I only want to trade instruments that trend well and frequently. I'm wondering if there's a way to mathematically examine, say, the past 30 days of a market's price action and determine how trendy or choppy a market is. Then, only trade that market.
I'm thinking in order to do this that price movement range, relative to price, is important, like ATR. So if an instrument is, say, at 1000 and it ranges 20 points a day on average, then it moves 2% (20/1000) a day. But I'm not satisfied with this, as it could be in chop all day then make 1 big 20pt move at the end of the day.
Another possibility is to look at intraday swing highs and lows, and determining the average size of each swing move. That might work. Markets that cycle could have many good swing moves and still be within a narrow trading range on a daily basis.
Another consideration is to measure chop (how to do this? see this thread () and see how much time each market spends in chop.
Am I on the right track? Anybody else have any other interesting ideas? Maybe some of sefstrat's DSP noise filtering methods would help, but I don't understand that stuff (yet).
I'd love to develop some sort of market "weather" report that tells me which market trended the most today/last week/last 30 days, etc then use that to trend-trade the most trending market I'm comfortable trading.
Comments? Has somebody already done something like this?