Trend-tracking trading systems - page 2
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
Everyone needs one, but no one has one. At least, those who have one won't admit it. And those who say that they have one are cheaters.
As a first approximation, the result of the advisor's work for the day should look something like this
As a first approximation, the result of the EA for the day should look something like this:
I apologise for my earlier question...
Are you an adult? If so, are you all right in the head or do you just believe in fairy tales?
SZZ: As for the point, train a neural network in historical sampling and you will get a similar chart, but ... But only on history...
But I've come to the conclusion that if NS were effective in trading, Google would be the richest one, because they have really working neural network solutions - but only in other areas, not in trading
I apologise for my earlier question...
Are you an adult? If so, is your head OK or do you just believe in fairy tales?
ZS: As a matter of fact, train a neural network in historical sampling and you will get such a graph, but ... But only on history...
ZZZY: About neural networks, I can't say that I'm strongly interested in NS now, I used to be, but I've come to the conclusion that if NS were effective in trading, Google would be the richest, because they have really working neural network solutions - but only in other areas, not in trading
Yeah, well, if they were profitable, taxi drivers would be the richest.
A neural network cannot be trained to win the lottery.
I apologise for my earlier question...
Are you an adult? If so, is your head OK or do you just believe in fairy tales?
ZS: As a matter of fact, train a neural network in historical sampling and you will get such a graph, but ... But only on history...
But I've come to the conclusion that if NS had been effective in trading, Google would be the richest one, because they have really working neural network solutions - but only in other areas, not in trading
Let's leave our age and our shady past alone, the task at hand is clearly more interesting!
Neural networks for forex trading - is unconditional overkill, everything is much simpler.
What are we dealing with? A limited number of elements of the system we are analysing,
movements, events, states, conditions, the usual processing. Some points
are not obvious, sort of "hidden" and not yet clearly defined, but they are quite possible
to be solved, if only one had the will and the "oil in the head"!
Forum on trading, automated trading systems and trading strategy testing
Trend following trading systems
Sergey Lazarenko, 2018.05.17 19:00
I don't know a Trend Advisor (that brings profit), I know Trading Strategies (manual) that bring profit... To make a good robot, it will take a lot of time and money... Because there are so many nuances, the primitive type, if the price is above the average and Stochastic reversed, it simply doesn't work...
how to describe it, elaborate...
You still have doubts that such exist... What proof do you need?
If monitoring, just mine... but he is not even two months old, description of TS here, the creator of TS Marat Gazizov, trades manually, the account is not monitored (because he likes to pomartingainichat), in July 2017 started trading with 6 000 dollars, year ended 35 000, started in January 2018 with 1 000 now in May on account of 7 000 dollars, all transactions describes every time in a daily review....
how to describe it, elaborate...
You still have doubts that such exist... What proof do you need?
If monitoring, just mine... but he is not even two months old, description of the TS here, the creator of the TS Marat Gazizov, trades manually, the account is not monitored (because he likes martingianit), in July 2017 started trading with 6 000 dollars, the year ended with 35 000, started in January 2018 with 1 000 now in May on account 7 000 dollars, all transactions describe every time in the daily review....
1. If monitoring, only mine but it's not even 2 months old
3. the creator of the account does not monitor
i got a chuckle, it's nothing new in this industry (trading) where some crooks look for other good-natured simpletons to either teach them to trade for their money or attract them as an investor. i used to show him how to make a million on MT4 and trade publicly ;) - Your example is not an example.
What are we dealing with? A limited number of elements of the system we are analysing,
movements, events, states, conditions, the usual processing.
Unfortunately, from your phrase I can only confirm one thing - the state of the market order, it is either profitable or unprofitable, and even knowing this order state - it is not clear to close it or not....
ZS: how many system elements are limited? - there will be 60 minute ones in an hour bar? - well almost so, if you seriously decide to do it, then time series statistics, yes - a relatively working model, but you will only come to the conclusion that the resulting statistics have probabilistic characteristics. Try to formulate some sort of statement that requires statistical testing, there you will see.... for example the number of bull bars and bear bars is the same in a year.... - I doubt it, .... the number of zigzag tops would be the same in a year? .... even assuming that as many sales as there are buys, the sum of the lengths of the up and down zigzag should be the same? )))))
HH: what is the number of elements in the system limited? .... )))))
As a rule, it is the current, the previous and the next (at first, partially hidden) trend, and if there is a "continuation" - a "compound" trend. In general, the controlled elements and parameters are: previous, current and next prices, the moment of beginning, end and duration of operations, the nature of changes - growth and reversal of quotes, pullbacks, reversals and continuations, false reversals, etc., external and internal balance (and imbalance) of current operations.
how to describe it, elaborate...
You still have doubts that such exist... What proof do you need?
If monitoring, just mine... but he is not even two months old, description of the TS here, the creator of the TS Marat Gazizov, trades manually, the account is not monitored (because he likes martingianit), in July 2017 started trading with 6 000 dollars, finished the year with 35 000, started in January 2018 with 1 000 now in May on account of 7 000 dollars, all transactions describe every time in a daily review....
I have no doubt there is no such thing. But also the rules of the strategy get lost when the market changes! Refinements appear, and it's already a different strategy. And in order not to lose money we switch on martingale. This is dangerous. Although it is possible to increase the deposit by means of martin. That is, the rules are as follows: if the strategy makes a loss, we open with a martin. And so we open with martin until the profit. That is, any manual system is relevant only for this market. The market has changed, and the strategy is already losing!
As a rule, it is the current, the previous and the next (at first, partially hidden) trend, and if there is a "continuation" - a "compound" trend. In general, the controlled elements and parameters are: previous, current and next prices, the moment when operations start, end and duration, the nature of changes - growth and reversal of quotes, pullbacks, reversals and continuations, false reversals, etc., the external and internal balance (and imbalance) of current operations.
Well, finally we come to the fact that your task is not formalized, you have already cited a figure with a zigzag, this figure can somehow be "stretched" to the concept of a trend:
"in an uptrend the subsequent peak on the chart should be higher than the previous ones, in a downtrend the subsequent declines on the chart should be lower than the previous ones (seeDow Theory). "
I.e., to find a trend, we should analyze a zigzag - as long as the zigzag tops update highs (lows), it is a trend
omit setting of the zigzag for the time being, let there be some kind of an ideal self-tuning ZZ
So, we have a trend following system based on a phase, and what? We have detected a trend - let it be anuptrend (bullish) ,
How many nodes in WP should be considered the beginning of the trend? - maybe the trend is over? ....
Well, to finish my vision of the topic, for I have looked it all up - trends exist only on the history, as well as patterns, you can not say that there are no trends, no patterns - yes they do, but the time code they will appear on the chart so far no one was able to predict (I lie - insider information can help)
Trends and patterns exist only on history!
They don't have any effect on the market, we just have to look at the chart.
They all have to be predicted. ZZZY: Level systems work somehow, trends work the same way as astrology, Gunn, coffee grounds.... - i don't disagree, it works for some people but it doesn't work for me