Shelley's multi-pair basket trader

[spotdespot]

I have to say I am a little confused too. I agree with Gary - it seems a bit odd to have a per trade TP, a per trade BE function and a static SL that effectively will give an unknown max risk/reward.

There are no per-trade tp or be functions. Some people are using mptm to create them, but that is their choice.

Sorry Steve - wasn't very clear. The adaptive TP and BE effectively makes the TP and BE per trade? 10 trades 25 pip Adaptive TP = 250 pip TP target or 25 pips per trade? Same for BE? Unless I have misunderstood?

Cheers,
Dave.

EDIT: Maybe better to have said "in ratio" to the number of trades.

[garyfritz]

Yes, it's not a separate TP for each open trade, but a basket TP that's scaled by the # of open trades -- which results in an "average" TP per position. Some pairs will have more profit and some will have less, but the basket averages X pips profit per open position. An "average" SL per position seems to make just as much sense.

Dave's right that you have an unknown risk/reward ratio if your TP is variable (scaled by positions) and your SL is fixed.

The "trades start out in drawdown due to spread" point is true for any trade, and yet SLs make sense for almost any system. If people are stupid and put their SL too close, that's their fault. Don't protect the stupid by penalizing the prudent.

[SteveHopwood]

The "trades start out in drawdown due to spread" point is true for any trade, and yet SLs make sense for almost any system.

Matter of opinion there, Gary. Mine is that sl's mostly benefit the crim, not the trader. I am a fan of trading tiny lot sizes and working out from there.

If people are stupid and put their SL too close, that's their fault. Don't protect the stupid by penalizing the prudent.

The prudent still have the stop loss features available. I am still not yet considering an adaptive one. As I am not a fan of sl's, you can perhaps understand my lack of enthusiasm for adaptive ones.

And, nothing will protect the stupid from themselves. Absolutely nothing.

[SWG123]

Fantastic. This EA has me more excited than I can remember being about one for a while. One of the main reasons NB 10.2 manually traded is so good, imo, is because it identifies the trend so well and this just piggybacks that beautifully. People have been saying this has worked so well because many pairs have been trending BUT if it only trades trending pairs, and it is rare that the markets are completely flat, now we have the adaptive TP etc this may be more consistent than people might expect. I hope I'm right

I'd like to think so too, spot. However, without any corroborating evidence whatsoever ( ) what I suspect we'll find is this - when we have 12, 14 or more pairs "qualifying" as trending (D1 TMA slope >0.8 or <-0.8), which is what has happened in the last couple of weeks, then the basket will perform spectacularly well. Why? Simply because we will only see these high numbers of qualifying pairs when there are significant global capital flows between correlated currency pairs.

The recent strong trends have resulted from large-scale liquidation of Euro and commodity currencies in favour of supposedly safer USD and, to a lesser extent, Yen acquisition (the good thing is, however, we don't really need to understand the underlying fundamentals; just looking at the composition of the basket should help to inform what's going on). My guess is that, when markets are more divergent, we will see periods when maybe only 3, 4 or 5 pairs will exceed TMA slope +/-0.8.

In these situations, without confirmation of trending across correlated pairs, I predict the performance of the basket will be a bit ... well, more wishy-washy.

All the above is based upon 0% research and 100% gut feeling, which I suppose comes from experience. See signature

[SteveHopwood]

All the above is based upon 0% research and 100% gut feeling, which I suppose comes from experience. See signature

I am tempted to add to my "Never trust my coding. I am a dimwit." signature this; "Also take serious note of SWG123's signature."

Except that you would be unable to resist changing your sig to something like, "Ignore Steve. He is a clown." Go on. Admit it. You would be unable to resist.

What I am hoping, with equally based research and gut feeling is this: Slope allows the EA to spot the trending pairs; however many or few of these there are, the adaptive trade exits will allow us to profit from them.

We shall see. Again.

[SWG123]

All the above is based upon 0% research and 100% gut feeling, which I suppose comes from experience. See signature

I am tempted to add to my "Never trust my coding. I am a dimwit." signature this; "Also take serious note of SWG123's signature."

Except that you would be unable to resist changing your sig to something like, "Ignore Steve. He is a clown." Go on. Admit it. You would be unable to resist.

What I am hoping, with equally based research and gut feeling is this: Slope allows the EA to spot the trending pairs; however many or few of these there are, the adaptive trade exits will allow us to profit from them.

We shall see. Again.

Steve - as usual, you are way ahead of me - I hadn't taken the adaptive TP into account as it's a relatively recent addition. And quite brilliant. I'm glad you adopted my brilliant idea - before I even thought of it!

[SteveHopwood]

Steve - as usual, you are way ahead of me - I hadn't taken the adaptive TP into account as it's a relatively recent addition. And quite brilliant. I'm glad you adopted my brilliant idea - before I even thought of it!

Ahh well, you see; here experience triumphs.

I am a PCA. PCAnessness conveys upon the holder of such PCAnessness the ability to read the thoughts of forum members before they have even thought them.

Very useful, as it allows me to see into the future. And the past. And the present whilst looking around corners. Or, more usefully, a glass of Scotch.

Now, I know that about 35,000 of you are already thinking, "Yes but what about that suggestion I made back in 20xx that you rudely ignored?" just bear in mind that you are reading this in whatever year it is in your lifetime that I anticipated back in 2012. It was a rubbish suggestion, OK? I hope that chasing it did not cost you too much money.

[garyfritz]

Matter of opinion there, Gary. Mine is that sl's mostly benefit the crim, not the trader. I am a fan of trading tiny lot sizes and working out from there.

Careful, or Khalid will lecture your ear off. Tiny lot sizes reduce but do not limit your losses, while similarly reducing your profits. Stops are IMHO an essential safety measure and an enabler of much higher profits. You cannot determine appropriate position sizing unless you know how much you're risking.

Too-tight stops are stupid and will cost the stupid lots of money. Sensible stops protect you from adverse events, black swan and otherwise, and enable you to take on higher position sizes with known and (short of disaster) controlled risk. Appropriately-chosen stops reduce your risk and increase your profits.

My guess is that, when markets are more divergent, we will see periods when maybe only 3, 4 or 5 pairs will exceed TMA slope +/-0.8.

And I would like to be able to specify the same sensible stop without having to prognosticate how many pairs will trade tomorrow.

[fxozgirl]

Matter of opinion there, Gary. Mine is that sl's mostly benefit the crim, not the trader. I am a fan of trading tiny lot sizes and working out from there.

I have to say I agree with Steve on this point...

and

What I am hoping, with equally based research and gut feeling is this: Slope allows the EA to spot the trending pairs; however many or few of these there are, the adaptive trade exits will allow us to profit from them.

I think the adaptive TP is going to prove to be one of the most inspired idea's ever

[Caillou]

I think the discussion can't be about SL vs Small Lots. SL is an exit strategy and Small Lots a position decision. What happens with Small Lots is that this allows you to apply a different exit strategy than SL, for example the Recovery Method, impossible with big lots.
In my opinion SL is the best invention for crims, I prefer the Recovery, more sophisticated than a SL, and more profitable IMHO.
Anyway it would be great to have here a deep discussion about Recovery, if there is enough interest, since I have not found any detailed thread in any forum.

Regards