Time vs tick

[They]

@They

The "VolumeSpeed" indicator is interesting to me... I'm guessing it's custom, not something out of the box? Since there is no "time", the trading day could be charted as a single bar or many hundreds of bars (depending on constant volume setting of course). So my question is what is the baseline being used for VolumeSpeed? Frankly, the omission of time, as in the case of constant volume charts, has always been a deal breaker for me.

This volumespeed thing is stoking my interest in volume charts once again. BTW, using time-based charts, I use volume pace... Take a volume measurement every n seconds, and calculate a pro-rata bar close total volume. For the majority of bars, it is clear well before bar close if volume is increasing or decreasing. In my world, volume is a prerequisite for price volatility, in context. So yea, volume pace can be identified and exploited on time based charts.

Thanks

tiddlywinks,

As stated there is no natural time segments in a daily bar, just an open high, low, close and volume. These are really the only market constants and thus as traders we need to know where they occurred. Volume bars are just one way of parsing that information.

The baseline is a fixed amount of volume/contracts. It is arbitrary just like time or tick based charts. The speed/frequency of those contracts being traded is another piece of information. Some may use the frequency of volume traded and some may just use volume bars alone.

I am sure that in most charting packages that allow for volume based bars there is a way to code an indicator/study that marks the start and end time of a bar. In Esignal the study is called tickbartime.

Do you want to know what is the average time of all of the up bars vs all of the down bars at any given time during the day or after a certain amount of volume has traded. Do you want to divide the range of the up bars against the volume of the up bars and vice versa for the down bars and then compare them to each other? Do you want to know the range vs the volume required to create that range vs the time required and compare that data against the downward range, volume and speed. How far down the rabbit hole do you want to go?

Your 'volumepace' study sounds interesting, kind of like volume profiling on individual bars.

 

[kj5159]

@They

The "VolumeSpeed" indicator is interesting to me... I'm guessing it's custom, not something out of the box? Since there is no "time", the trading day could be charted as a single bar or many hundreds of bars (depending on constant volume setting of course). So my question is what is the baseline being used for VolumeSpeed? Frankly, the omission of time, as in the case of constant volume charts, has always been a deal breaker for me.

This volumespeed thing is stoking my interest in volume charts once again. BTW, using time-based charts, I use volume pace... Take a volume measurement every n seconds, and calculate a pro-rata bar close total volume. For the majority of bars, it is clear well before bar close if volume is increasing or decreasing. In my world, volume is a prerequisite for price volatility, in context. So yea, volume pace can be identified and exploited on time based charts.

Thanks

I second the question on the bar speed indicator. I have ninja trader and they have volume bars but I don't think they have a timer for it. There is an indicator called "bar timer" but to my disappointment it just counts down the time until the next time bar, no connection to constant volume bars.

 

[Big AAPL]

Volume bars on NT have a countdown timer in the default indicators. Then they have this
:
http://www.freeindicators.com/ninjatrader/free-ninjatrader-indicators/bar-speed-indicator/

 

[tiddlywinks]

tiddlywinks,

As stated there is no natural time segments in a daily bar, just an open high, low, close and volume. These are really the only market constants and thus as traders we need to know where they occurred. Volume bars are just one way of parsing that information.

The baseline is a fixed amount of volume/contracts. It is arbitrary just like time or tick based charts. The speed/frequency of those contracts being traded is another piece of information. Some may use the frequency of volume traded and some may just use volume bars alone.

I am sure that in most charting packages that allow for volume based bars there is a way to code an indicator/study that marks the start and end time of a bar. In Esignal the study is called tickbartime.

Do you want to know what is the average time of all of the up bars vs all of the down bars at any given time during the day or after a certain amount of volume has traded. Do you want to divide the range of the up bars against the volume of the up bars and vice versa for the down bars and then compare them to each other? Do you want to know the range vs the volume required to create that range vs the time required and compare that data against the downward range, volume and speed. How far down the rabbit hole do you want to go?

Your 'volumepace' study sounds interesting, kind of like volume profiling on individual bars.

Thanks for the reply @They

Like almost every type of analysis in trading, volume is a rabbit hole, if you let it take you there.

Measuring a non-time based bar (constant volume, ticks, etc) in a time unit (ie seconds) by itself has minimal use, at least to me. Therein is my rabbit hole. It's like having a full tank of gas, but not knowing how many gallons the vehicle holds and the MPG to be expected. So this bar took 600 seconds to complete and that bar took 90 seconds to complete. Without some sort of benchmark (imo) to know if the 600 seconds is normal, excessivly slow, or maybe fast in a given environment, and what that means for price, it's merely 600 seconds, which means... it took 600 seconds, period. Vice versa with faster bar completion. In an overall context I know there is value in that data. That value has not been identified or exploited imo... that's the rabbit hole I fall in with constant volume bars.

Anyway, thanks again for the reply. I had not seen a VolumeSpeed/TickBarTime histogram before. Will check my platform (Sierrachart) for similar.

Thanks.

 

[SkyChef]

Thanks for the reply @They

... In an overall context I know there is value in that data. That value has not been identified or exploited imo... that's the rabbit hole I fall in with constant volume bars.

Anyway, thanks again for the reply. I had not seen a VolumeSpeed/TickBarTime histogram before. Will check my platform (Sierrachart) for similar.

Thanks.

I believe that value has been identified and exploited but you will need footprint (or number bar study in SC) to display that for you. Or if you are very good at DOM reading, you could see that info there too. In general, when a vol bar completed so fast, it is a result of market orders swept at several levels at once. It indicates how aggressive one side is until it hits a wall (absorption) then either violent reverses (trapped traders) or continues.

BTW, I don't think SC has a general study volume speed for general volume chart but they do have a tabular column data named Volume/sec within number bar study.

 

[manonfire]

Price over time is all I need. Find the lowest common denominator of the largest participants. On the floor of the CME a buzzer goes off every 15 minutes there is a reason for that.

At the same time don't think linear. Be able to see any 3 five minute bars as a 15 min bar, any 3 10 min bars as a 30 min bar, any 2 15 min bars as a 30 min bar, etc. Price action is not neatly defined on the :00, :15, :30 etc.

 

[treeman]

Price over time is all I need. Find the lowest common denominator of the largest participants. On the floor of the CME a buzzer goes off every 15 minutes there is a reason for that.

For us ignoramuses, what's the reason?

 

[manonfire]

For us ignoramuses, what's the reason?

Sorry I should have expanded on this. Before electronic devices were allowed on the floor it was for recording, or at least taking note of closing price, 15 minutes being the lowest common denominator. There is also a method of point and figure charting used by floor traders that has no time axis, nothing I ever grasped.

I would encourage everyone to master manual charting at least on 60 min timeframe at minimum, preferably down to 15. Learn to read price over time and you will never plot an indicator on a chart again.

 

[They]

Find the lowest common denominator of the largest participants.

The lowest common denominator on the largest participants is their volume. Time is actually subservient.

 

[pstrusi]

Overall thoughts? I’ve been a strict time based chart reader, in the vein of: if it ain’t broke, don’t fix it. Anyhow, I’ve been thinking a lot recently about how arbitrary time is, and how tick based charts are probably the most accurate reflection of the overall market structure. Have you transitioned to tick? Did you just slide right in? Any gotchas I should be looking out for as I toy with making the switch?

I'm aware that this subject can lead to some controversy but here my experience. Any chart different than tick chart is doom to fail in regards to system reliability for the long run. This is cause it's simply a changing "base structure" cause arbitrary limits such as Time or Price distorts the real market microstructure; therefore it's impossible to think to have a "reliable" signal system based on this kind of data.