my journal 2

[Yamato]

Travis

I know you are down right now.

But you have come a long, long way.

Listening to negativity on here about what you are doing won't help.

Only you have your experience. Somebody doing something totally different isn't in a good place to critique. Still - it's a very good way to make yourself look clever.

Like you said - this ratio & that ratio - well it's all the stuff of TA books. Who gives a fook about Sharpe? I tell you what - I know a few full time traders and if I asked them what their Sharpe ratio was, they probably ask me what on earth it was.

If you think you are close to getting this right - then go with that and stay on the path. You have already proven something. How many other guys here could write systems, attract financing and do the other stuff you did? Tell you what - the guy calling you a ginger headed kid couldn't.

This is the path you have walked. Don't deviate because of some dicks on an internet forum trying to make themselves look smart @ your expense.

Pete

Replying as I read.

I am not really down, or I would not be replying to all these posts, nor posting the precise picture of the situation (later today I will show the final equity line of this year and a half of trading). I am just busy documenting in detail my trading and my thoughts, as usual.

Yes, I agree that I have come a long long way, and also thanks to these investors.

I know this negativity won't help, but given that it is not "ur gay" and similar, I have a duty to listen to the devil's advocate painting the worst-case scenario and showing me my possible deficiencies. When I don't think it's the case, I will reply in this sense. So far the only one criticizing is bbmac, so I can bear it, and I need to address his subtle yet constant criticism.

It's a very good way to make myself or themselves look clever? I would interpret it best as "make myself look clever" because this criticism is thought-provoking, so the debate is healthy, or else I would not have replied. As I said, if he were writing "ur gay" or things like this guy, then I would not reply. But bbmac is writing detailed criticism, has been following what I've been doing for months, he writes on topic, in-depth posts, and puts effort into what he writes. So I will just complain about what he writes, but I must read his posts, and put some effort, given the efforts he puts into his posts. Yes, the content is unpleasant, and sometimes he quotes me out of context, and manipulates what I say, but I cannot say it's off-topic. And I cannot dismiss it quickly, because he still puts efforts into his posts.

Yeah, sharpe ratio might even be useful, but all these metrics distracted us from the fact that we overoptimized the back-tested combination, and gave us a false sense of security. We went too far with the scientific approach, and especially if I create so many excel workbooks as to be burned out, then I cannot be relied on in any other way: i will not be alert when there's potential mistakes going on. I've been signalling for months my process of burning out, but to no avail. Now I am so burned out that I almost do not feel so bad that this has all ended, because it gives me a chance to rest.

Yeah, the guy calling me things didn't go far enough yet to be banned. He puts too much intelligence and efforts into his posts to be ignored or banned. Given my experience the worst thing I could do to him is ignore him due to having exceeded his negativity quota. I cannot ban him, just because he's saying unpleasant things. When they'll get boring and unpleasant, I will ignore them. But right now they're only unpleasant and I have to show efforts in replying and make my points, and argue and debate. If he'll go on like this for another two months, then I'll start ignoring his posts.

I think I will keep going this way, and I am proud of showing all my failures. If there's something I can say is that I have shown my successes and failures in the same way, because a journal has to be honest, or else there's no point in writing it.

I enjoy being honest, and right now there's more enjoyment from exposing my problems than pain from experiencing them.

Besides, I met the investors on this journal, so I owe it to the forum and the journal to show my failures as well as my successes. And, finally, I don't consider this a failure at all. I worked my ass off for years, and this is just a step. Now we know what happens when we do so and so.

 

[bbmac]

Travis, if you cannot take some hard but well meaning critique alongside well meaning/properly considererd suggestions then I will stop posting to your thread- just as you cannot alter your inability to be hurt/offended by such and read into it something that is not there, I cannot alter my style either.

If hard work and diligence were all that it took to be successful re trading you would already be there. Your work rate is amazing and I'm not surprised you feel a little burned out right now.

I agree that it is easier to bet on a winner and less easy to do so on someone that has not failed but just not succeeded yet.

Anyways, G/L with the ongoing project, I hope you make it.

 

[Yamato]

Oh what a cynical view of the world the poster above has !

