My opinion....see "efficient Market Hypothesis" and "Random Walk Theory". These are pics from a random walk 1d generator on Wolfram. You can do 1000 of these and you will get a normal bell curve distribution of points. If you smoothed these out most wouldn't be able to pick out the random from the real any more often than they normally do. Brokerages asses risk usually around a 2 standard deviation move....and they do just fine. If it didn't work that way they would be put out of business at some point.
how you know high high? how you know dip dip? maybe high not high maybe lot higher. maybe dip not dip maybe lot dipper, yes? better think buy when cheap sell more less cheap. in china we say who say low who say high not know low not know high, better catch small slice than miss whole pie,
What you are talking about is a swing trading system, buy the dips, sell the rally before it corrects and then play the swing again once it has a correction. What people are saying is when you get a dip, it may continue down rather turn back up. That is a risk in all systems of any kind. So you have to have a point where you know the trade didn't work and you get out. You should know what this exit point is before you enter the trade then you can manage the amount of risk you are willing to take before entering the trade. The whole idea is to have an edge, which means what you are doing works more than 50% of the time, that is why people backtest their systems, to make sure there is an edge. You can make money even with a 50%/50% system provided your average win is greater than your average loss.
Swing trading is a very profitable way to trade, and fundamental analysis is not very applicable in a 3 to 5 day trade. But I do create my universe of stocks based on fundamentals, so the entire universe of tradeable stocks are all good sound companies, then I pick my trades from that universe of stocks using technical criteria.