There is an entry method for day-trading that can be used for both mean reversion and trend-following. This is because a trend trade intra-day may be a reversion on the daily. However, many would likely view this as a trend-following entry method, but they are not likely looking at the 'big picture'. Having discussed system development and strategies with traders and firms for quite sometime, I have yet to try to explain that we are trying to capture both RTM and Trend-Following in the same algorithm, but allow the market to guide us as to which to implement. We have designed strategies that vary entry methods based on where the market opens as well as other conditions (such as daily price patterns). Therefore, we are combining both "reversion to the mean" strategies with "trend following" strategies into a single algorithm for entry. This makes the systems much more diverse than just trading either 'mean reversion' or 'trend-following'. Are there any others out there that develop systems in a similar fashion?
I try to combine several trend and counter trend algorithms in the same strategy, for getting a smoother equity curve. But this is not really an algorithm that trades both ways. The algos are still developed separately, they are just combined in a single strategy.
yeah, rtm trading can get pretty boring and so can momentum trading.. diversity - that's the key ! I've always been of the opinion that profit is what we are after now I know it's also diversity! /btw, sounds like an advertisment to me/..
I can assure you, it's certainly not an advertisement. This simply proves my point as to how people see things differently. I am only implying that we all see something different in the market. And while I may demonstrate a strategy / entry method one may say well, you are certainly following a trend. However, I could say what are you looking at, because this demonstrates that I am trading a reversion to the mean. That's all
Actually, I am not referring to either of those methods. I am referring to a 'longer-term' intra-day trade, whereby the trader may be implementing either a reversion to the mean or a trend-following strategy, but entering using the same technique. This is of course, a position trade and not a scalp. What this comes down to in this case is that one may view this as a trend-following trade on an intra-day basis, but a reversion to the mean when looking at the daily data.
It sounds to me like you're trying to use the same setup for going trend or counter-trend. You can't do both at the same time. Who cares if a trend in one time-frame is a reversal in another? Trying to trade multiple time frames at the same time is a recipe for disaster.
I think it's important to understand why a strategy would work or not, that's why I think it's important to know what is happening in a larger time-frame. Not attempting to trade it, but realize why the method may / may not be profitable. Whether one is traditionally a reversal or trend-following trader, realizing that the other is profitable as well is key. Developing strategies that can take advantage of the 'phase' of the larger time frame, intra-day has been my goal, no matter what 'phase'. I am only implying that a trader could use an Intra-Day Entry Method to profit from both a Reversal and Trend-following phase on the daily. This applies to position style trading where a position would be held from minutes to hours. Does it matter which it currently is on the daily (other than being able to understand what may be happening)? No. That's the good part, but also speaks to the fact that it's certainly not trying to trade multiple time frames.
its more like following the pinball to next/previous bumper,whichever way it bounces,you need to know where the next bumper is and see how it reacts
What is the problem with trading multiple time frames at the same time? When algorithm A is profitable in a 30 minutes time frame and algorithm B is profitable in a 4 hours time frame, I see no reason why a system should not include both algorithms. Quite the opposite, trading different time frames will often decorrelate the results and generate a better overall Sharpe Ratio.