[TrentRothall]

Hi Scott

I am trying to workout IB DDE too at the moment just started working through it

[TrentRothall]

When combining two separate mean reversion systems I was hoping to increase trade frequency which in turn should equal more profits. The two systems are fairly similar so there are a lot of overlapping signals when they are run separately. My thinking was that if I could take the signals that triggered by system 1 then I could also take additional signals that triggered in system 2. While the number of trades increase by about 25% when the systems were combined the annual return only increased by about 2%.

So I went back and tested all the signals from system 2 that did NOT meet the entry criteria from system 1. Basically it is all the additional trades that show up in the back test when the two systems are tested together.

As I thought might be the case these additional trades are not as successful as trades that meet the criteria from system 1 AND system 2.

Winning percentage on these additional trades are about 5% lower than system 1
expectancy per trade is 0.92% compared to 1.42%
MAR is 0.57 compared to 2

So although the trades add a bit to the overall return of the system without adding to much to the maximum drawdown, and you might feel a bit better because you are placing more trades I’m not really sure if it is worth it in this instance.

It might work better if you are combining systems that don’t have as many overlapping signals, you might be able to tighten the restrictions on both systems so that you are only taking quality trades from both

[TrentRothall]

Yeah ok thanks

One of the systems i added to my original was a bit more consistent – the lower return years were closer to the overall average. i was hoping that this system might help pick up the slack of my main system when it is ‘out of sync’ with the market. it helped a bit but nothing fantastic.

The higher trades i was targeting too because like you said more trades should = more profit. The trade number went up 25% but CAR only 2%ish. so i figure that the trades are not as good quality if they dont have the original rules applied to the entry.

wrote:

The advantage of combining similar systems in my opinion are the following:
[ol]
[li]Similar over all market conditions[/li]
[li]Similar position size[/li]
[li]Similar exit mechanism[/li]
[li]easier to control priorities within systems[/li]
[/ol]

Agree here

[TrentRothall]

Yeah ok thanks

One of the systems i added to my original was a bit more consistent – the lower return years were closer to the overall average. i was hoping that this system might help pick up the slack of my main system when it is ‘out of sync’ with the market. it helped a bit but nothing fantastic.

The higher trades i was targeting too because like you said more trades should = more profit. The trade number went up 25% but CAR only 2%ish. so i figure that the trades are not as good quality if they dont have the original rules applied to the entry.

[TrentRothall]

Monday 14.3.16

1 Entry

6 Pending orders
2 Open Positions

Tested a couple of different systems that are a combination of separate MR systems, results are ok – no great improvements but i haven’t delved too deep into testing.

[TrentRothall]

Hi

Have you tred using Rate of Change? or maybe try RSI with a shorter look back period, then it might work better with a lower value. I think something like

PositionScore = -1000 + ref(yourRSI,-1);

should work i think

[TrentRothall]

Hi Said

I am also combining 2 MR systems together at the moment, i was wondering if the systems you are combining are similar or do they have alot of different signals. My systems are fairly similar ie alot of the same signals come up when running the systems separately . So far for me, it appears that most of the profit comes from trades that trigger on both systems, so while trade frequency and exposure increases (not a massive amount) when combining them, i am only gaining about 2% more on CAR and DD goes up about 1%.

I haven’t done extensive testing as yet though, but it might work better to combine systems that are more independent of each other. Interested to hear your thoughts

[TrentRothall]

Friday 11.3.16

No entries
1 Exit

4 Pending entries
1 Open position

System Weekly Wrap – to date

6 Closed trades
5 Winners
1 Loss

1 Open position

Quiet week really only 1 closed trade i think – hoping the system starts pumping out some trades so i can make sure everything is working correctly
my IB Margin account is finally setup so next week i will have access to a better paper trading setup.

[TrentRothall]

Hi Scott

try something like this

Code:

ATRp = Param(“Limit Price: ATR period”,10,1,200,5);
ATRmulti = Param(“Limit Price: ATR Multi”,0.5,0,100,.1);
ATRVal = ATR(ATRp)*ATRmulti;

BuyLimP = L – ATRval;
TickLo = IIf(L<0.10,0.001,IIf(L<2.00,0.005,0.01));
BuyLimVal = round(BuyLimP/TickLo);
BuyLim = BuyLimVal * TickLo;

LE = Ref(BuySetup,-1) AND L <= Ref(BuyLim,-1); LEPrice = Min(Open,ref(BuyLim,-1));

[TrentRothall]

Thurs 10.3.16

1 Entry

1 Exit signal
2 Pendng entries

Will be 1 open position

[TrentRothall]

[email protected] wrote:

The nice thing in this system is that the results are almost the same with and without the index filter.

Oh really? When i turn the filter on it hurts the system a fair bit. Are you testing on the XAO?

[TrentRothall]

Wednesday 9.3.16

No Entries

7 Pending entries

1 Open position

Coded up the MRV that Said mentioned in his Journal that uses the lowest low in x days. Changed it so close < LLV(L,7). Works well on ASX, pretty consistent with low DD

[TrentRothall]

Tuesday 8.3.16

Zero Entries

1 Open position

10 Pending entries

Finding that there are a fair few repeat entries, ones that don’t get filled often throw a signal up again the next day. Just an observation

[TrentRothall]

Have you calculated Max Wait of the systems to see how llong the DDs last? i guess with trading monthly some drawdowns can last years if you go to cash for ages.

PS i think if you press insert on your attachments they should be visible in the post

[TrentRothall]

I am not sure if you are entering on the open in the system above but if you are then maybe that is why there area wide range of results. With simple entry criteria like you said there would be a massive amount of signals and entering on the open means a massive amount of possible trades.

But yes i agree that simple seems to work best

[Author]

Posts