I'm new to trading with Tick Charts. I scalp following PATs trading channel on YouTube and I've noticed that his 2000 tick data is different from my 2000 tick data. So much so, that the signal bars and setups are often completely different. He uses NinjaTrader (unsure of data source) and I use Tradovate which uses CQG data.
Is this discrepency normal? If so, I don't see the benefit in Tick Charts as it seems you would want to trade the same chart everyone else is trading...seems like using the 1 minute time frame would be more useful. That said, I do like how Tick Charts react to the volume (transactions).
Anyway, just curious. I'm new to the forum and appreciate the wealth of knowledge available here. Thank you.
Justin
The following 2 users say Thank You to borch084 for this post:
Tick charts can vary because the time the platform was switched on in the morning is different between people so the platform starts counting from a different point on the graph, or the data provider bundles data say. Basically a number of reasons. Also you will quite often find that if you take a screenshot of your graph at the end of the session then close your platform and reopen, the data that reloads will be slightly different to the data that created the graph in real time so that will affect indicators and might change high/low relations between bars changing when a signal occurs. This would obviously makes back testing irrelevant if done as the real time trade graphs wouldn't match the same day reproduced afterwards in a back test.
Personally if you like the tick bar style then for futures I would use volume bars as you have a central exchange with proper complete volume data available to everybody. I always imagined that tick bars were really best used as an option for markets, such as spot forex, that don't have accurate volume data, but although that seems logical to me it could be completely wrong.
But at the end of the day you trade what is on your chart only, as with all the various available chart types and their choice of settings or timeframes there is no such thing as the "same chart everyone else is trading".
Trading, ideally structured, is a vehicle for expanding consciousness, not damaging it. - Brett Steenbarger
The following 18 users say Thank You to matthew28 for this post:
I totally agree with @matthew28's explanation, and also with this part, about the "same chart everyone else is trading" -- but I know there can be some controversy on this point, and that different traders, some very successful, may disagree.
So I'll just add this, which is my take on the "same bar" question: it really isn't that critical and in practical terms it doesn't matter. This is because my chart will be similar enough to your chart, and everyone's chart, so long as the underlying data is the same. So if my tick chart bars are a little different from someone else's using the same tick value, we'll still see essentially the same thing. They may be different according to whether they are longer or shorter term, or smoother or jumpier, but that will be the only real difference.
I would go so far as to say that if we use very different bar types or bar aggregation values (number of ticks, volume, time, range, etc.), we will still see substantially the thing and the trades that are indicated will be substantially the same, differing only in terms of whether they are longer or shorter-term, assuming we use the same method and apply it the same way -- which, of course, we really won't . This is because the market is the market, however you slice and dice it. There will, of course, be large differences between what is shown by a larger-period chart (say, 5 minutes) and a shorter period chart (say, 30 seconds.) But these are differences of resolution, of the level of detail provided and how grainy you want your view of the market to be. It's still the only market there is, just in less detail or more detail. The same applies to different tick or volume values, and to whether your tick chart starts where mine does or not. There's just one market. You may want to use a slower or a faster bar value based on your own preference for longer or shorter-term trades, and of course there are some methods that explicitly call for one bar type or value, which is fine if it works for you. But there is still just one market.
I wanted to give this view, which I personally have found to be the case, while also recognizing that there may be some objections to it by others who think one bar type/period is the best, or who do think that having the same chart matters, and who have also found their view to be the case in actual practice.... Essentially, if something works, then it is correct, or correct enough, and there is more than one way to slice up the chart data and to make decisions about it. You can expect to find good reasons to adopt one idea or another, and the best criteria for judging whether you should use this or that idea is simply whether you can make money with it.
Tick charts are created from time and sales data. Each record in the time and sales data is a "tick". So if you are aggregating 2000 ticks to create an OHLC bar, technically they should not be different. But, they may be for the following reasons:
The data provider is conflating time and sales. Conflation is a process employed by some data providers where if there are too many trades in a certain time period, they are merged into one before being disseminated. This is the most likely reason for tick charts being different.
Your platform has a bug. Upon startup, the historical "tick" data it queries is not an exact replica of real-time - its either out of sequence, incomplete (with gaps) or is conflated. Also upon startup, the platform needs to merge the historical stream with real-time and it may not do that correctly. This is an unlikely reason for tick charts being different.
