Mysterious inflated profit/loss value during backtest does not match chart

[st0rm]

Glad to be of service.

Very much appreciated, thank you. I'll return the favor to the community some day

[QuantKey_Bruce]

Glad to be of service.

[st0rm]

st0rm Is that -63 column the number of TICKS of profit or oss rather than points?

Oh my. Yes, those are TICKS! Ha! I had changed the display from Pts to Ticks without realizing it.

Mystery solved. Thank you!

[QuantKey_Bruce]

st0rm Is that -63 column the number of TICKS of profit or loss rather than points?

[st0rm]

Hello st0rm,

Welcome to the NinjaTrader forums!

If the strategy is using Calculate.OnBarClose, you can use High Order Fill Resolution increase the accuracy of fills.
If the strategy is using Calculate.OnEachTick or .OnPriceChange 1-tick intra-bar granularity and TickReplay is necessary for accurate fills.

Below is a link to a forum post that discusses both.
https://ninjatrader.com/support/foru...377#post773377

If you are interested in analyzing the data to understand what is causing a difference between real-time and backtest, see the post below.
https://ninjatrader.com/support/foru...nce#post100192

Thank you. I've been on quite a journey since this post but am still experiencing the same behavior. I've tried almost everything:

- Enable tick replay, add a 1-tick series to my strategy, and place orders using the proper EnterLong(1,<qty>,<entry-name>) format.
- Enable 'High' order fill resolution

I've noticed the 'profit' column is showing 4x the true profit/loss in every case. (see screenshot below)

I don't believe this is a problem with random fills, as the trades on the chart do not contain wild bars. Could there be any other cause? This is making my backtesting very difficult. Thanks.

[NinjaTrader_ChelseaB]

Hello st0rm,

Welcome to the NinjaTrader forums!

If the strategy is using Calculate.OnBarClose, you can use High Order Fill Resolution increase the accuracy of fills.
If the strategy is using Calculate.OnEachTick or .OnPriceChange 1-tick intra-bar granularity and TickReplay is necessary for accurate fills.

Below is a link to a forum post that discusses both.

https://ninjatrader.com/support/forum/forum/ninjatrader-8/strategy-development/94098-isfirsttickofbar-vs-onbarclose-for-backtest-live?p=773377#post773377

If you are interested in analyzing the data to understand what is causing difference between real-time and backtest, see the post below.

https://ninjatrader.com/support/forum/forum/ninjatrader-8/strategy-development/100192-comparing-real-time-historical-and-replay-performance#post100192

[st0rm]

I'm noticing extreme profits and losses while backtesting which does not match the open/close prices shown on the chart for the simulated trade.

The example below shows two /NQ contracts that experienced a combined -25pt loss. The same trade appears as a -100 point loss and inflates the statistics. The trade log records the appropriate open price (11351.50 and close price of 11339), resulting in a loss of -25 points. However, the 'Profit' column shows 100 points, MAE is 100 points.What in the world? Has anyone else experienced this?

I'm using stock data, so there are no custom data errors. 5-minute calendar, standard order fill resolution. The strategy is using a 12.5pt stop loss which I'm setting in code (and which works great on the charts and in real life.

Thanks for your help!