Why does TA not work (for you)?

Here is "a" blank log...

 

""There are few sophisticated technical rules that may explain why technical analysis works. For example, Treynor and Ferguson (JF, 2985) demonstrate the usefulness of past price information in making an abnormal profit. "" (from Surf's link)

I agree, it's very useful to many traders.

 

So the trend is not a high probability edge?

 

Buy and hold! When did that make a comeback. Hilarious. Do you trade against the trend? Since past price action means absolutely nothing.

 

Surf, the authors of that piece CLEARLY distinguish between two types of technical analysis, one which works and one which doesn't.

I don't need to make anything up. Here is the quote again:

"There are few sophisticated technical rules that may explain why technical analysis works. For example, Treynor and Ferguson (JF, 2985) demonstrate the usefulness of past price information in making an abnormal profit. They develop a Bayesian probability estimate using past price data to assess whether the market incorporates some firm-specific information, that is made available to investors. If the market does not have such information, and this is confirmed with past price data, the investor with this private information can make a trading profit. Technical analysis in the Treynor-Ferguson sense is beyond the scope of this paper, however."

It's "beyond the scope" of the paper because the authors want to focus on the reasons for the survival of the type of TA that does not work.

I can't help it if you don't read your own links.

Notice too, that the kind that does work depends on "private information", which is, as I alluded to earlier, one of the underlying requirements for making "abnormal profits" and can include "private information" in the form of a TA rule-set or an algorithm that captures some essential feature of the market.

 

I put your chart into paint and adjussted it.

I use five pages a day like this. (Bars 1 through 81)

I also have the end time for the bar number to the right of the bar #.

I'll go back to bars 1 through 14 of last tuesday and fill in the chart by hand. Look for the category test procedure and how it shows up as a set of arrows.

 

I skim through these types of threads looking for gems of information, but more often than not I'm disappointed.

I think both of the points above need to be repeated every few pages.

Just because you don't understand the meanings behind lines on a chart doesn't mean they are useless.

 

attached please find how to log the tuesday example.

review the category test procedure's three steps.

Start @ the dot on the arrows.

See the testing.

T gives you the answer.

If you reread the report I gave river on the keys to any system. you can see, now the 5 OOE's working together.

The system has no noise, no flaws, and no anomalies.

Using RDBMS and a convenient laguage, any programmer can drop in @ surf's office and explain to him how the TA world works.

As they say: Follow the money.

 

http://www.bigmiketrading.com/webinars/july11_2012/high_frequency_trading_telvent_dtn_nanex/

conclusion use longer time frames say longer than .10 milliseconds

 

Thank you for posting the report. I read and reread the document several times while vacationing on the shores of Lake Michigan; these multiple readings have helped in putting some of the pieces together.

Perhaps it’s my incomplete understanding of “quickies” that has me confused but I am having difficulty resolving two sentences in the report which, to me, seem to be somewhat contradictory:

"EE’s precede primary or secondary OOE’s."
and
"To be clear the Pattern is cast in stone."

I'm confused how End Effects can terminate the OOE before it's completion but at the same time the OOE (the Pattern) is cast in stone? I think I have missed something critical here. Could you help me to resolve my misunderstanding?

-river