Could someone address the following?
1. I have noted that in several instances PaperTrade fills market orders at prices that are well out of market for the time indicated. Upon further examination these prices are often up to 1/2 hour old. This naturally skews results quite a bit. If one is using a MA based trading strategy the net result is often more positive than it ought to be as the MA is naturally lagged somewhat and PaperTrade occasionally fills at prices closer to those at the beginning of a move. In any case, I would hope that this can be corrected soon. I find PaperTrade to be a valuable tool for strategy testing but am rather deflated over this issue.
2. What is the fill strategy is for PaperTrade?
3. In the PaperTrade.dat file, what is the exact base for the time numbers? I would like to convert these to times in Excel. So far I have approximated that the start date must be 01/01/1933 (?) while Excel uses 01/01/1900. So by using the following formula I arrived for a correct date/time given my time zone etc.: (TimeStamp/8640)+25568.8333333333 = correct excel time stamp which represents the number of whole days since 00/01/1900 + a decimal that represents the time. Apparently the decimal date represents a 20hr difference over the years or 1 day - 4 hours. Could you direct me to info re the time stamp format used?
4. Are there plans to include a more useful export function for PaperTrade (e.g. export to cvs)?
5. I'd like to request that winros be able to serve up the $playback symbol this would enable testing of strategies in excel, dynaorder and a number of applications that support the eSignal feed but not the $playback symbol via the tick replayer. I don't know if this is doable given your architecture, but it would be nice.
6. I have noted that in 7.4 there are times when undocked chart windows do not display the reload option for studies. This is a minor inconvenience.
7. We have a number of machines running eSignal. There have been a few instances where the charts displayed for the same instrument were very different (same compression and everything). What causes this? I would assume that for some reason the underlying data was different / corrupted? Is this the case?
8. We have a few windows XP machines. In PaperTrade these often display 0 gains as a very small number (scientific notation to 12 decimal places). How can this be prevented?
9. Is there some way for us to demo/preview the pro API to ensure that the functionality we require is available?
10. What about allowing market depth queries in efs? This would be nice.
11. Although there are a number of posts regarding this, we continue to have performance problems running efs scripts on charts displaying sotcks with a high/rapid tick rate. I may have missed a solution on this board or elsewhere. Could you direct me to a solution or indicate when a fix will be in?
12. I am aware that the structure of your tick data does not easily facilitate the downloading of a specific range of data. I would suggest however, that it would be helpful to be able to set the start and stop date/time for playback.
13. I would also like to request that the back testing module be able to import the results from PaperTrade for analysis. This request goes hand-in-hand with my prior export request.
I guess that's enough....for now
Regards,
Dennis
1. I have noted that in several instances PaperTrade fills market orders at prices that are well out of market for the time indicated. Upon further examination these prices are often up to 1/2 hour old. This naturally skews results quite a bit. If one is using a MA based trading strategy the net result is often more positive than it ought to be as the MA is naturally lagged somewhat and PaperTrade occasionally fills at prices closer to those at the beginning of a move. In any case, I would hope that this can be corrected soon. I find PaperTrade to be a valuable tool for strategy testing but am rather deflated over this issue.
2. What is the fill strategy is for PaperTrade?
3. In the PaperTrade.dat file, what is the exact base for the time numbers? I would like to convert these to times in Excel. So far I have approximated that the start date must be 01/01/1933 (?) while Excel uses 01/01/1900. So by using the following formula I arrived for a correct date/time given my time zone etc.: (TimeStamp/8640)+25568.8333333333 = correct excel time stamp which represents the number of whole days since 00/01/1900 + a decimal that represents the time. Apparently the decimal date represents a 20hr difference over the years or 1 day - 4 hours. Could you direct me to info re the time stamp format used?
4. Are there plans to include a more useful export function for PaperTrade (e.g. export to cvs)?
5. I'd like to request that winros be able to serve up the $playback symbol this would enable testing of strategies in excel, dynaorder and a number of applications that support the eSignal feed but not the $playback symbol via the tick replayer. I don't know if this is doable given your architecture, but it would be nice.
6. I have noted that in 7.4 there are times when undocked chart windows do not display the reload option for studies. This is a minor inconvenience.
7. We have a number of machines running eSignal. There have been a few instances where the charts displayed for the same instrument were very different (same compression and everything). What causes this? I would assume that for some reason the underlying data was different / corrupted? Is this the case?
8. We have a few windows XP machines. In PaperTrade these often display 0 gains as a very small number (scientific notation to 12 decimal places). How can this be prevented?
9. Is there some way for us to demo/preview the pro API to ensure that the functionality we require is available?
10. What about allowing market depth queries in efs? This would be nice.
11. Although there are a number of posts regarding this, we continue to have performance problems running efs scripts on charts displaying sotcks with a high/rapid tick rate. I may have missed a solution on this board or elsewhere. Could you direct me to a solution or indicate when a fix will be in?
12. I am aware that the structure of your tick data does not easily facilitate the downloading of a specific range of data. I would suggest however, that it would be helpful to be able to set the start and stop date/time for playback.
13. I would also like to request that the back testing module be able to import the results from PaperTrade for analysis. This request goes hand-in-hand with my prior export request.
I guess that's enough....for now
Regards,
Dennis
Comment