- Status
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
Yeah, I can tell you've been reading my 4000 posts. In my trading I make a distinction between back-testing and forward-testing.
We're talking about 120 systems that have been back-tested over the past 10 years and that have already been found to be profitable during those 10 years.
Then I also require them to be profitable and have a sharpe ratio above 2 in forward-testing during at least 10 trades. I can't ask for more trades, because some of them trade 3 times per year.
I can't wait until the year 2020 to get those hundreds of forward-tested trades you recommend using.
We're talking about 120 systems that have been back-tested over the past 10 years and that have already been found to be profitable during those 10 years.
Then I also require them to be profitable and have a sharpe ratio above 2 in forward-testing during at least 10 trades. I can't ask for more trades, because some of them trade 3 times per year.
I can't wait until the year 2020 to get those hundreds of forward-tested trades you recommend using.
tony1979
Active member
- Messages
- 112
- Likes
- 22
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
weekly update
No ****.
Who was it, counter_violent, who said that maybe the systems would have bounced right after being dismissed?
Look here (forward-tested systems, not traded with real money):
Who gives a ****? They dropped me? They don't deserve me.
Now i'll go on by myself. Hopefully they won't come back several months from now, asking me some details about my systems, because once it's over it's over.
What was wrong was the money management, of which I never considered myself an expert. That should have been their task to monitor me, instead of just being busy dishing out endless daily assignments, which ended up burning me out. Nothing is wrong with my systems. As the second picture shows I have plenty of excellent systems, and they simply cannot all fail at once. And certainly just because they lose at once during the same month... this is far from showing that they have failed or that they are unreliable, as someone claimed. ****ing bull****.
Also bull**** was that the line "but now I know where i failed" might not work next time. I didn't fail jack****. That is not my line, never has been, and I am sick and tired of opening up and having to debate with people manipulating what i say to discredit me. I will still open up but I will ban all people intefering with my efforts and abusing my sincerity.
My job was to create profitable systems, I did it, and I gave them away for free, to the same people now telling me that they're not reliable (which is obviously bull****, and they know it).
That is why from now on, no matter how desperate I will be, I will not give my systems to anyone ever again.
And now I am free.
Oasis - Whatever - YouTube
For the past few months, i've been loaded like a mule, and could not even think straight from the amount of excel work i had to process, and could not realize what was going on, the mistakes we were making. In a year we went from 2 excel workbooks to 7 of them. This is what you get, when you burn someone out. If he's doing all the work, and you're on vacation, who is monitoring the process once that person, that only mule carrying all the weight, is pushed to the limits, for over a year, and he burns out? No one, and you will get failure. And what does the mule get at the end of the process? What is he told after he gave a year and a half of hard work and gave his systems away for free from the first day? He is told his systems do not work, which is a lie.
No ****.
Who was it, counter_violent, who said that maybe the systems would have bounced right after being dismissed?
Look here (forward-tested systems, not traded with real money):
Who gives a ****? They dropped me? They don't deserve me.
Now i'll go on by myself. Hopefully they won't come back several months from now, asking me some details about my systems, because once it's over it's over.
What was wrong was the money management, of which I never considered myself an expert. That should have been their task to monitor me, instead of just being busy dishing out endless daily assignments, which ended up burning me out. Nothing is wrong with my systems. As the second picture shows I have plenty of excellent systems, and they simply cannot all fail at once. And certainly just because they lose at once during the same month... this is far from showing that they have failed or that they are unreliable, as someone claimed. ****ing bull****.
Also bull**** was that the line "but now I know where i failed" might not work next time. I didn't fail jack****. That is not my line, never has been, and I am sick and tired of opening up and having to debate with people manipulating what i say to discredit me. I will still open up but I will ban all people intefering with my efforts and abusing my sincerity.
My job was to create profitable systems, I did it, and I gave them away for free, to the same people now telling me that they're not reliable (which is obviously bull****, and they know it).
That is why from now on, no matter how desperate I will be, I will not give my systems to anyone ever again.
And now I am free.