I have read nothing but well meaning and genuine advice from posters on this thread. On an open forum and particularly with the style/content that Travis has conducted this thread it is perhaps inevitable that such comments/advice will come forward - but as far as I can tell such posts are all well meaning and well considerered and a lot of posters put considerable effort into the arguments/suggestions they put forward - often drawing on personal experience.

Travis will get a lot of advice from all quarters - it's up to him what he does with it (or not.)

There is only a dick posting on this thread !

G/L

I understand your negativity and I must be able to put up with some criticism (so far only from you, and in the past from brettus) given the equity line I have been posting. It would not be normal if there were no criticism after I show that equity line. And I'd rather get the criticism from you or brettus than from someone saying "ur gay" or "you suck".

On the other hand, I absolutely also need the encouragement from all those other people who have been offering it, like DionysusToast, and first of all from myself. Besides, I agree with the people encouraging me, because I can reason and I know that I have some good stuff cooking. I am not worried about the past 14 years of unprofitability. I keep repeating it, because I am not afraid of my track record. I have not made any money from the markets since I started, I have lost 40k since I started trading in 1997, and yet I know I am good, I know I have very good stuff, and of course I will keep going.

 

[wprins]

And, finally, I don't consider this a failure at all. I worked my ass off for years, and this is just a step. Now we know what happens when we do so and so.

Now, *that* is a winning attitude. :clap: There's a quote by someone to the effect that you only ever *really* fail when you quit. I think it's very true and is echoed in you comment above. Thomas Eddison also had to fail like 10,000 times to get to a lightbulb design that worked, and he said something similar to what you've said above. Hold onto that attitude.

 

[Yamato]

Nobody says there's neccesarily a downtrend starting when your equity curve breaks below the yield cuve (although there might be.) But (the point if you will, is this): if your backtests and forward tests over all available data has shown you that the equity curve should *always* stay above (for example) an adjusted yield curve as above, and it then unexpectedly breaks below it (and stays below it) then at the very least, amber lights should come on -- it implies something is, with a high probability, broken. (Whether temporarily or permanently is at that point not known and in any case besides the point. )

It is, in a word, unprecedented and therefore mandates careful scrutiny as to what's really going on, and taking whatever emergency actions was determined in advance to protect your capital, even if that action is simply to say the systems go back onto paper trading/monitoring for the time being.) To emphasise: The yield curve is after all an objectively defined model of what your system(s) behaviour is and should be with an extremely high probability, and any significant deviation from this expectation is therefore by definition not subjective/discretionary and a strong signal that something's wrong. What you *do* in this situation is a design decision of course, but no more discretionary than you picking parameters in advance for the rest of your systems parameters (market, time, levels, moving average periods/types, whatever) and then backtesting these choices to confirm the model "works".

As an aside, this is one of the unavoidable problem with mechanical systems IMHO -- you don't generally really know whether what you've found/captured in a system is due to sheer luck/randomness and/or indeed what the real nature of the edge that you've captured in a given system *really* is. Consequently, one also will in general not be able to know when the edge you have or had may be be disappearing or may have disappeared. However, my point is that such a loss of an edge (or temporary market behaviour alteration, or extreme random exception) will (obviously) show up in your equity curve (as it has in your case) so it's logical that defining some objective rules based on the behaviour of your equity curve is one way to monitor whether your system(s) as a whole is continuing to work and are continuing to exploit whatever edges they capture.

Lastly, I don't see why you say this doesn't fit in with any kind of automation process. The steps to automating this would be something as follows:
1.) Backtest (and forward test etc) your system or systems, and determine the actual equity curve data points for as much data as you have.
2.) Now calculate best fit (R^2) model for the equity curve dataset.
3.) Optimise/adjust this yield curve model so that it falls below your known equity curve for all data points.
4.) Incorporate this into your automated systems, so that they/it automatically calculates the expected minimum yield at every bar/data point in your system alongside all the other variables being calculated/monitored.
5.) On an automated/ongoing basis, compare the current value with the expected minimum according to your yield model.
6.) When the equity value falls below the model, automatically trigger the circuit breaker and stop trading the system (or all systems.)

