For those who are convinced that all EAs with a martin are losing out. - page 18
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
You're not - kind. (but bloodthirsty).
P.S. I do not understand how you can build a trading system, secretly hoping that "this" or "this" will not happen or will not happen before the doubling of the deposit. Or it won't happen before whatever else.
P.P.S. At a passing backlash such owl will close positions as soon as it will jump a little in the pluses, and at counterpoise it will go down completely. So much for strategies.
I have a question to the author of the topic khorosh.
In order to show us a nice ascending graph on historical data, do you optimize an Expert Advisor? Or do you select "winning" parameters of the Expert Advisor based on some "fundamental" considerations?
I want to warn you a bit. Suppose the optimization of a strategy on historical data (say, for the last year or two) has identified "winning" parameters, in which your strategy is very successful. I am sure that if the strategy has many adjustable parameters, it is always possible to find such, at which it will be fantastically profitable. But can you confidently say that with the same parameters it will be successful in the next future year or two? Have you carried out such research on your strategy? Or have you investigated the possibilities of the strategy on just one period between 1999 and 2013?
I advise you to optimise your strategy with historical data, for example from 2005, and then run a system test with the obtained "winning" parameters on the data from 2006. If it turns out that your system is "losing" as easily in 2006 as it was earning in 2005, then I can disappoint you: your system is not the "grail" at all, and the success in 1999-2013 is just the result of "well" adjusted parameters, which you probably can't analytically determine for future unknown time periods.
I have a question to the author of the topic khorosh.
In order to show us a nice ascending graph on historical data, do you optimize an Expert Advisor? Or do you select "winning" parameters of the Expert Advisor based on some "fundamental" considerations?
I want to warn you a bit. Suppose the optimization of a strategy on historical data (say, for the last year or two) has identified "winning" parameters, in which your strategy is very successful. I am sure that if the strategy has many adjustable parameters, it is always possible to find such, at which it will be fantastically profitable. But can you confidently say that with the same parameters it will be successful in the next future year or two? Have you carried out such research on your strategy? Or have you investigated the possibilities of the strategy on just one period between 1999 and 2013?
I advise you to optimise your strategy with historical data, for example from 2005, and then run a system test with the obtained "winning" parameters on the data from 2006. If it turns out that your system is "losing" as easily in 2006 as it was earning in 2005, then I can disappoint you: your system is not the "grail" at all, and the success in 1999-2013 is just the result of "well" adjusted parameters, which you probably can't analytically determine for future unknown time periods.
Before I got into EAs with martin, I tried a lot of different TSs using different indicators. When I checked them it turned out that they could not work on the interval outside the optimization period, they usually fail. However, Expert Advisors with martin behave much better in this respect. Sometimes we even find that optimization of some months leads to parameters that can make an EA profitable over several years. That's why in the recent years I've been focusing on Expert Advisors using martin. Of course, I use indicators for generating entry signals and sometimes also use them for determining exit points. The main task is to find the parameters needed for the Expert Advisor to overcome the longest undeclined trend we've ever seen in our history with some reserve. Of course, this does not guarantee to fail. But at least, despite the possibility of rare drawdowns, one can make some profit, because the profitability can be quite high.
Before I started using Expert Advisors with martin, I have tried many different TSs using different indicators. When checking them it turned out that they cannot work with an interval out of optimization period, they usually fail. However, Expert Advisors with martin behave much better in this respect. Sometimes we even find that optimization of some months leads to parameters that can make an EA profitable over several years. That's why in the recent years I've been focusing on Expert Advisors using martin. Of course, I use indicators for generating entry signals and sometimes also use them for determining exit points. The main task is to find the parameters needed for the Expert Advisor to overcome the longest undeclined trend we've ever seen in our history with some reserve. Of course, this does not guarantee to fail. But at least, despite the possibility of rare drawdowns, one can make some profit, because the profitability can be quite high.
I still don't understand from the answer: can this particular EA boast that with the parameters optimised, for example, for 2005, it will successfully run further (at least in 2006)? It would be interesting to see the results.
Have you already carried out such studies using this particular Expert Advisor?
I want to point out that the parameters that need an Expert Advisor to overcome the longest reverse trend will surely work badly in a "normal" reverse trend. Don't you think so? My point is that in order for an EA to be profitable, it is necessary to constantly find the right "winning" parameters. And of course it cannot be analytically determined in advance. As they say, if I had known better, I would have lived in Sochi.
Quote: "But at least in spite of possible rare drawdowns one can make money in general, because the profitability can be rather high" Probably that is what LTCM's developers thought too. ;)
I'm asking this for a reason. I used to study this kind of martingale (as I remember it was FAPTurbo) and came to the following conclusions: it is the result of work of short-sighted programmers-businessmen.
Before I got into EAs with martin, I tried a lot of different TSs using different indicators. When I checked them it turned out that they could not work on the interval outside the optimization period, they usually fail. However, Expert Advisors with martin behave much better in this respect. Sometimes we even find that optimization of some months leads to parameters that can make an EA profitable over several years. That's why in the recent years I've been focusing on Expert Advisors using martin. Of course, I use indicators for generating entry signals and sometimes also use them for determining exit points. The main task is to find the parameters needed for the Expert Advisor to overcome the longest undeclined trend we've ever seen in our history with some reserve. Of course, this does not guarantee to fail. But at least, despite the possibility of rare drawdowns, one can make some profit, because the profitability can be rather high.
The greater the failure rate observed in the history, the less profitable the Expert Advisor will be. Have I got that right?
I still don't understand from the answer: can this particular EA boast that with the parameters optimised, for example, for 2005, it will successfully run further (at least in 2006)? It would be interesting to see the results.
And have you already conducted such a research using this particular Expert Advisor?
I want to point out that the parameters used for the "Expert Advisor must overcome the longest reverse trend" will surely work badly in a "normal" reverse trend. Don't you think so? My point is that in order for an EA to be profitable, it is necessary to constantly find the right "winning" parameters. And of course it cannot be analytically determined in advance. As they say, if I had known better, I would have lived in Sochi.
Quote: "But at least in spite of possible rare drawdowns one can make money in general, because the profitability can be rather high" Probably the same thoughts were used by LTCM developers. ;)
I'm asking this for a reason. I used to study this kind of martingale (as I remember it was FAPTurbo) and I came to the following conclusions: it is the result of work of short-sighted programmers.
The greater the failure rate observed in the history, the less profitable the Expert Advisor will be. Have I got that right?
I will express my opinion: this is not exactly true. If the parameters are set correctly and the account size is large enough, the Expert Advisor can overcome any drawdown and can even obtain the increased profit in such cases (the size of compensating trades increases, and with it the possible profit). The difficulty is to determine the universal parameters, if they exist at all, which is unlikely.
Look at the balance chart on this page. If the balance has stepped, the orders are closed when the trend reverses or corrects. The step is due to the accumulation of many orders during the previous trend reversal. Thus, on the contrary, after a backward trend, if it is successfully overcome, the deposit's growth is always great. But measures have to be taken and parameters have to be adjusted correspondingly, which will guarantee that the drawdown at the irreversible trend will not exceed the critical value, it is desirable it will be at least less than 50%, of course it will be better less, but it depends on how it will turn out.
Can we see the equity? The balance graph doesn't tell you much. I'm sure the equity level is pecking away at just about the "critical value".
And it wasn't about steps, it was about the amount of no-backlash that the owl is able to overcome. So this value is inversely proportional to its profitability.