it looks like I will be utterly bored so...
...so I will do some money management stuff on the journal.
Yesterday I was saying how the systems ended their drawdown and might enter a profitable phase which will allow me to escape the office, Vito the hyperactive chimp and similar characters and annoyances.
In preparation for the moment when I will scale up the systems and the contracts, I've been reviewing 9 new potential systems, which I will now illustrate one by one here.
PSYCHOLOGICAL IMPACT OF DRAWDOWN AND BENEFIT FROM WORKING IN A TEAM
But first let me say a word about drawdown and its psychological impact on me. It was as exhausting as Vito the Chimp. How can you expect a compulsive guy like me to keep on turning the systems on day after day for 3 months, seeing them place about 130 trades, and still see no profit? The only reason I did this is because there were people I had made commitments to. These people and these commitments were a great benefit, because besides giving me the capital to trade, they taught me to trust my systems more than i would have with my own capital. In fact I was also trading the same systems with my own capital, but as soon as the drawdown started I blew out my account trying to recoup the systems' losses with my discretionary trades. So this teaches a big lesson as far as my previous concept that I could do everything by myself and that i did not need anyone's help. In fact, things were different, and with the help of other people (capital, knowledge and discipline) i went so far that i could not have gone in years by myself. The only problem is that it is not easy to find such perfect people to help you and that is why a person like me becomes antisocial and starts believing only in himself. Anyway many thanks to this journal, to myself for writing it, to the readers for reading it and writing on it, and to trade2win for hosting it, because it is on this journal that i met such helpful people.
The presently traded systems are these, with their trades:
ES_ID_3 2
YM_ID_2 2
GBP_ID_3 3
ZN_ON_2 3
ZN_ID_2 4
EUR_ID_5 12
YM_ON 27
CL_ON_2 36
GBL_ID 45
GBL_ID_2 0
They are 10 systems but one has not traded. I added it manually, because the pivot table didn't show it.
For one reason or for another, we will keep these ten systems. So let's not worry about them. Also, let's not worry now about how many contracts we will allocate to the 19 systems, because the biggest step is to decide whether we should trade them or not.
Here's the 19 systems I want to trade and their profit in the now 14 months of forward-testing (they're listed in descending order by profit). I will mark in
blue those we're already trading, so that we can focus on the new ones later.
GBL_ID_2 -435
YM_ID_2 -435
ES_ID_3 -41
CL_ID_3 755
ZN_ID 942
ZN_ID_2 1,367
GBP_ID_3 2,003
YM_ON_2 2,221
EUR_ID_5 2,931
YM_ON 3,444
ES_ON_2 3,713
GC_ON 4,256
ZN_ON_2 4,257
GBP_ID_2 4,665
GBP_ID_5 4,997
CL_ON_3 6,218
GBL_ON 6,783
CL_ON_2 7,971
GBL_ID 8,586
Now we can see some interesting things right away.
SECOND WAVE OF SYSTEMS SEEMS TO BE AS GOOD AS FIRST WAVE
First of all, the profit is pretty well distributed among the traded and non-traded systems, in the sense that I won't be implementing a second wave of crappy systems. This second wave of 9 systems is almost as good as the presently traded but didn't emerge until now and also it is not as good. They produce similar profit but with either more trades or with bigger losses. You know what I mean? For example, GBL_ON and GBL_ID produce similar profit but GBL_ON not yet implemented has a biggest loss of 1200 euros, whereas GBL_ID has a biggest loss of 700 euros. So this makes me consider it almost twice as good. I would allocate 2 contracts to GBL_ID but never to GBL_ON. In other cases of the new wave of systems, this is not the case at all, as for the CL_ID_3, which is a new system, and because of that, might be the best of them all (since I now know more than I knew 2 years ago, when I created the other systems).
THE EXCEPTION OF THE UNPROFITABLE OPENING GAP SYSTEMS
Second of all, one could wonder: what are those 3 systems on top of the list and why are you trading them if they lost money in one year of forward-testing? Those are the Opening Gap systems, they have a very high % of wins in the back-testing, they seem very healthy, they trade very rarely and all these reasons make me trust them despite the slightly unprofitable performance so far: they trade rarely so we still don't have a large enough sample to fail them. If they will fail for another year, I will certainly remove them.
