Julian’s Journal

[JulianCohen]

I just checked my MOC systems against the MCS for the last two months and I’m slap bang in the middle of expectations, which surprised the shit out of me to be honest

[JulianCohen]

Some immediate reversion is called for!

[SaidBitar]

Luckily I have this HERO
my system is good never misses one

[TrentRothall]

Yep, yesterday wasn’t kind

[JulianCohen]

Here’s a good tip!

When the market is closed for the day, don’t set your API up and running early in the day with the next days orders like I did yesterday.

I set the ASX orders for the 27th bright and early on the 26th and left the API running on the VPS. Of course at 3:45pm on 26th the API closed the orders and when I came to look at them on the afternoon of the 27th there weren’t any orders running.

Doh!!

[JulianCohen]

January ’17

Combined ASX ROC and SPY ROC MOC 1.82%
Combined ASX MR and SPY MOC 4.04%
Combined ASX MOC and RUA MR 3.63%
WTT US 0.88%
WTT ASX -0.57%

Overall Account 5.69%

The MR systems on the ASX were all small losers but the US MR systems were all good winners so everything balanced out very nicely for a good month.

I added a new ROC MOC system that looks quite nice and that is using the funds from the ASX ROC system in the evenings trading.

[ScottMcNab]

Great results Julian….I can’t quite get my head around how the account total is more than any of the 5 individual systems. I’m anticipating (incorrectly) that the monthly results would be an average of the 5 systems ….is it through a sharing of a central equity pool rather than treating each system with its own allocated funds ? Do you thing a shared equity pool would increase both returns and draw downs ?
Cheers

[JulianCohen]

It’s because I have been somewhat over enthusiastic with my asset allocation to each system from the pool. I effectively ended up over-leveraging at 1.8% or so.

I have adjusted everything down to more conservative levels for this month. I’d rather tread more carefully.

[SaidBitar]

Julian Cohen wrote:

I have adjusted everything down to more conservative levels for this month. I’d rather tread more carefully.

I think it is smart idea

For me it is real headache when i am thinking which strategies should be combined together and how much to allocate to each of them. Currently i am doing it based on risk.

[JulianCohen]

Said Bitar wrote:

Julian Cohen wrote:

I have adjusted everything down to more conservative levels for this month. I’d rather tread more carefully.

I think it is smart idea

For me it is real headache when i am thinking which strategies should be combined together and how much to allocate to each of them. Currently i am doing it based on risk.

Can you explain a little more about your methodology please Said?

[SaidBitar]

here is what i am doing not sure it is wise or not but sounded good at the time i thought about it .
For the WTT it has it’s own account so i do not bother thinking about it, regarding the others they are as follows:
ASX MRV1 10 pos * 10% each
ASX MRV2 10 pos * 10% each (ASX MRV2 has lower frequency than ASX MRV1)
so togther they can reach 200% of the capital in the account but it is very rare case when they are fully loaded both of them.
so they are sharing one account since in the US session i can reach 400% i am using US MOC2 with them 20 positions * 10% each
on another account i have US MRV 25 pos * 8% each and with it there is US MOC1 20 pos * 10% each
in this case i am targeting 400% on each account

but to be comfortable with this i ran back test for each system and added the results together especially the ones that are sharing the same account and i generated equity curve for the account and i checked the CAR, DD, longest DD but mainly i am caring about DD if the DD of the mixture is deeper than 30% the combination will fail and i have to combine them in another way.

I want the money to work for me as hard as i worked for it but also i do not want to have huge swings in the account. DD between 25 to 35% i can handle it but if it will start going in the 40% then it is tough.

there is nothing smart in this and if you think about it most probably it is the only way to combine them, for example in case i will trade ASX MOC then the only possible option is to have this system with the US MRV and the US MOC1.

note that all the systems are trading the total capital in the account.

[ScottMcNab]

Thanks as always Julian and Said. It seems quite a few people are using this approach so I might go and test.

[JulianCohen]

Thanks Said.

My method is similar but I’ll lay it out here.

The WTT systems are like yours with the funds accounted for separately.

The MR systems are run on the following lines. The MOC systems are 40 positions at 10% and the short term hold swing systems are 20 positions at 10%. Each pair of systems has 33% of the equity.

In each pair there is an ASX and a US system, one being an MOC and one a short term hold swing system. In theory there is a possibility that the ASX system could be fully loaded with 20 positions and then The US MOC system in the evening could then fill 40 positions leaving the system very exposed for that day, but the ASX systems are low frequency and the chance of this occurring is very slim.

Do you think this is an unreasonable risk?

[SaidBitar]

the best way to check if it is OK is Amibroker, run backtest for each and save the equity curve then create new ticker that is the summation of the equity curves of the systems that are sharing the same account and write small afl to buy and hold this ticker.

example :
WTT equity curve is ~~~WTT
ASX MOC equity curve is ~~~ASXMOC
ASX MRV equity curve is ~~~ASXMRV
US MOC equity curve is ~~~USMOC
US MRV equity curve is ~~~USMRV

~~TKR1 = ~~~ASXMOC+ ~~~USMRV
~~TKR2 = ~~~ASXMRV + ~~~USMOC
~~TKR3 = ~~~WTT
since you have equal cash for all then
the AFL will buy one position in each (~~TKR1,~~TKR2, and ~~TKR3) with same cash amount and then buy and hold the portfolio
check the report

just one comment if the MOC is 40 pos * 10% then if there are any positions that are held by the overnight MRV system you cannot achieve the 40 *10% since you will be aiming at higher than 400% so after you reach the maximum amount of cash the orders will be cancelled due to insufficient funds . that is OK but I think here you are introducing selection Bias
why you do not aim with MOC 40 pos * 5% or 20 * 10% (from my testing i found out that 5% is not good position size due to the fact it is small) .

[JulianCohen]

Well with the MOC I’m trying to maximize the CAGR as at 40*10% I’m not getting a very high MDD so I don’t mind running 40 and 10.

I realise that if the ASX MR is holding positions then in theory I should use less for the US MOC, but in fact I’m using the full allocation, so in reality I’m leveraging a little more. This is because I’m pushing the funds to work at full extent. I’m counting on a very slim possibility of every system being fully loaded at the same time, and should that rare occurrence happen, then the worst case scenario is that IB will not let me fill all the orders for one day on one or more of the MOC systems. I don’t think that will adversely affect performance

[Author]

Posts