How to end the search for the holy grail?
- Thread starter FireWalker
- Start date
If one could improve their predictive ability over 50/50 (which way market will go in the next x time) in the long term, it would cause a change in the market characteristics as more people discover the method and the opportunity gets traded away. The key is long term, you can get a slight edge in certain situations in the short term but usually by the time you find the 'setup', it is already being traded away. Ask hypostomus about how his mechanical systems find opportunities and how often they must be updated/changed/re-optimized to keep him profitable.
ig0r:
>You are profitable due to well thought out money
>and position management.
Ok, I'll agree that the MM and PM strategies are carefully thought out, but if applied to randomly generated price/volume data, these same strategies are 50/50 (before transaction costs). We have tested over hundreds of thousands of years of this data and are quite confident of this finding.
The results of this testing would seem to invalidate your above claim.
ig0r:
>If not, then you are simply taking advantage of a
>temporary market characteristic that will soon be
>gone, impossible to predict when
This claim I will be unable to prove or disprove -- it just depends on your definition of the word "temporary". Is a system profitable since 1992 gaining only a "temporary" advantage? Of course it can be argued that is it, but I would be hard to argue that such a system is not stable enough to profit from during this "temporary" phase.
If a system can be devised such that through rolling optimizing and forward testing can profitably withstand over 50,000 stock years of real data (from only price/volume) and survives all tests against the random, I'd have to say that in my world the statement that price and volume isn't enough it just plain wrong.
JB
>You are profitable due to well thought out money
>and position management.
Ok, I'll agree that the MM and PM strategies are carefully thought out, but if applied to randomly generated price/volume data, these same strategies are 50/50 (before transaction costs). We have tested over hundreds of thousands of years of this data and are quite confident of this finding.
The results of this testing would seem to invalidate your above claim.
ig0r:
>If not, then you are simply taking advantage of a
>temporary market characteristic that will soon be
>gone, impossible to predict when
This claim I will be unable to prove or disprove -- it just depends on your definition of the word "temporary". Is a system profitable since 1992 gaining only a "temporary" advantage? Of course it can be argued that is it, but I would be hard to argue that such a system is not stable enough to profit from during this "temporary" phase.
If a system can be devised such that through rolling optimizing and forward testing can profitably withstand over 50,000 stock years of real data (from only price/volume) and survives all tests against the random, I'd have to say that in my world the statement that price and volume isn't enough it just plain wrong.
JB
my goodness!
You dig out this thread after it was buried for 17 years.
And I don't recognize those nicks who participated in this thread.
Are they still around?
Or have they stopped looking for the holy grail?