my journal 2

Status
Not open for further replies.
automated systems' rankings

In preparation for next week, when I'll take the loan, I am screening which systems I should be using, according to several parameters. With the exception of the high performance opening gap systems, the rules will be more or less: a system has to be in the top 50% by forward-testing performance in:
1) profit by trades
2) profit by maximum loss
3) profit by maximum drawdown

Snap1.jpg

View attachment Book1.xls
 
Last edited:
Re: self-control and automated trading updates

As far as discretionary trading, I've lost more money, and it looks like i will have to take that loan so I can quit it altogether, and resume my automated trading.

Once I invest the loan's money and start investing with my automated systems, I better make money fast because in less than two months, I have a holiday coming up with my highschool friends, and I'll need to run my systems from a server overseas (at a friend's), which will cost me about 1000 dollars, as he'll have to buy a server and run it for me.

Or just stop trading and enjoy the holiday?

Or is it a case of no money from the system, no holiday?

Coop
 
Last edited:
No, this is a totally rational decision: i need the forward-testing to go on for those two weeks of holiday. I cannot stop the forward-testing, ever, whether I am using real money or not.

Pretty soon I'll try automating it so well that I will only need to restart it once a week. The major problem to solve for that, is to keep TWS from shutting down once a day. There's an application that someone built that does the job: they talk about it here. Once my friend's server is running and this little problem is solved, I will only have to start it on monday morning.
 
Last edited:
No, this is a totally rational decision: i need the forward-testing to go on for those two weeks of holiday. I cannot stop the forward-testing, ever, whether I am using real money or not.

Pretty soon I'll try automating it so well that I will only need to restart it once a week. The major problem to solve for that, is to keep TWS from shutting down once a day. There's an application that someone built that does the job: they talk about it here. Once my friend's server is running and this little problem is solved, I will only have to start it on monday morning.

My partner and I had the same problem. He wrote a little java app that clicks the login button when required. However, IB are daft enough to sometimes change the text on that button, so you know what happens then!

On another note, when selecting your systems, why not include consistency? After all, some models have returns coming from very few trades. those can scare you as you are never sure if the next equity jump will be a fall. One simple way to do this is to look at the median number of trades / week (to ensure a spread) and also measure the sharpe ratio. This generates a number based on the smoothness of your returns, and should say quite a bit about consistency.
 
Thanks, I'll keep it in mind, even though I am not familiar with some of the things you mention, and in general I am not good with formulas, so I'd rather keep it simple and do well something basic, rather than do poorly something more complex (like sharpe ratio, which is too complicated for me).

Make sure you check out that TWS START application I mentioned (see link), because it will solve all your TWS auto-logoff problems (if I am assuming correctly that you had any).

Getting back to how I will select my best systems, in order to be able to start with a small capital (from the loan) next week, I will consider, in order of importance:

1) how much they made in the forward testing period
2) what's the biggest loss in the forward testing period
3) what's the maximum drawdown in the forward testing period
4) what's the average profit per trade in the forward testing period

I will not consider the back-tests, because things get too complicated and even unreliable to some extent. I will worry about consistency once I've got all of the above figured out. I will now proceed to review all systems, one by one, with comments on each one, and decision on whether they pass the test or not.
 
growing up gotti

I finished watching all the episodes of growing up gotti. Very well made show. The family was interesting, not an ideal family, I wouldn't want to live with these guys, but it was a very interesting show. Well made because of the crew, director and stuff. Very well made.

http://www.youtube.com/user/walker88337#p/u
http://www.imdb.com/title/tt0405545/
http://en.wikipedia.org/wiki/Growing_Up_Gotti

My favorite clip is this one, hilarious:


Especially the last part of the video, where a lady who keeps on getting her apartment burglarized writes to victoria for help.

This is a great parody of the show:

 
Last edited:
selection of best automated systems

Getting back to my selection of automated systems to trade with a small initial capital.

In choosing what's best to trade, I will use these three ratios (all from forward testing):

1) profit / max loss
2) profit / dd
3) profit / back-tested dd

Snap1.jpg

View attachment Book1.xls

All these systems will be traded:

Because of their forward-testing performance:
GBL_ID
CL_ID
EUR_ID_6
YM_ON
CL_ON_3
CL_ON_2
EUR_ID_3
EUR_ID_2
JPY_ON
ZN_ON_2
ZN_ID
CL_ID_2
GBP_ID_2

Because of their back-testing performance BUT little forward-testing data (they trade rarely):
GBL_ID_2
ES_ID_3
YM_ID_2
 
Last edited:
another step ahead in measuring performance

I created an excel formula for my profitability index:

= - FT profit / AVERAGE(FT dd, BT dd, 2*max FT loss)

It translates:
Forward-tested profit divided by average of backtested and forward tested drawdowns and largest loss (the latter is multiplied by 2, in order to have the max loss count 50% in the index).

