Technical Indicator Fixation

Purple Brain

Experienced member
Messages
1,613
Likes
180
I’ve invested so much time into technical analysis and into finding what I think works that when I found a combination which I felt gave me sufficient edge to trade profitably, I was confident that this was the backbone of my method and while research is always enjoyable, any changes or additions I decided to make would be minor tweaks compared to that central basis and criteria. Much as in the same way an investor gets married to his position and ends up not being able to see the wood for the tress, I realise I have been guilty of the same myopia in my technical analysis methods.

I’ve used the exponential moving average over 49 periods on my 15 minute trading charts for a long time now. I have always taken the price in relation to the moving average as a good indicator of bias: Long or Short. I fully understand that any moving average is a derivative of the price and not the other way round and any reference I make to price ‘reacting’ to the moving average or ‘bouncing off’ the moving average are teleological, but convenient descriptions of the price action with respect to the moving average. When the price appears to ‘respond’ in the same way on a frequent basis to a moving average or other technical level regardless of the physical mechanism involved (many other traders also responding to that same level would be a good reason I imagine), it is a phenomenon which can be traded. What hit me just last night was that the 49-er was not technically a potential support or resistance level, which is how I have been using it, but a centre of gravity around which the price rotates. That’s obviously what all moving averages are, quite by definition, but the significance of that must have been lost in the excitement of my Eureka moment in discovering my very own blessed, if not quite Holy, grail. So although it mostly serves me, at that periodicity there are times when it swings about giving me sudden changes in bias which are nothing of the sort. If anything, the move counter to what I considered to be trend, if viewed from a somewhat greater distance provided a better perspective that would have provided me not with a failed counter move but an entry into the previous much more solid underlying move at a far better entry level.

While I have retained the 49-er for all its uses are still valid in other setup criteria qualifications, it makes sense to move out to a longer period exponential moving average to determine what I am going to consider as denoting a long or short bias on the simple basis of its slope. Spent this morning checking historically against all the charts I trade and I can’t understand why I didn’t see what was right in front of me all the time. The moving average in question (a 131 exponential) has been on my charts since the beginning purely to ‘provide context’ as I like to term it. However unconscious was my awareness of the significance this context provided, I wish I had been able to summon into consciousness before now.

We live and learn.
 
I’ve invested so much time into technical analysis and into finding what I think works
.......................................................
, I wish I had been able to summon into consciousness before now.

We live and learn.

A thoughtful and worthy statement at post # 1. Having been down similar paths myself it confirms my belief that the more you think and study your market and your tools the better you will come to understand it and be able to trade it. The Holy Grail?
 
No holy grails in forex. Keep your eyes on the market and watch the candles and the patterns the form.
You will definitely learn more from this.
 
. . .What hit me just last night was that the 49-er was not technically a potential support or resistance level, which is how I have been using it, but a centre of gravity around which the price rotates . . .
Hi PB,
This is a contentious issue as there are many members who will fight to the death defending your former opinion! Not the 49MA per se, but that MAs in general act as S&P. My view is that your new way of viewing MAs makes much more sense. If you've not seen it already, it might be worth your while taking a look at the The 3 Duck's Trading System - as the way it uses MAs is clear, simple and, I think, complements your latest realization.
Tim.
 
The vast overwhelming majority of indicators have zero worth.
The few that have some worth are still marginal at best.

Oscillators - too many early false signals, and they get slaughtered in trends.
Trend indicators - lag too much, and they get slaughtered in a range.
They are all based on price anyway, so use price alone, or at least
treat price as the most influential factor.

That isn't conjecture, I've tested most of them in an automated setting.
With no discretionary filters, they are at best worthless, at worst they will
gradually drain your account.

A few do have some worth, but it certainly has nothing to do with
their traditional application.
 
You're a clever guy, and if you don't mind me saying, an interesting read. But here's my question.

What exactly did you learn?
Not to believe that what you have right now is the best it could ever be. Not to get complacent with your methods, and to be especially suspicious of them if there's no real reason to be unhappy with their current performance - that's what lulled me into relative laziness. If they'd be outright poor performers I would have been working much harder to resolve them.

Not that this is something I've just learned but something I've always subscribed to: Keep on researching and testing on the expectation sooner or later another bit of the puzzle will get filled in.
 
Not to believe that what you have right now is the best it could ever be. Not to get complacent with your methods, and to be especially suspicious of them if there's no real reason to be unhappy with their current performance - that's what lulled me into relative laziness. If they'd be outright poor performers I would have been working much harder to resolve them.

Not that this is something I've just learned but something I've always subscribed to: Keep on researching and testing on the expectation sooner or later another bit of the puzzle will get filled in.