Oasis - Whatever - YouTube
For the past few months, i've been loaded like a mule, and could not even think straight from the amount of excel work i had to process, and could not realize what was going on, the mistakes we were making. In a year we went from 2 excel workbooks to 7 of them. This is what you get, when you burn someone out. If he's doing all the work, and you're on vacation, who is monitoring the process once that person, that only mule carrying all the weight, is pushed to the limits, for over a year, and he burns out? No one, and you will get failure. And what does the mule get at the end of the process? What is he told after he gave a year and a half of hard work and gave his systems away for free from the first day? He is told his systems do not work, which is a lie.
Last edited:
counter_violent
Legendary member
- Messages
- 12,669
- Likes
- 3,787
Re: weekly update
So was this the case that the equity chart recovered?
Can you update the chart as if the systems had not been turned off?
I don't think theres any surprize here considering that the 26th showed price extension extremes across a broad range of instruments !
So was this the case that the equity chart recovered?
Can you update the chart as if the systems had not been turned off?
I don't think theres any surprize here considering that the 26th showed price extension extremes across a broad range of instruments !
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
You're seeing it. I posted the chart of the forward-tested systems, which is almost exactly the same as the real money chart. Except it stretches further back. You can even compare the last parts of the chart i posted with the chart i posted previously and you will see that they match for over a month (the combination of systems traded got changed, so the forward-tested chart shows the last combination throughout the available period).
Yes, it bounced. But even if it hadn't, and if we kept falling for the next month, one still could not say, in good faith, that my systems have failed or that they are not reliable.
Yes, it bounced. But even if it hadn't, and if we kept falling for the next month, one still could not say, in good faith, that my systems have failed or that they are not reliable.
counter_violent
Legendary member
- Messages
- 12,669
- Likes
- 3,787
Well look, lets not beat about the bush here. This was exactly what I tried to convey earlier in the week. The combined systems in play at the time suffered from the LTCM effect of giving the appearance of diversity under normal market conditions, but when subjected to whole market extreme conditions, too many of the systems behaved the same way.
Is this a fair assessment or not ?
And yes...the 26th was in terms of your drawdown timing and turning off....very unlucky indeed...however as you rightly pointed out...this market condition may have persisted and gotten worse...so on balance I expect you did the right thing by turning them off as per agreement.
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
"lets not beat about the bush"? Too quick and too dismissive, if not even offensive. I didn't write 4000 sincere, detailed, explicit and transparent posts to be told that I beat about the bush. First read all my posts and then accuse me of beating about the bush. And I haven't worked my ass off for 7 years on this to be told this stuff about my systems. I didn't create 120 systems to be told that they all behave the same way.
We did not dose/choose them correctly (systems/contracts selection), and here is why. As I said several times in my last 20 posts, we over-optimized the combination of systems, by looking up hundreds of combinations and finally finding a magical combination which returned great profits with very little drawdown, which could not be reliable due to being curve-fitted.
The mistake we avoided making in the creation of systems by using the out-of-sample, we did make all the way, when we looked for a combination of systems to trade.
We did not dose/choose them correctly (systems/contracts selection), and here is why. As I said several times in my last 20 posts, we over-optimized the combination of systems, by looking up hundreds of combinations and finally finding a magical combination which returned great profits with very little drawdown, which could not be reliable due to being curve-fitted.
The mistake we avoided making in the creation of systems by using the out-of-sample, we did make all the way, when we looked for a combination of systems to trade.
Last edited:
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
done but didn't work
I have tested it over and over again.
It doesn't work.
Or rather: the automation can be easily implemented but I come up with terrible results.
Is it really what I want?
What I have done is:
1) get those above 2 of sharpe ratio
2) get those with more than 9 trades
3) multiply the contracts based on their max loss
Yes, this is a perfectly automated method.
But the results are terrible compared to what I had been used to. But the previous results were awesome, yet I think they were over-optimized and curve-fitted.
Maybe it's ok to optimize by combined drawdown, but then I must not use the max combined drawdown but all the drawdowns.
But I am really lost in the sea of possibilities.
Maybe I should focus on the combined sharpe ratio, but only of the systems that have a forward-tested sharpe ratio above 2.
