Shelley's multi-pair basket trader

[SteveHopwood]

V 1j is in post 1. The minutes thingy could not work - a bit of miscoding.

Those of you who can, go to start() and find the block of code that begins:
int h = TimeHour(TimeLocal() );

Copy this block over the existing one, or make the two changes manually. These are:
if (h == BasketSendHour && !BasketSent)
if (m >= BasketSendMinute)

Code: Select all

      int h = TimeHour(TimeLocal() );
      int m = TimeMinute(TimeLocal() );
      if (h != BasketSendHour) BasketSent = false;
      if (h == BasketSendHour && !BasketSent)
      {
         if (m >= BasketSendMinute)
         {
            ReadTable();
            LookForTradingOpportunities();
         }//if (m >= BasketSendMinute)
      }//if (m != BasketSendMinute)

Some newcomers to the thread are confused about the trading system. As a rule, folks, when you join a thread late you need to read it to know what has been going on. At the very least, you should read the thread-starters posts - especially when that is me; I tend to be a touch on the rude side to people who are too lazy to do their own reading and thinking. Here, you would have discovered that Shelley left a big bit of information out of her original method description because she thought it so blindingly obvious that it did not merit mentioning. I have added a little bit to post 1, so return to it and read the first paragraph of 'Trading' for clarification.

[vivaldi]

I want to warn for too much optimism with the use of Shelley's EA. What can be seen is this : Actually Shelley creates a super-currency, consisting of around 10 different basis-currencies. When we look at the chart (which we don't have) of this supercurrency (SC), we see probably an analogous behavior as in single currencies. This means , there will be bottoms and tops. At the moment , we are in negative territory and there are 10 downsized and 2 upsized currencies. On average, Shelley will take a short position now today in the morning. But let us look to the chart of the SC: where are we ? did we reach the trough already or not ? When we just passed it, there will be a number of days, that the SC will be recovering, but Shelley will daily advise us to shorten the SC. This will result in a number of severe losses for a certain number of days. I just want to warn you for the possibility of this event. When Shelley has not yet reached the trough, the profits will be huge , but there will also come reversal days and then some of the profits will have to be given back.
On the flipside, when reaching a top, a similar process may occur on the long side.

Vivaldi.

"The slope also has to be in agreement with the WOH 2x1. I had 3 tests running at different times - one opened 3 trades, one 4 and one 5 - all were correct."

The above remark of Spotdespot changes the whole analysis for a great deal. I had not taken this condition into account. In practice, this will mean, that reversal days will be eliminated much faster than only by application of the TMA mechanism alone. This enhances of course the winning chances for Shelley's EA.


Vivaldi

[Zypip]

Just a thought.
I realise that I was more confortable trading when the TMA slope was above/below 0.8/-0.8 but raising i.e. - in D1 for a buy trade - today's TMA slope value is above 0.8 and > of yesterday's TMA slope value and WOH 2x1 is green. There may be some issues with the repainting of the slope indi though.
Could this be added as a filter to the way Shelley is trading? Again, just a thought.
Cheers,

[spotdespot]

Just a thought.
I realise that I was more confortable trading when the TMA slope was above/below 0.8/-0.8 but raising i.e. - in D1 for a buy trade - today's TMA slope value is above 0.8 and > of yesterday's TMA slope value and WOH 2x1 is green. There may be some issues with the repainting of the slope indi though.
Could this be added as a filter to the way Shelley is trading? Again, just a thought.
Cheers,

I had a similar thought but slightly different. It seems to me that the D1 slope rising (for longs) is perhaps less important than the H4 slope rising as we are looking at more of a shorter term view of the immediate trend momentum. Maybe this is something to look at if the EA starts to have issues.

[garyfritz]

Steve, reading through the latest... my understanding is that the Adaptive stops are a "per symbol" value, since e.g. you set JumpingStopPips = AdaptiveJumpingStopPips * OpenTrades. In the default extern values you have AdaptivePipsTakeProfit=25 and AdaptiveJumpingStopPips=50. That jumping stop could never be hit, could it?

I was a bit surprised to see in JumpingStopLoss() that you increment JslPips by JumpingStopPips when PipsUpl > JslPips * 2. I always thought the JSL followed along a fixed distance from the max open profit level, but this says it only runs at about half the max open profit level. Learn sumthin' new ever' day...

[SteveHopwood]

Steve, reading through the latest... my understanding is that the Adaptive stops are a "per symbol" value, since e.g. you set JumpingStopPips = AdaptiveJumpingStopPips * OpenTrades. In the default extern values you have AdaptivePipsTakeProfit=25 and AdaptiveJumpingStopPips=50. That jumping stop could never be hit, could it?

Defaults are not suggestions. Something has to be a default. It is up to users to find their own best ones.

I was a bit surprised to see in JumpingStopLoss() that you increment JslPips by JumpingStopPips when PipsUpl > JslPips * 2. I always thought the JSL followed along a fixed distance from the max open profit level, but this says it only runs at about half the max open profit level. Learn sumthin' new ever' day...

If a jumping stop is, say, 50, then the profit has to reach 100 to lock in 50, then 150 to lock in 100 etc

[SWG123]

Shelley closed today's basket CHFJPY, EURJPY and GBPJPY at +99 pips.

[garyfritz]

If a jumping stop is, say, 50, then the profit has to reach 100 to lock in 50, then 150 to lock in 100 etc

JslPips initially gets set to BreakEvenProfitPipsLock. If we say BreakEvenProfitPipsLock is 60, then when PipsUpl > 2 * JslPips = 2 * 60 = 120, you add JumpingStopPips so now JslPips is 60 + 50 = 110. Now PipsUpl has to be > 220 to move JslPips to 110 + 50 = 160, > 320 to move to 160 + 50 = 210, > 420 to move to 210 + 60 = 260, etc. The JSL falls further and further behind the max PipsUpl.

It's clear that you'd better understand exactly how JSLs work, because there are plenty of innocent-looking values that wouldn't work very well. E.g. if you set BreakEvenProfitPipsLock to the same value as JumpingStopPips (or lower), the first time the JSL moves, it moves to the current price (or beyond it) and you'll get immediately stopped out.

Is that what you wanted? Or did you want the JSL to trail a fixed amount behind the max PipsUpl? I assumed JSL's worked something like "if (PipsUpl >= JslPips + 2*JumpingStopPips) JslPips += JumpingStopPips;" -- so the JSL is always between 1* & 2*JumpingStopPips behind the max PipsUpl.

BTW I see you store the JSL in a GV -- but you don't in JumpingStopLoss().

[Zypip]

Hi again,

I had another run with Shelley - NY open start time - and it traded the same 3 pairs as this morning. It is still running at the mo but it looks like that it is going to miss my BE and the basket is now reversing. I thought that it could be useful if the EA could automatically exit any trades for which the TMA slope falls below the 0.8 value. I think this is connected to the issue I have raised before in which you should consider entering the trade when the slope gets above 0.8 and raises (for longs) and not enter at the end of the move.

[mackaozy]

Another great basket this morning...manually closed with over 100pips profit.