Quote: Originally posted by admin  | I am experimenting with the idea of trade say 2 contracts, but 7 systems. So if 3 systems trade at once, you dont take the third contract.
There are lots of issues.
If 3 systems trade at once, you many not know what systems will execute.
You will tend to get higher ration of loosing trades. This is due to big (winning) moves trigger all systems, choppy moves trigger a few systems.
I think this is a bad idea if you trade 2 or more systems on 1 contract. You get less profit that 2 systems with 1 contract each, but the same draw
down.
Here is a comparison.
I suspect it will work best with say 3 contracts, and 6 or so systems.
One advance its it limits the risk on a specific day.
|
Im not doing work right now on this. Basically we just take all trades that occur if we are allowed. You cant really sort this as you just have to
take the trade if it occurs.
Should multiple trades occur in live trading, that comes down to execution method. In real terms trades are not likely to occur at the same time. They
are mill seconds apart. Any ranking at this point would have to wait (delay) a trade to see if other trades occur at the same time. Thats complex to
do, and I dont like the idea of any sort of delays in trades.
As I have a custom execution app (iblink), I can program no more than x contracts to trade. Im not using the feature right now as my focus is still on
methodology and building systems
Hi Peter, have you gotten further with the research on the above "issue"?
It is an interesting one.
Have you tested with different ranking methods for prioritise the signals when its more than the amount of contract allowed? such as perhaps rank by
average trade over x time, NP/DD, Sortino, etc..
You say that it would most likely best on 3 ctr and 6 or more systems, what is your logic with the above..
Thanks |