Try looking at eminiz.com - John Ehlers latest incarnation (the MESA guy). Check out its performance (not good). This is about the biggest argument against detectable cycles I have seen...
Ehlers cycle indicators are not meant to be traded. He designed them to detect trending vs non trending markets. He doesn't trade his own cyclic indicators. He uses them as filters. Whats your beef with traders that want to discuss technicals?
I think that Fibonacci numbers, as applied to trading, are just crutches. And that would be fine if they were not imaginary, magical, mystical crutches.
Man, it's great talking shop with you guys. Here's a link to Thomas Stridsman: ThomasStridsman He's a Swedish trader/researcher who writes numerous articles on improving one's trading through various analytical methods. I attended a seminar of his at the recent NY Expo where he discussed using the median of the price to keep traders in a trade (vs. a moving average). Maybe it'll spark some ideas, maybe not, maybe you'll have 10 15 or 100 failures before you come up with a workable methodology. Who cares? for a process oriented person, it's the iterative process which develops the knowledgebase, not the results. Good trading, Jimmy J
You should read "new trading systems and methods 4th ed" Perry Kaufman. He covers this topic(and 100`s more) including the proper math. Also has all the excel code for most of his book included in a CD
Thanks for the info. A lot of the traders didn't understand the concept, and thought he was just refering to the average number (as in moving averages), he wasn't. The median price reprsents the true middle price, and he showed charts where it actually held the trend much better than a MA would. I've been trying to figure it out, but can't quite get there. I'll have to give the book a read through. Good trading, Jimmy Jam
There is alot of math. You wont be reading it quickly...I promise. If you know someone in NYC thats a CMT or a candidate they should have it to borrow as its required reading for the level 2 exam.
That is one way to use them. Imo what you want to take from this experience is that you don't design systems using filters, you design systems based on models of markets (e.g., CAPM or APT, Black Scholes, etc), then you use DSP on the theoretical prices that these models give you so as to get a greater signal to noise ratio. Trying to use DSP directly as a model on the market to discover correct theoretical prices is fruitless, unless it is calibrated by market logic. nitro
Jimmy, I think the value area in market profile (or price historgram) is very close to the median price he talked about, it is 70% (or 90% - you can change the settings in charting programs) of the trades take place at a particular price at a given time, this price is the battle ground or Point of Control (POC) between shorts and longs. If longs win, they'll pull the price to the upper 5-10% boundary. If shorts win, they'll pull the price to the lower 5-10% boundary, it is dynamically changing with the time, the POC is always the gravity, it is the mean price of a normal distrubution on the trading volume, it acts as the support for uptrend and resistence for downtrend on any given timeframe. It can be calculated by volume or price. Some readings on market profile trading: http://traderfeed.blogspot.com/2007/02/market-profile-best-practice-in-trading.html Here are market profile examples on QID for the day in 15 minutes chart. The mean price is at 48.28 as of 2:30 PM on chart one , prices goes over the POC, bulls were in control, they push it to the 10% top area which is around 48.64. Bulls push the price to the upper 10% boundary: In the process, the mean price is going up, the price is pulled to the new center of gravity (POC).