Some signs of the right TCs

New comment
 

Market patterns do not change in the cases of

  • Multiplication of symbol prices by a non-zero constant.
  • Symbol flip (1/Symbol).

As a conclusion, proper TS should give identical trade signals when run on any custom symbol derived from the original action described above.


For example, we took the EURUSD. We ran the TS, getting a series of entries.

Then we created the symbol 100/EURUSD. Then we ran the TS. Entries should coincide with the original ones.


If this does not happen (99%), the TS is not written correctly.


How TS should react to the symbols, raised to some degree - I do not understand.

 
i would add the main thing:
  • has no period-dependent parameters.
  • operation of the TS does not depend on the current chart timeframe
  • the operation of TS does not depend on the symbol-instrument
  • the whole setting of TS is only the setting of risk management (the size of the used deposit)
 

I'm afraid it's not that clear-cut.

The appearing (or disappearing) difference in roundings plus the "butterfly effect" may well lead to very different results, even though the TS will be written correctly.

 
Georgiy Merts:

I'm afraid it's not all that clear-cut.

The appearing (or disappearing) difference in roundings plus the "butterfly effect" - may well lead to very different results, although the TS would be written correctly.

Or the "hippo effect": - you don't have to be agile and run fast, you just have to be thick-skinned and have a big mouth).

 
khorosh:

Or the "hippo effect": - you don't have to be agile and run fast, you just have to be thick-skinned and have a big mouth).

And mice, so it's no big deal)

Sabker, what's the crazy idea behind this, what's the point?
 
Georgiy Merts:

The appearing (or disappearing) difference in roundings plus the "butterfly effect" may well lead to very different results, although the TC will be written correctly.

To clarify, we are talking about mathematical operations, not computer operations. I.e. one third is one third. The root of the PI is the root of the PI.


I will also add that TC profitability/loss estimates are irrelevant to the topic. I'm sure everyone who has risen well with algotrading has used the wrong TS.

 
fxsaber:

I'm sure everyone who has risen well with algotrading has used the wrong TS.

There is nothing wrong with thatParadox Parrondo

fxsaber:

For example, we took EURUSD. We ran the TS and received a number of entries.

Then we created the symbol 100/EURUSD. We ran the TS. Entries should coincide with the original ones.

If this does not happen (99%), the TS is not written correctly.

Well, multiplication by a constant should not check anything at all

I would have thought of more complicated tests. I suspect that if we add something periodic ( sinus?) to correct TS input data then at least the number of deals should remain the same ? - generally there is something to think about,

 
Igor Makanu:

well, multiplication by a constant should not check anything at all

This is the easiest/initial way to test your TS.

I would think of more complicated tests, I suspect that if the correct TS is mixed with the input data, something periodic ( sine ?) then at least the number of trades should remain the same ?

It is normal for the results to differ on different data.

 

If multiplication by a constant gives a "strategy failure", then what should be in this strategy, I can't even think of it))

w.s. oh I got it - circular levels

 
Aleksey Mavrin:

If multiplication by a constant gives a "strategy failure", then what should be in this strategy is interesting, I can't even think about it))

For example, the binding to pips. In particular, take/stop in pips.

 
People who really make money in the market, not people who spend their lives making theories, they don't give a damn.
12345678910111213141516171819202122232425262728293031323334353637
New comment
Reason: