Positive expectancy is a combination of 2 things , without the 2 inputs no conclusion can be made .... stupid question ..... how long is a piece of string ?? probability curves also need consideration after positive expectancy is defined , its not as simple as you making it out to be . Close this thread down its heading nowhere .
More reward per risk on the 10% trades as long as you stay in. Fatter tails. There's been a number of traders/funds to go this route. Very difficult to manually trade though. If you get kicked in the face 9 times my guess is the 10th time you'll be expecting the same
With a 10% trade success rate over a large enough sample you are going to get kicked in the face >20 times in a row , this is where probability curves come into play
Exactly. 50% is the centerline; it is the easiest to develop or find. Ie. the saying "50:50" means that case. And the following relation is true as well: Developing a strategy that is correct on average in 10% of the cases is harder than developing a strategy that is correct on average in 60% of the cases. (this might sound counterintuitive, but 10% in this context means in reality 100%-10% = 90% when inverted, as we've already seen in the previous postings; see also above quotation) [One can here also substitute "finding" for "developing"]
With no tax game, such as war, most game is 50:50. However, with tax (commission) such as security trading and blackjack, successful trader might be ONLY 10% or less in the long run. Most zero-sum game may be around 50:50. Many research paper support it, such as http://facultybio.haas.berkeley.edu/faculty-list/odean-terrance/ If you have a chance, please see reality at “Just How Much Do Investors Lose from Trade?” with Brad Barber, Yi-Tsung Lee and Yu-Jane Liu, Review of Financial Studies, 2009, Vol. 22, 2, 609-632. "The Internet and the Investor" with Brad Barber, The Journal of Economic Perspectives, Winter 2001, Vol. 15, No. 1, pp. 41-54. "Do Investors Trade Too Much?", American Economic Review, Vol. 89, December 1999, 1279-1298.
What a good answer! It reminded me of this article I read the other day by someone who ranks high in understanding trading and probabilities and whose blog I highly recommend. Look how from the four definitions of probability in the article only the last one which is about beliefs is maybe applicable to trading. The frequency definition is also applicable but only when the appropriate sample size is available. I think all traders should understand this article and dismiss all irresponsible arguments that about probability being the key. If it were that easy, you would only need to get a phd in math. Get this straight and stop wasting time with books and authors who insist trading is probability and statistics. They manage to push their hype and profit from book sales and presentations but where are their trading records? They are taking advantage of ignorance to sell their snake oil. Edit: The author of the article I mentioned is in favor of portfolio backtesting instead of statistical analysis when analyzing performance. He explains that in his book Fooled By TA, which I highly recommend because I learned a lot from it. This is also the method that he uses in his data-mining software.
The most valuable book I have read after Market Wizards. Fooled By Technical analysis is a true gem. It offers step-by-step methodology for reducing data-mining bias among other things. An eye opener.
Yep, i made a module to reverse exact buy/sell entries in MQL and tried it with many losing strategies. Unfortunately most losing strategies are scalpers aiming only for few pips. You lose on spread. To reverse strategy you have to have losing trades with many pips above spread amount.