For discussion, useful topic.... what criteria do you look for? There's slow steady trends, sharp brief trends, consolidations, reversals, pullbacks that continue and more.... My thoughts: Swingtrading: I look at daily charts and check to see if the current day closes above prior days' high vs below pd low. If a previously uptrending chart trades down 2 or more days in a row I consider the short term uptrend broken. Can also use candles and ma lines. Daytrading: much more specific patterns, using whole number s/r, pd OHLC, trailing stops based on prior trend price action, orb atr etc Thoughts? Eg if S&P has a red day or two does it need to stay under 3k for a couple of days before considering uptrend broken...
That's one way to do it. Unfortunately I've noticed that this technique, while conservative, usually produces a lot of false signals, when the market prints this 2-day low pattern and then resumes its uptrend in a big way, leaving the trader behind. For exit I mainly use a close below a simple moving average. This allows me to milk the trend until the end, in most cases. Plus I can always re-enter the trend later, if the exit was premature.
Check the relative performance of comparable assets against the trend/chart. I love using the risk adjusted ratios of competing instruments (dollar-volatility adjusted ratios). (price an asset in terms of another -- using hedge ratio) Examples are silver/gold, nasdaq/dow, dow/russell, etc. If you see action in an important spread like this, you know traders are positioning. For example, on Friday everything was rallying (nasdaq, spx, dow, russell) but then the russell stalled out and the dow just kept grinding higher (relatively). Part of that was institutions buying the dow and selling the russell as a hedge (that spread hit an intraday low around noon).
How to Tell when a Trend has ended? You can tell anything you want, anyway and anytime. What matters is how you structure your bet around this speculation. We might break a trend line but not the previous low. Does the trend's broken ? The line is certainly invalidated. We may test the previous low and consolidate around it. Does the trend's broken ? The momentum has certainly weakened. Does it excludes the scenario of a sudden spike, reintegration and moon flight ? Nope ... It's a matter of probability. What if we break the consolidation on the upside ? What if it's a fall break out ? What if ? What if ? How do I position myself ? How the other players are positioned ? Are shorts trapped ? Is there any interest anymore ? It's not because we break a trend line that a continuation is impossible. It's probability has certainly weakened. So does our expectations.
Decide how a trend is defined and then you can define when it began and when it ended ... almost always after the fact.
If it is a "strong" trend and I'm swing trading, a two or three day pull back is an entry point for me. Assuming the whole market hasn't gone to crap.