Regime Changing... How's it done? Theres a reason why 90% of traders fail and 10% succeed. The 10% can correctly identify the turn of regimes. So I wanted to open up this conversation to regime changing. How do you do it? Technical indicators? Quant methods?
Double tops , bottoms, trip tops, bottoms, Hand$ , inverse hand$ 2br reversal, 3 bar Triad Long upper shadows, long lower shadows. A,V, N,M,W .................................. Just patterns that 90% know about . What do you use? Are we talking politics or change of direction in price?
Regime change, do you mean finding correlations between current price and history? If trin closes > 2.0, rally next day xx% of the time? Data Mining? Fed Chiefs? My guess is you mean, trending or ranging, the age-old question. Volatility is one old idea: Build your own Regime Switch Indicator using volatility. Regime switching using volatility rising/falling and at what levels. But bolly bands are better. or for the wannabe-quantish: http://quantivity.wordpress.com/2010/01/24/review-trading-regime-analysis/ Book review with it's very own Trading Regime Indicator (TRI), all this for $60 bucks. A mini regime switch right now with the ES if it breaks over 1555 here. IF not the current regime owns 46.
I started off using TA but don't any more, minus acouple of indicators i have in my systems. I don't believe in the patterns that prices form though I am talking about change in direction of price. Politics is a qualtitative story that does lead the market in my opinion but for someone who wants to trade with limited direction they help paint the big picture theme but not the parameters for one to trade off of. Thus i want to focus on price
This is exactly what i'm talking about. Ideally i want to research and backtest regimes that last a few days to a week. How do you come to the conclusion that the ES is a mini regime switch if it breaks over the 1555? I'm not questioning your logic, i believe it, how do you come to that? I'm looking at the screens right this instant. Es broke 1555, is now heading back and theres now buying pressure in the bonds. Sure you can label that a regime change but i want to research and backtest things that will last a few days rather than micro scalping
m2c is that it doesn't matter. You can define a regime however you see fit based on your particular constraints, views, and conclusions. Pretty commonly people use things like moving averages (bull/bear), the vix (high/low vol), or a given strategy's success in a certain period (breakouts have worked well). At the end of the day you still have to make a buy/sell decision with that info. So what if it's a low vol mean reverting regime...are you going to sell rallies and buy dips hoping it continues or buy rallies and sell dips hoping it changes? HTH.
The tape has always turned when POMO / QE was scheduled to end. Then it turned back up as a new schedule was released. This week marks the end of the current schedule, but Bernanke's testimony for keeping rates low for the indefinite future implies there is more to come.... So all you suckers calling the top, wait until they announce the suspension of QE. Otherwise you're all just contrary indicatorz :eek:
Niederhoffer's group has done extensive work regarding regime changes. That would be. Wise place to start.