So, I'd use as candidates all systems showing more than 9 trades and a sharpe ratio above 2. Then I'd allocate contracts so to get the best combined sharpe ratio.
But then I would have to use that software again. And that software is now extremely slow, now that they've made me add thousands of cells, including the calculation of margin needed.
I am really lost in the sea of possibilities.
I need an expert cook. A money management cook to advise me on the recipe. But where do I find one? Especially where do i find one who has my same exact problems?
I need to talk to my mathematician friends for this.
I have tested it over and over again.
It doesn't work.
Or rather: the automation can be easily implemented but I come up with terrible results.
Is it really what I want?
What I have done is:
1) get those above 2 of sharpe ratio
2) get those with more than 9 trades
3) multiply the contracts based on their max loss
Yes, this is a perfectly automated method.
But the results are terrible compared to what I had been used to. But the previous results were awesome, yet I think they were over-optimized and curve-fitted.
Maybe it's ok to optimize by combined drawdown, but then I must not use the max combined drawdown but all the drawdowns.
But I am really lost in the sea of possibilities.
Maybe I should focus on the combined sharpe ratio, but only of the systems that have a forward-tested sharpe ratio above 2.
So, I'd use as candidates all systems showing more than 9 trades and a sharpe ratio above 2. Then I'd allocate contracts so to get the best combined sharpe ratio.
But then I would have to use that software again. And that software is now extremely slow, now that they've made me add thousands of cells, including the calculation of margin needed.
I am really lost in the sea of possibilities.
I need an expert cook. A money management cook to advise me on the recipe. But where do I find one? Especially where do i find one who has my same exact problems?
I need to talk to my mathematician friends for this.
Last edited:
counter_violent
Legendary member
- Messages
- 12,669
- Likes
- 3,787
Travis, I am not trying to offend nor be dismissive and certainly not be offensive, what I am trying to do is get some answers to what I consider to be legitimate questions in order that I and others may be able to help in some way and explore for solutions to the problem(s)
If you don't want me asking questions...thats fine...just say so.
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
The problem is that you like asking questions but don't have the time to read my answers. I have already answered your questions before you even asked them. Then you tell me that I am beating the bush, but you don't even read everything I write.
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
I am lost. Total distraction by the people writing here, and i can't even ban some of them.
This is not working. The journal.
I cannot think out loud here.
I am lost in the sea of possibilities and implications. I don't know how to proceed, but let us record here that:
1) I know I don't have it figured out
2) I know the systems are ok, and the whole problem is with money managament and systems' selection
3) I know what is not right, which is what we did. But I don't know precisely how wrong it was, nor a better alternative.
4) I know the issues and elements at stake, but i don't know how to dose them together.
I need someone who has exactly my same problems, but I am not going to find him.
What matters now is that I do not do anything until I have figured out a univocal method for selecting systems. Proceeding by guesstimates is going to leave room for curve-fitting, which has happened and will happen again and again.
I need a math person.
I need to sit down with him for hours.
Let me read a little of that wikipedia entry again:
http://en.wikipedia.org/wiki/Modern_portfolio_theory
Are the books by these Nobel Prizes of MPT going to be any good?
http://en.wikipedia.org/wiki/Harry_Markowitz
Do I really have to read all this crap?
http://en.wikipedia.org/wiki/Present_value#Background
Why don't I just get a guy who has read it all, to tell me about it?
More from:
http://en.wikipedia.org/wiki/Harry_Markowitz
Ok, I am going to need not a book for traders, but rather a textbook for these students. That's the way to go.
Even easier way to go is watching videos:
fascinating stuff...
The famous paper they mention is here:
http://www.math.ust.hk/~maykwok/courses/ma362/07F/markowitz_JF.pdf
16 pages full of formulas. I don't like it at all, but it is the closest thing to my problems right now.
