CCI (700) indicator/strategy idea

Thanks Simon

multiple data references sounds easier to implement than I'd imagined. I will perhaps give it a go.
I understand what you mean about defining S & R in EL code. I am not aware of any method or technique that allows this to be done. Accurate S&R interpretation does seem to be more of a discretionary visual thing perhaps.....
 
Last edited:
Hi

I have not coded anything yet, but referencing higher timeframes before deciding to take a trade in a certain direction, certainly does help to filter out and reduce the taking on of non-profit making trades.

For example, with a 2 minute chart and my CCI 700 idea, referencing the 15, 30, and 60 minute charts on a discretionary basis, would help prevent me from taking on long trades, if the longer term direction is not saying UP. I'll perhaps look into coding the multiple longer timeframes into my strategy-indicator frameworks.

Cheers
jtrader.
 
jtrader,

You're coming at the problem the wrong way.

There are two ways to create strategies.

1) Long term, trend-following, 'fat-tail' approaches.
2) Strategies that take advantage of a particular phase/mode of the market. e.g. breakouts work in volatile markets, overbought/oversold signals in congested markets etc

1 is easy. You just get a couple of moving averages or do x day breakouts. The problem is that it's slow and expensive.
2 requires that you watch the market and identify which mode/phase it is in. Then apply the correct strategies.

Which of these is your CCI(700) looking to exploit? IMO neither. You've come up with a random indicator and looking to apply a random filter that will make your random indicator look less random. Why would CCI(700) work? You can make a case that something like MA(20) or RSI(14) would work because you know other people are looking at them. But there's no-one looking at CCI(700).

Go back to your charts. What market mode/phase are we in? Trending, congestion, highly volatile? Identify short term opportunities. Are the markets always down on Mondays? If you take the morning breakout does it always make 10 points before failing? Then design a strategy that exploits something you can see. Applying random indicators will get you nowhere.

not a bad post my friend (y)..........I like CCI's and incorporate a lot of the principles into strengthmeter trading

N
 
Top