要使用聊天,请登录。

If the open trades are locked...

Elkart
Elkart
Jan 04 2010 at 04:01
941帖子
They definitely were not heroes or wizards. But certainly cleverer than I am, and probably you are. You don't just roll out of bed and get Nobel prize.

My point is simply that some really clever people, very well resourced people, have been trying to do this for a very long time. Majority of them failed.

Elkart
Elkart
Jan 04 2010 at 04:24
941帖子
Just want to point something else out.

I've been where what's his name is twice now. You build a system. You test. It looks good, you run it live. Works for a while then Bam!! And you're out.

I made two years twice. Making and losing small fortunes in the process. Think about that, 24 consecutive successful months twice...

I presume now that any system I (or 99% of people) will develop will fail. Sooner rather than later. There's no point in trying to plan and program for everything. It's just not possible.

So instead what I do is draw heavy and for clients I force them to take out initial capital as soon as they possibly can. Stuff long term growth, safety first. It's in the contract and agreed to upfront. Client doesn't want to do it, I don't trade for him.

I've also written shut downs into the systems. Mostly the conditions that cause the failure only exists for short periods of time.

Anyone, who thinks they cracked forex after a year or three is in for rude surprise. Only way to handle the external shocks is to do it outside of the system, so failure parameters need to be well defined and in the system before you even think about putting down a trade.

It takes about 5 years I think to see all the fx cycles. If a system made 5 years, now that's impressive.

Elkart
Elkart
Jan 04 2010 at 05:01
941帖子
Put it another way.

The collective system development happening globally is basically a bell curve. Some people (very few) are going to do very poorly at it, some are going to lose a little bit of money, some are going to make a little bit and very, very few are going to be extremely successful at it.

We know from broker stats more than 90% of people will give up, so of the 10% left, they'll have their own bell curve. Probably less than 1% of people trying will develop a long term working system.

That's the reality of what we're doing.

bizWiz
Jan 04 2010 at 17:07
397帖子

Elkart posted:
 
I presume now that any system I (or 99% of people) will develop will fail. Sooner rather than later. There's no point in trying to plan and program for everything. It's just not possible.



you're probably not referring to complete failure as it wouldn't make any sense to have clients with such approach?

i would like to know how do you define failure:

- a system that worked for a year a stopped working?
- a system that lost money from day one?
- a system that needs to be constantly developed and improved?

and on that note, how do you define a system to be successful? after 5 years of positive figures yearly?


Sleep is for the weak.
Fill or Kill (cateful)
Jan 04 2010 at 19:50
87帖子
biz0101 posted:
i would like to know how do you define failure:

- a system that worked for a year a stopped working?
- a system that lost money from day one?
- a system that needs to be constantly developed and improved?

and on that note, how do you define a system to be successful? after 5 years of positive figures yearly?


Just as John Maynard Keynes once said, 'in the long run we're all dead'.
I don't believe a 'system' could ever beat the market in the long run. It's pointless to define the success of a system by its longevity. Market is dynamic, what worked in 1998 may not work in 2010, what will work in 2010 probably had lost money from 2005 to 2009.

And this not only applies to automatic systems. Even a discretionary manual approach is prone to the risk. Considering this, the sterling lost value when BOE jacked up rates in 1992 in order the maintain the ERM, GBP went down and my dear lord didn't Soros make a bundle in that trade. When BOE again raised rates from 2004 to 2007, GBP instead went up on carry trade buying. So you can't say for certain the GBP is going to go up whenever BOE comes out and hikes the rate a few bps. Because there were so many other factors. British economy was in a mess in 1992, the exports were down and and the Germans were raising rates to combat inflation which created pressure on other ERM countries. This was a different situation from 2004-2007 when the global economy was growing rapidly and rising rates were justified fundamentally. So a trader must look at the big picture and deal with all available information. Also I hear a lot of times people tell each other that history repeats itself. Yes, it often does repeat itself but in a subtly different way. A trader needs to understand the variations.

For automatic traders, if they have a talent of finding suitable systems for different market times then they will make money. So in the end it's all up to the trader, whether you are automatic or not.

