Market theory - page 66
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
Gentlemen! Having doubts about the algorithm, I ran an experiment on it without changing anything in the settings and instructed it to analyse the entrepreneur's trading process, when he bought $10 worth of goods and started selling them in a competitive market at C prices, earning income of D. The algorithm performed a complete analysis of the entrepreneur's work and passed the test with flying colors, accurately calculating profit according to both the traditional methodology, as the difference between income and expenses and according to my profit formula, as well as, the levels of virtual prices Ts1, Tsopt, Tsr, Ts2 and Tspr (!) that we know:
The calculation of profits is a perfect match:
Conclusions of their analysis of the process of trade: The businessman begins to make a profit at once. Having established the sale price Ц > Цпо = 10 $, the maximum profit is received at Ц = Цот = 10,91 $, trade stops at Ц = Цпр = 11,90 $, that, completely corresponds to logic of process of trade and is confirmed by Fundamental representation of profit as difference between income and all kinds of expenses.
The calculation of the levels (our chart of virtual levels), indeed, shows that, the market is led by the Bulls, as the price has been increasing:
Conclusion: The algorithm reflects the market situation absolutely adequately and its results can be trusted.
Yusuf, your algorithm is very sensitive to input data changes. This is indirectly proved by the huge spikes - remember the small delta fission effect - it's very similar to these spikes. You don't do any data preprocessing, but there is noise, and the first and most obvious one is the sampling noise.
The algorithm gives different, sometimes opposite results/recommendations, if it is calculated on a daily candle, but at different times of the day. It proves that the algorithm is not coarse, or in other words, it does not have robustness properties. But to ensure roughness of the algorithm, we should again eliminate/eliminate noise and separate the signal itself from the mixture of "signal+noise".
Yusuf, your algorithm is very sensitive to changes in the input data. This is also indirectly indicated by the huge spikes - remember the small delta fission effect - these spikes are very similar to similar effects. You don't do any data preprocessing, but there are noises, and the first and obvious one is the sampling noise.
The algorithm gives different, sometimes opposite results, if it is calculated on a daily candlestick, but at different times of the day. And it shows its coarseness, or in other words the algorithm does not have robustness properties. But to ensure roughness of the algorithm, we should again eliminate/eliminate noise and separate the signal itself from the mixture of "signal+noise".
It's not about the excel, now we're figuring out at what point it's best to give a signal. I propose to buy when the price breaks away from the yellow line upwards, and you give when the red line reaches the price. Your signal is later than mine.
The main problem is determining when the event you are talking about is happening, for that you need to organise online market tracking.
Doesn't the chart in the post2015.05.21 23:44 show this?
Now I understand you, the point is that the yellow line won't fall until the red line strikes, and the yellow line falls by holding on to the last data with the hinge. So we're talking about the same thing. The main thing, as long as you manage to pick out adequate entry and exit signals, as in the case of the BAY at the beginning of Friday'ssession https://www.mql5.com/ru/forum/58256/page68, which is now working out.
Market status at 13-00 MSK. Calm , competitive , bullish market:
Now I understand you, the point is that the yellow line won't fall until the red line hits, and the yellow line falls by holding on to the last data with the hinge. So we're talking about the same thing. The main thing, as long as you manage to give adequate entry and exit signals, as in the case of the BAY at the beginning of Friday'ssession https://www.mql5.com/ru/forum/58256/page68, which is now being worked out.
I cannot tell you that the Bears (yellow line) let the price go and fell (not went down, they fell) because they were hit by the Bulls, at the time of the hit, or, immediately after the hit, I gave the signal to BAY. We are talking about the same thing.
You posted the chart when the red line touched the price. Isn't that right? If we were talking about the same thing, you would have posted the chart earlier, when price broke away from the yellow line. I don't know how it really is, but you can see on the chart that these events are not happening at the same time. Unless on the horizontal we have the time axis.
Oleg, hi, I am strongly against dividing a signal into a main signal and noise. There is no such thing as "noise" worth billions of dollars, perhaps the noise is the useful signal.
Yusuf, there is no "" noise" worth bill ions of dollars", there isn't.
And if you think"noise is the useful signal", then by that logic, you should be tracking every tick. And there isn't.
I'll try to make my point clear with this picture:
The signal is clearly highlighted in this section - the line moving DOWN. But the bars themselves, superimposed on the signal line, show a mixture of "signal+noise".