So, once implemeted I submitthere's no discretion at all in the above, it's all automatic and objective. The only discretion involved at all, if you want to call it that, is deciding exactly how far to adjust the yield curve while setting all this up, and what the nature of the circuit breaker should be. But I would suggest this is no different the many other decisions made about other aspects of systems during the design and optimisation of those systems.

How's the systems performing currently? (The current market realities are probably likely to kill/hurt many systems -- silver's recent moves for example has probably hurt an huge number of traders...)

Replying as i read, but I have to be quick, because this post is huge.

I have a problem with implementing what you are saying. There's hundreds of unforeseeable implications in this method of yours and of many others, who have suggested it in the past, including the investors. It's not that simple to just plot an equity curve and see how it behaves. We're talking about 120 different systems on 16 different futures.

The way i see it instead is you pick what has worked in forward-testing and keep on trading it. That's all.

Your method is too complex, has too many unexpectable implications, and to do it right would require me too many hours of work and if I can't do something right I don't do it.

Your idea has crossed my mind before, too, but I have discarded it after giving it careful thought.

Your 8 points (thanks for the thoughtful post by the way) show why I am not doing it. If you consider 120 systems and all the implications your simplified 8 points have, you will see, or at least I see why I am not doing it. Just reading the words you write, even without understanding everything you write, makes questions and objections pop up in my mind relentlessly.

Yes, once implemented and automated, it is automated. But I cannot do it. I cannot automate this.

I cannot go down this road. I have to go down the opposite road. We have already made things too complex. They've been made so complex that they're counter-productive.

The systems are doing ok. Most of my systems have six months of drawdown. One month of drawdown is nothing. Not my fault if it happened on so many systems all at once.

 

[Yamato]

Travis, if you cannot take some hard but well meaning critique alongside well meaning/properly considererd suggestions then I will stop posting to your thread- just as you cannot alter your inability to be hurt/offended by such and read into it something that is not there, I cannot alter my style either.

If hard work and diligence were all that it took to be successful re trading you would already be there. Your work rate is amazing and I'm not surprised you feel a little burned out right now.

I agree that it is easier to bet on a winner and less easy to do so on someone that has not failed but just not succeeded yet.

Anyways, G/L with the ongoing project, I hope you make it.

Keep writing and don't be hurt either by my counter-arguments. If I feel offended, I will tell you. If you feel like writing negative stuff, write it. As long as it is well-documented, I will keep reading it and replying it.

Thanks for the (rare) compliments. Very appreciated.

Whenever I read negative and unpleasant stuff in your posts, I replied in detail, so now you know exactly what bothers me and what I am thinking about every little nuance in your posts.

 

[Yamato]

Now, *that* is a winning attitude. :clap: There's a quote by someone to the effect that you only ever *really* fail when you quit. I think it's very true and is echoed in you comment above. Thomas Eddison also had to fail like 10,000 times to get to a lightbulb design that worked, and he said something similar to what you've said above. Hold onto that attitude.

Thanks for the encouragement. It's impossible to quit now. I know my systems are profitable, including the combination that has had that sharp drop in the equity line I keep posting, and which I will post today and it will look even worse.

That is just a drop. They will keep on growing. Yes, precisely those systems that seem to have failed. It's 50 systems that have been working for years. It's totally unlikely that they have all stopped working for good.

 

[shewolf]

I am not worried about the past 14 years of unprofitability. I keep repeating it, because I am not afraid of my track record. I have not made any money from the markets since I started, I have lost 40k since I started trading in 1997, and yet I know I am good, I know I have very good stuff, and of course I will keep going.

With all respect, Travis, what makes you so sure that you are destined to be a successful trader?
The fact that you have 14 YEARS of unprofitability would indicate to me that I was either totally unsuited to trading or that my trading strategies were completely off-target.
I respect that you are determined to keep going, but at what point do you acknowledge that it is a lost cause?

 

[Yamato]

I keep going because I know where I am. I know the map. If you were climbing mount everest and knew you were near the top, would you stop? But if you don't know where you are, you might feel lost and give up or die there. But I know where I am.

If it takes 10 years to become a doctor, do you stop at the ninth year, because people tell you "hey, you've said you'd become a doctor for the past 9 years and you still haven't made it... just quit"?

