Page 1 of 1

The ol plain Jane 50% rule................

Posted: Fri Jun 10, 2011 6:51 pm
by aziamaziam
Apologies if this subject has been aired before but I wanted to share something that many, if not most, of us have a familiarity with. Its the tendency that price seems attracted to revisit its 50% level.
I have no stats on this but its just an observation, drawn from thousands of hours screen time, that I'm comfortable with.
I've looked at various price action strategies - such as the 3 or 4 bar reversals - and have noticed that, quite often, a better price is obtained by entering on a 50%, or more, retracement.
Attached is a 5 min chart of Crude Oil showing a simple reversal strategy on a pretty typical day on CL.
Does it work all the time on all markets and time frames. Almost certainly not.
In fact the odds of a basic RR 1-1 return are no better than 60% at best. Consistent... but not exactly jaw dropping.
However - and this is my quandry - the signal often establishes significant daily highs and lows which can be very frustrating if you've left the train go on its way without you.
I believe it was Mighty One who suggested that a good solid reversal system - maybe he was commenting on the rat reversal - can actually run much further after the initial shakeout.
So comments and constructive criticism are very welcome. I will detail the actual parameters in the next post.

A few thoughts on the market behavior

Posted: Sat Jun 11, 2011 10:09 pm
by Paul&Paul
This sort of behavior I find common in terms of a chaotic system and its determinants.

One expects the energy (of the trigger) ideally to have a one direction vector.

Unfortunately, triggers can be dual and they can act in a dual way. The same energy can ignite a movement in the "wrong" direction.
You name it a retracement, a drawdown, whatever. But the "right" direction will be remembered and the "right" target will be reached eventually.

Placing a stoploss does not help at all.
Let me explain it in a simple example.
I find a trigger up of the size of 10 pips. It should make 91 pips in its fractal expansion (9.1299 times 10).
It does indeed, but before that it collapses exactly 91 pips! Exactly the same 9.1299 times 10 but in the opposite direction.

Is there any protective stoploss for that? Ney.
Why this happens? I do not know. It happens so precisely that it is really jaw dropping.

Posted: Sun Jun 12, 2011 12:42 am
by aziamaziam
Thanks for your response. I am not a mathematician..in fact the high point of my education in math was achieved by mastering long division. I do not pretend to understand anything more advanced than what I can eyeball on a chart.
However, if I knew that a certain trigger would produce a 9 - 1 return,even not knowing its direction, I would be able to profit from that knowledge...wouldn't you?

It is not so easy

Posted: Sun Jun 12, 2011 10:24 am
by Paul&Paul
You have got a trigger say of 10 pips. You are lucky to have discerned it from the chart. It may produce some fractal expansion, the most typical being 4.669 of those 10 pips.
The first question is which direction? Knowing that there exists duality.
Most of the time it will be a "normal" trigger, though. Yet it is not the end of problems.
Expansion 9.1299 is nearing a local crash, while 14.208 is a local crash in fact.
Vide Nikkei225 before and after the earthquake. The index crashed 14.208 of the trigger down.
What if you are used to take typical +4.669 while some positions are giving you -9.1299, and some even -14.208?
What if you chose a trigger which yielded "only" 3.5699? And you know it after, not before. The game becomes really very complicated and it is not a game for a small account. Playing chaos you cannot make mistakes. Yet making no mistake does not guarantee that your account grows in a geometrical way.
Remember that Mandelbrot set comprises all possible Julia sets which translates into layman's language that the market can play infinite numbers of dirty tricks on you.
Chaos shows numerous examples when no stoploss strategy really makes sense.
For weakly trending markets the use of 4.669 is very profitable. For strongly trending markets 9.1299 is better than 4.669.