You might have mixed two things. Positive or negative correlation is still somewhat equal in terms of their prediction values. No correlation (or â0â correlation) suggests that there are no dependencies of any sort. That is my conclusion so far. Unless, of course, staying true to your self you are being facetious and making an exquisite joke in which case the markets should care about everybodyâs predictions.
You overestimate me. I wasn't trying to be clever. My point was that a trader is more likely to be "right" if he chooses a general direction than if he not only chooses a direction but also a specific price projection to be attained at a predetermined time. The likelihood of all those ducks getting into a precise row in an environment of uncertainty strikes me as very low. That premise leads me to conclude that the more precise your attempt to project, the more likely you will be wrong. Hence the likely negative correlation between specificity and outcome. Conversely, the less demanding you are of future price movement to accommodate a specific whim or "projection," the less likely you are to be disappointed, all else being equal.
Any statistical analysis depends on the sample. Ask all americans if they want to have an african american as president vs the opposite. You may get an indifferences. As each group separately, and you may get different conclusion. So what you see may depend on where you sit. In order to see everything, you must sit everywhere. There is a difference between correlation and causality. Causality implies correlation. The other way is not always true.
Yes, I agree. You are making a very valid point. In the world of Math it translates: A conditional probability of two independent events is always lower than the individual probability of each of those separate events.
Silent readers, If you wonder whether I have tools or how did I come up with my tools. Read the following books (the originals, not what others told you about the original content): 1. Isaac Netwon's Principia 2. Laplace's book 3. A book by a great mathematician of 800 century (AlKhawarizmi, from Persia/Irak) 4. Bachelier 5. A paper by a british nobel laureate in physics on the topic of kinematic waves. Good luck everybody.
I think it goes beyond the notion of joint probability. Even first order specificity in an environment of uncertainty is likely to be met with an unhappy outcome. The joint probabilities only make it more so. I think that the further out you go with a specific time/price prediction, the more likely that you are to be wrong, all else being equal.
Do you think that RFT realizes you are just toying with him and that he is just playing along like a good sport?