I know the map and I know where I am. Let alone the fact that there are people who are profitable for 5 years, lose everything and then quit. None of us trading has a perfect map, and you could be profitable while doing the wrong things or unprofitable while doing the right things, but I have a rough idea of the map and I've been seeing land on the horizon for a long time, like Columbus travelling to America for the first time. I don't have an exact map, but I know there's land in that direction and that it can be reached, and I can see the birds, so the land is near.

 

[counter_violent]

Hi Travis,

The majority of the most recent posts appear to latch on to the same / similar theme.

Your systems appear to be acceptably diverse and sufficiently profitable to everyone including yourself for the majority of the time, however, it seems clear that when MrMarket throws you a curveball (and it will ) that the systems are not sufficiently robust as to come through the drawdown (unrealised loss).

I'm not suggesting for one second that you should re-engineer your systems which appear to perform well enough, most of the time. What I do suggest is to find a pre-determined point where you can enable a further set of rules which will at worst neutralize your overall exposure, or at best, if well designed could add to your P and L bottom line for as long as the curveball is in play. Only you can determine how to measure the conditions and when to invoke rule set 2.

cheers...keep on keeping on

 

[Yamato]

Thanks for the feedback, but I disagree with some of the things you say (everything except "hi travis"):

1) that the majority of the last few posts say "that the systems are not sufficiently robust as to come through the drawdown"

2) that I should "find a pre-determined point where you can enable a further set of rules which will at worst neutralize your overall exposure".

As i said before, there is nothing wrong with my systems. None of them has exceeded its drawdown, and even if some of them did, they're expected to perform worse in the future than in the past. The problem was either with:

1) the investors unwilling to put up with the drawdown
2) me underestimating the drawdown

We still could have done everything right and have been unlucky. We were ready for some of the systems incurring in the drawdown, but not for most of them to incur their drawdowns all at once. This is even far from being the case, but what I am saying is that there were either bad luck or problem with money management. There is nothing wrong whatsoever with the individual systems. The 120 systems are just as healthy as they were before. The only problem is that we did not expect them to all be in a drawdown at the same time.

This problem doesn't get solved by halting the trading. This is too complex to back-test, has too many implications (what systems should we disable and why? Dude, there's 50 systems being traded), and I am not doing anything I cannot completely understand.

What maybe we should focus on is: how much are we going to lose if all systems incur their worst drawdown at the same time?

If we did that, we would see that we're still far from seeing that happen.

My systems have not failed. We have just underestimated their drawdown. That is all. Keep asking me as long as you want and I will keep repeating this, as I have been doing for the past few days.

 

[Yamato]

There were two entities telling me, all these months, something like: why are you doing this thing with your current investors when you could do things much more comfortably and conveniently with my money?

It was my father and the other entity was this long-time friend.

As I said, I am fearing that they will not be interested any more now that I need them.

I have asked the long-time friend and he said he has to think about it. LOL. This really makes me laugh, when he has been telling me repeatedly why I didn't just do with him until just a month ago.

Then there's my father. I am not even going to ask him. As soon as he asks me about it, I'll tell him that I am not trading any more and see if he says "hey, I am willing to give you money" as he said repeatedly in the past.

These two ****-sucking entities are the only entities that would not force me to reveal my code. My father doesn't care and the US ****-sucker already has the code, as I am running my systems on his server.

If the ****-sucking entities do not come through, I will be left with just my dick in my hand.

I will still keep renting the server, and will still forward-test my systems, but I will not have any money to invest.

First thing to do though is to make sure I get the 3 ****-suckers to tell me clearly that they are not interested:
1) the current investors (they have one more month to decide)
2) the server running US ****-sucker
3) the father ****-sucker

Then I don't want to hear in the future that they are interested and why I didn't ask them first... I never want to hear such bull**** again. Because when I needed them and asked them directly, they said they weren't interested.

Then there's one more guy. This Italian guy who wanted to give me 500k euros. This is more complex, because he doesn't know anything about trading, so it would all be entirely in my hands.

I would rather not do this.

I would rather work with people who know about trading.

