Hi Everyone, So I have implemented my own TD Combo, and did some backtesting myself on the daily data of US, AU and Japan stock markets. I wouldn't say the results amaze me. Well, tbh, it's totally garbage. I have searched through other threads and the general sentiment regarding Demark's stuff sounds like it has little-to-none value. The other thing is even there are a bunch of documentations detailed its procedure and claim possible awesomeness, I don't think anyone has given a concrete example that this signal has worked in what market. Before I come to conclude it doesn't work, can anyone that have it working explain how you implement trading based on its buy/sell signal? Where is your exit point after you commit into the position?
Jason Pearl in his new book published by Bloomberg describes how to set risk level if you trade perfected TD Setup or at TD Sequential/TD Combo countdown completion. If you want the 1 sentence version, it's essentially finding the True Range of the lowest (if you are in Buy setup/Countdown) bar, and then subtract that True Range from the low of that bar. Tom Demark has made it perfectly clear TD Sequential and TD Combo is NOT a trading system. I use TDST, TD Sequential and TD Combo together. I would describe myself as a discretionary position trader in the daily timeframe. I don't automatically enter into trades when the daily TD Sequential or TD Combo produces a buy or sell signal. I always check the daily signal in the context of the weekly and monthly count, so I can determine when to enter the trade and how long to stay in. A good example would be the SPDR Retail XRT -- the daily Combo and Sequential Countdown Sell signals were produceds days ago, but in the context of the weekly Sell Setup, the count was not completed yet, so I waited and did not short XRT-- and that patience saved me some anxious and stress had i prematurely shorted the stock. When the daily, weekly, and monthly TD Sequential/Combo lines up, the signal can be quite powerful in identifying the vicinity of a major trend change. The value of Sequential and Combo is to force me to sit on my hands and be patient, and not prematured jump the gun. That in itself is worth major dollars to me.
An article in SFO magazine several months back made these indicators sound like the holy grail. But as you have found out, the indicators are as good as using a coin flip to make your trades.
I have been using TD Combo/SEQ for a a few years. Although they are not the HG of indicators, they have certainly made more than a few good calls over the years. The daily, weekly COMBO seems to be the most consistant, yes you really need to watch every timeframe for them to have any value.
I firmly believes in Gann's observation that Time Turns Trend. In other words, price level by itself is not enough tell whether the item is "oversold" or "overbought". It takes time to exhaust a trend. The trick is figuring out "how long". Demark's work simply gives weight to the time element in measuring a trend. In my opinion, there is not a hell of lot of magic in the 9 count for Setup (momentum), or 13 count for Countdown (trend). In other words, I don't believe in the mysticism of the fibonacci counts (4-day look back + 9-day Setup = 13, and the 13 2-day lookback Countdown). But they do impose discipline and check your id and ego at the door.
Thank you all for the useful input to my question. I'm skeptical to the magic count idea myself. I think you guys are right to the point that use it only on a specific frequency probably doesn't really tell much. Another problem I think is use a mechanical testing method, it doesn't necessarily reveal the true power of the signal as that's not a fixed set of trading rules.
Pretty compelling evidence here of Demark nailing the March 2009 bottom and the April 2010 top. http://seekingalpha.com/instablog/6...owed-us-the-complete-bull-phase-of-the-market