There could be other reasons but the most likely one is conflation.
The following 12 users say Thank You to Hulk for this post:
This is known for some data sources. Off hand, I can recall ToS (Thinkorswim) and IB (Interactive Brokers) being mentioned with this issue (sometimes called "bundling" or "filtering"), and I'm sure there are others. It is deliberate, because it saves bandwidth and processing at the user's computer, but most traders regard this as a low-quality type of data feed. Most do not do this.
The issue mentioned by @matthew28, where the platform starts building tick bars depending on when it starts receiving data, rather than resetting at a standard time, such as the session open, has also been seen, but it should be a user's choice (no idea why you would choose it ). For instance, here is a setting in Sierra Chart, for volume bars, but the same applies for tick bars too. Checking the box makes each day start with a new bar, so every bar after that should be the same as anyone else's volume (or tick) bars that use the same option, until it starts over with the next session. It is possible that you don't have such an option with your platform, or haven't set it, or that the other person doesn't have it, or hasn't set it.
I think that not having this is also not a good thing, because without it you don't know when the software starts to count the ticks and you don't know when the bar will be fully formed. However, in the big picture it is not really that important, as has been discussed.
Bob.
When one door closes, another opens.
-- Cervantes, Don Quixote
The following 8 users say Thank You to bobwest for this post:
Thanks for the detailed reply. I use Tradovate as my platform which uses CQG data. I've emailed them about this and I'm waiting for a reply back. As far as I can see there are no options to specify the start time of the tick/volume data. They seem to keep things as simple as possible, so there aren't a lot of back-end options for the data/charts.
Thanks again,
Justin
The following 4 users say Thank You to borch084 for this post:
CQG data is not bundled, so that is not a problem. Truly, few data feeds are.
I don't know how Tradovate builds the tick bars, never used it.
I see you are doing PATs. I recall, not having watched any PATs vids for a few years, that he commented on this difference in bars a few times (I think... if memory is still functioning .) As I recall, he was not too concerned about it either. Just trade the setups you see on your chart is the best advice.
As to why to use tick charts (or volume bar charts, which are like tick charts but are based on volume -- number of contracts traded -- rather than simply number of trades), it's the kind of thing that you have to just decide for yourself. For instance you could have a short-term time bar chart up (experiment with different time lengths) and compare to your tick chart. You may see that the tick chart gives you more detail, although this will depend on the tick setting and the time setting you are comparing it to. But this level of detail may be good or bad for you, because you may want a less-detailed, more overall type of view. Just experiment.
By the say, the PATs method is loosely based on Al Brooks' method (which is much, much more detailed and some find more difficult), and Al uses only time charts -- mainly 5-minute bars, with a look at 15 and 60-minute bars. Compare that to your tick charts and you will see a big difference.
Also, the "pre-market" or "pre-open" period before the open of the NYSE at 9:30 Eastern time is portrayed very differently if you are using tick or volume bars, vs. time bars. The time bars give you the same number of bars for every hour, for instance, but the tick bars will change the number of bars depending on the activity level, which makes the overnight market action a much smaller part of your chart, since fewer trades are taking place.
You just have to try the different choices and see what you like.
Bob.
When one door closes, another opens.
-- Cervantes, Don Quixote
The following 7 users say Thank You to bobwest for this post:
You're incredibly helpful. Thank you for all the great information. I'll keep experimenting with tick vs volume vs time and see what I like best. So far, I've seen benefit in keeping both a 1-min time & 2000 tick chart open. But, I also don't want to get too caught up checking back and forth before making a trade. "Just trade your chart" definitely resonates with me...that's all I can do. I will check out Al Brooks' method and see what kinds of new insights I can pull away.
Thanks again,
Justin
The following 4 users say Thank You to borch084 for this post:
. This is because the market is the market, however you slice and dice it. There will, of course, be large differences between what is shown by a larger-period chart (say, 5 minutes) and a shorter period chart (say, 30 seconds.) But these are differences of resolution, of the level of detail provided and how grainy you want your view of the market to be. It's still the only market there is, just in less detail or more detail. The same applies to different tick or volume values, and to whether your tick chart starts where mine does or not. There's just one market. You may want to use a slower or a faster bar value based on your own preference for longer or shorter-term trades, and of course there are some methods that explicitly call for one bar type or value, which is fine if it works for you. But there is still just one market.
). For instance, here is a setting in