I posted several charts yesterday of the ES futures, in which each bar represented a fixed number of contracts traded, not a fixed time period. Years ago, Richard Arms made a comment that stuck out in my mind. He said that the market's clock is expressed in units of volume, not time.
That notion has some deep implications. Concepts such as cycles, for example, take on a different meaning if we are to measure cycle periods in units of volume, not time. (That notion gets even more interesting when we standardize the volume units and express them as a percentage of a stock's float.)
Setting the x-axis of a chart in volume units also untangles the relationship between volume and volatility. Each bar in the Market Delta chart above represents 150,000 contracts traded, so that we can see Friday's action clearly. Note how some bars are quite a bit larger than other ones, despite the fact that all the bars represent the same number of contracts traded. Volatility is no longer a function of volume; we see it purely as a function of the depth of the order book. When we have lots of buyers and sellers around a set of prices, we will chew through volume in a narrow range. When buyers and sellers become unbalanced in the book, that range will expand. That makes breakout trades from bar to bar very interesting.
Finally, however, there is a psychological benefit to denominating charts in volume units. We know that volume forms a kind of smile pattern during the trading day, with volume (and volatility) highest in the first and last hours of trade, lowest in the midday hours. That typically leads active traders to overtrade those midday hours. When we denominate charts in volume, each bar is the same. If we are trading the relationship of one bar to the several preceding (as in a transition pattern), then we will wait as long as needed for that bar to form.
If we click on the chart above, we see that the amount of time it takes to trade 150,000 contracts is highly variable. Our holding period is not measured in time any more; it is measured in bars. As long as we wait for those bars to form, we can't overtrade slow market periods. We will tend to trade most when markets move the most, because that's when we'll have more bars and more setups.
I will be illustrating these and other ideas related to volume-based charts in coming days.
.
Posts
7 comments:
nice post.
Hi Brett,
what's interesting is, that all charttypes form distinctive, tradeable patterns. You have good patterns on Volume, but as well on Range (each bar covering the same range) or Tickcharts (each bar represents the same amount of trades) and of course on minute charts. If there were only one contract to trade Volumecharts are the way to trade imho. But the moment you watch multiple charts, eg trading the NQ, but watching the ES, the YM and maybe also the DAX and FTSE as well, then Volumecharts provide you with a problem: They are not in sync to one another.
The 5min 12:35 bar starts at 12:35:00 and stops building at 12:39:59. That bar is the same on the ES, the YM, the NQ. But the moment you use Volumecharts you can't compare the YM to the ES to the NQ Volumechart, as they all trade a different total Volume within the day.
Sure you can approximate: Say on the ES you use 150.000 contracts, on the YM 50.000 on the NQ 100.000, but it's not exact. Which means the patterns forming are not the same. If you trade watching multiple markets you look for confirmation of a pattern over a number of markets, they all doing the same, some faster or stronger, others maybe lagging, but not diverging.
Using Volumecharts, you can't do that, which is the reason I abandoned them after a while. I now use small Tickcharts to finetune an entry or exit. But the maincharts are back to minute charts as that's the only way for me to see at a glance, how markets are doing in comparison to one another.
Best regards,
Chris
Hi Dr Brett,
Interesting post and view of market action. If you were to add Net Day Delta it showed a minus -66000 which suggests distribution, though Quad Witching might skew this view, but it it worth observing on a daily basis.
Thanks for all your insight.
David
To Dr. Brett, I was the person that first introduced Volume Bar Charting to the industry in 1996 but it wasn't until 2003 that the first software charting company, Ensign, started to offer a Constant Volume/Share Bar Chart, the only accurate for of the charting. There is a huge difference between Volume Bars and Constant Volume Bars so be careful. Some of the charting companies like Tradestation float the number of contracts or shares per bar so they aren't exact.
To Globetrader, there is no exact or consistent correlation between any of the indices so trade each independently and trade each based on its perfect and natural cycles that the Constant Volume Bars create for you. When you view the indices through the eyes of variable charts such as daily, minute, tick or range bar charts there is some inconsistent correlations from time to time but those become idiosyncrasies the longer you monitor the Constant Volume Bar charting. The volume bars give you beautiful consistent natural cycles if you have the patience to look for them.
Wm. "Bill" Schamp
Hi Brett,
I see the chart but I dont know how to read it. This makes it impossible to follow the points you are making about it. Could you give a post that explains the basics of a chart like this? I am accustomed to my time-based candlesticks and bars. Thanks.
Hey Doc
Would you consider posting the chart definition on chart hub? I am a market delta user.
thanks
Dave
Sorry, this is Dr. Brett's blog and I use bar charts not Market Delta.
Post a Comment