I'm posting a well composed comment by Lonely Trader after my posting a couple of weeks ago regarding discretionary versus mechanical trading.
Did it occur to anyone that discretionary trading also includes mechanical models? And did it occur to anyone that automated trading has its own drawbacks? I think as traders we tend to chuck our models too quickly, without really studying them -- and, more important, without really putting in the personal effort at executing our strategies according to our rules. I am not saying this about you, Rich, but I think most discretionary traders let themselves down by not putting in enough effort at studying their markets and following their rules -- and then after all that wasted time, thinking that letting their software make all the decisions will bring the profits they dream about. Of all the models that I've seen, the discretionary models have the higher returns without necessarily increasing risk. The reason for this is automated systems tend to truncate profits because of an inability to intelligently react to market changes. (Granted, there are some very sophisticated programs that now use adaptive neural networks, but these are out of most retail traders' price ranges...by like millions of dollars!) I firmly believe that discretionary rules-based trading is the most effective way for retail FX traders to make money. Algorithms help, but as decision-aides. And I also think that most traders tend to fail in the execution of those models because either they are risking too much money or they are just lazy about following their rules. It's also a matter of trial and error. Most people who think they are looking at a discretionary model, or some aspect of it, are unwilling or unable to conduct a thorough empirical evaluation of it. They rarely put in the necessary time to quantify aspects of their model and to quantify aspects of their behavior in executing it. And so they either chuck their models out of frustration (before empirically proving whether it works or not) or they move over to automated trading models -- and get caught up in the system-fetishism that plagues so many. Either way, each of us has to make a decision. Will we continue to fiddle around with charts, backtests, and flavor-of-the-month indicators, or will we really put our noses to the grindstone, study the markets, study ourselves, and then execute the plan? It doesn't matter, ultimately, what course one chooses to follow. Just execute the damn plan!
