Data Box data

[Dolfan]

Every morning I log in and update my spreadsheets with data from the day before and the opening price for the day. I use a daily OHLC chart for this with the Data Box open.
The problem is, the opening price for the day will change tomorrow. For example, while the Data Box data for a day is not complete at 5am, one can reasonably assume that the opening price for the day is accurately depicted at this time. We do not REOPEN at a later time, I assume. However, when I log in tomorrow and complete today's data (high, low, close) and log tomorrows opening price, today's open price will have changed. Why is that?

Thanks in advance!

Best regards,

Dolfan

[NinjaTrader_JessicaP]

Hello Dolfan, and thank you for your question.

I am providing a publicly available link to an InvestoPedia page which I believe provides a great answer to this question.

http://www.investopedia.com/ask/answers/139.asp

Please let us know if this wasn't quite what you were asking, or if there is any other way we can help.

[Dolfan]

This still does not make sense to me. Looking at this verbage,

"These fluctuations are also why closing and opening prices are not always identical. In the hours between the closing bell and the following day's opening bell,"

I might assume that the opening price has not yet been set and will be set at the opening bell (9:30am ET). Would that be a correct assumption? With that in mind, how does the price that I see on a daily bar at 5am relate? It is not the same was what closed yesterday and it is not the same as settlement for the day.

With all of this in mind, if I wanted to write a strategy around this data and an accurate opening price is critical to setting daily targets, how and where would I get an accurate number? Do I tell my strategy to wait until 9:30am ET and pull a price at the close of that minute bar?

I have attached a picture file (CL) for your viewing pleasure. I am not yet sure that it is important for it to be that accurate but, being the Six Sigma kind of guy that I am, I would like for my data to be as accurate as inhumanely possible.

http://screencast.com/t/6FkW2gyc9

Thanks for your timely response.

Best regards,

Dolfan

[NinjaTrader_JessicaP]

Hello again Dolfan,

It sounds like this is the situation that Session Templates were designed to resolve. A session template lets Ninja know when the market opens and closes, so that it can filter incoming data accordingly and provide you with correct opening and closing conditions and build candles starting at the correct place.

For the CL instrument, the default session template trades 6 PM - 5 PM EDT.

In order to use a session template, you do need minute data instead of daily data. In order to build daily data out of a minute chart, so that you can apply session templates, you can simply multiply 1440 times the number of days per bar, since there are 1440 minutes in a day. This is because minute data and daily data are provided to Ninja differently.

If you have done no configuration at all, your session templates will probably apply as soon as you switch to 1440 minute charts.

Please let us know if there is any other way we can help.

[Dolfan]

Thanks again for your timely response. I will take a look at it in the early AM tomorrow and see what's what. Did this question make sense?

" Do I tell my strategy to wait until 9:30am ET and pull a price at the close of that minute bar?"

Best regards,

Dolfan

[NinjaTrader_JessicaP]

Thanks again for your timely response. I will take a look at it in the early AM tomorrow and see what's what. Did this question make sense?

" Do I tell my strategy to wait until 9:30am ET and pull a price at the close of that minute bar?"

Best regards,

Dolfan

I am not sure I do. I took that to mean you planned on waiting until just before the markets were open, and checking the price during that bar. I am unclear whether you wanted to check this bar while it is 09:30 still, or whether you intend on checking 09:30's bar at 09:31. I am also not quite sure what we are testing for.

What I do know is that a correctly applied session template will ensure that the data and candles in NinjaTrader accurately reflect market conditions.

[Dolfan]

In retrospect, I suspect that checking the price at the time that the market opens will subject me to wild variances and unreliable results. It is also important to have the strategy active early in the morning, not wait until later in the day or trend.

Thanks for your insight.

Best regards,

Dolfan

[Dolfan]

Jessica,

I am still seeing a fluctuation of data (from 1440 min charts) that causes a variance of open price depending on whether I pull it late at night (when I update my spreadsheets) and the next day when I update again. It is not a large variance and I am not sure that it matters that much at this time but worth noting. In creating my spreadsheet I note that also on the bottom of the data box is the daily ATR (if that indicator is active). I have compared the data from the data box and find something else that I need explained to me. The ATR number listed at the bottom will often times be different from the difference between the high and the low of the day. I can understand a little variance but sometimes it is quite large and when the ATR is more than the difference between the high and the low, well that makes no sense at all to me. Please explain to me where the ATR data point comes from and if need be, how I can compare that to the daily high/low difference. Is the Average True Range weighted by volume of orders at different levels? It seems to me that it would be impossible for the average range to exceed the actual range. Thanks in advance your you assistance.

Best regards,

Dolfan

[NinjaTrader_PatrickH]

Hello Dolfan,

Thank you for your response.

So I may investigate this matter further please answer the following questions.

• What version of NinjaTrader are you using? You can check under Help -> About (Example: 8.0.0.X)
• Who are you connected to for realtime data? Who are you connected to for historical data?

I look forward to your response.

[Dolfan]

Here is my version - 8.0.0.10 64-bit (Standard) - BETA
I connect to Dorman for trading (realtime?), not sure about historical data, you'll need to enlighten me on that.

Thanks in advance.

Best regards,

Dolfan

[NinjaTrader_PatrickH]

Hello Dolfan,

Thank you for your response.

So the only connection used would be your NinjaTrader Continuum connection, correct?

[Dolfan]

This is correct

Dolfan

[NinjaTrader_JessicaP]

Hello Dolfan,

I am investigating on my end to see if I can reproduce what you were observing.

As far as the Average True Range, just by way of explaining where Ninja's ATR data comes from, I am including some information from our code, as well as a definition from a publicly available website.

For an ATR definition, we have this from Investopedia, http://www.investopedia.com/terms/a/atr.asp .

As far as how ATR is defined in Ninja, in (My) Documents\NinjaTrader 7\bin\Custom\Indicators\@ATR.cs, we find this :

[FONT=Courier New]
                double trueRange = High[0] - Low[0];
                trueRange = Math.Max(Math.Abs(Low[0] - Close[1]), Math.Max(trueRange, Math.Abs(High[0] - Close[1])));[/FONT]

Converting that from computer code into math,

[FONT=Courier New]
true range = greatest of
  |current low - previous close|,
  current high - current low,
  or |current high - previous close|
[/FONT]

An n period ATR is an average of the true range for n days. For example, ATR(14) would average 14 days ATR values.

If the difference you are observing is larger than all three of these at the beginning of the day for the previous n days (e.g. 14 days for an ATR(14)), please let us know.

I will return soon with what I find on my end.

[NinjaTrader_JessicaP]

Hello Dolfan,

Please see the attached spreadsheet for the ES 09-16 instrument for June. Columns A-F are data I recorded from Ninja, columns G-K were ones I calculated.

On my system, I observed the 14 period ATR reported by Ninja stayed bounded between the lowest and highest calculated TR values, and remained close to the calculated June ATR.

Some of my numbers may have been affected from some things unique to my system. If your numbers are significantly different from mine, or there are are any other questions we could answer, please do not hesitate to reach out to us.

[Dolfan]

Thanks Jessica. I will pull this down and import my data this weekend for a side to side comparison and get back to you Monday with my findings.

Best regards,

Dolfan