How do you determine when a strategy is valid and profitable?

I think the title is a little misleading. The martingale hedge is still more of a money management thread than a strategy. The idea is that you have a higher TP than SL, and depending on how much larger your TP is, you only have to partially martingale your position size to make up for the losses. Rather than a 1/1 TP/SL, you create a 2/1 ratio and perhaps you can make up for the previous losses without having to increase your risk per trade as fast as a 1/1 TP/SL martingale.

The system does not have to be always in. You could place trades weeks apart from one another, but using the adjusted lot size depending on what the TP/SL is, and still follow the rules of the system.

FxTraderPro just structured his method to play a breakout of a tight range. Someone trading MA crosses, RSI bounces, or any other method could still implement the same money management presented in the semi-martingale thread.

What your posts bring up, though, is that a more important way to judge the insurance of a system is strictly by it's money management. If the money management is strict enough and can feasibly account for sigma 5-9 scenarios on paper, then almost any reasonable system should be able to be traded to profit. The idea is we can probably trust a system without doing extensive forward testing so long as the system can handle 20 or more SL's in a row without much DD or a margin call. I guess this depends on the MM, as a strict 1:1 R/R with no variance in possible profit would ultimately fail with terrible win ratio's.

This is why there is truth in the idea of 'let winners run and cut the losers.' It can be difficult to overcome a string of losers with strict TP goals or position sizes. Or simply, make sure the winners are bigger than the losers. This is moving the thread in a new direction, but to keep it on track I would say that we can generally trust a system that has variable profit goals. If the system is designed on paper to have variable position sizes or larger TP than SL requirements, then the system inherently has a propensity to be profitable when it can survive sigma 5+ scenario's.

I guess that argues the idea of not needing to forward test at all. Again, I mean this in the aspect of a reasonable system. I'm not talking about a system that has 300 TP, 3 SL with 2+% risked in those 3 pips. It's not as simple as that.

If that is true, you have to judge money management systems and not EAs. If that is true, any entry signal - even complete random ones - could be profitable. Right? So what exactly is the right money management system?

Nope, the forex market is too efficient, that whole good money management = profit thing is just a myth. You need a edge, it doesn't have to be big, think about it like a casino, a tiny edge which is exploited over and over. Money management is completely necessary, but comes second.

Yes, that is why I said any reasonable system will be traded to profit. I'm not going to touch the topic of random entry in this thread, but I believe simple systems like breakout models, price action over a MA, or other simple systems have just enough edge to work. My belief is that they will all work over time if you give them that chance with MM, but some will be quicker to materialize bigger profits than others.

I disagree with you that MM is second though. Taken to the extreme, a very high win ratio system will lose if you risk too much per trade or if you don't let the winners make up for the losers. To me, MM is first, the edge second.

This is sort of my problem aswell:

If I happen to find a system that I can trust to make 20% / year and max 25% DrawDown, I would retire today.

However very rarely do you see a system that can accurately predict a systems Profitability. But then again, this figure (the ROI%), doesn't help you 1 single bit in determining if you can use the system yes or no.
The real figure you need is RISK and DRAWDOWN. How much money can you lose if all goes against you, and some more? Any more then 30% drawdown is a killer for both spirit/mind/soul and your financial distress.

So your real question should be: How do you determine a systems & my own DrawDown percentage?
And your 2nd question would be:
How do you determine a systems profitability and is the systems profitability worth the risk I take (DrawDown% & STDev).

For instance: I'd rather have a system that has 15% Return and 10% MaxDD then a 50% Return and 35% MaxDD. Very Very easy choice.


As far as determining a systems MaxDD & ROI%, you would need at least 100 trades. However its not the # of trades you should look at, but more the StDev. The higher the StDev (thus the more Variable the outcome of a trade) the more Trades you'd need to make to come at a high confidence factor.

Good Point

We have to recognize current Market condition, and have a proper System for. Most of the time, market is moving in 3 different ways:

Trending
Counter-Trend (sideways)
Breakouts

If you are trading only one of the market conditions, or you have only one Trading system which works perfectly only in one of the conditions, then there is an important question: what you are doing in rest of the 70% of market conditions. You are 70% of the Time WRONG or you are CHARTWATCHER.
No one trading system on this world is bullet-proof and works perfectly in all 3 market conditions. As a traders, we need to develop (and test of course) each Trading system for different Market condition. Even sometimes the 2 Systems may give you an opposite signals, that's fine, with controlled SL and good Win ratio, you are always Winner.

http://www.forexfactory.com/images/misc/progress.gif

money management = waste of time
risk profit ratio = waste of time
finding the best "system" = waste of time

finding the pattern =
finding the general assumption of the money market =

What is the law of conservation of money?

Money can not be destroyed.
It can only be transfered from one form to another.

think about money flow, if people sell gold what would they buy?

Hi Matt,

I suppose it depends a lot on how good your system really is. If it makes 50 trades and 35 (70%) are winners with an R:R of 1:1 or better then that's pretty "heavy" statistics in your favour already - considering aspecialy that it was an automatic EA so discretion didn't mess up the results in ANY way.

Second, if your system is optimized for a small TF like 1m and a few pips per trade for example then I suppose that a couple of weeks of testing should do the job of introducing it to most if not all of the "local" market conditions that are relevant. Maybe you'll even notice times when it had bad streaks and know how to improve your system to avoid those.

In short, run it on demo (the fact you got just 200 bucks in your account doesn't mean you have to give them away easily, does it ? )and see what happens after 50-100 trades or a couple of weeks. If it's success is not abvious then run it longer to verify and try to see when it failed the most. If it's loss is abvious, I guess just let it go and find something else.

Good Luck

Take a look at the thread EA Beginner Help, post #4. This explains my concern towards using EA\'s. With that said do you have plan of attack when the way the EA is written no longer works with the current market conditions?

I imagine that is just as difficult as determining when a strategy works.