Simple Trading Philosophy - Daily Trend Following

Has anyone studied the % of time the market follows the previous day's trend vs. retraces from the previous day's trend? Many claim that the market ranges more than it trends; however, this is a different question. For example, the eurusd could go down 500 pips this week and then go up 550 pips next week. The record would show that the eurusd did not have a trend over the coarse of two weeks. However, in those 10 days the eurusd may have experienced 5 occurences where the following day had the same direction of the prior day and 4 occurences where the following day had the opposite direction of the prior day. Or perhaps the ratio was 7:2, or 2:7 for trend:reversal.

I believe the incidents of follow-through trend vs. reversal is about 50%:50% (or close to it). Applying this probability to risk mgmt (i.e. risk/reward ratio) & money mgmt (i.e. trailing stop with x% of lot size) may give the edge one needs to be very successful at forex trading.

Any thoughts...

thanks,

Walt

Yea, around 50% over time, I'd imagine.

But if you ask questions like "after a large move down yesterday, if price moves X pips below yesterday's low, how often does it also close below that?" you'll get different results. Even then it's only about 60/40, but careful trade management (trailing stop, mostly) can make it work.

You might be ready for:
http://www.forexfactory.com/showthread.php?t=38423

We use a 2-day approach like the one you're talking about.

Ironically, I recently began investigating the benefits of a 2-day pattern. With a good risk:reward relationship, such as 1:2, even if you lose 60% of the time, you would still be profitable. When you add the additional dimension of using a trailing stop for a small percantage of the lot size (i.e. have a hard tp for 80% of the lot size & a trailing stop for 20% of the lot size), then I believe that'll give you a stronger edge.

Thanks for the referral to the other thread. I'll certainly check it out.

Walt

Nod. The problem w/ a 50% move is that you'll have equal distribution above and below the line. So a 2:1 return, while certainly possible, is half as likely to occur as a 1:1. So if you've got a 50% move w/ 1:1, then you actually end up with a 25% chance of a 2:1.

This is why money management isn't enough to win a random game. The trick then is having an edge that gives you some directionality bias. Hence the need for a pattern that's tight enough to have an edge, but lose enough to apply to more than 1 situation a year.

I agree with your assessment. However, I think the effect of the periodic breakout would skew the 1:2 risk:reward ratio to a greater rate than 25%. When the breakout occurs, if it's against your position, then your loss is limited; whereas, if it's with your position, then your gain & the trailing stop will benefit accordingly.

Nonetheless, having a directional bias is best. I guess utilizing a given indicator (i.e. support/resistance) and utilizing temporary retracements against a given trend may help give the directional edge. Once I reveiw the "Market Profile" thread, then perhaps I'll have a better idea on properly discerning market direction.

Walt

Nod, assuming you can limit your trades to when there's a higher probability of a breakout occurring, yes. And that would constitute an edge. Otherwise the spikey ranges will get very expensive.