The systems allowed are the same as in the previous post, but now i have a very simple index that will always tell me what's good. I am adding the three Opening Gap systems because, even though I have few forward-tested trades, the back-tested data is too good not to trust them and trade them.

I am getting rid of one Weekly Bias system, because even though it looks good, the potential loss is too big.

Here's the new stuff:

Snap1.jpg

View attachment Book1.xls

 
Last edited:
more progress on identifying my best systems

Here's what i've found out. If I require my systems to score above 2 on that profitability index I've written about, the systems that pass the test are only 7: 3 CL, 3 EUR, and one GBL (the best one).

These 7 systems make a profit of 90k, which is two thirds of the total profit. Just one sixth of systems make two thirds of the profit. This speaks in favor of using those seven systems, regardless of their drawdown. Hoping to get lucky and not encounter their worst drawdown all at once: -5000 dollars. If I will, I will not have lost everything, but I won't be able to trade them all at once any more. I'd get from 13k to 8k, and only be able to trade most of them, but not all of them. On the other hand, if I don't come across their worst drawdown (so far, in seven months of forward tests), then I will keep trading just those systems, and will probably double my money pretty fast, since, combined, they make about 13k per month. On my first month, I'd double my capital, with a somewhat reckless money management. On the following months, I'd keep my investment constant, so I'd be making only 50% (13k on 26k is only 50%). That is IF I get lucky and I don't lose everything on my first month, which could happen if the maximum drawdown is exceeded. Then I'd have to pay a debt for the next two years.

Snap1.jpg

Snap2.jpg

Now, an interesting question that arose in my mind is: what would happen if I traded the worst systems, those scoring less than 2 on my profitability index. The answer is I'd make a profit of 55k using 33 systems, which is not bad, for the sake of diversification, but which is much worse than 90k with 7 systems. The drawdown would increase from 5k to 13k:

Snap3.jpg

Now, what would happen if I traded them all together? Profit of course would be over 140k, the sum of 90k with 55k. Drawdown I would have predicted it to be an average of the two groups. Instead it increases actually to 14k, which means it happens in the same period for both groups. On the other hand drawdown does not happen exactly in the same period, or the new drawdown would be 5k plus 13k which is 18k. So it would make sense to say that, in percentage terms, drawdown decreases the more systems you run. In absolute terms it increases, but that may not always be the case. Very good non-correlated systems could very well eliminate all drawdown and never have a negative day. I would be happy with never having a negative week.

Snap4.jpg

Needless to say, if all this works out, I am planning to quit discretionary trading altogether. After trying for over 12 years, I have found that the unknown factors that can affect your profit are too many for me to keep them all under control, and something will always get away from my control and cause me to lose everything. I cannot control my mind well enough to be profitable at discretionary trading.
 
Last edited:
Reading this thread I'd recommend two things.

1) Don't start until after your holiday. Have a break, have some fun, clear your head and you will come back more focussed. Seriously if you are going to do this you need to be on it 100%.

2) Don't go all in at first. I'd start with at most 50% of the money availabe until you at least a few months under the belt. If there has been no foirward testing there will be issues and bugs. You don't want to be taken out by a stupid mistake.
 
Of course I can't expect you to read all of my 2000 posts, so let me sum it up: I've done forward testing, almost 8 months of it. (I write about it every other post). I will take into consideration the other things you wrote. Thanks for the advice. Any feedback is appreciated.
 
Sorry - didn't realise. I thought you were discretionary trading right now. If the systems are fully tested then yes ignore point 2. I do mean fully tested even down to opening and closing the trades live - not just calls - you don't want some strange bug opening a trade that is far too big, or opening the same trade 100 times and trashing your account by mistake.

I still say take a break though - if you have had a hard time in the markets recently it will do you the world of good. you will feel so much fresher and ready for it when you come back.
 
Last edited:
Hey, it's ok. It happens to almost everybody. I am always complaining about you guys not taking the time to read all my posts, but I've learned to accept it. Thanks for the advice, even though everything was done safely as you are recommending. Other than that, taking a break would be good, but I need to make some money now. I want to. Besides, automated trading is quite safe: I am not even required to do anything, so there's little danger of screwing up.
 
regarding your systems and choosing which ones to trade: are you sure you are not curve-fitting? If you pick the best ones, they might have put in their peak performance already and the less good ones could be revving up to show some good returns now. I depends on how closely related your systems are. Are the systems all variations on the same theme? If so, then it smells really strongly of over-optimisation. If the systems are based on different algorithms then sure do it.

