Now let's be clear, do patterns work? Let's discuss) - page 7
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
If there is NO pattern there should be roughly 50% winners 50% losers from each trade, and in both cases the DC earned on spreads and other bullshit. Moving on. OFFICIALLY we have 5% winners and 95% losers. QUESTION. WHY? So there is a pattern.
Where does it follow from your post that there is a pattern?
Maybe from the fact that if there were none, we would have 50% winners 50% losers? But this statement is only true for the case where there is no DC commission. In any other case (when there are no patterns) we would have 0% winners 100% losers for a long time interval. Is it an obvious statement? It seems so, because the presence of any spread, (even 0.01 points per transaction), will sooner or later ruin any player in a random market.
And the fact that according to official statistics we have 5% winners and 95% losers tells us two likely facts:
1. There is no regularity, and the time frame in which players go bankrupt is not long enough for reliable statistics.
2. Regularities are present, as evidenced by the 5% of traders who have survived.
Thus, it is impossible to draw any conclusions about the market from analysis of its players' success/failure, without additional information about their getting rich/poor dynamics. In general, I would say that this is a very perverted way of estimating latent regularities about time series.
Classic time-tested pattern....
The trend is more likely to continue than to close. I open any daily chart and make sure... I take the last 300 (for example) days, if it didn't work, the trend would change 300 times (up and down) but we see that it doesn't. Can anyone object to this?
I take the last 300 (for example) days, if it didn't work, the trend would change 300 times (up and down) but we see that it doesn't. Can anyone object?
And it doesn't have to change 300 times on a random walk chart either. Visually, it is difficult to tell where the trend continues and where it reverses. There is, for example, a numerical estimate - the Hurst coefficient. On a long history it will be about 0.5 - i.e. no persistence, but it does not mean that it is impossible to identify individual moments when it exists.
Classic time-tested pattern....
The trend is more likely to continue than to close. I open any daily chart and am convinced...
Your assertion is very easy to formalise and then verify.
Indeed, continuation of the trend, is equivalent to the fact that the colour of the next candle will be the same (on average) as the previous one. This is true both for the uptrends and for the downtrends. Now let's take a long enough period of time containing many candlestick bars, and count the number of candlesticks of the same colour standing side by side. Obviously, if this fraction is larger than 1/2, we have a trend market, or as you correctly noted, "the trend will continue rather than close". If we get a value smaller than 1/2, then we have a pullback character of the price movement and "the trend is more likely to change than to close".
Now let's take this coefficient on the daily bars for all currency pairs and calculate it. And what do we see? Everywhere, this value is a little less than 1/2, i.e. the condition "the tendency is faster to change than to close" is usually fulfilled for all symbols and you, RomanS, are wrong in your statement "The classical time-proved regularity ..."! - it was true under Tsar Gorokh in the second half of the last century.
Your statement is very easy to formalise and then verify.
Now let's take and calculate for all currency pairs this coefficient on daily bars. And what do we see? Everywhere, this value is slightly less than 1/2, i.e., the condition "the trend reverses rather than closes" is usually fulfilled for all symbols, and you, dear RomanS, are probably wrong in your statement "The classical time-tested regularity ..."! - it was true under Tsar Gorokh in the second half of the last century.
A trend is the possibility of certain events evolving in a given direction. One can argue on this topic for a long time and everyone will have his own opinion. Let's take a daily chart of the oil for the last (for example) 2 years... You say that the trend has changed many times, and I say that it was increasing until July 15, 2008, and decreasing after that, i.e. the change over 2 years has been only one - July 15. But I do not argue with you.... everyone has an opinion. It's all a matter of which way to look at it
Yes, why is it so complicated?
You define a logical framework on which to base your analysis and we will do it. My suggestion to you is to consider adjacent bar colours as an indicator of the trend. Is that logical? - Yes. Did you do it? - Yes! Is it trending? - No! What else do you want?
Do you want to count something other than the trend? You're welcome! The main thing is to define what you want and how you want to count.
Otherwise, opinions... this and that. There is a truth and it is one for all.
Yes, why is it so complicated?
What's so complicated about it? The topic of the thread is "Now let's be honest, do the patterns work? Let's discuss them)" The most important question is whether they work or not? If you think that they do not, then why even trade? The point of any TS is to identify the trend, open a position and expect that this trend will continue for some time..... Another question is how to identify it and that's a separate topic.
No, no, I think there is a pattern! And that is why I have been studying this question for more than a week.
As for the way of its detection, I believe that nothing can solve this problem better than NS (unless, of course, there is a priori knowledge of the process, which there is not).
-I've lost it!
Where?
-Yes there .... y market
- then why are you looking here under maths
- it's lighter here