Is SC real-time data actually delayed?

[terr]

ATP (EOM) = prior chart bar reads 10:11:00 minutes and current chart bar 10:11 (seconds in-progress)

ATP (EOM) = prior chart bar reads 10:12:00 minutes and current chart bar 10:12 (seconds in-progress)

So you have two bars in the chart that have the same timestamp? I think that would be quite confusing to the user.

SC is consistent - every candle is stamped with the starting time of the candle.
ATP is not consistent. Some candles (the ones before the last one) are stamped with the ending time of the candle, while the last candle is stamped differently.

I really don't see any reason to do it the ATP way. If you say that the "close" price on a 12:30 candle in SC shows not the price at 12:30 but the close price at 12:31, that's true. But its "open" price is at 12:30, and all the candles are consistent.

Each candle has open, high, low, close. Obviously they all occurred at different times, so the timestamp of the whole candle cannot reflect all four of the timestamps of those values. The start-of-candle convention is at least consistent throughout, unlike the end-of-candle one.

 

[Betterthanamonkey]

Yes. And, why is that confusing? That is the way time has been calculated since time began, is it not? Look at your clock and count the seconds off. When sixty seconds hit at the top of the hour it advances the minute. Now keep counting the next 60 seconds and you will note that you're still on the same minute until 60 seconds advances again.

 

[Betterthanamonkey]

And, we don't consider a chart bar that's "in-progress" officially a time stamp until it's completed at which time it becomes a time stamp. But that could be subjective if you deal in seconds which we do not. So it's not really two chart bars with the same time stamp. It's one chart bar with a time stamp and one chart bar in-progress. Hence, there are no historic chart bars with the same time stamp.

And, respectfully, I've never found anything inconstant with EOM time stamps which is the industry standard and for a reason. But you can clearly see how BOM has caused much confusion and probably a lot of systems not as profitable as they could be had they been analyzing the correct historical chart bars. Just saying!

 

[terr]

Yes. And, why is that confusing? That is the way time has been calculated since time began, is it not? Look at your clock and count the seconds off. When sixty seconds hit at the top of the hour it advances the minute. Now keep counting the next 60 seconds and you will note that you're still on the same minute until 60 seconds advances again.

Maybe. But the candle consists of ticks within the minute *excluding* the 60 second mark. So marking it historically with the timestamp of a tick that is not included in it is misleading.

 

[easymon1]

probably a lot of systems not as profitable as they could be had they been analyzing the correct historical chart bars.

Wowsa.
Seems like backtest fiends would be pretty PO'd about this.
How much difference are we talking here re: Backtest Accuracy?

 

[studentofthemarkets]

Wowsa.
Seems like backtest fiends would be pretty PO'd about this.
How much difference are we talking here re: Backtest Accuracy?

I don't think accuracy is really the problem, because the tests themselves should be accurate. It's the interpretation. I thought I was looking at a certain time of day for # of profitable trades, but the backtest had labelled it differently than I was used to on Sierra Chart, so I was off by as much as 1 hour because that was the time frame I was using. But the error was mine for not adjusting to the timestamp difference Ninjatrader used.

 

[Betterthanamonkey]

Wowsa.
Seems like backtest fiends would be pretty PO'd about this.
How much difference are we talking here re: Backtest Accuracy?

As long as a trader is using only one trading software, which is what most retail traders use, there should be no problem.

However, if a trader is using two or more trading software platforms, as I do, then, it can become a MAJOR issue which could result in SIGNIFICANT losses or, at the very least, diminished returns if the trader using two or more software programs doesn't make sure of the following: a) all the trading software utilizes the "same" time calculation, BOM or EOM; or, b) make the necessary adjustments in their code for the one-minute, historical, chart bar, time difference between BOM and EOM, if they're going to use two or more software programs with different time calculations.

As far as the difference in accuracy, if you're using two or more trading software programs with different chart bar time calculations? It's one-minute for each and every chart bar which would make your losses cumulative meaning, you're analyzing the wrong chart bar each and every time (one-minute chart bar) so, the losses will continue to build. Just like a gear cog that jumps one space; it will remain that way until you fix it.

 

[Betterthanamonkey]

Maybe. But the candle consists of ticks within the minute *excluding* the 60 second mark. So marking it historically with the timestamp of a tick that is not included in it is misleading.

If you're dealing in seconds, I assume you're talking about current (in-process) chart bars. If so, as I've stated before, I'm not a scalper and, therefore, I'm not referring to a time stamp in seconds within a "current" one-minute chart bar which scalpers might use to their advantage. I'm only referring to historical chart bar analysis and, as I understand it, for the most part, only minutes are used for historical chart bar analysis as I don't see the advantage of analyzing seconds for the prior intraday chart bars or, analyzing seconds for the chart bars for prior days/weeks/months. But if you can get that data and make use of it, to each is own.