Quant analytics: Algorithmic trend detection methods
Trend detection is the corner-stone of many trading strategies. In my research I have found that accurate classification of the current trend (up, down, sideways) is as important as the entry/exit rules on which strategies are based. Currently I am using a naive approach to trend detection – the gradient (or 1-period change) in a 50 period simple moving average with surprising levels of success. What other methods have you used or where would you recommend I look for more advanced methods?
I am a trader and I take stops. Stops are not necessarily based on regression or whatever mathematical method you want to name, like, or not like. For a trader stops mean, my view is wrong, therefore I need to reassess, watch or maybe reverse or reenter.
Stops are not necessary based on lines or retracement levels they can be based on other factors, I will mention one:
A sudden reverse in tick. Say tick moved to +1000 suddenly from a much lower level. then within a couple of seconds collapsed to some negative number. I might stop out of a long position based on that alone, most of the time I look at other factors, but there are situations in which that alone might be the trigger to stop a long. I think some trader call that formation a “tick hook”
If you want to go with a simple solution, I like to use the RSI set to 51 periods. Above 50 is uptrend and below 50 is downtrend.
Any and all technical indicators are empirical in origin and nature. They have no scientific basis to them.
I think they are good for observation of trends, but needs to be backed by human experience to make decisions.
I am sure there are pure mathematical algorithms to detect/determine the trend, with some scientific logic. In case there is scientific logic, human experience can rest
NOTE I now post my TRADING ALERTS into my personal FACEBOOK ACCOUNT and TWITTER. Don't worry as I don't post stupid cat videos or what I eat!