Profit % vs CumProfit % Trades Report Differs

[fercho]

Hi I've exported trades from Strategy Analyzer report and found that Profit differs from CumProfit in a multi Instrument Percentage Report. In a few trades is not a big problem, but in 3000 trades Ninja Trader CumProfit stats could differ form summing each trade profit by 20% or more...

Bug?

Thanks in advance

[NinjaTrader_Bertrand]

fercho, this would not be a bug, but an expected difference as the cumulated profit in percentage mode would reinvest - are you saying you're seeing those differences only in a Multi Instrument report?

[fercho]

Single instrument too.

yeap also in sigle instrument, sorry didn´t notice that.. (attached)

I've spent several hours on that but can´t figure out why it happens

any idea ?

Thanks for reply.

[NinjaTrader_Bertrand]

fercho, thanks - yes because of the reinvesting aspect of the Cumulative Profit in % I would expect differences. You cannot compare both by summing the percentages individually.

[fercho]

Why reinvest ? Suppose that I buy 100K AUDUSD and I made 0.02% then again, buy 100K and loss -0.01% Cum Profit: 0.01%... Ninja will say 0.011% but I never reinvest my money

Sorry but I can´t understand.

The things is that CumProfit it's used later for calculate a lot of stats and I'll not be confortable thinking that Ninja CumProfit is not working as expected.

Thanks in advance.

[NinjaTrader_Bertrand]

fercho, the reason here is that it's potentially working counter intuitive from what you expect - here's an example we had also posted on other threads on this topic that attempts to clarify :

Cumulative Profit % mode can be thought of as if you were reinvesting your complete account including winnings back into the trade again and again.

Consider you start with $1 and you made 3 trades at profits of 25 cents, 50 cents, and 75 cents. That would mean the trade's percent profit was 25%, 50%, and 75%. What % mode does is it weighs each of these percentages in from a reinvestment standpoint.

Trade 1 = 25%. 25% of $1 = we made 25 cents.
Trade 2 = 50%. 50% of $1 + $0.25 = 50% of $1.25 = $0.625
Trade 3 = 75%. 75% of $1.25 + $0.625 = 75% of $1.875 = $1.40625

This means after all 3 trades we would have made $0.25 + $0.625 + $1.40625 = $2.28125.
Since we started with $1, winnings of $2.28125 is equivalent to a Cum. Profit % of 228.125%

You can see that 228% cumulative profit is not equivalent to Total Net Profit which is just $1.50 or 150% if taken from a summation only standpoint.

This would apply to DD calculations in % mode as well - we are aware of requests to report those items differently and track those in our product management lists, however for now this is the implementation used.

[fercho]

Thanks Betrand for your fast reply.

Well, in my personal opinion ninja complicate things. Simpler solution will be always the best.

If I'd reinvest I'd like using a Money Managment scheme, like Ryan Jones Fixed Ratio, and ninja should post percentage profit/loss in function of USER capital invested, not a Ninja supposition that every $ profit is reinvested.

Working with single instruments, I can use "Currency" as a partial solution, meaning "partial" 'cos one dollar made today from a 1,34 EURUSD cross is not the same as one dollar made in 2006 from a 1,50 EURUSD cross. So, that solution is partial, and stats lacks of precision.

If you add a Multi-Instrument scheme, things get more complicated using Ninja Currency scheme, for example in Days Distribution Chart is funny watch how you can get a Monday that loss money when in percentage terms could be in fact a winning day. USDJPY and USDCHF are far from being used in currency scheme together.

MAE/MFE Analysis as well. Is a mix of percentage Y axis and currency X axis, very confusing.

So back to CumProfit Percentage in a multi instrument forex scheme, Ninja is far to be usefull to get proper stats to work and get accurate results.

My personal opinion is that you can´t work with Ninja if you want to go further. Is a very expensive license if you compare what you really get.

Probably edition 8 takes seriously this aspect and erase the supposition about reinvest USERS money, and fix all the percentage scheme.

I'll be more than happy to pay a big price for a big tool that get my strategies forward.

By now I feel that I'm stuck, and I'll may come back to my personal excel spreadsheet to get accurate results.

Thanks a lot for your understanding and support.

Best Regards,

Fernando.

[NinjaTrader_Bertrand]

Thanks for sharing this detailed feedback with us Fernando, I will make sure to route it into the appropriate channels. So for your application, you would seek a cumulative mode that would not reinvest in any manner and just adds the individual trade results in percentage? Or provide an input on which capital / currency to base it...

[fercho]

Yeap, the right thing is to get a cumprofit percentage from individual trades profit/loss.

Reinvesting is a personal decision and is always subjet to a money managment policy system, like Delta, drawdown, trading cycle, etc.

Also Ninja should have informed at Help file about that Cumprofit situation in Percentage terms is in fact a reinvesting result.

It's not the same 1,000 usd from a 100K trade than 1,000 usd from a 10K trade. First is a 1% percent profit, second it's a 10% profit. Far away different...

So I can imagine that for Ninja marketing team, it's better showing a reinvesting CumProfit in percentage terms, than the ordinary way, results are "quite biger", aiming traders to put their strategy on realtime faster.

I hope these serious matter will be fixed as a patch asap. I can't work seriously on a strategy using Ninja in those terms.

Anyway it's my personal opinion, and it's only focused to improve a great software that probably is just one step behind to achieving a goal in the strategy development scheme.

Great Coding Platform, Great Database Managment, Great Backtesting Simulator, but unreal stats from all above at the end.

Bertrand, Thanks again for your help and patience.

Regards,

Fernando.