Anyone use anything else than time based charts for day trading?

I don't use raw volume either, @Scataphagos. That is, volume under my time based bars. And to adress your original concerns: my system is built on time bars and most of my charts are time based.

The volume chart is just my finest degree/resolution and can also help to 'tune in' with what's going on in the market. I don't read volume per se, but like that these charts reflect activity/trades taking place.

Ir I had to make a choice, I would choose time based only. As it is, I can supplement and use both.
 
wiesman02 said:
Hmmm very very interesting. When you say volume considerations, you are referencing all forms of volume not just volume bars, correct ?
More...

Correct. Volume is not reliable. Let's say you're ready to make a trade... "I'm thinking of pulling the trigger on this play, now let me consider the volume". If you think volume is supporting the trade, you take it. Or you don't take the trade because you don't like what the volume is saying to you. If you could log and review all of those "didn't takes because the volume put me off of it", you'd find that you blew LOTS more profit opportunity than losses you avoided.

And as far as using "volume bars" on the x-axis... a volume bar is also "the amount of time" it takes to cross a volume of stocks/contracts. Volume in this case is just another word for time. (Using volume this way doesn't hurt anything, but it doesn't help anything either. A side effect might be... thinking if you're "using volume in some fashion", then volume is important.... and being important here maybe means it's important elsewhere.) Easiest way to deal with volume... as far as whether or not to take a trade... is forget you ever heard the term*.

So... where do I come by this view of volume? Years ago I'd concocted a couple of "volume oscillators" in the custom indicator part of MetaStock. They correlated well with aspects of the market for what seems like quite a while. Then one day I noticed they no longer seemed to correlate like they did before. In fact, they often were disagreeing so much as to be the OPPOSITE of what they used to be. Totally out of synch. At that point, I just took volume out of the equation entirely.

And also, Joe Granville. Surely you've heard of him. He got consumed with his notion of "On Balance Volume". I guess it correlated with market turns or something at one time, but by the time I'd learned of it and checked into it... seemed to have lost all relevance. I can recall Granville being on FNN/CNBC and making his "where is the market going from here?" forecasts based upon OBV. He was waaay wrong almost every time. He clung to his OBV beliefs until his dying day. Probably tarnished his legacy a bit.

*You still want to consider the handle on an issue. That's a matter of "how big can I go" rather than whether or not you should.
 
Last edited:
I'm in agreement with everything Zcat is saying, as for OBV, i spent years coding, experimenting on this indicator thinking it was a holey greyl but it was a rabbit hole which led me into a foxes asshole.
 
I'm surprised no one has mentioned "information driven bars". Marcos Lopez de Prado covers them in his financial machine learning book (not a great overview...but suffecient, if you've never used something like this). There are an infinite variety of information driven bars, but the premise is that a bar is considered complete when an objective function exceeds its expected value (computed dynamically). Because the objective function is (can be) a function of price, volume, number of ticks, etc... each bar is not a fixed time,volume, # of ticks, etc. For example, if the objective function is SUM( TradeSize ) ...than a single very large trade would contain the same amount of information as a bunch of small trades.
 
Last edited:
Scataphagos said:
Correct. Volume is not reliable. Let's say you're ready to make a trade... "I'm thinking of pulling the trigger on this play, now let me consider the volume". If you think volume is supporting the trade, you take it. Or you don't take the trade because you don't like what the volume is saying to you. If you could log and review all of those "didn't takes because the volume put me off of it", you'd find that you blew LOTS more profit opportunity than losses you avoided.

And as far as using "volume bars" on the x-axis... a volume bar is also "the amount of time" it takes to cross a volume of stocks/contracts. Volume in this case is just another word for time. (Using volume this way doesn't hurt anything, but it doesn't help anything either. A side effect might be... thinking if you're "using volume in some fashion", then volume is important.... and being important here maybe means it's important elsewhere.) Easiest way to deal with volume... as far as whether or not to take a trade... is forget you ever heard the term*.

