[Sorry, I'm late to the party here! But hopefully there're still some people around and I can at least get myself a drink ...] I wonder if the above two observations are related? "Time frame is important", on the one hand; and "FX is different", on the other. What is the time frame for FX (24x6 market, with several overlapping sessions Asia/London/NY every 24 hours) that corresponds best with another instrument that just trades mainly in a single fixed time slot 5 days a week?
What? Who are you, what's your agenda? You are wrong in every single aspect, a strategy is highly adapted to a specific market nature. If you were going to have a strategy work for multiple markets you would do exactly as all other noobs that think optimization is required to have a robust system and your pockets would soon dry out of money.
I disagree that you need to build from "the ground up" for a new market. One example of this in the public domain is the ACD method. If your strategy framework is generic enough, it can be adapted to new markets. The issue is that there are very few of these kinds of frameworks (ACD, Demark, and Elliott Wave come to mind) and the ones in the public domain obviously need to be customized for the individual trader. Anyone capable of developing a framework at this level or customizing it profitably is definitely not a noob. While my experience is limited to about three months, the examples of Crude, the Euro and the ES are already showing me some differences within them, even though the overall framework is the same. Compared to the ES, Crude has a bigger payout ratio and is less sensitive to the value range of my primary parameter, for example, and the Euro gives far more false positive signals and appears to require an additional parameter to help filter them, while both create sets of trades which, when filtered, can be taken in reverse, while the ES does not. So, those are some examples of how the framework needs to adapt to different markets, but it was definitely not necessary to start over in trying to add these markets to my trading. I was not sure if it would be when I started, so I had an open mind, but as the data is coming in, it seems to be on the side of affirming the hypothesis that my framework is generic enough to trade multiple markets. Obviously, this could all change in a short period of time, but if something were going to go terribly wrong, I would think it would have started to already.
Completely disagree on that statement. If anything you are penalizing over fitting by forcing the strategy to work across several markets. That is one form of regularization designed to combat over-fit optimization. The more you adapt it to a specific market nature, the more you likely you are to converge on an over-optimized solution.
Here is a swing trading system that has never been optimized and is the exact same parameters on each market. Trades about 3x per month per market...so about 21 trades a month for these 7 markets. Anyways, while some markets do better than others, they all swim up stream. Also, missing data in yen, bonds and swiss from 2000-2003 (as you can see on the chart). I consider this fairly robust.
He is certainly wrong but he is pushing his own agenda like everyone else in here. Very few people in this forum actually trade. Most are computer types, a good number of them from East European countries who have spent time to learn forum slang. JCL is a computer type. He thinks he understands the markets but he does not. It takes more than 10 years of professional trading in a well-regulated environment under strict guidance to understand the markets. Amazing! Amazing! Quick research just showed that he/she also goes by the name Rita. This is disgusting really... http://www.elitetrader.com/vb/showthread.php?s=&postid=3472387&highlight=#post3472387