If everything else fails, I will just have my salary, which I can use to pay the server, forward-test my systems, and... maybe I can stop riding cabs, save money for a few months, and invest my own money again.

But I doubt all these potential investors will stop being interested at the same time. Actually I think I can afford to still be picky and not disclose my systems to anyone ever again.

I am fed up with the risk of people getting my code and then telling me "hey, we don't like your systems", and then who knows if they're actually using them. I will never again get so desperate as to give my code to people who ask for it. I'd rather not trade, than give my code to anyone ever again.

 

[counter_violent]

Thanks for the feedback, but I disagree with some of the things you say (everything except "hi travis"):

1) that the majority of the last few posts say "that the systems are not sufficiently robust as to come through the drawdown"

2) that I should "find a pre-determined point where you can enable a further set of rules which will at worst neutralize your overall exposure".

As i said before, there is nothing wrong with my systems. None of them has exceeded its drawdown, and even if some of them did, they're expected to perform worse in the future than in the past. The problem was either with:

1) the investors unwilling to put up with the drawdown
2) me underestimating the drawdown

We still could have done everything right and have been unlucky. We were ready for some of the systems incurring in the drawdown, but not for most of them to incur their drawdowns all at once. This is even far from being the case, but what I am saying is that there were either bad luck or problem with money management. There is nothing wrong whatsoever with the individual systems. The 120 systems are just as healthy as they were before. The only problem is that we did not expect them to all be in a drawdown at the same time.

I'm really not sure how you can say i'm wrong when in fact you have admitted here in the above statement that you did not forsee drawdown in a number of systems at the same time. This rather makes the point that i and others are trying to convey. So if we take this a step further...you are trading on the hope that the drawdown will turn around, rather than having a proactive plan that mitigates or eliminates some or all of problems.

This problem doesn't get solved by halting the trading. This is too complex to back-test, has too many implications (what systems should we disable and why? Dude, there's 50 systems being traded), and I am not doing anything I cannot completely understand.

I'm not suggesting for one second that you should disable or halt the systems and I have not said this. What I did say was...pick a point in the collective drawdown of the systems where you feel that a second set of rules may need to be implemented that would support continuing running the systems as is.

What maybe we should focus on is: how much are we going to lose if all systems incur their worst drawdown at the same time?

Yes that is exactly the point...you will need to work out solutions before a smallish problem has the potential to escalate into an insurmountable problem.

If we did that, we would see that we're still far from seeing that happen.

My systems have not failed. We have just underestimated their drawdown. That is all. Keep asking me as long as you want and I will keep repeating this, as I have been doing for the past few days.

I wish you wouldn't be on the defensive and perhaps take the time to consider what is being said without the need to be so dismissive.

It appears to me that posters to your journal are in the most trying to be helpful.

 

[Yamato]

Yes, the posters' attitude is great, and I feel very grateful to all of you. And also to this great forum which allowed me to speak freely for years without any type of censorship or pressure.

However, I have to be defensive and dismissive, not towards you but towards your suggestions, because that's what you need to do to keep things simple and coherent.

For example, with the investors, given that they had the money, I was available to work and create files for a year, and this has lead me to an overcomplex infrastructure (7 files with dozens of worksheets) and to being burned out. Your solution and that of many others of changing things when the drawdown happens is not a linear and simple solution. It messes up the automation of systems. Systems have to be run as they are and cannot be tampered with. The only problem is appraising the potential drawdown and being ready for it. The problem is not to devise a way to tamper with the systems. The systems must be traded as they are. If you don't do that, things go even worse. And most of all you should not do that unless you're ready to comprehend all the implications of such changes, which are countless and huge. I cannot handle all the implications, but I can certainly see them immediately, and I cannot get into something where all implications are not being comprehended and handled. This is univocal and automated trading, and not a discretionary thing where one can proceed by guesstimates. Everything has to be back-tested and univocal.

 

[Yamato]

final equity line

On this occasion of posting my final equity line of a year and a half, I must admit that the investors did their job of sticking 100% to the systems until the uncle point and beyond it. I definitely was shown by them how to stick to systems with discipline from the first to the last day. It was definitely a good lesson. They stuck to my systems for a year and a half more than I was ever able to do for just one week, and they were my systems. It was a great experience. We ran everything. We paid off cops. We paid off lawyers. We paid off judges. Everybody had their hands out. Everything was for the taking. And now it's all over.

