Avery's absolutely right regarding the next hour's bar being a 50-50 proposition as to the bar's direction up or down.
Suppose the previous three bars opening price is also the high and closes at the low, down 20 pips. Three consecutive bars each straight down indicating a down 'trend' for those three hours. It is just as probable that the next hour's bar will go up as it is that it will go down. Just because the preceeding 3 bars were down is no assurance that specifically the next bar will also be down nor will it be limited to only 20 pips - could be 50 or 100 pips.
For the most part it is true that if the price is moving in a direction that it will continue in that general direction relative to the specific time period used to establish the direction of the price trend (i.e. the price is down for the day and down for the last 3 hours - but that same price could be up for the week and up for the month - different time trading segments). But the fact is as to specifically the direction the very next bar will take is an unknown and unpredictable. However it is a fact that only 3 things can occur and 1 of them will actually occur which are the price goes up, goes down or stays flat.
If I proposed to you that we flip a coin so if it comes up heads I'll give you 99 cents and if it comes up tails you'll give me 1 dollar. You'd probably tell me to kiss off because it is so easy to see the disadvantage that you'd have and the advantage that I'd have. Mathematically I would be beating you out of 1/2 cent every time the coin was flipped and after 100,000 flips you'd have lost $500 to me. Well this is essentially the exact same proposition offered by the forex (actually any trading instrument).
Assume that you:
(1) Always take a long position at the opening price of an arbitrarily selected new hour's bar, regardless of the preceeding bar(s) visually apparent directional trend.
(2) You ridgidly adhere to restricting your net stop loss amount to 6 pips (your risk or bet) and re-enter a long position should the price re-bound to your target entering price within the bar stopped out of.
(3) If you are stopped out and unable to re-enter as in #2 and a new hourly bar opens, you will follow #1.
(4) Once you have taken a position you will stay in the position, regardless of the number of subsequent bars, until you are stopped out per #2 or the price moves up where you would
net 27 pips.
You'll likely find your results are in the range of a net positive expectation of 10% with an
average winning percentage of 20% and losing 80% (1 out of 5). This doesn't mean that if you lose 4 in row that you'll win the next one, its an AVERAGE! You could lose 12 in a row (7% probability of occuring), win 1, lose 2, win 2, lose 2 and win 1. It takes a large number of trials for probability to reach true/real numbers.
The same concept can be used using a smaller gain of pips other than 27. You could use 20, 15 or 10 but what happens is the cost of the transaction goes up percentage wise relative to the smaller gain in pips. In addition, the percentage of wins to loses increases substantially. I could show you the optimum numbers but to tell the truth, I'm getting tired of typing