Another thing to think about. If you don't have enough money to trade all your systems all the time, you could consider trading them all, but only taking trades when you can afford the margin. If you are marginned out, then any further signals have to be ignored until you close a position or have made some extra profit in the account. For some systems this worked - e.g. the Turtle system. But I would imagine the simulation and back-testing is difficult to run to see if it would work.
 
I have heard similar objections before. Thanks for the advice and the feedback.

About the risk of overoptimization, I will say this. As I mentioned in a previous post, we're basically assuming the past will repeat itself, in both back-testing and forward-testing. It's unheard of that you take a bad system because you expect it to do better in the future, and discard a good system, because you expect it to do poorly. You rather do the opposite: use a system that's been doing well, because you expect it to continue.

Regarding my systems all being different versions of the same algorithm, it's partly true: those 40 systems are based on about 8 different ideas and their variations and customizations (settings change depending on which future the system trades). But even if it were totally true, which is not the case, I wouldn't say that it is a cause of overoptimization and unprofitability.

About using all systems as far as margin allows instead of selecting the best ones, what I said above applies to it as well. I believe some systems are better than others, because they've done better in these past 8 months, and I believe they will continue to do better.

But yes, if I had the money, I would definitely use all of them together, for the sake of diversification. Maybe I'd just exclude those that didn't make any money.

I could be wrong, so thanks for bringing it up and testing my ideas and convictions. But right now I am pretty confident about this.
 
Last edited:
self-control update

It's somewhat getting better. At least by the values I assigned to each of my rules. I don't feel any more stable than I was before. Or maybe yes: the lack of trading (I am still in a multi-day trade and planning to take a loan and resume my automated trading) has affected me in a positive way. Not trading discretionary, whether profitably or unprofitably, is preserving me from a lot of destabilizing stress. If there's destabilization, then I can refrain from giving in to more urges. In other words, resisting some of the urges on the list helps me resist the others.

Snap1.jpg

2.JPG

Another interesting thing is that if I buy sweets (cookies and similar), I get more satisfaction than by smoking and drinking together. This way I still hurt myself, but maybe less than by smoking and drinking (together), and also I can keep my self-control rating higher, as a consequence. Of course, during the summer, I get satisfaction just by swimming or playing frisbee, so my score then is even better. Essentially this scale/rating is not merely on self-control but rather on my ability to refrain from actions that hurt me (and in keeping others from hurting me) by using self-control. Of course now I am feeling my belly getting bigger every minute, as I eat the cookies and the other sweets, so I that is not a permanent solution or I am screwed worse than if I did other things.

Anyway, monitoring self-control is a useful thing to understand and improve myself. Of course there's other people on this forum, writing healthier journals, about their profitable trading in the meanwhile. I envy them. They not only manage to control themselves, but even to control themselves while trading. I am at a prehistoric stage compared to them. I have to accept my limits and build on what I have, and do my geeky self-control charts and everything, while brad pitt is out screwing beautiful women and stuff. After all, all these successful actors who have it all, deserve it, because they've mastered their self-control by giving it their best shot when it was needed. I didn't, I played it safe, and here I am: having achieved nothing.

I always chose the comfortable choice, the easiest route. I never wanted to risk looking bad, or failing at something, so I never went for a long shot, in everything. Jobs, women... everything had to come to me. The consequence was that I rarely failed but I achieved nothing. Those other guys who went for long shots, might have failed more than I did, but when they succeeded they got much more than I got with everything I've achieved. I guess we could define this in terms of risk/reward and % of wins. I had few losses but small and rare winning trades, they had huge wins and many small losses. I don't like taking losses, so I couldn't use their method unless I changed.

Ever since I was in junior high, teachers wrote on my report that I did well, but that I could have done better. I never tried. I wanted to win without trying my best, as if there was to be ashamed of giving it your best shot, as if I lost my composure by giving it my best shot. Instead I simply didn't achieve much. Even now I can't change it. If someone asked me to learn to dance, I'd say "no, man, I don't want to look ridiculous", i don't want to risk failing at something. The opposite of a "go-getter" essentially. Rather than "go-get it", I've always been a "stay-forget it" kind of person. I've never begged anyone for anything. I've never fought hard and I've always disliked the saying "no pain no gain".

