Hearst index - page 46

 
 
nikelodeon:
By the way, there is such an indicator, the invariance index. It looks at the slope of the regression line answering thus the question if there is a trend in the market or not. I have not got time to use it, but I want this index to estimate not the price movement in the market, but the balance curve movement. It would allow to see if there is balance change or not. We determine the direction, if it is going upwards, we mirror the signals, if it is going upwards, we repeat it. It seems to me that any system may be pulled out by the ears with such a helper .... I think any TS can be pulled out by the ears with such a helper... What do you think?

It is difficult to explain how this indicator is interpreted: >=0.5 - flat <0.5 - trend - does not correspond to reality - in fact approximately >0.39 is a flat, but it not always identify the flat... In some cases, it coincides with some others, it lags, in third ones it shows nothing - but after the thorough analysis you understand that the indicator tries to adapt to different periodic structures of the flat, but unfortunately it does not succeed.

 
I'm afraid the invariance index, as well as Hirst's, won't tell you anything. Since you don't bother (as far as I understand) to ask for which window to build it for? Usually they build it for a fixed one and then try to draw it into the analysis...
Have you tried to build Hearst (or invariants) - for a smoothly increasing window? If so, you've probably seen that this indicator will float as it pleases from 0 to 1 (or a little bit Already the interval) . So - what period will you take? For example for 60 minutes you will have 0.2, for 120 minutes = 0.8. Which one do you prefer?
Just in the light of this it's amusing to hear phrases like "not true" and "in real life this"...
 
what if we apply the invariance index to the balance curve??? that was the question.....
 
IronBird:
I'm afraid the invariance index, as well as Hirst's, won't tell you anything. Since you don't bother (as far as I understand) to ask for which window to build it for? Usually one builds it for a fixed one and then tries to draw it into the analysis...
Have you tried to build Hearst (or invariants) - for a smoothly increasing window? If so, you've probably seen that this indicator will float as it pleases from 0 to 1 (or a little bit Already the interval) . So - what period will you take? For example for 60 minutes you will have 0.2, for 120 minutes = 0.8. Which one do you prefer?
Just in light of this, it's funny to hear phrases like "it doesn't correspond to reality" and "in reality it does"...
there are areas that correspond but for short-term flat values, when the trend is either changing or continuing the trend often does not fulfil the condition of the state interpretation...
 
nikelodeon:
what if we apply the invariance index to the balance curve? that was the question .....

Instead of ivar you can use Pearson coefficient indicator - which is much better at identifying trend/float without the dips that iVar has formed