It starts like this:
So far so good. Very related to the concept of sharpe ratio. Oh, look:
http://en.wikipedia.org/wiki/List_of_Nobel_laureates_in_Economics
It actually shows that Markowitz and Sharpe won the nobel prize together. Oh, and look here:
http://www.stanford.edu/~wfsharpe/art/djam/djam.htm
Anyway, I have got "stage 1" totally figured, and I precisely want to focus on stage 2. The problem is that this paper gets more complex very quickly, at least for me and my dislike for formulas.
Well, at least I will watch these videos over and over again. It's easy:
http://www.youtube.com/watch?v=gWj3Bim7_ls&NR=1
This Markowitz guy is precisely addressing the problem of univocal selection of systems, which bbmac brought up. And he is showing why I found it so hard. You have to automate it, and you have to do it right, and Markowitz breaks apart the principles for doing it.
In the paper he obviously does not make it seem as easy as he does in the videos, so the first thing I do is focus on all the videos I can find on Markowitz, and with him or on him. This is going to be pretty easy, and i can do it.
If I fully focus on this guy I might be able to come out of this maze. I need to start working, reading, watching, and get to the bottom of this thing once and for all.
This is not working. The journal.
I cannot think out loud here.
I am lost in the sea of possibilities and implications. I don't know how to proceed, but let us record here that:
1) I know I don't have it figured out
2) I know the systems are ok, and the whole problem is with money managament and systems' selection
3) I know what is not right, which is what we did. But I don't know precisely how wrong it was, nor a better alternative.
4) I know the issues and elements at stake, but i don't know how to dose them together.
I need someone who has exactly my same problems, but I am not going to find him.
What matters now is that I do not do anything until I have figured out a univocal method for selecting systems. Proceeding by guesstimates is going to leave room for curve-fitting, which has happened and will happen again and again.
I need a math person.
I need to sit down with him for hours.
Let me read a little of that wikipedia entry again:
http://en.wikipedia.org/wiki/Modern_portfolio_theory
Are the books by these Nobel Prizes of MPT going to be any good?
http://en.wikipedia.org/wiki/Harry_Markowitz
Do I really have to read all this crap?
http://en.wikipedia.org/wiki/Present_value#Background
Why don't I just get a guy who has read it all, to tell me about it?
More from:
http://en.wikipedia.org/wiki/Harry_Markowitz
Ok, I am going to need not a book for traders, but rather a textbook for these students. That's the way to go.
Even easier way to go is watching videos:
fascinating stuff...
The famous paper they mention is here:
http://www.math.ust.hk/~maykwok/courses/ma362/07F/markowitz_JF.pdf
16 pages full of formulas. I don't like it at all, but it is the closest thing to my problems right now.
It starts like this:
So far so good. Very related to the concept of sharpe ratio. Oh, look:
http://en.wikipedia.org/wiki/List_of_Nobel_laureates_in_Economics
It actually shows that Markowitz and Sharpe won the nobel prize together. Oh, and look here:
http://www.stanford.edu/~wfsharpe/art/djam/djam.htm
Anyway, I have got "stage 1" totally figured, and I precisely want to focus on stage 2. The problem is that this paper gets more complex very quickly, at least for me and my dislike for formulas.
Well, at least I will watch these videos over and over again. It's easy:
http://www.youtube.com/watch?v=gWj3Bim7_ls&NR=1
This Markowitz guy is precisely addressing the problem of univocal selection of systems, which bbmac brought up. And he is showing why I found it so hard. You have to automate it, and you have to do it right, and Markowitz breaks apart the principles for doing it.
In the paper he obviously does not make it seem as easy as he does in the videos, so the first thing I do is focus on all the videos I can find on Markowitz, and with him or on him. This is going to be pretty easy, and i can do it.
If I fully focus on this guy I might be able to come out of this maze. I need to start working, reading, watching, and get to the bottom of this thing once and for all.
Last edited:
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
more from youtube
Harry Markowitz in Michael Covel's Documentary; Classic Line! - YouTube
Damn.
Youtube is not enough. I am going to have to start reading and taking notes, and identify the issues.
It would be better to study it with someone else. I can't usually make myself study this crap. Too much technical jargon, too many formulas.
But i need to do this. Without this I cannot go any further.
1) the systems are created
2) the funds will be there again, one way or another
3) the disciplined execution should be there, hopefully (it's been there for a year)
4) the money management according to a reliable automated formula is lacking: I cannot go any further without this
I need to start reading markowitz and his fellows boring academics, and I need to discuss this with others, in my usual simplified way.
If anything, i should proceed on the web, with wikipedia or blogs. Videos are no longer useful.
Harry Markowitz in Michael Covel's Documentary; Classic Line! - YouTube
Damn.
Youtube is not enough. I am going to have to start reading and taking notes, and identify the issues.
It would be better to study it with someone else. I can't usually make myself study this crap. Too much technical jargon, too many formulas.
But i need to do this. Without this I cannot go any further.
1) the systems are created
2) the funds will be there again, one way or another
3) the disciplined execution should be there, hopefully (it's been there for a year)
4) the money management according to a reliable automated formula is lacking: I cannot go any further without this
I need to start reading markowitz and his fellows boring academics, and I need to discuss this with others, in my usual simplified way.
If anything, i should proceed on the web, with wikipedia or blogs. Videos are no longer useful.
Last edited:
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
googling portfolio selection formulas
http://www.google.com/search?q=portfolio+selection+formulas
Good ones:
http://www.seykota.com/tribe/risk/index.htm
http://web.mit.edu/15.415ab/www/mid2form_all.pdf
http://cgi.di.uoa.gr/~vassilis/aee/MPTTextbook.pdf
http://webpage.pace.edu/pviswanath/notes/investments/assetalloc.html
This stuff is totally boring. I am going to try and read it from work. Or find someone on skype to discuss it with.
http://www.google.com/search?q=portfolio+selection+formulas
Good ones:
http://www.seykota.com/tribe/risk/index.htm
http://web.mit.edu/15.415ab/www/mid2form_all.pdf
http://cgi.di.uoa.gr/~vassilis/aee/MPTTextbook.pdf
http://webpage.pace.edu/pviswanath/notes/investments/assetalloc.html
This stuff is totally boring. I am going to try and read it from work. Or find someone on skype to discuss it with.
barjon
Legendary member
- Messages
- 10,752
- Likes
- 1,863
travis
I have followed your journey with interest although most of it sails wildly above my head. Nonetheless I wonder if I can offer some "babes and fools" comments that may (or more likely, not) give you a different perspective to your thinking.
Firstly, it seems to me that you have a single Master System whose component parts are the portfolio of Minor Trading Systems (MTSs). At different time the positive performance of the Master System will be affected by a different set of those MTSs so to seek to intervene in the mix may prove as counter-productive as productive. It may be, for example, that turning off those with the deepest drawdowns may be well be turning off those who are just about to turn the corner and bounce back more strongly than the others (or suffer further but at a less severe rate than the others).
So, if you are happy with your MTSs (and you say you have full confidence in those) why not just concentrate on your Master System per se so far as your money management is concerned?
I have drawn a few lines on your overall experiment picture (below) which are pretty meaningless in themselves except to illustrate what I'm going to say. I think there may be room to protect by reducing your total amount invested (would need reduced position sizes in MTSs) at various drawdown danger points and the increase it again once that danger is past (as evidenced by new performance high, or whatever). Should a second danger point be breached you can turn the whole lot off until it's back on track.
I dunno, just a kindergarten thought, but with the confidence you have in your MTSs, the crucial task seems to keep well away from "uncle".
jon
I have followed your journey with interest although most of it sails wildly above my head. Nonetheless I wonder if I can offer some "babes and fools" comments that may (or more likely, not) give you a different perspective to your thinking.
Firstly, it seems to me that you have a single Master System whose component parts are the portfolio of Minor Trading Systems (MTSs). At different time the positive performance of the Master System will be affected by a different set of those MTSs so to seek to intervene in the mix may prove as counter-productive as productive. It may be, for example, that turning off those with the deepest drawdowns may be well be turning off those who are just about to turn the corner and bounce back more strongly than the others (or suffer further but at a less severe rate than the others).
So, if you are happy with your MTSs (and you say you have full confidence in those) why not just concentrate on your Master System per se so far as your money management is concerned?
