"Volume buzz"

[Boris Tsipenyuk]

Hello all, I'm looking for a way to replicate something comparable to volume buzz/volume sizzle that you see in other products, I've rolled my own version of this by downloading tick volume data I have in Medved but it's a cumbersome process obviously - wondering if anybody has had any success of doing this natively via the screener or api somehow, thanks (the following is from TC2000's explanation) 

A stock's volume for the current day (let's say at 11:32am) is compared to its average historical volume for the same percentage of the day. The volume buzz tells you how far ahead or behind the stock is based on its normal historical activity. A volume buzz of +250% means the stock is trading 250% more than normal for this portion of the day. A figure of -50% means the stock is trading at only half of its norm for this portion of the day. A figure of 0 means nothing unusual is going on in either direction. 

 

thanks

 

 

[L W]

More information and formula.   Is this the same as you are referencing?

http://forums.worden.com/default.aspx?g=posts&t=59308

 

Volume Buzz is based on the 100-Period Simple Moving Average of Volume over the previous 100 days (so not including today).

The current Volume is compared to a percentage of the average volume which is directly proportional to the percentage of the trading day which has already passed. So at 10:15 AM ET, the current volume would be compared to about 11.54% of the 100-Day Simple Moving Average of Volume ending yesterday. If the result is +25%, it would mean that current Volume is about 14.42% of the average volume instead of 11.54% of the average volume.

So at the end of the day, you could approximate Volume Buzz using a formula like:

100 * (V / AVGV100.1 - 1)

This formula would also have the same or a similar order to Volume Buzz over the course of the day even though the values would not match.

Volume Buzz can be used as a WatchList Column and sort, but cannot be used as an EasyScan Condition.

[Boris Tsipenyuk]

Yes this is the same thing basically - I’ll give your examples a shot 

[Boris Tsipenyuk]

ok so - can you point me in the right direction of how this would look like in the scanner ? thanks  

[Mike Medved]

Hmm. This would be a bit expensive to calculate in MT. MT calculates indicators in an "incremental" fashion - that is, there is a "full" recalc of the indicator, then the state of the indicator is remembered for the last completed candle, and the subsequent candle(s)' value is calculated from the last completed candle state.

This does not lend itself very easily to this. But I will put this on my list. I would basically have to keep N pointers (where N is the # of days) into past data and advance them all together in order to recalculate the average volume.

[Boris Tsipenyuk]

I rolled my own poor man’s version because any stock in a portfolio that’s back filled has individual tick data in medved - so I have to query each say 20 last days up to current time - add up the volume ticks and weigh for time of day - very ugly and quite slow (and isn’t integrated at all) but in theory the data is already there 

I guess the right way to do this now that I think of it is export those 20 previous days in advance into a database 

[Boris Tsipenyuk]

Sorry for multiple posts had another thought - if I am not mistaken the existing vol % compares average daily value for 20 days to today - so really - all I would have to do is take that value and devide it by the % of the day that has past - every (rounded off) 4 minutes of the trading day is another 1% of the trading day 

Edited August 3, 2020 by Boris Tsipenyuk

[Mike Medved]

Not quite The volume is not distributed evenly throughout the day.

 

But you may want to look at the Pace indicator. It compares the latest N-minute period's volume to the average of all N-minute period volumes for the last X days.

[Boris Tsipenyuk]

I will look at that - I get that volume distribution is like that but if current vol % say is 30% but you are only 10% of the day done that would probably be useful and that’s what I am trying to do (among other things) thanks 

[Boris Tsipenyuk]

so out of sheer ignorance, is pace a % value ? thanks

 

[Mike Medved]

Yes it is %. The interesting thing is that it compares let's say the last 10 min to the average of all 10 min volume periods over the last let's say 2 days. Not just 9am to 9:10am and 9:10 to 9:20. But 9:01 to 9:11, 9:02 to 9:12 etc etc.

[Boris Tsipenyuk]

Right and I’m just trying to compare volume at 10:30 today to avg volume at 10:30 the last 20/30 days.. so I guess it’s down to figuring how to code that “% of the day” value via some kind of code related to time 

[Mike Medved]

Boris, I will try to carve out the time to do that indicator. It is a non-trivial one What should it be called? "Time of Day Volume %"?

[Boris Tsipenyuk]

time of day % sounds good Thanks 

[Oliver]

Hi,

 

I am looking to compute something like this in an historical scan. Can you help by correcting the code below? On an histoical scan, Timestamp canno't be used to get the current time and the function Now is not recognized in the code. The min function is not working either.

 

 

Thanks

 

var SessionInfo = GetTradingSessionInfo(Timestamp);

if (Now>SessionInfo.SessionStart && Now<SessionInfo.SessionEnd)

{ TodayRatioVol = SymbolData.Volume/SymbolData.AvgVol / ((Math.Min(Now,SessionInfo.SessionEnd)-SessionInfo.SessionStart)/(SessionInfo.SessionEnd-SessionInfo.SessionStart)) ; }

else

{TodayRatioVol=SymbolData.Volume/SymbolData.AvgVol;}

[Jerry Medved]

I emailed you