Jem's Uber Basket Trading Thread

[Jemook]

effluvium the way I see it is that there is no way to 100% fool proof a system and get every single trade right. If I get stopped out of this basket I lose 2% but I'm not that fussed when I've won 5 consecutive 2% trades in a row. It's up to you if you want to employ a safety feature like that but I'd prefer to work on other things with a system that is right so often.

Anyway my point is that if a stop-out freaks you out then your money management needs tweaking. You will never be right 100% of the time no matter what you do.

[nonlinear]

Jem's correct as usual. It is all about expectancy. We cannot be right on every trade. Even high probability entry systems may still lead to a negative expectancy. Without proper mm you can still be in the red. Are any of steve's EAs close enough to this system for purposes of testing the different exits? The ones that I have seen all look slightly different, but admit I have not been through them all.

[Commodore]

It would seem as if 'August Trading' is already setting in!...

...if this is so then I'd expect the TMA D1 values to drop to ranging. Let's see how what the market tells us over the next month...

Indeed. It will be interesting to see how D1 values react in the weeks ahead.

[tx413]

In light of today's reversals and subsequent drawdowns, I'd like to mention something I always keep in mind...

The best trading systems have EXIT signals in addition to SL / TP. After entry, they continue to monitor the market and have signals that say "the market is not going the way we expected" and exit the position. Essentially saying "that was a bad entry signal."

This applies to manual and automated systems as well. Just something that might be considered at some time...

And please, I'm not trying to say that Jem's basket system isn't any good. It's freakin' awesome as we all well know. I love it and it has my full attention right now.

[forexripper]

I think I had said it before that if we have 3 good basket days it is highly unlikely to get the same results on the next 2 (exceptions are there but just as a general rule). So I f I have 2 bad days initially and trends remain strong I enter again and expect next 3 to be successful. Now, am I following my own rul. Noooooooooooo!!! I should and will actually.

[Zool]

wingate, this is brilliant! We are trading the *basket* -- the sum of profits of all pairs. To date we've only had Baluda's iExposure, Tickets "what is the P/L for each pair now" indi. But this approach lets us see the equity curve of the entire basket!!

I would modify the indi to take "long pairs" and "short pairs," and to calculate & sum up the #pips open profit in all pairs. That would be an exact analogue for what the NB basket systems use. In fact, now that I think of it, we might not even have to mess with offline charts! I'm not certain but it might be possible to write an indi that does the above -- sum up the open profits for each of the basket pairs -- and plots that in history. I think that would be a fabulous tool for understanding the basket behavior!

Of course, if you calculate only the open trade profit, you couldn't use it as an entry signal like you are. Hm. Maybe just summing up the prices is the right idea. That would definitely be do-able in an indi, without offline charts. By summing them together you hide the behavior of individual pairs, so you reduce your ability to cherry-pick the best pairs to enter -- but really, with this basket approach, the assumption is that they're all going to run together anyway. Seeing a total-basket equity like that would make it really easy to see when the basket as a whole has its pullback and resumes the trend. I like it!!

EDIT: I bet you could even adapt Baluda's SlopeValues code to select out the pairs >X and <-X and have the indi automatically plot the "equity curve" for all the right pairs.......

I wasn't able to follow the forums in the last 2 weeks, and reading through the 500+ replies in this thread, and reached this post. Earlier int Sehlley's EA thread I mentioned the indicator Squalou made for a correlation strategy. It can be used to monitor a 2 pair basket's profit (in the correlation strategy you buy one pair and sell a highly correlated other one). It puts a vertical line on the chart and it can be used to mark the entry point of the basket, and shows the profit on a line graph in an indicator window.
Maybe it can be modified to use more pairs
I attach a screenshot off the indi, and the mq4 file.

[Zool]

I made a little backtest on the basket profit indicator I posted. Since it can be used only by 2 pairs I've chosen 2 pairs with high slope values, and searched some entry points for this 2 pair basket. I've marked the entries by blues and green vertical lines.

The basket profit indicator is on the bottom chart. The black line is the equity curve. I marked the zero balance line too with horizontal lines (where the vertical line and the black line intersects, there we start with the basket at 0 profit).

Here comes the interesting part: I've added a SMA (period 5 shift 1) to the equity curve, and the exit point would be the point where these two crosses each other. I've marked the exit points with red vertical lines. With these entries and exits the first basket's profit is around 228 pips, and the second basket's profit is around 100 pips - 2 trades 328 pips. With a 50 pip adaptive take profit (OOB ShelleyNow setting) only 200 pips would be made, that's 128 pips less.

I don't know if it's possible to modify this indicator to use more pairs, but that would be awesome if someone could make it Maybe tomorrow morning I will look into the code, but I'm really novice in coding these kind of programs (I have coding experience in sql and that's really a different language).

[garyfritz]

Nice work zool!

I've looked at the code, and making it handle more than 2 pairs would be a total rewrite.

I'm no expert in the basket-management area, but my first basket went a fair ways into drawdown, then reversed and went 300 pips (50 pips/pair) into profit. So exiting on an MA hit wouldn't always be a good thing, but it would preserve your capital well as long as you didn't hit it too often/early. It would have been very useful in my next basket, which took out the SL for an 800 pip loss.

We'd have to take a look at a lot of examples to see if it would be a winner overall. I suspect it might. 5ma might be a little tight.

But... I don't think the MA really provides anything that an appropriate adaptive Jumping Stop couldn't do. The JS creates a step-function stop on the basket:

JS.gif

Seems like that would do the job just about as well, and we've already got that.

[Jemook]

Zool, personally I'd love to see an indicator that graphs the open equity by magic number. That way I can test multiple baskets for patterns. If anyone wants to have a crack at this be my guest!

The best trading systems have EXIT signals in addition to SL / TP. After entry, they continue to monitor the market and have signals that say "the market is not going the way we expected" and exit the position. Essentially saying "that was a bad entry signal."

My first basket in a long time got stopped last night, these things happen! It's funny to see the old trader inside me want to immediately 'fix it' and change the strategy but I know better than that now. I do have a few ways of telling me when a basket is about to go against me but I won't be implementing them unless I can prove my findings. Anyway just wanted to say that I am thinking along the lines of what you said but my point is that I will not change a single thing on the actual strategy until I have seen it proven over time.

[garyfritz]

Nice to know The Master got whacked last night too. I feel a little better.