Non-fitting system - main features - 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
to Svinozavr
Honestly, I don't want to trample on your own beloved carrot field. It is more correct to remind that everyone has his own opinion, and I respect yours, at least I still do, despite the harshness that I made in your address, but again, not on a level playing field.
К тому, что при решении задач по геометрии в школе поток котировок не используется. Вы это все время как-то упускаете
Easier to agree, indeed you are right. Schoolchildren will never build this in their lifetime:
https://www.metatrader5.com/ru/terminal/help/objects/fibo/fibo_arcs They can still manage it somehow in the woods, on the lake, but no way on the quotes. Really missing out, as it uses the most sophisticated analysis of the quotation process. Forgive my unprofessionalism. Amateur, what can I say.
Been drinking?
No, but I like your approach, you're trying to get right to the root of it:
Bullshit? ))) Then what's TA to you? All right, then. If you think TA is just MA, then yes, bullshit. Why am I saying that? It's mathematics - it's not TA. Then it's not clear at all. Oh... I get it - TA doesn't exist! No. It doesn't work either.
You know, there's an old joke about it. Young Jew comes to an old, elderly man and asks "Tell me, how is your health". He grunts, looks at him attentively, and replies, "Eh, young man, do you have the time?" What is TA to me? It's four years of wasted time and realising it's bullshit. And for you, it's still:
Easy. You stop bullshitting and we come to a compromise.
Well, okay, so that you finally understand - have as much fun with your analysis as you like. Who's stopping you?
The signs of a working system are that it optimises almost identically on both sampling and double sampling with a constant lot. For example, we take 10,000 bars. We divide it approximately in half, i.e. by 5000 bars. Optimize it for 5000 bars. Then we apply it to 10 000. If the profit factor remains almost unchanged or has insignificantly changed (becomes independent of the sample length), the system is likely to pass the forward test. Of course trades should be about 600 - 1000 on 10 000 bars (300 - 500 on 5000 bars).
Unfortunately I only have a printout
"Checking statistical significance with chi-square test".
The chi-square test is one of the most accurate statistical methods for determining whether a given indicator reading is trustworthy. The formula is: (/a1 - e1/ - 0.5 )^2 : e1 + (/a2 - e2/ - 0.5 )^2 : e2
/ abs.value
a1 - observed frequency of the market (e.g. growth) result 1
e1 - expected (theoretical) frequency of the market of result 1
a2 - observed frequency of the market (e.g. growth) of result 2
e2 - expected (theoretical) market frequency of market result 2
Example: market went up 669 Mondays
market went down 865 Mondays
Total Mondays 1534
But the market grew 52.1% of days (not only on Mondays) that is 799 days, so the first term equals (/669 - 799/ - 0.5)^2:799 +(/865-735/-0.5)^2:735 = 129.5^2 :799 + 129.5^2 : 735 = 43.81
This is a highly significant result with 99.9% confidence level.
Unfortunately it's only Dow ;)
Any doubts about this method?
I.e. you need to learn how to highlight the rightness/ wrongness contexts for each and every signal. To be more precise, you should make such a selector, teach it (set it up) and use it intelligently. That's all.
The only thing to do with higher probability "sometimes tells the truth, sometimes lies , sometimes speaks the truth, sometimes rubbish". It would be much easier, but I think we have to learn "to forecast the market" with the same probability, and if we can do it, why do we need the "set of indicators (or TS - whatever at this stage)"? We go back to the starting point, having made a detour and made our work more difficult...
The matter is to do it with probability more "sometimes tells the truth, sometimes lies, sometimes speaks the truth, sometimes speaks nonsense ". It's easier, but I think that for this we must learn "to forecast the market" with the same probability, and if we succeed in it, why do we need this "set of indicators (or TS - it makes no difference at this stage)"? We go back to the starting point, having made a detour and made our job more difficult...
You see, if I end up with a set of indicators whose predictive value is not 50% but, say, 60, I don't think my further work will get any more complicated.
I might even do a couple more of these hooks. :) I'm quoting realistic figures, by the way. That's what I've got.
But of course you can also flip a coin (like MACD or RSI). Or meditate on the chart. I wish you good luck, if it's OK. :)
It is not the first year that people have been relatively successful in using the overnight trading strategy inside the channel on cross pairs. The optimizer improves the performance, as a rule.
The optimizer has nothing to say about the robustness of the strategy. But this strategy is still working (judging by profits) and we do not know how long it will remain so.
The above is a vital example of how optimisation cannot guarantee anything. However, it is able to have an impact on profitability figures.
There is also a special ALWAYS profitable system - arbitrage. This strategy has unique properties:
1. Does not use TA (price history analysis).
2. No input parameters.
3. No concern with the nature of the market. Its inefficiencies are exploited.
I.e. there is at least one robust strategy that cannot be optimised.
You see, if I end up with a set of indicators whose predictive value is not 50% but, say, 60, I don't think my further work will get any more complicated.
I might even do a couple more of these hooks. :) I'm quoting realistic figures, by the way. That's what I have.
But of course you can also flip a coin (like MACD or RSI). Or meditate on the chart. Good luck, if anything. :)
Good luck to you too :), but I disagree again)
Indicators have no predictive value, indicators are opium for ATS. Indicators (MACD and RSI) are a kind of relic of manual trading (sorry if I am hurting someone's feelings). ATS is able to work with the source - the price. If you may not agree with me at once, but yes, it's your business, and I have no such goals.)
Good luck to you too :), but I disagree again)
Indicators have no predictive value, indicators are opium for ATS. Indicators (MACD and RSI for sure) are a relic of manual trading (sorry if I am hurting someone's feelings). ATS is able to work with the source - the price. I don't think you would agree with me right away, but yes, it's up to you, and I'm not intending to do that.)
Of course I won't agree, you're absolutely right. I was having a laugh. I did not put indicators and TS next to each other. What difference does it make? Indicators also work with the source. Or do they? I tried to formulate the layered principle of building a robust context-independent system, and you tell me that the input of the system is price.
I'm not against it at all. The input is the price, and the output is the profit. And what is in between? :) :)
The systems under discussion must have one, very simple attribute - a complete lack of optimisable parameters.
+1, except perhaps for the MM and risk parameters.