From theory to practice - page 264
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
I agree with the first part.
If this is true, then the solution to the problem should be built exactly the opposite.
I.e., if now I work in a moving exponential time window, and in it I count real ticks and intensity of trades, then it will be so - I should work with the moving volume of tick sample and count for what time this tick volume "gathered", and already from there to find intensity...
But, to do it, one must prove the exponentiality of time intervals between ticks. Prove it convincingly.
This will be a real breakthrough in solving the problem.
It's a little different:
there was a sell signal, but price keeps going up for a while... we wait for the muvings to cross above the sell signal, i.e. we reduce the chance of catching knives
same thing with stop orders, we put a sell stop at the sell signal level, if price went against us by n-pips, and trails behind the price until it triggers
in both cases we get a better opening price for the signal, but we may miss some if the price goes immediately in the direction of the forecast. But in this case we can open in portions - some at the current signal and some at a better price. It may seem like a small thing, but sometimes it significantly increases the probability of winning.
I like the second suggestion. I didn't understand at first what you mean about entering with a stop order. Such an entry can indeed give a better price. However, I'm not sure it would work as a filter for losing trades. And we should also take into account that all logic of trawl has to be implemented in an Expert Advisor because in this case we will not be able to use pending orders. There is a StopLevel and the terminal has no trawl mechanism for such orders. But as an idea to enter at a slightly better price than the system enters now - it is a good suggestion!
About the muwings - in most trades there will not be a crossover above the level where the diffusion signal was received. The system is counter-trend, which means that at the place where the signal comes in, the price has probably already moved away from the truck. Before there will be a MA crossing, the price itself will have to cross a slow moving average in the opposite direction - because the price is MA1 - the fastest moving average. Actually, even if you don't have a tester, Alexander can test how the filter will work by crossover of moving edges! After all, he has a couple of months of history of trades and... The family! They are waiting for the grail - they can help! You can send them to their terminals - put trades on the charts, overlay the slips, and manually filter the last months trades. Routine and people make mistakes, but carefully and slowly, motivated by the grail - they will do it! Then the result of filtering can be compared directly in dollars.
Opening in portions, averaging, and refilling are all ManiManagement. It is a good thing, but first, you should have a strategy and then you can implement MM in the bot. Mani Money Management is never a filter - it is only a booster of expectation of a trading system.
Then the result of the filtering can be compared directly in dollars.
What a great motivational phrase!
Serge, you have really brought colour and power to this thread! My respects.
Many interesting things could be done with these transformed series, and not only counter-trends could be traded
It would be possible to create a custom custom symbol on transformed series, and use it to back up any system, including NS.
But there is no article or anything, only whissim waving to us slyly :) and alexander chuckles, putting bags in a pile
2. I WANT a stream of events (quotes) without an aftereffect, i.e. with exponential time intervals between events.
Alexander, what makes you think that by changing the interval between ticks you get rid of aftereffects? How does one mean the other? I don't see any connection at all, it's like mixing green and sour)
I've jumped into the grail chase with the members of the "Machine Learning..." thread. While Misha has gone into a rampant stupor there, we need to get ahead immediately. There are plenty of simple fake Grails, I agree. There's only one gold one. That's what we're running after.
Alexander, what makes you think that by changing the interval between tics you get rid of the after-effects? How does one mean the other? I don't see any connection at all, it's like mixing green and sour.)
Well, it seems to be written everywhere that a process with exponential intervals between events is Markovian, with no aftereffects. No?
We still can't completely remove the "memory". We get a kind of emulation - a pseudo-Markovian process.
I've jumped into the grail chase with the members of the "Machine Learning..." thread. While Misha has gone into a rampant stupor there, we need to get ahead immediately. There are plenty of simple fake Grails, I agree. There's only one gold one. That's what we're running after.
I stumbled on a gold mine myself, posted about it a couple of times in a thread - no one checked out the Klondike, now I'm getting it right with my own efforts and choosing an apartment in Australia
"Everywhere" - where? Can I get just one link?
Without consequence is when a future price change does not depend on previous price changes. I don't think intervals had anything to do with it at all.
"Everywhere" - where? Can I get just one link?
Without consequence is when a future price change does not depend on previous price changes. I don't think intervals had anything to do with it at all.
The first one I found.
http://stu.sernam.ru/book_rop.php?id=49
http://poisk-ru.ru/s32113t2.html
AndNovaja gave some links somewhere in the thread.