Stops - page 14
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
In the age of OOP, C++, VB.NET, API, MAPI, SHMAPI, MQL4 & 5, ....... I don't understand how this is even possible???----------------------------------
No, just imagine a similar "technical failure" in the control system of an aircraft, nuclear power plant, power supply, armaments, etc.
----------------------------------
Z.U. For a modest fee, I can write a code that eradicates such technical malfunctions. )))))))))))))
Here. Turns out the satellites crashed recently at your mercy.
Because you didn't write them a code. You let the entire globe down... eh
The speech is not that of a boy, but... not a girl. )))
Agreed. But is that all there is to it?
Well, if they want to continue, let's go on.
There is also the opinion of learned men:
I deliberately truncated the post, because what I said applies to any type you can think of. The task of setting stops is equal in complexity to developing the profit part itself, but with the additional requirements - it is even more difficult than it seems at first glance.
This seems logical and true, considering the fact that stops are more a part of the trading system than anything else. Stops determine critical exit points, which are very lucky to find. Based on even one such point, you can easily build a trading system with a positive MO. However, I can argue that any, even an ineffective stop (i.e. a stop that does not use a critical market point) will have a positive effect, albeit a very small one.
I have always been a proponent of stops, although I have not understood why they work. Tests have invariably shown their effectiveness, albeit very weak (the MO of positive TS increased, though not by much). At the same time, the stops I used and use now are not effective in and of themselves. Their breakpoints are no better than any other random point. But for some reason, they worked poorly, yet. The question is why?
My theory is simple. As mentioned above, on the scale of a normal distribution, the width of a stop will be balanced by its probability of execution. That is, the wider the stop, the greater the loss it will incur, but the less likely it will be executed. The narrower the stop, the less losses it will generate, but the more frequently it will trigger. As a result, whatever way you look at it, we have a zero effect from its use. But it is true only for normal distribution, for example on the random walk charts. In reality, markets are not random number generators, they obey economic laws and their distribution is not normal, though very close to it. The distribution of any market will be elongated and there will be tails along its edges. I specifically drew by hand a basic, very rough and hypertrophied diagram of the distribution of any market. I drew this picture on my knee and built it by eye (by the way, this is how VAZ cars do, judging by their advertisements), so do not judge strictly:
What does this scheme mean? First of all, we have to be very much afraid of these very tails. They are rare, but if we catch them from the wrong direction, we will fly in very quickly. We don't know when they're coming or their strength or vector, but we know they're coming. The only thing we can do is to cut them off with hard stops. There's no more effective tool than a simple stop: it' s guaranteed to shave those tails off, keeping them out of bounds. You can take any value as a stencil for the haircut: a certain number of standard deviations or an ATR value or a fixed value expressed in pips. Whatever you want, it doesn't matter as long as they shave those tails.
The second difference between this pattern and the normal distribution is that it is stretched, i.e. the market shows the property of reversion. Price is more likely to reverse than continue its movement. Since stop orders are "buy" or "sell" orders above or below the current price, it is more likely that the price will reverse before it reaches the stop (for the maximalists, "most likely" does not mean 75% chance by 25%, but rather 51% chance by 49%). If so, the stop will generate slightly less loss than it should on a normal distribution. This is the second advantage of using protective stops.
For the same reason, using profit levels (Take Profit) is much less effective than using protective stops. Firstly, we never catch the tails that develop in our favor (because we also shear them), and secondly, the probability that the price will not reach the necessary amount and turns around is lower than normal, so we will lose more on returns than we would with normal distribution. For the same reason, pulling up the stop is also inefficient. The price will more often than not randomly reverse and execute our pulled stop, therefore more often than not we will not make profit.
Overall, this theory very accurately predicts the effects discovered by the masters Jeffrey Owen Katz, Donna McCormick and Larry Williams in their books "The Encyclopedia of Trading Strategies" and "The Long Term Successes of Short Term Trading". In particular, almost all the strategies in which the entries were replaced with Limit orders have shown much better results than with stop orders (think for yourself why). Larry Williams has a clear answer too: any profit limit (Take Profit) reduces the final result, his practical tests confirm it (see chapter 3 "The key secret of short-term trading").
Well, as long as we have to continue, let's continue.
This is also the opinion of pundits:
This seems logical and true, given the fact that stops relate specifically to the trading system rather than anything else. Stops define critical exit points, which are a great deal of luck to find. Based on even one such point, you can easily build a trading system with a positive MO. However, I can argue that any, even an ineffective stop (i.e. a stop that does not use a critical market point) will have a positive effect, albeit a very small one.
I have always been a proponent of stops, although I have not understood why they work. Tests have invariably shown their effectiveness, albeit very weak (the MO of positive TS increased, though not by much). At the same time, the stops I used and use now are not effective in and of themselves. Their breakpoints are no better than any other random point. But for some reason, they worked poorly, yet. The question is why?