So... where do I come by this view of volume? Years ago I'd concocted a couple of "volume oscillators" in the custom indicator part of MetaStock. They correlated well with aspects of the market for what seems like quite a while. Then one day I noticed they no longer seemed to correlate like they did before. In fact, they often were disagreeing so much as to be the OPPOSITE of what they used to be. Totally out of synch. At that point, I just took volume out of the equation entirely.

And also, Joe Granville. Surely you've heard of him. He got consumed with his notion of "On Balance Volume". I guess it correlated with market turns or something at one time, but by the time I'd learned of it and checked into it... seemed to have lost all relevance. I can recall Granville being on FNN/CNBC and making his "where is the market going from here?" forecasts based upon OBV. He was waaay wrong almost every time. He clung to his OBV beliefs until his dying day. Probably tarnished his legacy a bit.

*You still want to consider the handle on an issue. That's a matter of "how big can I go" rather than whether or not you should.
More...

Volume information in a time based chart does correlate with something though. Volatility. On a given day, when volume is constantly printing higher than average, the volatility can be observed to be higher than average. Most (not all) inside bars form on lower volume than previous bars. When volume is too low, it’s more likely ranging, running stops, etc. than trending. Using that for setups I guess is another matter.
 
Scataphagos said:
"Almost EVERYBODY" uses "time" on the X-axis. Your objective as a trader is to "tune in to the market" and do what it's doing. How are you going to do that if you're trying to consider something the market is not? Sheesh!

For all you "volume bar" junkies... did you ever consider that your "1000 volume bars" are also a reflection of the "TIME it takes to trade 1000"? IOW, a volume bar is just a "time bar by another name". (I know, I know.... some say "volume" is important, so you MUST include volume somehow in your decisions or you're some kind of trader-dickhead!) What if volume is actually irrelevant and you're deceiving youself giving it any weight? Just a thought.

Seems to me that everyone on this thread is spinning their wheels trying to "reinvent the wheel".

KISS baby, as always. :)
More...

So, under your logic, since "everyone" uses time on the X-axis, then everyone "tuning into that market" methodology should be trading with relatively the same success. The problem is; that documented success rate consistently hovers around 5% - 8%. Proof that time artificially skews the view of price flow or action one need only to look at typical overnight trading when trade volume is usually thin. During RTH trade volume increases, in normalized markets. One way to better see the complete view of price flow or price action is using any charting structure where every bar contains the exact same number of contracts or shares traded. The fewer indicators one uses to see this price flow the better. A constant volume bar is absolutely not a "time bar by another name". Here's a visual explanation:

Personally I consider myself a perfectionist when it comes to viewing price action not a junkie. If you think watching a football game, movie or other program on an analog TV is the same as watching the same event on a 4K TV then there are larger issues. If that makes me a "dickhead", so be it. If a clearer view of price action, like watching a 4K TV is reinventing the wheel. . . . then I will continue using my better wheels.

Constant volume bars aren't the ONLY way to have success in viewing price action and pulling profits from these markets but compared to raw time based charts, in my opinion, there are obvious advantages.
 
Last edited:
tiddlywinks said:
The problems with Volume-based bars, that I have determined for MY use are twofold...

1) The elimination of time. One bar may take 2 minutes to complete and the next bar may take an hour to complete. Without looking at start-stop details of each bar, either in hindsight or witnessing in real-time, for me, usefulness greatly is reduced.

2) Without tracking bid-ask price and/or spread during volume based bars formation, the bar removes important pricing factors... lifting offers, hitting bids, immediate support/resistance levels etc. all are missing! While the same can be said with time based bars, each price bar and volume data point represents a uniform time interval which can be analyzed visually, mathematically, with or without additional data, and most important (to me) in a universal way, bar-by-bar.
More...

I would love to see someone come up with a constant volume bar that addresses the internal bid/ask balance or bias of each bar. If they exist, I wish someone would share. :-)
 
Last edited:
I guess the phrase "to better see the complete" wasn't correct. I should have said " to see a unique and balanced". I don't want to imply "my view" is the best view.

If by "kiss" you mean to keep it simple, I am.
 
You must log in or register to reply here.
Back
Top