3 or more consecutive days? Should I have to separate LEV SEV based on the whole three 1 minute days?
Do I have to set = 0 any of the three 0900 min or do I have to start a new separation for any day? If so, I should have a gap between last day value and the next day.
As last question, I wonder if I can have a feedback on my data: attached are two txt data both 1 min time frame, would be very happy to have a xls or txt result with EV, TEV, LEV and SEV separation using the equi power method.
Data is in ddmmyyyy with ; fields delimiter and . as decimal delimiter: I can upload as different date and delimiters
My result for 1 day (EniJul24) is as follow, all three plotted