I have drawn a few lines on your overall experiment picture (below) which are pretty meaningless in themselves except to illustrate what I'm going to say. I think there may be room to protect by reducing your total amount invested (would need reduced position sizes in MTSs) at various drawdown danger points and the increase it again once that danger is past (as evidenced by new performance high, or whatever). Should a second danger point be breached you can turn the whole lot off until it's back on track.
I dunno, just a kindergarten thought, but with the confidence you have in your MTSs, the crucial task seems to keep well away from "uncle".
jon
Attachments
Yamato
Legendary member
- Messages
- 9,840
- Likes
- 246
I will reply as I read.
First of all, thanks for the feedback.
Second of all, let's make the premise that there's always, for all of us users, an extra problem with moderators (similar to the one I had with investors), which is the fact you don't know if you're risking anything (getting penalized in some way) by telling them what you think and there's the problem that you can't ban them (they can ban you instead), but with you it's different because I know that we get along, from discussing with you since a long time ago, before you were moderator. So I won't need to kiss up to you, like I did with the other moderator who told me I am beating up the bush and whom I could not ban.
Third of all, I am seeing your drawing and i must tell you that i don't trust trendlines: they can always be drawn nicely with hindsight, but there's no univocal way of drawing them nor, accordingly, any way of backtesting them (unlike for moving averages). So I see that drawing you attached, and I am immediately prejudiced and know i will probably not agree with you. In other words, seeing a trendline implies I am dealing with discretionary trading (as there is discretionary drawing in drawing the trendline, whether you realize it or not). So I am expecting you to now propose some discretionary intervention to me, and if you do, I will be against it. Having said, I appreciate the effort and time you took for your post, that went to the extent of quoting my equity line and modifying it and posting it. This is a lot of work, so I thank you, as I also thank the others who have put efforts into their posts, despite the fact that I had to momentarily ban some of them, due to the excessive poison and criticism in their posts.
Thanks for taking the time to give me your perspective on my experience.
Well, however you consider it and call it, Master System or Combined Performance, let me clarify that I am just trading a bunch of systems at the same time.
Correct: the dosing of systems (enabling and with how many contracts) affects the combined performance. One important thing to stress out is that I never enabled/disabled systems quickly. I mean: it takes pondering and it is a stable and usually permanent enabling. It's not like I turn them on and off several times per year. If things go as expected, and usually they do, once it is enabled, it stays enabled, and the only change can be an increase in contracts.
That is correct. I am indeed concentrating on my Master System. That is what I have been saying for the past two weeks. The problem was not in my systems, which are good enough. The problem was money management.
Oh, ok, I see your point. You see: this would initially sound good. It would sound good to reduce bet size as you lose, and increase it as you win. That is the kelly criterion rationale and so on:
http://en.wikipedia.org/wiki/Kelly_criterion
But there are two problems:
1) with futures it is not possible, because the contracts are too large to be able to reduce bet size with the limited capital size i have. I either trade or not trade. I can't possible reduce bet size, given that i am usually trading 1 contracts.
2) The other problem is pointed out by the kelly criterion wikipedia entry:
So, all things considered, I do not believe in the approach of reducing bet size, nor in monitoring the trend lines of the equity line, nor the moving averages crossovers in the equity line (and disabling the systems when there's a crossover).
I believe we must keep trading constant throughout the period. And most of all, I believe so for the reason that we have backtests telling us the entity of potential drawdown, and all these back-tests would be worthless if we altered the investment during the drawdown.
So I have listed 3 reasons to not do this: the first that I can't do it, and the other two that i don't believe in it and it's not scientific.
You see, I can't "turn the whole lot off until it's back on track" because there is no evidence, no back-testing, no univocal proof that profit behaves according to appearances, and that it gets back on track according to the trendlines (which cannot even be back-tested), moving average crossovers (which could at least be back-tested, but it's too complex) that we see, as a mirage, on the equity line, which among the other things, is totally biased by the fact that it represents the profit by 11 different combinations of systems, traded thanks to 4 different levels of margin available.
