Very much a newbie - I’ve been trading with a practice account for several months and have learned quite a bit.
I realize that this is kind of a general question: Is there any consensus on what constitutes a trend or (even more generally) what period of time one applies indicators and analysis to predict behavior of a currency? By that I mean that, for example, if you look at EUR/USD for the last three months - there’s a strong trend upwards. Look at it for the past two weeks, and still strong trend upwards. But look at it for, say, two days, and one sees many down areas and no real trend for those days. This may not be the best example, but the general idea is that a shortened time frame (1 day, 1 week) may look like a trend in the opposite direction from a long time frame (3 months). So what do people look at to reveal a trend and to apply indicators (Bolinger bands, Fibonacci, etc.) to trade more predictably?
A trend is witnessed when the price is making higher highs
& higher lows for an uptrend & vice versa, lower highs &
lower lows for a downtrend.
Dependant upon the type of trader you are will determine
which trend you follow. Monthly, weekly, daily, hourly etc.
I agree with daydreamer65. Whatever timeframe you trade, that’s the timeframe which defines trends for you.
Yep! Me too!
I agree with Daydreamer65 and Rhodytrader.
There is wisdom in the wise counsel of many especially when they all agree.
OK. Now I’m getting worried. 
I agree too with the wisdom of the wise counsel of three… haha.
One point i would make is just to be aware of the overall longterm trend.
Cheers
Hi kb_mike
It is an accepted fact that trends exist at all scales in the markets, from many years down to seconds. It is also widely agreed that it is worth paying attention to a range of scales above the one on which you are trading, as these influence what can expect from the smaller trends. Smaller trends against a bigger trend are likely not to go as far (until the big trend eventually fails, but it is so easy to predict this prematurely many times before being right!) Eventually, with a huge difference in time frames, the big trend becomes of little importance. For example, the trend on the monthly chart is of little interest to a daytrader using a 5 minute chart, but the big trend of the day is, and the context on the daily chart is of some relevance.
You may find the module on multiple time frames accessible via the “School” tab above is very relevant to your question.
Hi KB,
Perhaps a more profitable question to ask is " Which time frame trend will give you on average “a certain number of pips” that meets your “Reward” criteria of your Risk:Reward ratio in your trading plan?
If your trading plan calls for you to have a 1:5 Risk:Reward ratio and you have a 20 pip stop-loss, then you need to be looking for trends on time frames that give you the most 100 pip trades. If you’re stop-loss is 40 pips then you need to be looking for trends on timeframes that give you the most 200 pip trades.
You can always use smaller timeframes for entry points, but you should know, on average, which timeframe will give you your reward for the risk that you are taking.
Enjoy your trading…
Yep me too. (or three) 
This is not a thought process I would recommend.
Pick a trend timeframe based on your trading timeframe, not the other way around.
Also, if you’re trend trading you have to toss out the idea of risk/reward ratios on a per trade basis. It will hinder your performance, not help it.
Thanks for all the replies, and I’ll definitely go to the relevant module that was recommended. Do successful traders look at all time periods (15 min vs. one month) to base their analysis on? I think that Elroch came pretty close to defining the logic (look at the “range of scales above the one on which you are trading,”). From what I’m seeing, technical analysis is more relevant in forex than in stocks and futures for a number of reasons. But the big question is: what analytics and factors does one look at to create a more-or-less predictable trading strategy? I see the definition of a “trend time frame” as an important issue, because at least some of the factors that affect movement are different hour-to-hour from month-to-month.
I have a relative who does a lot of stock trading. She understands technical analysis very well and has been using it for three or four years. She has really not done well - she might have broken even so far, but she certainly has not gotten the returns that would justify spending her time on it as if it were nearly a full-time job. Granted, that’s only one data point, and it’s a different market, but the big question that’s at the back of my mind is “Does technical analysis yield profits, and what part of technical analysis is the most important to start looking at?” So, if you’re making money, and you feel it’s because of technical analysis, is the selection of a trend time frame an important part of your strategy?
Technical analysis works. Fundamental analysis works. Combinations work. Other stuff works too. But none of it works consistently for the person who doesn’t believe in it. If technical analysis doesn’t make sense to you on a personal level, then find something that does.