Fill or Kill (cateful)
Jan 04 2010 at 20:40
87帖子
I'll try to explain a live example. Right now I'm starting to get interested in the euro pound cross. It has been trapped in a 300 pip range for the past three months. It was in a 700 pip range for 4 years before it broke loose in 2007. There will be a rate decision later this week and there is very little chance that they would raise rates. Right now I'm clueless as where it's going like everybody else but if you stick a gun to my head I would probably say it is going to break down rather than up. But most likely it will do nothing and sit there just like it did in 2003-2007. There are people looking for parity on this pair but I just don't see that happening in the current environment. My bet is that ECB will lack behind BOE in ending crisis measures and starting to raise interest rate. I need more information before making a decision to start a campaign on pound sterling. I will have no problem flipping to the long side if new information comes out supporting this view. But now my view is mostly balanced with a slight downside bias.

Elkart
Elkart
Jan 05 2010 at 00:48
941帖子
Biz,

I am referring to a complete failure yes. I'd love to think after all these years I managed to develop this ninja system that can cope with anything, but the odds are against me. Of all the people on this forum probably 1 or 2 will. Is it me? I don't know. Is it you? Don't know, we can check in five years or so and see.

In the meantime, I need to manage the risk of a failure. It makes sense to have clients if they take out enough so if there is a failure it doesn't really matter. But no one should expect to write down a few lines of code, test for a month or seven, and then sit back for the rest of their lives sipping Pina Colada's. And someone with enough experience in systems to manage money will never sit around touting the latest greatest result. They will know how fragile systems are.

Someone will do it. Someone always wins the lottery or the jackpot, but will it be you or me or PC8 or someone else? Odd's are normally 1: 1 000 000 000 or so it's me. So I'm going to presume it's not me and work for my money until proven otherwise.

PC8 doesn't understand this yet, so unless he's the lottery winner he's in for a hiding. Added to that he used his good times to test and develop. The longer the system runs the higher the risk of a failure, unless you are that one lottery winner. So you're in effect increasing the risk to your clients as you test away, not decreasing it.

Defining failure, now there's a thing. That takes some doing. Parameters to tight, you can't get through a bad streak. To loose, and you can lose the lot.

Thing is you can't just define it. You have to define it in terms of your system. If you normally have 70% winners would 50% constitute a failure? Or if you're still doing 70% but your equity hasn't moved for a week, is that a failure? What do I do a month? If I lose a months profit is that failure? If I did 10% a month and I lose 11% will I close out? To me 10% loss in a month is ok. 25% isn't.

It's something you'll have to work out for yourself in terms of your system.

But what I won't do is keep developing systems. I have one that works in most conditions. If it stops working, it switches off and stays off till I think it will work again. Go for a holiday or reduce the trade sizes to a very small portion of what it used to be, run it through the bad period and when things go green again turn it right back up. There are ways of dealing with it. Long as you planned for it. And to do that one needs to be realistic about the prospects of success.

If I make 5 years, I reckon if I get through that then there's a good chance I have system that will work for the rest of my live. Or at least long enough to buy my toys, a mountain of beer and small harem.

Elkart
Elkart
Jan 05 2010 at 00:58
941帖子
A good read for this is would be Taleb's Black Swan: https://www.amazon.com/Black-Swan-Impact-Highly-Improbable/dp/1400063515

It's not something I fully understood till late last year. Especially in relation to system development.

Some exerts here:

https://www.fooledbyrandomness.com/imbeciles.htm

Elkart
Elkart
Jan 05 2010 at 01:10
941帖子
Reminds me of the guy with the fake box.


'but this pair been doing this for a long time'

Sure, that just makes it more likely to stop doing that. Like the 26 year bond bull market. Stopped last year. Dead in it's tracks. If you weren't ready for it, that's you after 26 years wiped out. Old and poor.

And that's a very obvious example. In fx the changes are a lot more subtle and very difficult to pick up before it's to late. System performance is the first clue that something has changed.

Any working fx system has to be based on change. Not lack of change.

Elkart
Elkart
Jan 05 2010 at 02:30
941帖子
Give you a practical example.

I have three levels of exits (not talking tp's here, just ways of handling losing positions).

1. My normal trading. So if my parameters are met I do a new trade and close out the apposing trade.

2. An equity exit - if the total sum of trades go against me X percent I get very aggressive in trade reductions and turning pairs. Cause then I have positions that aren't working, so I need to swing them the other way and then keep going. The idea is to get back on the right side of the market and at least get back to my previous equity high, as apposed to trying for new equity highs.

3. At X x X I will close out completely, cause then I'm having a series of positions that aren't working after working for a long time, so I stop, somethings obviously gone wrong and something changed. The system will then send me a mail to come look.

请登录以发表评论。