Now what's ahead is studying markowitz and other portfolio selection theorists, together with you and/or with the other few readers left and who have not been banned yet.
The next paper we should all read to work on my task is this:
http://www.math.ust.hk/~maykwok/courses/ma362/07F/markowitz_JF.pdf
By the way, you guys should all stop drawing and seeing things on charts: anything that cannot be easily be turned into a univocal formula, is not good for my trading, so, at least as far as I am concerned, i stay away from it from the start. I mean: I cannot forbid you from seeing things on a chart, but I advise you to do what i did and learn to not see on charts anything that cannot be univocally coded. At least that is what i do with my automated trading. And that is why I let bbmac remind me several times a week that I had a problem with my money management not being automated. He bothered me, but he had a point, so I could not deny he had a point. Everything must be automated, and if it isn't, then you might have a problem and suffer from curve-fitting, which is absolutely evident when someone engages in drawing trendlines on an equity line and, based on that, decides to enable/disable systems (it might work in the past, but is not a guarantee it will work in the future, and, if you really used a formula to draw those trend-lines, you might find out that it doesn't even work in the past).
Sorry, guys, but you don't have the scientific approach developed and ingrained in your minds, and you're polluting my rational automated trading with your discretionary ideas. I am not trying to shame you, but I need to be alert, identify the dangers, label them as dangers and mistakes, and stay away from your dangerous ideas.
Having said that, if you guys want to go on with me, you say it. You don't, you make your move.
Of course, moderators won't have the time to read this paper:
http://www.math.ust.hk/~maykwok/courses/ma362/07F/markowitz_JF.pdf
so that assignment is only intended for myself and my three other readers.
First of all, thanks for the feedback.
Second of all, let's make the premise that there's always, for all of us users, an extra problem with moderators (similar to the one I had with investors), which is the fact you don't know if you're risking anything (getting penalized in some way) by telling them what you think and there's the problem that you can't ban them (they can ban you instead), but with you it's different because I know that we get along, from discussing with you since a long time ago, before you were moderator. So I won't need to kiss up to you, like I did with the other moderator who told me I am beating up the bush and whom I could not ban.
Third of all, I am seeing your drawing and i must tell you that i don't trust trendlines: they can always be drawn nicely with hindsight, but there's no univocal way of drawing them nor, accordingly, any way of backtesting them (unlike for moving averages). So I see that drawing you attached, and I am immediately prejudiced and know i will probably not agree with you. In other words, seeing a trendline implies I am dealing with discretionary trading (as there is discretionary drawing in drawing the trendline, whether you realize it or not). So I am expecting you to now propose some discretionary intervention to me, and if you do, I will be against it. Having said, I appreciate the effort and time you took for your post, that went to the extent of quoting my equity line and modifying it and posting it. This is a lot of work, so I thank you, as I also thank the others who have put efforts into their posts, despite the fact that I had to momentarily ban some of them, due to the excessive poison and criticism in their posts.
Thanks for taking the time to give me your perspective on my experience.
Well, however you consider it and call it, Master System or Combined Performance, let me clarify that I am just trading a bunch of systems at the same time.
Correct: the dosing of systems (enabling and with how many contracts) affects the combined performance. One important thing to stress out is that I never enabled/disabled systems quickly. I mean: it takes pondering and it is a stable and usually permanent enabling. It's not like I turn them on and off several times per year. If things go as expected, and usually they do, once it is enabled, it stays enabled, and the only change can be an increase in contracts.
That is correct. I am indeed concentrating on my Master System. That is what I have been saying for the past two weeks. The problem was not in my systems, which are good enough. The problem was money management.
Oh, ok, I see your point. You see: this would initially sound good. It would sound good to reduce bet size as you lose, and increase it as you win. That is the kelly criterion rationale and so on:
http://en.wikipedia.org/wiki/Kelly_criterion
But there are two problems:
1) with futures it is not possible, because the contracts are too large to be able to reduce bet size with the limited capital size i have. I either trade or not trade. I can't possible reduce bet size, given that i am usually trading 1 contracts.