And now a little music break:
Let's now focus on the new 9 systems, and their profit:
CL_ID_3 755
ZN_ID 942
YM_ON_2 2,221
ES_ON_2 3,713
GC_ON 4,256
GBP_ID_2 4,665
GBP_ID_5 4,997
CL_ON_3 6,218
GBL_ON 6,783
Grand Total 34,551
In fact their profit is even higher than the presently traded 10 systems. But let's now see the number of trades:
system Total
CL_ID_3 1
CL_ON_3 24
ES_ON_2 26
YM_ON_2 29
GBP_ID_5 52
ZN_ID 53
GC_ON 77
GBP_ID_2 119
GBL_ON 142
Grand Total 523
There you go! To make the same money they need to trade 4 times as much as the 10 presently-traded systems! No wonder I had not selected them.
Let's now look at the new 9 systems by strategy families:
ON bounce:
GBL_ON
GC_ON
overstretched:
CL_ON_3
GBP_ID_5
WeekDay Bias:
ES_ON_2
YM_ON_2
WITH ID trend:
GBP_ID_2
ZN_ID
CL_ID_3
With the exception of the Opening Gap, these families are already represented in the presently traded 10 systems. So we might as well list the 19 systems by family and by overall profit to see how much each family contributes. This time I need a picture because the pivot table is hard to copy paste:
Before analyzing the families, since I've had some depressing thoughts, I need another music break.
When I fall in love it will be forever
Or I'll never fall in love
In a restless world like this is
Love is ended before it's begun
And too many moonlight kisses
Seem to cool in the warmth of the sun
When I give my heart it will be completely
Or I'll never give my heart
And the moment I can feel that you feel that way too
Is when I fall in love with you
And the moment I can feel that you feel that way too
Is when I fall in love with you
http://en.wikipedia.org/wiki/When_I_Fall_in_Love
http://en.wikipedia.org/wiki/One_Minute_to_Zero
Ok, back to the families and their profitability and how the new 9 systems will fit into the families.
The most profitable family - within the 19 systems (which are basically my good systems so far) - is the ON Bounce family. Yet there are... let me post another picture which will show what I am getting at:
The picture above about says it all about the
Five Families we're going to be trading. Let's review them one by one.
The
ON Bounce (organized crime) family says there will be a bounce between tonight and tomorrow morning. It makes a lot of trades, trading very often. The trades are not that profitable but the profit always shows up, with great regularity. It's the family that trades the most, that makes the most money, the profit is small but it's steady. This family is reliable. Small edge but reliable. Since it's reliable and yet has small wins, I think all its systems should be allocated 2 contracts, or at least those with a small leverage like GBL and YM (the others are too powerful maybe).
The
Opening Gap family has not paid off yet, but it only made 23 trades so it doesn't count. When it will have made 50 trades and still be unprofitable I'll turn it off. This is a somewhat unreliable family so for sure there should not be more than 1 contract allocated per system.
The
Overstretched family is clearly the best performing family, with just 100 trades producing 15k of profit. Yet it trades rarely because the point is precisely that we need a lot of overstretchedness to trigger a trade. Since this family is so profitable and trades so rarely I think all its systems should be allocated 2 contracts or at least those with a small leverage like ZN and GBP (CL is too powerful maybe).
The
Weekday Bias family is almost as performing as the Overstretched one (but this comes more from big profitable trades rather than from a high percentage of wins, like for the overstretched). In this case we can allocate 2 contracts to ZN and EUR because their trades only last a few hours. ES and YM make the same trades, but sometimes (they are different futures) one doesn't make the trade, so we are trading them both but we definitely cannot allocate more than 1 contract or it'd be too dangerous, especially because their trades last 48 hours.
The
WITH ID Trend family is like the twin brother of the Overnight Bounce family, in that whereas one guy says there will be a bounce at night, the other guy says: whatever the trend is, it will continue during the day. It performs just like its twin: many trades, small gains, but steady. Likewise we should allocate 2 contracts to the smaller leverage futures and just one to the CL.
Basically all systems get allocated 2 contracts for the smaller leverage futures, except for the Opening Gap systems. ZN_ON_2, being so good and with a small leverage, should be allocated 3 contracts.
CONCLUSIONS
Some final thoughts can be that these 19 systems seem to be pretty reliable, and yes I have 61 (20 are still being forward-tested but not trusted because they're new) and probably it sounds like a failure that less than half of them will eventually be traded and less than half of them are profitable right now, but that's the beauty of my work. I've built so many of them that i can afford to throw away half of them (I still would forward-test them, without throwing them away, ever). Hopefully - but i am confident about this - the 30 that were profitable won't become unprofitable next year and viceversa, because in that case i really have nothing.
Later edit:
Continuing discussion here:
http://www.trade2win.com/boards/trading-journals/85510-my-journal-2-a-148.html#post1263740