please look at the difference close price on 6/10 between the two pics below...
the first is captured on the open of 6/11.
and the second is captured on 6/12...
can TS compensate my loss due to the error?
Or what can I do??

slicent 

please look at the difference close price on 6/10 between the two pics below... the first is captured on the open of 6/11. and the second is captured on 6/12...

I'm not a heavy Tradestation user and I've discovered several cases of bad data. When notified they quietly fix it behind the scenes with no notice or record there was ever an issue. I believe this is why when you rerun a backtest months or years later your can't replicate exactly. If I've found that many others must have found a lot more. As for your screenshots the bars definitely look different but without more information it's impossible to tell.

slicent 

can TS compensate my loss due to the error?

That's an easy question. No.

slicent 

Or what can I do??

Well first you'd need to verify it was bad data and not something happening you don't understand. (Like LAST vs SETTLE on a daily bar). If it is a continual problem though the only thing you can do is switch software.

I don't think this is an issue. The close you see is not always the actually daily close. The daily closing price almost always changes due to settlements.

Example:

• Market closes at 3901.5
• Settlements occur by the exchange during market close time and the close is actually determined to be 3902

It is just the nature of the markets.

~vmodus

vmodus 

I don't think this is an issue. The close you see is not always the actually daily close. The daily closing price almost always changes due to settlements. Example: Market closes at 3901.5 Settlements occur by the exchange during market close time and the close is actually determined to be 3902 It is just the nature of the markets. ~vmodus

the first pic is captured "after" the open of 6/11 and the close price on 6/10 is changed on 6/12!
the same things has happened several times and cause additional loss. But this time I have left the screen, leaving evidence.

SMCJB 

I'm not a heavy Tradestation user and I've discovered several cases of bad data. When notified they quietly fix it behind the scenes with no notice or record there was ever an issue. I believe this is why when you rerun a backtest months or years later your can't replicate exactly. If I've found that many others must have found a lot more. As for your screenshots the bars definitely look different but without more information it's impossible to tell. That's an easy question. No. Well first you'd need to verify it was bad data and not something happening you don't understand. (Like LAST vs SETTLE on a daily bar). If it is a continual problem though the only thing you can do is switch software.

After the open of 6/11, the close price of 6/10 I see is SETTLE?

the data of rithmic will be more reliable?

You understand what @vmodus is saying. The settlement price of say Crude ("CL") is determined based upon trading between 1:28 and 1:30 Central Time. The exchange doesn't close or reset for the day until 4pm Central Time. So if you look at your bar at say 3:59pm the bar is going to show the last print NOT the settlement price. At some point after 4pm Tradestation updates your previous daily bar from the last price to the settlement price. This "should" happen in the market down time, which in CL's case is between 4pm and 5pm, but in my not-extensive experience of Tradestation, this does not always happen in a timely fashion. This could well be one of those situations where that did not happen.

To avoid this problem some people don't use daily bars and instead use 1440 min bars (1 day = 24*60 = 1440 mins). This bar will have the same open, high and low as a daily bar, but will always use the last trade and not the settlement as the close. The problem with this is that the last print is not nearly as important from a volume perspective as the settlement print.

SMCJB 

You understand what @vmodus is saying. The settlement price of say Crude ("CL") is determined based upon trading between 1:28 and 1:30 Central Time. The exchange doesn't close or reset for the day until 4pm Central Time. So if you look at your bar at say 3:59pm the bar is going to show the last print NOT the settlement price. At some point after 4pm Tradestation updates your previous daily bar from the last price to the settlement price. This "should" happen in the market down time, which in CL's case is between 4pm and 5pm, but in my not-extensive experience of Tradestation, this does not always happen in a timely fashion. This could well be one of those situations where that did not happen. To avoid this problem some people don't use daily bars and instead use 1440 min bars (1 day = 24*60 = 1440 mins). This bar will have the same open, high and low as a daily bar, but will always use the last trade and not the settlement as the close. The problem with this is that the last print is not nearly as important from a volume perspective as the settlement print.

Using the 1440 bars (for data2) does solve that issue. There are other solutions, but SMCJB's is probably the most elegant.

~vmodus

vmodus 

Using the 1440 bars (for data2) does solve that issue. There are other solutions, but SMCJB's is probably the most elegant.

Makes me think - you could have Daily as data1, 1440 as data2 and check them to see if they are different!

SMCJB 

You understand what @vmodus is saying. The settlement price of say Crude ("CL") is determined based upon trading between 1:28 and 1:30 Central Time. The exchange doesn't close or reset for the day until 4pm Central Time. So if you look at your bar at say 3:59pm the bar is going to show the last print NOT the settlement price. At some point after 4pm Tradestation updates your previous daily bar from the last price to the settlement price. This "should" happen in the market down time, which in CL's case is between 4pm and 5pm, but in my not-extensive experience of Tradestation, this does not always happen in a timely fashion. This could well be one of those situations where that did not happen. To avoid this problem some people don't use daily bars and instead use 1440 min bars (1 day = 24*60 = 1440 mins). This bar will have the same open, high and low as a daily bar, but will always use the last trade and not the settlement as the close. The problem with this is that the last print is not nearly as important from a volume perspective as the settlement print.

This "should" happen in the market down time...
How about other data feeds like rithmic?

slicent 

This "should" happen in the market down time... How about other data feeds like rithmic?

Don't know everybody probably does it differently. I do know Trading Technologies push the price well before the 4pm close but they are not your traditional 'bar based trading software'