the whole idea of position sizing seems a bit flawed to me. ignoring the idea of scaling in or out. if a TA gives you the size of the position that should be originally taken, that could be considered a confidence indicator. rather than having the position size proportional to the confidence indictor, intuitively there would be a cutoff point which maximizes expected profit using 100% capital each position? I haven't bothered doing the math for this but my gut feeling is that the position sizing doesn't cause the trader to gain anything in the long run. does anyone have any good arguments why position sizing works?
Yes, edgeable... but whatever is edgeabel must now subject to individual trader interpretation and makes it no better than a coin flip. TA is NOT anything like: If the 20 day MA cross 50 day MA, and when Stochastic is X, MACD is Y than you would have a winner. Or, That if you got an assenting triangel... you do this and you would have a winner. or That one can clearly define mathematically an accenting triangle, a head and should, a double top.... I remember sitting in a TA workshop, the instructor ask what do the students see in a specific period. Eight students come up with 6 different anwers. TA is definitely not.... the holy grail.
TA or no TA. Money management/Risk Management is the most important factor in keep you in the business of trading.
But like a "trend", there is no evidence that something that has happened in the past will continue in the future. In fact, history will likely show that anything that is actually worth pursuing (and not random/noise/incidental) will eventually fail. How can you tell if something has gone from statistically valid and profitable to no better than random? How will you tell the difference between a temporary fluke of losses and a sign that your concept no longer holds in producing profits? What is the standard length of time or number of trades after which one can draw a definitive conclusion -- especially if what you are looking at has just gone from "edge" to "coin-flip" or barely-profitable?
look, all that is necesasry (it's simple - but not easy) to be a successful trader is to have a methodology that has an edge BEYOND the amount you pay in commissions and spread (if you pay the spread- part of how i make the money i make is making the spread not paying it) and applying that using PROPER position sizing/risk management that's it TA is simply a (very large field) of modeling past transactions into some sort of framework a chart is a model of price over time so is a MP histogram so is any lagging indicator (and I use TA to scalp futures, but i use no lagging indicators) so is market delta so is advance/decline line so is trin so is tick etc. TA is much like genetics. generally speaking, it is not deterministic - it is PROBABILISTIC iow, certain setups have always existed that have positive expectancy. those setups tend to, over time, stop working if too many people use them -see: market efficiency but every single floor trader is using TA - a model of price action and if u don't think institutions use TA, you are crazy do you think the algo's, black boxes, etc. DON'T use TA?
{ Poor Entries and Exits } + { Great Money Management } = Slow Death { Great Entries and Exits } + { Weak Money Management } = Still a winner unless you really bet too much (2% rule will keep you afloat)
People should remember that a win rate of 60% (or even 80%) will still lose money if the losses are substantially bigger than the profitable trades. A much better benchmark is the ratio of wins to losses (total profits/total losses). A trader with a 40 % win rate can be very profitable.
This is a money/risk management problem trade management problem if you have more winning trades than lossing trade but still lost money. This is not about TA. The trader must be winning little but lossing big AND cut profit early and let losser run. But we are talking about if TA can produce winning trade better than 50%.