In fact, when there was pain involved, I've always declined the activity. "Polar bears" activities (getting up early to go swimming in icy waters) and similar crap was never for me. I remember i was in a cycling race once, and there was a hill, which was quite steep. And all of a sudden i said to myself "what the hell am I doing this for?", so I stopped pedaling, got off my bike, went back down, said bye to everyone and went back home. The next day on the highschool newspaper, they had the list of the race and they wrote "travis bickle quit" or something like that. I've never found out for sure if there was any irony in writing it or not. I guess americans don't have much irony and had to write down everyone's name as a rule. But yes, basically I've been a quitter at everything they've imposed on me. I've never quit things I chose to do simply because I only chose to do things I was going to finish, so I am not a quitter literally, but I am someone who rarely begins doing anything, which maybe is even worse.
 
Last edited:
I don't think you are that different from most people. Not many people enjoy facing the pain. I know a few who do - but they really are few and far between. I don't think it is that quality that makes people good at something though - it's usually persistence and hard work doing something within their comfort zone that pays off more than anything. The autotrading looks promising, has obviously taken a lot of work and is in your comfort zone - so keep going at that and you may surprise yourself.
 
Last edited:
Re: self-control update

It's somewhat getting better. At least by the values I assigned to each of my rules. I don't feel any more stable than I was before. Or maybe yes: the lack of trading (I am still in a multi-day trade and planning to take a loan and resume my automated trading) has affected me in a positive way. Not trading discretionary, whether profitably or unprofitably, is preserving me from a lot of destabilizing stress. If there's destabilization, then I can refrain from giving in to more urges. In other words, resisting some of the urges on the list helps me resist the others.

View attachment 78554

View attachment 78556

Another interesting thing is that if I buy sweets (cookies and similar), I get more satisfaction than by smoking and drinking together. This way I still hurt myself, but maybe less than by smoking and drinking (together), and also I can keep my self-control rating higher, as a consequence. Of course, during the summer, I get satisfaction just by swimming or playing frisbee, so my score then is even better. Essentially this scale/rating is not merely on self-control but rather on my ability to refrain from actions that hurt me (and in keeping others from hurting me) by using self-control. Of course now I am feeling my belly getting bigger every minute, as I eat the cookies and the other sweets, so I that is not a permanent solution or I am screwed worse than if I did other things.

Anyway, monitoring self-control is a useful thing to understand and improve myself. Of course there's other people on this forum, writing healthier journals, about their profitable trading in the meanwhile. I envy them. They not only manage to control themselves, but even to control themselves while trading. I am at a prehistoric stage compared to them. I have to accept my limits and build on what I have, and do my geeky self-control charts and everything, while brad pitt is out screwing beautiful women and stuff. After all, all these successful actors who have it all, deserve it, because they've mastered their self-control by giving it their best shot when it was needed. I didn't, I played it safe, and here I am: having achieved nothing.

I always chose the comfortable choice, the easiest route. I never wanted to risk looking bad, or failing at something, so I never went for a long shot, in everything. Jobs, women... everything had to come to me. The consequence was that I rarely failed but I achieved nothing. Those other guys who went for long shots, might have failed more than I did, but when they succeeded they got much more than I got with everything I've achieved. I guess we could define this in terms of risk/reward and % of wins. I had few losses but small and rare winning trades, they had huge wins and many small losses. I don't like taking losses, so I couldn't use their method unless I changed.

Ever since I was in junior high, teachers wrote on my report that I did well, but that I could have done better. I never tried. I wanted to win without trying my best, as if there was to be ashamed of giving it your best shot, as if I lost my composure by giving it my best shot. Instead I simply didn't achieve much. Even now I can't change it. If someone asked me to learn to dance, I'd say "no, man, I don't want to look ridiculous", i don't want to risk failing at something. The opposite of a "go-getter" essentially. Rather than "go-get it", I've always been a "stay-forget it" kind of person. I've never begged anyone for anything. I've never fought hard and I've always disliked the saying "no pain no gain".

In fact, when there was pain involved, I've always declined the activity. "Polar bears" activities (getting up early to go swimming in icy waters) and similar crap was never for me. I remember i was in a cycling race once, and there was a hill, which was quite steep. And all of a sudden i said to myself "what the hell am I doing this for?", so I stopped pedaling, got off my bike, went back down, said bye to everyone and went back home. The next day on the highschool newspaper, they had the list of the race and they wrote "travis bickle quit" or something like that. I've never found out for sure if there was any irony in writing it or not. I guess americans don't have much irony and had to write down everyone's name as a rule. But yes, basically I've been a quitter at everything they've imposed on me. I've never quit things I chose to do simply because I only chose to do things I was going to finish, so I am not a quitter literally, but I am someone who rarely begins doing anything, which maybe is even worse.


Jesus, this is honestly like you've written about me :eek:
 
Status
Not open for further replies.
Top