Simulator reports

[NinjaTrader_Ray]

Lets keep things simple -

- Real-time or simulation --> PnL calculations internal go through the same code
- Easiest comparison is then MBT Navigator to NT
- As per attached image
--> If you ever see that these two values are not identical when trading live positions (you could use MBT demo if you wish) please let us know since surely we would want to fix it but as per our thousands of tests --> It works as expected.

[BIGRED]

Lets keep things simple -

- Real-time or simulation --> PnL calculations internal go through the same code
- Easiest comparison is then MBT Navigator to NT
- As per attached image
--> If you ever see that these two values are not identical when trading live positions (you could use MBT demo if you wish) please let us know since surely we would want to fix it but as per our thousands of tests --> It works as expected.

The only way to compare MBT trade P/L to Simulator Trade P/L would be to run the trades directly to the MBT Demo so I would have access to the MBT trade reports.

You are showing an open position P/L, not a completed trade. The problem I have found is in completed trades.

My understanding is.....that a pair such as USDJPY has a static point value and will not adjust to the dynamic changes in pip values. If that is correct, I expect each P/L result to be off by a small amount. That the user can manually adjust the pip value if required. Is this correct?

No matter the answer to my last question, it seems I have failed to choose the correct words and or use them to express the issue in such a way that they have meaning and or create understanding. So I will try 2 more times, differently to see if it goes "click".

1. The amount of P/L difference between what is showing up (on some trades) and what should show up means....the pip values would have to change by more than 10% which did not happen in the foreseeable past, let alone within minutes of one another.

2. You could take either the highest or the lowest pip value for the past week and multiply it times the number of pips in the trade profit and it will not come close the the P/L being reported.

Finally,
I feel like I am trying to argue a point that should be accepted gratefully. I am a potential customer on the one hand trying to make something work and on the other report what is not so that I have something that will work for me and have your company wind up with a better product in the end. This is not to say that anything I point out means there is a problem, of course I will make mistakes in application and or not understanding the meaning of a something being communicated. I do however have 20 years of computerized financial market software experience, ran 15 2000i packages in a trading company with the staff to support them, had Bill Cruz as my support person at Omega.

I am not trying to toot anything, but I am trying to point out that I am qualified. That I am willing to work with you and share my experience.

The other option for each of us is obvious.

Regards
Don

[NinjaTrader_Ray]

Don,

It is being accepted gracefully. In programming, we need reproducible scenarios. We do follow up on reports. All day yesterday I ran about 10 trades in MBT. Ran a performance report on the closed trades and all PnL calculated identical to what MBT reported and what I would expect thus it works as execpted. Also, we have had the math calculations for 1.5 years and no one else has reported making it more difficult to find out why your experience is different.

That being said...Your report is not falling on deaf ears, we have logged this as a reported issue, have investigated but can't find anything wrong. This happens. Hopefully more people will report and sooner or later if there is something wrong we will find it.

Thanks for your understanding and cooperation.

[BIGRED]

Ray,

I appreciate your situation and the dilemma.

If I had 100 dollars for every problem I found that went unnoticed for a long period of time. I wouldn't trade for a living.

I am not sure where to go with this issue at this time because I am stuck not being able to confirm my trading costs (slippage, spread and trading issues).

On my end there is no doubts that the Simulator Trade P/L is not correct (at times).

The model I am working with generates between 500 and 1000 trades per day, per currency pair and so running spreadsheets is not a practical solution. The hyphen at times cause cells to shift a column. I can't use the trade price without removing the hyphen (for fractional trades). I can't keep up with new trades coming in...I have tried.

As a non-subscriber to your service, the simulator becomes the source of all trading information. As a paid subscriber, can I bypass the simulator and pull my trade information from MBT/MBT Demo (as it is now, I see no trade activity at MBT/MBT DEMO only in the simulator).

And if the above is done can I pull the data from the simulator as well, so that I can show you a side by side with the problem? I found the concept of using the simulator a very appealing/important feature that a competitor does not have. I had hoped to compare the theoretical(chart) against real world(both demo and trading)
Finally,

I can't run the competitors software at the same time as yours. Would it help to see the MBT (but not along side yours) confirming no trade P/L issues of the like I am seeing in the simulator?