My theory is simple. As mentioned above, on the scale of a normal distribution, the width of a stop will be balanced by its probability of execution. That is, the wider the stop, the greater the loss it will incur, but the less likely it will be executed. The narrower the stop, the less losses it will generate, but the more frequently it will trigger. As a result, whatever way you look at it, we have a zero effect from its use. But it is true only for normal distribution, for example on the random walk charts. In reality, markets are not random number generators, they obey economic laws and their distribution is not normal, though very close to it. The distribution of any market will be elongated and there will be tails along its edges. I have deliberately hand-drawn a basic, very rough and hypertrophied diagram of the distribution of any market. I drew this picture on my knee and built it by eye (by the way, this is how VAZ cars do, judging by their advertisements), so do not judge strictly:
What does this scheme mean? First of all, we have to be very much afraid of these very tails. They are rare, but if we catch them from the wrong direction, we will fly in very quickly. We don't know when they're coming or their strength or vector, but we know they're coming. The only thing we can do is to cut them off with hard stops. There's no more effective tool than a simple stop: it' s guaranteed to shave those tails off, keeping them out of bounds. You can take any value as a stencil for the haircut: a certain number of standard deviations or an ATR value or a fixed value expressed in pips. Whatever you want, it doesn't matter as long as they shave those tails.
The second difference between this pattern and the normal distribution is that it is stretched, i.e. the market shows the property of reversion. Price is more likely to reverse than continue its movement. Since stop orders are "buy" or "sell" orders above or below the current price, it is more likely that the price will reverse before it reaches the stop (for the maximalists, "most likely" does not mean 75% chance by 25%, but rather 51% chance by 49%). If so, the stop will generate slightly less loss than it should on a normal distribution. This is the second advantage of using protective stops.
For the same reason, using profit levels (Take Profit) is much less effective than using protective stops. Firstly, we never catch the tails that develop in our favor (because we also shear them), and secondly, the probability that the price will not reach the necessary amount and turns around is lower than normal, so we will lose more on returns than we would with normal distribution. For the same reason, pulling up the stop is also inefficient. The price will more often than not randomly reverse and execute our pulled stop, therefore more often than not we will not make profit.
Overall, this theory very accurately predicts the effects discovered by the masters Jeffrey Owen Katz, Donna McCormick and Larry Williams in their books "The Encyclopedia of Trading Strategies" and "The Long Term Successes of Short Term Trading". In particular, almost all the strategies in which the entries were replaced with Limit orders showed much better results than in the case of stop orders (think for yourself why). Larry Williams has an unambiguous answer: any profit limit (Take Profit) will decrease the final result, his practical tests confirm it (see Chapter 3 "The Key Secret of Short-Term Trading").
. The only thing left to do is to add code to make it show profit, at least a little bit. But that's it. :)))
I have something from the old one. I will dig around, maybe I will find some pictures (without stop and with stop). The code itself with a positive MO is not in my plans to lay out, so do not judge.
I had some of the old ones. I'll dig around and see if I can find some pictures (without stop and with stop). I don't plan to post the code itself with a positive MO, so don't be discouraged.
Yes, you don't need positive MO. It is necessary that at least it was visible. :))
A good account would require a special expert to investigate the effectiveness of random protective stops. Targeted trading systems are not quite suitable for this purpose, as the results may be distorted by non-random properties of the trading system. The work of this Expert Advisor should be based on a random number generator. In general this is a pretty simple solution, then I really cannot spend a few days to create it. If anyone is interested in practicing programming and actually supporting the conversation between scientists, but I'm ready to give the terms of reference. When using stops in it, we should observe certain effects predicted by theory. In general who is interested in povodit for the sake of science - write here.
p.s. Honestly, I don't know how sound this theory is, but it has indirect practical proof.
Some people recommend Stop Loss, others shout Stop-Limit Order, while others say that lock is evil!
Imho - it depends on the strategy, if the strategy is upside down - then Stop loss, if the strategy has take outs of 10-20 pips - you can use lots too.
Well, if they want to go on, let's go on.
There is also the opinion of learned men: ...
I still don't understand what exactly you were arguing from my statement.
The second difference between this pattern and the normal distribution is that it is elongated, i.e. the market exhibits a return property. Price is more likely to reverse rather than continue its movement...
It depends on the time series classification you adopt. If you investigate reversals based on EZ approaches, then yes, there will more often/always be a reversal. But if you use other classifications, like stopping at arbitrary points, or compare the ZZ segment statistics - you'll get about 50/50.
Is your diagram a returnee statistic, i.e. x(n)-x(n-1)? If so, what does what you say have to do with returnees? Or do you have a transaction duration of less than one count? It seems to me correct to use the structural function of the process or the law of the double logarithm for such reasoning, rather than the GLISTogram of first counts.
PS: and especially don't trust the surnames Katz, those surnames, may have very different purposes :o)
I still don't understand what exactly you were arguing from my statement.
It depends on the time series classification you adopt. If you investigate reversals based on the ZZ approaches, then yes, there will more often/always be a reversal. But if you use other classifications, like stops at arbitrary points or compare the statistics of the ZZ segments - you will get about 50/50.
Is your diagram a returnee statistic, i.e. x(n)-x(n-1)? If so, what does what you say have to do with returnees? Or do you have a transaction duration of less than one count? It seems to me correct to use the structural function of the process or the law of the double logarithm for such reasoning, rather than the GLISTogram of first counts.
PS: and especially don't trust the surnames Katz, those surnames, may have very different purposes :o)
Would you like to call Katz?
-- You know, Shura," Panikovsky went on, "I somehow stopped trusting Bender. He's doing something wrong.
-- Well, well! -- Balaganov said threateningly. --You weren't asked.
-- No, seriously. I have a great respect for Ostap Ibrahimovich: he is such a man! Even the Pound, - you know how I respect the Pound, - said about Bender, that he is a head. But I'll tell you, Shura:
Pound the Donkey! By golly, he's such a fool. He's just a miserable, insignificant person. And I have nothing against Bender. But there's something I don't like. I'll tell you, Shura, as one of my own...
No, I don't need Katz. I'm my own Katz.)