 

[brettus]

Re: final equity line

On this occasion of posting my final equity line of a year and a half, I must admit that the investors did their job of sticking 100% to the systems until the uncle point and beyond it. I definitely was shown by them how to stick to systems with discipline from the first to the last day. It was definitely a good lesson. They stuck to my systems for a year and a half more than I was ever able to do for just one week, and they were my systems. It was a great experience. We ran everything. We paid off cops. We paid off lawyers. We paid off judges. Everybody had their hands out. Everything was for the taking. And now it's all over.

I think you've done well to get investment and not intervene with the systems for over 18 months. I don't think anyone on this forum is actually capable of that. Personally i give you a lot respect for that than having made a bunch of money by intervening.

This is where i think you stand in your journey to becoming a profitable mechanical trader. You've removed any discretionary input. You've stuck to your plan. You're pretty much there. I think your next step is to speak to a successful mechanical trader who can give you some input, which i hope you would take seriously. I have a friend who successful only through learning from someone else who is successful and funny enough he recently came across the same problem of over optimization by system combinations. He didn't trust the fact that his systems individually could have a worse drawdown than a combination of systems. He works for a hedge fund which specialize in black box currency trading. I think you need that input from elsewhere. Unfortunately advice coming from me doesn't go very far in your opinion because i'm not a mechanical neither have i actually built a system. Netherless i hope you take my advice seriously.

 

[Yamato]

Thanks for the feedback and compliments, and that idea about your friend sounds good. If there is a possibility, tell him to get in touch with me, after warning him that I am pretentious, arrogant, self-centered, argumentative, and all that.

If he doesn't, it is still ok though. A characteristic of me is that I am persistent, but slow, and the reason why is that I have to do everything by myself, and do not listen to other people very much. Somehow the investors managed to get me to listen because they suggested things very subtly instead of telling me "you must do this".

I have never listened to any single person more than I have listened to the investors. Then there were other people who helped me, but with the others it was all help that came because of questions I asked. With the investors, I benefited from advice I hadn't asked for.

Regarding your remark that my tampering is gone, this is true. I hope to have gotten rid of my compulsive gambling for good after this year and a half of strict discipline.

 

[Yamato]

brainstorming on what to do next

1) The investors still have 4 weeks to go in order to decide whether they want to resume their investment with me, or just end the partnership completely.

2) If they resume, then I'll have the usual exclusive relationship with them and won't do anything with anyone else. How likely is it that they will resume? I think it's 50% likely. If they resume, probably we'll change the combination of systems. Other than that, everything will keep going as usual, except that we'd be starting at -11k (cfr. final equity line in previous post), so it'll be a while before I see any profit.

3) If they do not resume, then I will be totally free, but without any capital. If they resume, the plan is clear. Let us focus on what to do if they do not resume. I will be totally free, but without capital.

4) the number one priority is that this time I do not want to give my systems to anyone.

5) the favorite person, the server-running long-time friend, told me he changed his mind. Fine. I obviously cannot force him to give me his money to invest.

6) the next favorite person would be my dad. I don't want to ask him, but within a month, the topic should pop up in a conversation ("how's your trading?"), and I will ask him if he still wants to give me some money. If it happens before the end of October, it would not be a problem because our current agreement (with investors) does not exclude that I invest my own money on my own systems. This would be ideal, because i'd have complete freedom and complete secrecy. I would be on my own, completely.

7) if #6 fails as well, there's the Italian friends. One already disappeared a month ago, after I told him that I was burned out and I would not work on his system. He had told me "do it if and when you have time". I said I was burned out, and he was so offended that he disappeared. Very coherent. That's like saying "how are you?" and not wanting to hear any answer other than "fine, thanks", which is pretty much what people do. So much the better: I am rid of another idiot.

