Ultimately, as is often the case, the point of pinning down the definition of trend specifically is to be able to measure it. Only this will determine if the trend is tradeable. As pointed out in my thread on measuring trend, determining trend direction is easy. The hard part is determining how much trend is present.
If determining trend direction were easy, then everybody would be able to do it. However, without indicators of some sort, few traders can. In any case, I disagree that the point of defining trend is to be able to measure it. Rather I suggest that the point of defining trend is to determine its direction and trade in that direction.
If most traders can't determine trend direction, it's because they're as lazy and sheep-like as the genpop. Just look at the endpoints (trends are defined over specific windows of data). If the last point is higher than the first point, the trend is an uptrend, obviously. Now consider the following price sequences: 10, 11, 12, 13, 14, 15. 10, 15, 10, 15, 10, 15. These two sequences are identical in every significant way except one: trendiness. What happens between the endpoints is what determines if you even have a trend and if so, how much of one. The Hurst exponent measures trendiness but it is unsuitable for trading. ADX lags too much to be a useful measure of trendiness. And so it goes.
Actually they're not identical at all. But thousands of posts have been made with regard to trend and whether or not it exists and if it does how to determine it and how to trade it and so on and so on and so on. If the OP has some specific question, I suggest he ask it. Otherwise, trend is the line of least resistance. If one has no idea what that means, he will lose a great deal of money.
To this can we also add that its crucial to determine when the trend is over? I'm finding that its difficult to always get the best entry without price ever moving against you by a tick. But if you have a way to identify where for you the trend is most likely over, then knowing this is just as important as knowing that you are in fact in a trend.
The two sequences I posted have identical starts, identical finishes, identical highs, identical lows and identical durations. That's a lot to overlook. My point was that what keeps them from being exactly identical can be summed up succinctly with one word: trendiness. Sure, we can bring in other differences like standard deviation but that won't help in determining whether or not there is a trend in either sequence. If a trader can measure trendiness, then Bob's his uncle.
LOL.... I love it! I think after more rigorous backtesting... all I will end up proving is that getting in on a retracement is the best entry. Add in some filters such a line breaks, or breaks out of something such as a hinge or trading range to get in in the opposite direction or continuation, once again, after the retracement, and there you have it, a complete trading plan! (SLA in a nutshell!)
%%%%% A line,[ straight line ]or i like a 200 dma area also nicely defines an uptrending bull market, as IBD uses;As Alan Farley said bears live below 200 dma [down trending market ] As far as ''know for certain'' thats not a helpful phrase in trading- Investing. Sure , certain amount of truth in ''rising tide lifts all boats'' but i just saw in IBD,;BAC + C sector turned negative.Ever seen a super ball bounce down[downtrends down ] stairs??? Trends are not predictions- '' know for certain'' Good question