the basic flaws in TA

[kiwi_trader]

Quote from cnms2:

kiwi_trader,

I think you meant MAs are low-band pass filters (filter out higher frequencies). This makes the MACD a band pass filter. Using higher values for EMAs (dropping to a lower time frame for the same price action wave), brings them closer to SMAs, hence better filters.

Yes, when i think about it, I do mean low pass ... but my real points were that:
- the momentum interpretation is not wrong
- tuning, as you suggested, is also not wrong
And, despite all the crap talked in this and other threads, very little in trading is "wrong." The challenge is simply for the individual to use it in a way that generates a positive expectancy - the nay sayers simply couldn't do it themselves (including probably my own rejection of elliott waves).

I'm not convinced that the output of one low pass filter minus the output of another lowpass filter is a bandpass filter. If a 12 period ma is filtering out frequencies above 1/12 (I cant be bothered going back to the math to find out what the response of an ema or sma might be) and the 26 period ma is filtering out frequencies below 1/26 I don't know how that implies that lower frequencies get attenuated strongly (which would be required for it to be a bandpass filter - ie attenuates high and low frequencies to allow a band through).

Now you have my curiosity piqued. Can you show me how a lowpass minus a lowpass is a bandpass filter?

 

[kiwi_trader]

Quote from Thunderdog:

The irony here is that if you are looking at historic prices (and/or volume and/or breadth and/or open interest, where applicable) in the calculation of your "statistical edges," then you are resorting to TA. That is what TA is. It need not necessarily include either tired, boilerplate indicators or subjective chart patterns. The use of historical data of price et al implies the use of TA. Unless your statistical analyses include fundamental data (econometrics), then you are simply resorting to a variation of TA. Unless, of course, you are looking to redefine the core definition of TA.

LOL. Yes. Yes. Yes.

 

[cnms2]

If you're interested, there's an interesting article in the January 2006 issue of "Technical Analysis of Stocks & Commodities" magazine: John Ehlers's "SWiss Army Knife Indicator" (SWAK). In the Traders' Tips section you'll find its coding for several platforms.

Regarding the difference between two low-pass filters: their spectral difference is a true band-pass filter, but even if you do their difference in the time domain you'll get a pretty good filter as long as the two lags are close.

Intuitively, filtering with a MA results in a smoother price like curve, with some delay. The longer the MA, the smoother the curve. Subtracting one from another will emphasize the smoothness difference between the two curves.

Quote from kiwi_trader:

...
I'm not convinced that the output of one low pass filter minus the output of another lowpass filter is a bandpass filter. If a 12 period ma is filtering out frequencies above 1/12 (I cant be bothered going back to the math to find out what the response of an ema or sma might be) and the 26 period ma is filtering out frequencies below 1/26 I don't know how that implies that lower frequencies get attenuated strongly (which would be required for it to be a bandpass filter - ie attenuates high and low frequencies to allow a band through).

Now you have my curiosity piqued. Can you show me how a lowpass minus a lowpass is a bandpass filter?

 

[cnms2]

I modified the Wealth-Lab SWAK script for an EMA difference, and I applied it to a chirped sine wave. See attached.

Quote from cnms2:

If you're interested, there's an interesting article in the January 2006 issue of "Technical Analysis of Stocks & Commodities" magazine: John Ehlers's "SWiss Army Knife Indicator" (SWAK). In the Traders' Tips section you'll find its coding for several platforms.

Regarding the difference between two low-pass filters: their spectral difference is a true band-pass filter, but even if you do their difference in the time domain you'll get a pretty good filter as long as the two lags are close.

Intuitively, filtering with a MA results in a smoother price like curve, with some delay. The longer the MA, the smoother the curve. Subtracting one from another will emphasize the smoothness difference between the two curves.

 

[kiwi_trader]

LOL. I think you might be right (thought experiment of subtracting one filter from another giving a notch between the top and bottom suggests that you are) but to prove you wrong I'd have to spend hours getting out my old text books which is too hard. I'll concede

Except of course on the issue of whether or not the macd is a form of momentum measurement

 

[thebuzzkill]

I know full well that TA is not the be all and end all ,,,but for short term trading which can be very profitable if you cannot make TA work for you then you need a new job. Simply bashing others who trade TA just shows your desperation at the markets in general. My suggestion would be to find a new line of work. here lets practice,,, "I'll have the veal marsarla with garlic mashed potatos and my wife will have the roasted lamb shanks". Hope that your new job pays wel

 

[nononsense]

the basic flaws in TA
You tell me what is meant by TA and I will tell you what are its flaws.

nononsense

 

[smallfil]

The debate probably will end if everyone traded the same way. On the contrary, each trader has his own biases, uses different indicators, sees the same chart differently, etc. So, why expect the same results?
On the fundamental analysis side, the figures are only as good if they are accurate. Remember the accounting scandals of top companies? If you were relying on that information, wouldn't you be wrong? Take Enron for instance.
Now, if you were looking at Enron's chart, while, Ken Lay and the other executives were egging on the small investors to buy----they were dumping stock!!! The insider figures would have told you (fundamental analysis) and the charts (technical analysis) would have told you that these guys were lying!!!
I use a combination of technical and fundamental analysis for my longer term trades and just technical analysis for my shorter term trades.

 

[Buy1Sell2]

Quote from smallfil:

The debate probably will end if everyone traded the same way. On the contrary, each trader has his own biases, uses different indicators, sees the same chart differently, etc. So, why expect the same results?
On the fundamental analysis side, the figures are only as good if they are accurate. Remember the accounting scandals of top companies? If you were relying on that information, wouldn't you be wrong? Take Enron for instance.
Now, if you were looking at Enron's chart, while, Ken Lay and the other executives were egging on the small investors to buy----they were dumping stock!!! The insider figures would have told you (fundamental analysis) and the charts (technical analysis) would have told you that these guys were lying!!!
I use a combination of technical and fundamental analysis for my longer term trades and just technical analysis for my shorter term trades.

bingo

 

[ifinitis]

Quote from Buy1Sell2:

What I meant about crowd sentiments is that indicators when used together, will show when the crowd is getting more hesitant about buying or when the public has gotten on board ie that final blowoff in way overbought territory. These types of thing show fear greed panic etc. As for chart patterns --I dont use them at all --they are way too subjective--one may see an asecending triangle while another sees a head and shoulders. I disregard any patterns and only use indicators, they are much more objective.

I agree with this as well, the first part about indicators. If the chart backs up what the indicators are showing you, then is not your decision to trade re-inforced by that?