Page 1 of 1

### Spurious indicators, or uses thereof

Posted: Fri Jun 29, 2012 9:55 pm
In one of the Market Wizard interviews I remember reading a quote to the effect that "angle techniques are no technique at all," I presumed at the time referring to Gann angles, on the grounds that angles drawn on charts are entirely dependent on the scale to which the y-axis is drawn, which of course can be arbitrarily changed. This was brought to mind recently as I watched a webinar of a self-proclaimed forex guru, part of whose "method" consisted of entering/exiting a position when an oscillator line was at a "10 o'clock or 2 o'clock angle."

What other spurious TA does the forum have opinions on? My two cents worth would be the slope of a moving average. The Slutsky-Yule Effect states
that the moving average of a random series may generate oscillatory movement when no oscillations exist in the original data
which would imply that any changes in slope are mere artifacts and not representative of any underlying change that warrants action, on the presumption of course that prices are mostly random.

### Re: Spurious indicators, or uses thereof

Posted: Sat Jun 30, 2012 3:38 am
babelproofreader wrote:In one of the Market Wizard interviews I remember reading a quote to the effect that "angle techniques are no technique at all," I presumed at the time referring to Gann angles, on the grounds that angles drawn on charts are entirely dependent on the scale to which the y-axis is drawn, which of course can be arbitrarily changed. This was brought to mind recently as I watched a webinar of a self-proclaimed forex guru, part of whose "method" consisted of entering/exiting a position when an oscillator line was at a "10 o'clock or 2 o'clock angle."

What other spurious TA does the forum have opinions on? My two cents worth would be the slope of a moving average. The Slutsky-Yule Effect states
that the moving average of a random series may generate oscillatory movement when no oscillations exist in the original data
which would imply that any changes in slope are mere artifacts and not representative of any underlying change that warrants action, on the presumption of course that prices are mostly random.
I think angle matters.....
A near virtical rise ( fall ) after a big Rise ( Fall) almost always means the trend in going to end ( at least in short and medium term )..
Just because one market wozard is blind to this fact doesn't mean others cannot see that

Posted: Sat Jun 30, 2012 5:53 am
you do realise this is one of those never ending debates that occurs on every forum ad nauseam......

My two cents and this is all completely subjective - as most indicators are generally based around simplifying the underlying price data usually just to help represent it as a number or a visual representation, then they are largely useless on their own. The underlying data still tells you the same info.
They may have some validity in helping add context, but i think back testing usually debunks them as being totally reliable - ie; you still need discretion, or will still be stuck with the pot luck of did the inidicator work or did it just seem to work.

I do like the idea of building indicators based on other things other than just angles etc; which may help.
eg; changes in volatility, changes in relative price movements, correlations.......
Maybe its all much the same.....and beyond my pay grade.
As for them being predicitive ???

As for angles.....again you can show lots of things, but can it be reliably back tested, and shown to help make money, or is it just another clue as to help with probabilities. Most remember the internet bubble - it was fundamentally over priced, it was going vertical for a long while and it was always likely going to crash......yet the shorts got killed. I dont know many people who made money both on the up and the down...

as for other spurious TA - anything involving the use of "over bought", and "over sold" - i am working on one right now, but it will cost you to see it and use it, and it will involve an upfront payment, no money back guarantee and it will work brilliantly 30% of the time.

As someone says somewhere on the net (paraphrasing a couple of things) -
all you need is a horizontal line - price goes, up price does down....
-
its all system dependent.

(tough day yesterday, luckily Monday is a new week)

Posted: Sun Jul 01, 2012 12:20 pm
I always find it strange when people plot currencies on a linear scale and then draw trend lines on them. Surely, if the trend line has any validity, it must give exactly the same signal on the inverse of the currency? So they should only ever work on log-scaled graphs.