8) There's two other Italians who wanted to give me their money. One wanted to give me 500k and knows nothing about trading and the other is a system developer, and I would prefer him, for this reason, but the problem is that he'd want to see my systems, most likely, and this from now on is not going to happen. They can see the trades, which is already a lot, but they cannot connect to the server nor see my systems. So maybe it is better if I opt for the other non-trading guy. But with the trader, I can get in touch with him and ask him, but with the non-trader, I cannot do it, because I don't know him well enough to go and ask for money.

9) so the next steps are:
1) curent investors resume/end relationship (one month to decide),
2) father,
3) Italian trader,
4) Italian non-trader.

If everything fails, I will just keep forward-testing. Three rules are fixed:
1) I will not show my systems to anyone ever again and
2) I will not resume my trading unless I am sufficiently capitalized, which means having at least 20k. And
3) I will not take out any loans from my bank.

 

[Yamato]

Let's now focus on what my new money management will be.

The sharpe ratio, I have got it now, I have all the reliable values for all 120 systems. No point in throwing it away. It shows me not only that a system has been profitable but also how efficient it is. It is more or less like combining profit and profit factor (sharpe ratio and profit factor are quite close anyway).

So I will use sharpe ratio.

I will not use any complex combinations of systems and assess how promising they are. Yes, I will use the relativized drawdown, once my combination of systems is ready, but I will not use it to come up with a combination. We see what happens when you do that. No optimization of combinations and by all means, like bbmac kept repeating to me, I need to automate systems' selection, to avoid any danger of discretionary over-optimizing of combinations.

Here's what I will do.

It has to be simple and stupid, foolproof.

I will select for trading all the systems that have traded in forward-testing more than 10 times and that have a forward-tested sharpe ratio higher than 2.

If I do not have the funds for them, because my funds do not cover the relativized back-tested drawdown multiplied by two, then I also need an automated way of removing systems.

Here's what I will do. I will keep the best ones and remove the worst ones until I can afford the doubled relativized back-tested drawdown.

So let's say I can't afford to lose twice the relativized back-tested drawdown if trading all systems with sharpe ratio above 2. Then I will consider all systems with sharpe ratio above 3. If I cannot afford that either, then I will consider all systems with sharpe ratio above 4.

This still leaves us with a problem: is it really reasonable to treat a silver contract the way you treat a british pound contract?

Of course not.

But how do I automate the solution of this problem?

If I had endless capital, I would decide the contracts based on the max back-tested loss by the systems.

So, if GBP had a sharpe ratio of 3 with a max back-tested loss of 1000, and silver had a sharpe ratio of 3 with a max back-tested loss of 8000, I would trade 8 GBP contracts and 1 silver contract.

But this cannot be done, because I will not have all this capital.

If it could be done, every system should be treated the same and be allowed to take the same risks. If a Silver system can risk 8000 per trade, so should the GBP system.

But in reality, we do not have the capital allowing us to make bets of equal sizes on all futures.

So this would lead us to not only removing systems based on their too low sharpe ratio, but also removing them based on their too high maximum back-tested loss.

Let us recapitulate the requirements and rules:

1) forward-tested trades >= 10 or higher (depending on capital available, make rule stricter)
2) sharpe ratio >= 1.5 or higher (depending on capital available, make rule stricter)
3) max loss (formula to be defined)
4) some capital available must cover twice the back-tested relativized drawdown (if it doesn't, make #1 and #2 stricter)

I am now wondering if I should double the GBP contracts so their max loss will equal the max losses by all other systems.

I feel that every system, all other things being equal, should be allowed a max loss that is as big as the other systems.

Damn, is this hard. And, damn, did we screw up.

Somehow I felt that we could not be so unlucky as having to worry about the worst possible outcome, but the fact is that by over-optimizing the back-tested combination, I came up with a combination of systems which delivered a great profit while keeping the drawdown low, and that was guaranteed to not happen again in the future, like all products of over-optimization and curve-fitting.

I was recommended to use the out-of-sample in back-testing individual systems, and benefited from it, but we were all aware that we were over-optimizing back-tested combinations and not applying the same methodology for finding combinations.

That is why I cannot ban bbmac, despite all the negative and repetitive criticism of his posts: he identified a couple of problems that were important. We must automate systems' selection and 2) why did we change combinations?

Of course he didn't offer any solutions, none that were intelligible to me (I got lost in his kilometric posts).

