edge and luck

That is certainly a fair point. Obviously the question of what even constitutes a recovery sequence (or necessary percentage return) where you would resume live/full risk trading is debatable. Again, this will largely be an issue of risk tolerance, but one would presumably need to see a recovery that would be statistically similar to how your system would perform if your edge was still functioning. For me, I would want to see a new equity high in the paper traded results. Moreover, I would be concerned not only with the the account reaching a hypothetical new account high, but equally with the specific makeup of the trades and the time required to achieve it. All of this would be important in making a determination of whether your system was still operating with an edge.

Of course, this could take a long time, or perhaps never happen at all. If your edge had deteriorated or never existed, then you would paper trade the system into further drawdown and never trade it live again.

But, to your example, imagine that the system completely recovered on a paper traded basis - complete with every statistical verification that would give confidence that your edge was still functioning. Then imagine that when you resumed live trading you hit another drawdown that was outside of expectation. What then? Well, in that situation I would again stop at a preconceived point, and would be very weary about continuing to trade that system again. Yes, you could always reach that point of drawdown, stop live trading, paper trade a subsequent recovery, resume live trading only to have another drawdown that was statistically unsound. That system, at its most catastrophic, would have lost the trader x% from the first drawdown and a subsequent x% from the second drawdown. It would be very unpleasant, but at least it would be controlled, and certainly not catastrophic. The more conservative your risk management, the less you could potentially lose even in both drawdowns. If, say, 20% was your statistical cut off, you would lose a worst case total of 36% from your peak equity. If this occurred on day 1 (just to make the example even more improbable) you would have lost 36% of your principal, rather than merely giving back profits.

If it sounds like I am agreeing with you, it is because I am - its just that it doesn't really agitate me. Sure, it could happen, but with a good edge, a good system, enough data to formulate reasonable expectations, and conservative enough position risk, it becomes increasingly unlikely. The trader that prepares himself properly and still completely fails - having been slowly bled from unsound drawdown after unsound drawdown, only to devise new systems and have it happen all over again until he is broke - would be extremely unlucky indeed. So, there you go, you need some luck, but adequate preparation surely negates the need for a constant fixation on it. Most who fail in this industry are not like the unfortunate character I just described above; they fail for a myraid of other more inane reasons. And most that do meet those characteristics, I would imagine, haven't failed - or at least not yet...

http://www.forexfactory.com/showpost.php?p=2506456&postcount=578