Is this everything? I hope not, otherwise you've really missed the most important bit.
 
random + random = ?
2*Random if you leave it alone. Random squared if you b a s t a r d i z e it?

I appreciate professional traders don’t trade in any way the same way at all as do individual retail, but surely it’s better than random even for those of us trying to trade purely directional plays? If it wasn’t, how would any of us be making better than 50:50?
 

Attachments

  • random.png
    random.png
    293.3 KB · Views: 477
surely, random is like infinity.

infinity + infinity = infinity.

ergo, random + random = random.

random + random = ?

Purple, I believe Kimo was pointing out that indicators at best are random.
Actually, they are quite often worse due to the lag bias built in.
Random entry + Random exit = set fire to your money its quicker :LOL:
Random entry + Managed exit (not random) = potential.

Managed exit at the most basic rudimentary level could be trailing stop,
or something a bit fancier like triggered step trailing stop.

Thats is still extremely basic and you would need to manage the exits
a bit more than that in reality, but its a pointer...:)

Of course you can refine entries as well, but most people's attempts
at that are just falling for sucker plays, that doesn't mean it isn't possible though.
Same for exits.
Depends how you want to do it.
The above is a simple approach, but there are as many ways to skin that cat as you want.
Momentum increasing or slowing - in price on chart, or tape.
Consolidation and levels and so on.

The only thing to remember is there is no such thing as a cast iron outcome,
no matter what approach you use.
 
Last edited:
Here’s the random math the way I see it here at t2w (with a little latitude :innocent:)

Assume integers, solve for random numbers
--> odd random + odd random = even random

substitute any t2w members for odd random
--> odd t2w + odd t2w = even random

subtract odd:
--> t2w + t2w = even random -2odd

divinde by 2:
--> t2w = even random/2 - 2odd

some even numbers divided by 2 = odd number. Substitute
--> t2w = odd – 2odd

result
--> t2w = odd :whistling

:smart:
 
Here’s the random math the way I see it here at t2w (with a little latitude :innocent:)

Assume integers, solve for random numbers
--> odd random + odd random = even random

substitute any t2w members for odd random
--> odd t2w + odd t2w = even random

subtract odd:
--> t2w + t2w = even random -2odd

divinde by 2:
--> t2w = even random/2 - 2odd

some even numbers divided by 2 = odd number. Substitute
--> t2w = odd – 2odd

result
--> t2w = odd :whistling

:smart:
Is that taking into account the boolean law of distributivity though :LOL:
 
Why are you using a MA? What is it telling you about price that you couldn't ascertain from the price movement alone?

You're not far off a few lightbulb moments here now - just a little push........
 
Why are you using a MA? What is it telling you about price that you couldn't ascertain from the price movement alone?

You're not far off a few lightbulb moments here now - just a little push........
Why use an MA? Audusd on a 15 minute chart with a 49 for entry timing and a 131 exponential moving average for trend bias plus the standard pivot points on yesterday’s action. There were two touches to the 49, one when I was just heading off to bed at midnight and one at 09:15. I took a long position at 09:45 entry at 0313 after the up bar at 09:30. Had a MAE of -4 pips and eventually bailed out at a resistance level when the price touched it at 0333, getting out at 0331 for +18 with an MFE of +2. It’s not a killer of a trade, but it is a win. If I didn’t have the 49 to ‘suggest’ an entry when the price ‘bounces’ off it, I would potentially be looking at any up bar as a potential long entry. During the prime trading session I acknowledge picking any up bar at random I would have been unlikely to have picked an awful entry. There are a few, but not as many as there would have been favourable entries. So in principle, yes, I could just trade off price alone with a basic buy on up bar method and used any of my exit methods to good effect. But it is entirely possible I have picked audusd on a good day when the pullbacks were shallow and the trend well developed. I suspect this wouldn’t work on all pairs every day. It surely can’t be that random?

I’m sure this wasn’t the direction you were trying to lead me Robster and I have likely misunderstood your hint in which case please persevere and hammer it home as bluntly as you like.
 
If I didn’t have the 49 to ‘suggest’ an entry when the price ‘bounces’ off it......

No problem trying to shove you but you do have to make the leap yourself :)

Why did you need a lagging derivative of price to tell you that the price was low enough to buy at?

Why did you delegate responsibility for the entry decision to this mechanism?

What have you achieved by not being engaged in what price may or may not be doing now in it's own right?

Were you really engaged in what the market was doing?
 
No problem trying to shove you but you do have to make the leap yourself :)

Why did you need a lagging derivative of price to tell you that the price was low enough to buy at?

Why did you delegate responsibility for the entry decision to this mechanism?

What have you achieved by not being engaged in what price may or may not be doing now in it's own right?

Were you really engaged in what the market was doing?

Stop being so cruel and spill the beans Robbie:)
 
Top