And another point that he made me think about (among the endless negativity) was the nature of what we are doing. Why aren't we simply trading all the systems? How about that?

But even that would not be right.

Let's say that we start with the approach of trading all the systems that have a sharpe ratio above 2 and more than 10 forward-tested trades. This would be ok and there would be no bias until here. No risk of curve-fitting. We're simply expecting the systems that have been doing the best to keep on doing the best. In most cases we will be right.

But there is still a problem, and it is: GBP has a much smaller contract than Silver. How do we equalize those contracts?

Based on the max back-tested relativized loss by the system?

Based on the max back-tested relativized drawdown by the system?

Based on the back-tested profit?

Based on the margin?

Until here we have been fair to every system: if you behaved well in forward-testing, we're willing to trust you. But why are we investing so much less on the GBP systems than on the others?

This is going to be really complex to solve, but I would opt for using the max back-tested relativized loss. Then, given the limited capital, we cannot just increase the GBP contracts until the GBP systems until equal the silver systems. So we would have to exclude the silver systems, but is this even reasonable?

What if instead we disregard this difference until there is available capital to increase the contracts on some systems?

This is so complex and the potential solutions are infinite. No wonder I gave up several weeks ago when bbmac suggested to automate it.

But some points are now clearer. Automation of systems' selection excludes in itself the unhealthy search, in the back-tested sample, for magical combinations where the success of one system compensates the failure of another, which is something we have done heavily and that has lead to the future performing much worse than the past. In this last occasion too much worse.

Systems should not be selected based on their combined performance, but based on their individual performance, and this should be done based on fixed criteria. Furthermore, each system should be allowed to weigh the same in the basket of systems and its contracts should be increased as capital allows. But this still has to be defined. I don't know how it should be done. Whatever we do, systems' selection should be automated. The big problem with all this is that I don't have money or, in other words, the future contracts are too big. There is no way to finetune this automated selection because you can't practice it, and because you cannot allocate 8 contracts to GBP systems to make them equal the silver systems: the implications in terms of drawdown, margin and therefore capital required are immense. I would need a million dollars to be able to do things right.

Having said this about the useful input by bbmac, if there's one more negative comment by him, despite acknowledging that he's not a troll, I will still have to ban him because his continuous criticism destroys my confidence, and I need a certain level of confidence to keep working. I will then unban him when I am feeling ok, much like I did with brettus, on and off my ignore list in the past 2 years. I received too much criticism from my father when I grew up, so anyone resembling my father in that area, has to make sure that I can put up with their criticizing PLUS all the criticizing I have already received in my life from my dad, and as a consequence I get fed up pretty quickly with criticism.

 

[Yamato]

You know, i still can't believe we blew it so quickly. I can't believe we could blow it at all, and I cannot believe all these ****-suckers who just a month ago were telling me they wanted to give me their money are now telling me they're not interested.

My systems are just the same and they have not failed whatsoever. Most of them have six months long drawdowns, and one month of drawdown is nothing. Just the fact that it happened all at once indicates it was clearly a problem of money management.

And it's also obvious that dozens of excellent systems cannot fail all at once.

Whoever tells me the contrary is just out to get my systems for free.

That is why I am not giving them to anyone ever again no matter how desperate for capital I am.

There's been no more gambling in my life for several months, and if I keep it that way, I will not feel that thirst for capital, which is what makes me turn to people and give them my systems.

On the other hand, as I said before, I am not someone who ever markets himself or who knocks on people's doors, so I won't do that either.

So I really wonder if ****-sucker dad and ****-sucker server-running long-time friend don't give me their money, who else will do it?

Maybe, given all these requirements in my mind (not giving my systems, not asking for money, not asking for a loan, not engaging in discretionary trading, not selling my systems, not contacting banks and such), it is time for me to find another source of income and trade with the money coming from that source. If I cut down on taxis and restaurants I could count on 600 euros per month. But that would only be enough to get started with compulsive gambling again. So I am better off being lazy and riding cabs.

Or maybe I will just keep forward-testing my systems and see what comes down from the sky. Yeah, no more desperate moves. If anything, I'll keep still and weigh my opportunities and offers.