2) The other problem is pointed out by the kelly criterion wikipedia entry:
In other words, if we lose money today, it is not like rolling a die, but it may come in a specific sequence, so that we might hurt our investment by betting less the second time. Let's picture a case where you lose today and make money tomorrow for a few times. You would not end up even but broke, because you lose 1 today, make 0.5 tomorrow (after reducing bet size), then you double bet size again and lose 1 again, but then you reduce it, and gain 0.5 the next day... you get my point.
So, all things considered, I do not believe in the approach of reducing bet size, nor in monitoring the trend lines of the equity line, nor the moving averages crossovers in the equity line (and disabling the systems when there's a crossover).
I believe we must keep trading constant throughout the period. And most of all, I believe so for the reason that we have backtests telling us the entity of potential drawdown, and all these back-tests would be worthless if we altered the investment during the drawdown.
So I have listed 3 reasons to not do this: the first that I can't do it, and the other two that i don't believe in it and it's not scientific.
You see, I can't "turn the whole lot off until it's back on track" because there is no evidence, no back-testing, no univocal proof that profit behaves according to appearances, and that it gets back on track according to the trendlines (which cannot even be back-tested), moving average crossovers (which could at least be back-tested, but it's too complex) that we see, as a mirage, on the equity line, which among the other things, is totally biased by the fact that it represents the profit by 11 different combinations of systems, traded thanks to 4 different levels of margin available.
Now what's ahead is studying markowitz and other portfolio selection theorists, together with you and/or with the other few readers left and who have not been banned yet.
The next paper we should all read to work on my task is this:
http://www.math.ust.hk/~maykwok/courses/ma362/07F/markowitz_JF.pdf
By the way, you guys should all stop drawing and seeing things on charts: anything that cannot be easily be turned into a univocal formula, is not good for my trading, so, at least as far as I am concerned, i stay away from it from the start. I mean: I cannot forbid you from seeing things on a chart, but I advise you to do what i did and learn to not see on charts anything that cannot be univocally coded. At least that is what i do with my automated trading. And that is why I let bbmac remind me several times a week that I had a problem with my money management not being automated. He bothered me, but he had a point, so I could not deny he had a point. Everything must be automated, and if it isn't, then you might have a problem and suffer from curve-fitting, which is absolutely evident when someone engages in drawing trendlines on an equity line and, based on that, decides to enable/disable systems (it might work in the past, but is not a guarantee it will work in the future, and, if you really used a formula to draw those trend-lines, you might find out that it doesn't even work in the past).
Sorry, guys, but you don't have the scientific approach developed and ingrained in your minds, and you're polluting my rational automated trading with your discretionary ideas. I am not trying to shame you, but I need to be alert, identify the dangers, label them as dangers and mistakes, and stay away from your dangerous ideas.
Having said that, if you guys want to go on with me, you say it. You don't, you make your move.
Of course, moderators won't have the time to read this paper:
http://www.math.ust.hk/~maykwok/courses/ma362/07F/markowitz_JF.pdf
so that assignment is only intended for myself and my three other readers.
Last edited:
barjon
Legendary member
- Messages
- 10,752
- Likes
- 1,863
fire away, travis - no probs.So far as trendlines are concerned (yours in red) I said they were meaningless and just to illustrate the possibility of protecting at various times. It should be possible to come up with some mathematical "drawdown protection" formula so would not necessarily involve discretionary intervention.
jon
L
Liquid validity
If you could code a stop loss, I take it you would use stop losses then?
Even if they were just disaster stops for the recent volatile movements.
Most of the time your basket is essentially hedging itself as long as market
correlations are maintained, meaning stops are not as important.
Even so, with stops your margin requirements per trade could be lowered and protect against outlier events.
If it isn't easy to do this in your current combination of Tradestation and excel,
maybe a different platform would help.
I know that is a massive change and a lot of work with 120 systems though...
Worth looking at Ninja Traders capabilities (connects and executes fine with IB TWS)
http://www.trade2win.com/boards/tra...uels-money-making-journal-11.html#post1690444
- Status
Similar threads
- Locked
