The big smile at the end of your post doesn't detract from the fact that your statement is wrong. Position sizing is quantified AFTER evidence of the edge has been established. One can merely assume fixed position sizing when testing for the edge. Then, using win%, max drawdown, etc...etc...a risk management schema (which includes position sizing) is formulated. Position sizing need not be considered until evidence of "the edge" exists. To do otherwise would be an utter waste of time.
Exactly my point. Apparently there is some semantic disagreement here on what is meant by "position sizing." My point is that your results are dependent on what you assume as the trade size when you test for the edge. As the wrong-odds coin-flip scenario illustrates, assume the wrong size and you can miss the edge completely. The edge is still there, but you've effectively neutralized it by overbetting, or overtrading, as the case may be.
Uh-huh, sure. And not only that, but with a suitable money management strategy appropriate for each and every one of those "permutations," right?
I think you need to see this:<object width="425" height="350"><param name="movie" value="http://www.youtube.com/v/ihD93bGoXXA"></param><param name="wmode" value="transparent"></param><embed src="http://www.youtube.com/v/ihD93bGoXXA" type="application/x-shockwave-flash" wmode="transparent" width="425" height="350"></embed></object>
When I test for an edge, I use fixed position sizing - an equivalent bet on each trade. Thus, if an edge is found, it is not something that results from money managing hocus-pocus - rather, it is a viable setup upon which a trading strategy (including position sizing / risk management) can be formulated. The edge has already been identified before the issue of proper dynamic position sizing is tackled. On any given trade, the relationship between position sizing and returns is simply a 1:1 linear correspondence. This relationship is wholly irrelevant to determining the existence of the edge itself. Right. You have just stated that the EXISTENCE of the edge is not dependent upon other variables. You're coming along nicely.
The DETECTION of the edge is dependent on other variables, specifically, money management. Look, all we're talking about is the difference between arithmetic growth and geometric growth. If you choose arithmetic growth, fine, you'll find an edge if there is one AND your trade size is significantly larger than slippage plus fees. If you choose geometric growth, you better be careful or you can completely counteract the edge, even without factoring in slippage plus fees. The TA researchers are going all in. Guess what type of growth that is.
You're off your meds. That wasn't covered before I posted it. Clearly further attempts at communication with you are pointless. Have a nice day