If none of this seems to make sense what do you suggest I do?


Thank you
Don

[NinjaTrader_Ray]

Unfortunately there is nothing to do...

As you have seen, I have confirmed our results are identical to MBT so your suggestion. I know you believe that you are convinced there is something wrong. I don't dispute that you believe it nor do I deny that you do experience something that I do not. Until I or my development or any other user runs into this issue that provides us an idea what may be happening there is nothing further to do.

Thanks for understanding.

[BIGRED]

If I subscribe to NT, with my trading broker being MBT, how does the simulator play a roll if at all? Being that I am having problems with it on non xxx/USD pairs (and until a solution is found) I want to use NT automation, MBT reporting and Excel to do my analysis.

Thank you
Don

[NinjaTrader_Ray]

If you have have a live subscription you can submit orders to a simulation account internal to NT or a live MBT account.

[BIGRED]

Ray,

In the case of the ...or MBT live account, does it then bypass the simulator or go through it?

Thank you
Don

[NinjaTrader_Ray]

Bypass since you are trading live. You place orders in whatever account you have selected...Live one or a simulation one.

[BIGRED]

I subscribed to NT and ran it directly to MBT so that the trading is taking place at MBT and being reported on NAV. I had hoped this would have eliminated the problem I had been experiencing in the simulator.

I am finding trade price differences and a couple of trades in the simulator that off by many hundreds of dollars. The total P/L in NAV and the simulator do not match and is off by a couple of thousand dollars.

I was not aware that the simulator would have trading information if traded directly with MBT. If this is suppose to take place then I am experiencing trade price and P/L issues when compared to MBT. If the simulator is not suppose to present MBT results... I will not pay attention to it.

If you would like pictures of all of this, Could I please have an email address to send it to, as the pictures are large.

Thank you
Don

[NinjaTrader_Ray]

The simulator can not be compared to MBT results since they are *not* the same trades. Do the same comparison that I have done. Trade in MBT live account (demo or a cash makes no difference), then NT has the identical executions as MBT since it comes from MBT and are not simulated. Then compare the results.

--> In all of our tests and the many users we have the result have always been identical.

[BIGRED]

Ray,

That is what was done. I traded in MBT and NT did not have identicle trade prices to MBT.

[NinjaTrader_Ray]

For clarification -

- You submit an order in MBT Navigator window
- The price of each or any execution in MBT does not match the execution price reported in NT executions tab in the Control Center window?

If this is the case, could you provide me a screenshot of this next time it happens and I would then provide further instructions. This should not be the case.

[BIGRED]

By far the worst trade comparison took place last night on the USDJPY. I was trading on MBT live, generated the account performance and looked under trades this morning.

I went short 1 mini lot at 99.950 and covered at 99.725, for a total of 22.5 pips. The profit reported for the one this one trade in NT was ($7,650.00).

I am going to spend my day entering individual MBT trades into Excel. I am not at all happy about it, but in an effort to keep moving forward I have little choice. Something is causing this and all though I find it impossible to believe I am the only one. I do know I am the only one reporting at this time. If in fact it is possible to have an installation take place without this issue, then there is something about the combination I have (be it hardware or software) that makes it possible to wind up with the mess I have.

I don't think the forum is the place to layout my credentials and if taken private I will be happy too. Suffice it to say that be it the Intel floating point error in the Pentium 90 or TradeStations automation switching between mini and full lot positions in 8.3 or the hundreds of "impossible to exist bugs" in between I have found that which no one else has, I know something is wrong as NT has and continues to produce p/l errors.

Because I could not get NT to run on Vista (I had to re-install it almost every time re-started it) I built a new computer with 3gigs of memory, running a fresh install of XP, all the software's required of NT. My current activities are on the XP computer.

Don

[NinjaTrader_Ray]

Please send your logs/traces via Help Mail To Support and tell me exactly the time and entry/exit price of this trade. Attention to Ray.