Performance Stats: real MAE and MFE ?

[koganam]

Once the trade is closed the stats for the first contract are recalculated based on what happened after the contract was exited. (i.e., the MAE for the first contract actually includes a further adverse excursion that occurred after that contract was closed out). The same goes for MFE.

A direct result of using the entry as the delimiter of the trade, rather than the exits. If one uses the exits, then naturally, the number of trades is the number of exits, and calculations will, as I demonstrated, not adjust anything about the previous trade, that happened before the current exit.

Which, as my example clearly demonstrates, cannot happen if you use the trade exit as the trade delimiter, because calculations ignore, as they should, what happened to contracts that are exited before or after any other contract is exited. And to boot, it would be the only way to track per contract, as one is always only analyzing what happened to the most recent contract that was exited.

[koganam]

That approach will of course give the price MAE/MFE for each contract, though there is no way of deriving the position MAE/MFE from the final per contract results.

Actually it will. Those values are the max values of MAE/MFE while the trade was in effect, however one chooses to delineate the trade. If one insists on treating a scale out trade as one trade based on the entry rather than as the logically multiple trades based on the exits, one then has to track all the scale outs until the initial entry would be flat from all the exits, and then take the relevant max values, only after the trade is flat, or alternatively track and evaluate that max value until the trade is flat.

Hence, in the example trade that I gave, the MAE would be 2 and the MFE would be 8. Truth to tell, that tells me nothing about the original 3 contract entry position. After all, in this particular case, when both MAE/MFE were met, there was only one contract in play, so why would I care about the 2 contracts that I had exited.

Regardless, the point really that I was making is that one may have to have to track this oneself for the simple reason that each of us determines how we see trades, and how we look at what is a drawdown. For example, it is perfectly valid to look at MAE/MFE to be the total at the time they are hit with however many contracts are still in play at any given time. That would change the arithmetic of the example that I gave, albeit, the results in that particular case would still show the same max values.

If you wish, I am perfectly willing to look at a more complex scale out example with you on here. I am sure that for those of us who are not trading geniuses who never lose, we shall all benefit from a deeper look at these ideas.

Thanks for forcing me to take another look at this whole question. I was not quite as done with it as I thought.

[palinuro]

[QUOTE=koganam;445193 If one insists on treating a scale out trade as one trade based on the entry rather than as the logically multiple trades based on the exits, one then has to track all the scale outs until the initial entry would be flat from all the exits, and then take the relevant max values, only after the trade is flat, or alternatively track and evaluate that max value until the trade is flat.)[/QUOTE]

Yes, that's what Position MAE/MFE does - it's also a perfectly logical method. Without it you don't ever know the actual maximum risk of the position.

In any case, my point is that there are two ways of calculating MAE/MFE, and NT8 currently fails to do either unless you trade all-in all-out. It should do both. I have code for doing both for NT7, and I can obviously port it, but it seems to me correct calculations of such basic stats ought to be built in.

Moreover, as indicated in the examples from my last post, the MAE/MFE calculation is in any case mathematically wrong.

[koganam]

Yes, that's what Position MAE/MFE does - it's also a perfectly logical method. Without it you don't ever know the actual maximum risk of the position.

In any case, my point is that there are two ways of calculating MAE/MFE, and NT8 currently fails to do either unless you trade all-in all-out. It should do both. I have code for doing both for NT7, and I can obviously port it, but it seems to me correct calculations of such basic stats ought to be built in.

Moreover, as indicated in the examples from my last post, the MAE/MFE calculation is in any case mathematically wrong.

Agreed on all counts, and you are evidently tracking it yourself already anyway, as am I.