FYI, there are numerous code blocks with explanation available here: https://www.prorealcode.com/snippets/

Hi,

• There is essentially nothing wrong with (Close – Close[1]); it shows the change in price…

However, it is not the same as an actual position…

For example, a real position is opened at the “Open” of a candle and involves slippage, costs, etc…

• Values of variables (for example, MyBalance) are carried over from one bar to the next…

PositionPerf is not suitable for your purpose, as it is a ratio/percentage and therefore not expressed in monetary value or points…

However, perhaps PeterSt could show how he intends to calculate the Open PnL using PositionPerf?

This is the code to calculate the Profit or Loss each bar (when OnMarket):

// this line detects when a new Trade has been opened (in order to reinitialise some values later)
NewTrade = (LongOnMarket AND ShortOnMarket[1]) OR (LongOnMarket[1] AND ShortOnMarket) OR (OnMarket AND Not OnMarket[1]) OR (OnMarket AND (StrategyProfit <> StrategyProfit[1]))
//
//    Profit Calculations when On Market (updated to the current bar, i.e. the bar that just closed)
//
IF OnMarket THEN
   Pips  = PositionPerf * PositionPrice / PipSize            // if you need PIPS
   Money = PositionPerf * PositionPrice / PipSize * pipvalue // if you need MONEY
ELSE
   Pips        = 0
   Money       = 0
   CurrentGain = 0
ENDIF
//
// To tell the Gain or Loss in the c utrent bar
IF OnMarket THEN
   IF NewTrade THEN
      // in case it's a new trade, there's no prior value to subtract
      CurrentGain = Money                  // or CurrentGain = Pips
   ELSE
      // subtract the prior value from the current value
      // if negative, then it's a loss
         CurrentGain = Money - Money[1] // or CurrentGain = Pips - Pips[1]
   ENDIF
ENDIF
graph Pips
graph Money
graph CurrentGain

JS wrote: PositionPerf is not suitable for your purpose, as it is a ratio/percentage and therefore not expressed in monetary value or points… However, perhaps PeterSt could show how he intends to calculate the Open PnL using PositionPerf?

I take it that Roberto’s example sufficiently speaks ?

But similar to what I said before, his example doesn’t contain all. For his example the easy to understand lacking entity is NumberOfContracts. Less visible elements are in fact all the elements that a symbol can bear in the realm of points (they are really *all* necessary, depending on the instrument and also broker (like IB vs IG). Then there are also plainly lacking elements like ContractSize (this lacks in PRT but really is necessary too – try Forex in BB vs IG).

As far as I recall the good example of meeting all the elements in one instrument (type) then try Futures of Currency – and then within the IB environment.

Of course one should not omit the Commission which is in order for IB, which is a very special pain because it is in Money while often you will need it in Points. Or the other way around : you use the Commission in Money and thus need the other amount(s) in Money too. Then there is a difference between the Open Position and just the Portfolio because the Open Position formally contains only half of the Commission which Backtest really applies (IB) (without you knowing). More pain comes in once you know or notice that the closing Commission is applied by PRT by means of a dummy small trade after the original position was closed, which make equations with StrategyProfit fail (did it change ? then …).

Apply this all to in-bar trades where your code can’t reach the exit and entry (Pending Orders – Limit/Stop) and by the time you solved that too, well, you may have lost some hair. 🙂

Hi Roberto – not to hang out the teacher (most will be obvious for you) :

A Pips/Points based solution is *not* allowed to use CountOfPosition (the number of contracts involved). This is because your whole “regime” will be based on points. Commonly this regime is applied throughout, where profit and exit choices etc. are based on points. This is then not related to the number of contracts in the open position.

If MyProfitPoints > 10 then
   // Exit
endif

In the end this is so because indeed PositionPerf is based on a percentage – see the documentation. Btw in itself I find this hard to all grasp, especially if the understanding is required when for example converting to Money. For fun you could indeed try Commission as a profound example, which is of course normally not in order for IG, so never mind for most people. Still, would you focus on needing to earn at least the commission (otherwise you can’t make profit), then the system of Spread comes down to the very same : you would need the Spread in points.

And so the Money regime may come more handy because you don’t need to deal with these conversions. Well, that is what one would think, because it depends;

When you want to earn 10 points because 10 points is equal to USD 500, then make that 11 and the profit you will make (before going out) will be 550. This is sufficiently easy to work with. But when you trade with the microstructure of the instrument, you will really need the TickSize of say 0,25 (the minimum step in points) or else you really won’t know what you are tuning. Thus, you would not want to do something like

If MyProfitMoney > 12.50 then   // This implies more than one tick
  // Exit
endif

because it implies that > 15 would also depict logic – which is not true because the next “step” will be > 25 (2 ticks). And thus will this be the more convenient setup :

If MyProfitPoints > 0.5 then   // Points/Pips. 1 Point = USD 50.
  // Exit
endif

Thus any Parameter involved (Optimisation) will increase e.g. a take profit level by means of steps of 1 (tick). And not by means of some random amount of money.

In the end in a self-respecting system both regimes will be used, I think because working with Points (Pips) throughout is far more convenient and decent (portable to other instruments with the proper parameters). This, while Money is just required because of the cost element which is the most crucial to about any system, *if* only a somewhat higher frequency of trades occur per time unit (like 20+ per day) – and the lower the profit per trade, the more crucial/impacting the cost element is (the cost element is fixed per number of contracts involved).