Volume--- how to use it and WHY

Quote from MandelbrotSet:

CHART 3
More...

Without knowing the volume you created the chart in I would say #3 or #1 are the closest of them, if the chart wasn't created in Sierra or Ninja then they aren't accurate to begin with. The only software that offers them is Ensign, MultiCharts, Sierra & Ninja. Volume isn't consistent on each bar in other software packages. Candlesticks are a distraction anyway since they need to be interpreted or read on there own.

#2 is a minute chart.

If I know the volume I could tell you specifically which is which.
 
all the different types of bars... they just look different.. highs and lows are going to be the same for the same trend end points... so you break the question out, what can you see with one that you can't with the other? Not all chart packages will display volume as ticks...

Time Bars: you can see the ticks, volume and price range

Volume Bars: you can see the ticks, time, and price range

Tick Bars: it's volume, price range, and time

Range Bars: Ticks, volume, time...

What will change is your geometrics if you draw on the charts and possibly your perception of something or another... so many questions, so incredibly much work to code up things and so little time....

Speaking of time, I have an indicator that I developed. It is based on 8 bars. It involves candle colors and the relation of volume higher or lower compared to adjacent bars... I got it from just watching charts and it is only useful in a specific situation, and there is more that has to be done after the pattern shows up to complete the transaction... there are 16384 combinations of the above parameters for each of the four bar types not counting the follow on pattern.. what are the chances somebody with a backtester or a data mining thingy is going to find such a pattern, find the follow on pattern, get them in the right order, get them applied to the appropriate situation and find out what it applies to best.... I'd say it's not going to happen any time soon, that's all I'm saying...
 
Quote from MandelbrotSet:

To do ProfLogic justice, I don't know fully how his charts are created, nor am I familiar with his way of viewing price action, so I can't accurately re-create his charts without knowing his system.

A. I use a fractal approach to measure the movement of price action in accordance with a multi-time frame (or multi-range of price action, or multi-volume of shares traded ... as you see from the charts that I posted, it is all relatively the same to me).
B. He uses a relatively small amount of volume to form each unit of measurement of his charts, and then uses two indicators to measure (i) price oscillation and (ii) strength of price movmenent.
A. My method incorporates time when creating charts, either directly or indirectly.
B. His method looks only at shares traded at multiple fractals of ech other.
A. My method is linear.
B. His method is circular.
More...

I use an amount of volume as it relates to the increment I trade inside, Intraday, Swing or Position. I also only use a single indicator not two. My method is both linear and cyclic.
 
Quote from marketsurfer:

every TA method incorporates time directly or indirectly--- the wizard refuses to admit this fact--- since doing so will monkeywrench the entire philosophy.

i was speaking of finding charts posted by the wizard previously, not recreating same. sorry for the confusion.

surf
More...

I wake up at 6 am and trade from 7 am till noon so if that is what you mean by indirectly, yes I use time but only as a period I trade. None of my trades are based on the time of day. The time of day has no meaning to WHEN a trade sets up. I trade it WHEN it sets up not because it is a certain time of day.

Your time based charts artificially skew your view of the markets. You like that, so be it. There are just others ways to view price. Just like some of us like the way HD TV improves the view of our programming. If you prefer analog black and white . . . that is up to you.
 
Quote from marketsurfer:

:confused: what am i missing?

an SMA needs to be created from periods-- constant volume bars are formed from a set volume number that is created by default at various time intervals therefore time is incorporated by default. you can't avoid the time factor inherent within market measurement.
More...

Sorry Smurf but Constant Volume bars aren't created at various time intervals, they are created in real time, as they happen. The are created naturally not artifically in your time based charts. This is why your moving averages are worthless on my charts.
 
Quote from MandelbrotSet:

1. CONSTANT VOLUME BARS - use VOLUME as their fixed element and RANGE and TIME as their variable elements in their creation.
2. CONSTANT TIME BARS - use TIME as their fixed element and RANGE and VOLUME as their variable elements in their creation.
3. CONSTANT RANGE BARS - use RANGE as their fixed elemment and TIME and VOLUME as their variable elements in their creation.
My charts show this implicitly, and in fact, someone has already figured-it-out and sent me a PM to that effect ... :D

Like surf says, it's all TA, and all ways of understanding how price action is working ... no method is beter or worse than the other, it's all about how you use it.
***
It's always fun talking about trading with you guys.
More...

1. The range of the bar is naturally based on the natural range of price within a specific volume increment. The time it takes to create a bar varies on the liquidity at any given period in the market. This is why volume makes a better bar because it is grounded on the natural flow of price. Markets are traded in volume NOT time or Range.
2. Correct
3. Correct

Any you can't say which is better till you test it which you are incapable and unwilling of doing.
 
Quote from euclid:

Just a little.

It's quite simple. Tick, Range, Volume, Time bar charts all show how price changes over time. The only difference is the determination of the time interval for each bar. They each show the highs and lows at the same price and in the same order. There is not much to choose between them.

However, if you are interested in looking at how volume changes over price and time, then one of those chart types is much less useful. In fact the only reason I can think of for promoting constant volume charting is as a marketing gimmick.
More...

Allow me to clarify; Tick, Range, Volume, Time bar charts all show how price changes over time. The only difference is the determination of the time interval for each bar. They each show the highs and lows at the same price and in the same order but the Constant Volume Bar charts show the natural cyclic oscillations devoid of the spikes in the time charts, the inconsistency of the tick charts and the constant consolidation of the range charts. Simply the flow of price is smoother and easier to read for strength and direction.
 
Quote from jprad:

Surf, using time as a reference point to synchronize the dataset is entirely different then being an integral part for computation of the method.

Let's say there's a horse race you want to watch live and you know that it's going to start at 6pm EDT on some Saturday in May.

While you certainly need to know that information, as well as the channel it's going to be televised on in order to watch the race, time itself is irrelevant to the outcome. All that matters is which horse's nose crosses the finish line first.

However, if you want to know the duration of the race or the average speed that the race was won then you certainly need to know the elapsed time.
More...

DAMN WELL STATED!!!!
 
Wow, how incredibly ignorant. Backtesting only gives "fudged" signals if you do it wrong and is orders of magnitude more efficient than forward testing. Take stocks for example... one can backtest a strategy on thousands of stocks over years of varying market conditions, to include OOS data, in minutes... while forward testing would take years. And there's no guarantee that the period you forward test on will be any more representative of future market conditions than the backtested data. In fact, just the opposite because forward testing is far more limited.
Quote from ProfLogic:

I don't believe in back-testing just forward testing. I only trust that which I can utilize in real time. Back-testing can give too many signals that can be fudged. It takes a lot longer period of time to make sure your method is consistent but in the long run it pays off.
More...
 
Quote from jprad:

Thanks for biting...

Other than synchronizing the start of the race, time is certainly irrelevant.

Go back to my first example. Let's say that the SMA dataset starts at the 100th bar of the session.

But, instead of giving you the entire session worth of tick data, I remove the ticks representing the first five bars and re-index the remaining ticks to start at 1.

Because the data is still synchronized to the 1K share bars you can still construct the bars correctly and compute the SMA correctly.

All that's changed is the frame of reference. From your perspective, the SMA starts at the 95th bar, not the 100th.

Still think that time's relevant?
More...

Careful, when he is backed in a corner . . . out comes the opossum. You can tell because he will start calling you names.
 
You must log in or register to reply here.
Back
Top