Fibonacci - How does it work - Does it really work?

I promised so here it is. 16 try outs of 1000 coinflips put on one screen. They don't really look like chartpatterns, rather like mountain ranges or clouds.

I agree that if one looks hard enough he can discover certain patterns everywhere, but as it was pointed out the stockmarkket is not an even distribution game like a coinflip....
 

Attachments

Quote from agin1415:

Fibonacci is the way things grow.

It can discribe the growth of sea shell chambers, or the growth of a galaxy.

Its described by some as the "Natural Ratio".

Technical Traders use it to predict price targets and pivots, and if you use it properly you can trade with about 70% confidence that a trade will work out according to retracements or extensions that reach Fib ratio lines.

At least thats the way my mentor was taught and he was taught by a successful London Currency Trader.

Then others will say they tried it and it helped them Lose money.
More...

"Then others will say they tried it and it helped them Lose money." ... I guess that was the other 30%? :D

I'm trying to expand my TA tool box here..I'm learning to read candles like NihabaAshi :D but I want to expand. I know some use fibs, and some use piviots. I cant decide which one to dedicate my time too.

Should I just study both?

your thoughts...

thanks

- nathan
 
Here I go...

In my opinion traders would do much better to limit their "tool box" to a few things that work very well and then look for geniune extreme readings in those things rather then--as most seem to do--try to find a stock/currency/future contract that has half the indicators under the sun sortof going for it. Don't equivocate trading sophistication with trading success.

There's nothing magical behind a chart: it's just a recorded history of people deciding to trade a given possession for other person's cash. Be thoughtful of what studies you subject to it; it should be logical and based on how opinion changes in population and according to price--not some half-baked approach that supposes that people consistently change their minds when dips/spikes of x% occur.
 
I have not read every single post in this thread so if this is rehash, my apologies. I personally do not pay much attention to fib numbers. Guys who really like these numbers tend to get really carried away with all types of retracements, extensions, fib time, fib price, etc. The only professional traders who I talked to have said they are aware of certain fib levels because eveybody else is watching them. So in a way, that may make them useful. Maybe a self fulfilling prophecy or perhaps it really is the secret of the universe. I have no idea.

What I think would settle the argument once and for all is to use historical data to try to figure this out. I do not mean back testing a fib based system, what I would propose is to take historcal charts from 100 years ago, 75 years ago, or even 20 years ago and compare them to charts today. The goal would be to see if most retracements did indeed stop at or around .328, .618, etc. Now, obviously this is made difficult because people could disagree on where to calculate the retracement from. However, I think if you just kept it very basic, we could put this argument to rest. If the markets in the 1920's tended to retrace to .382 or .618 and reverse as often as they may appear to now from major highs and lows, you would kind of have to think there is something to this. If however, the charts reveal that Fibs became magnets after everbody had a computer and charting software on there desk, well, then we can assume that the Fib penomenom is of our own making.

My Charting software only goes back to the early 1980's, but a very simple look into this shows exactly the opposite of what I expected. This is anecdotal at best, but it appears that some of the fib based movement was actually cleaner back then. Not very scientific, but somewhat interesting.

If anybody has easy access to historical data, and is better at designing studies then I am.......this might be a project.

Comments appreciated, I don't need to get shredded though.....if my idea is dumb, just say so. Not looking for a debate......:)
 
I did just see this post and it made my day. Very funny.

QUOTE]Quote from Thunderdog:

If Leonardo Pisano (Fibonacci) were alive today, rather than using his number sequence to describe the reproductive mathematics of rabbits, he would have used his number series to describe the proliferation of baseless trading methods.

P.S. I just can't get over the 3rd decimal point "accuracy" of the ratios.
:D [/QUOTE]
 
Let me through a few opinions:
- backtesting can be only as good as the understanding of the system by the person who does it
- backtesting can't prove that a system works, it can only prove that it doesn't
- backtesting separate elements of a system (i.e. entries, exits) won't prove much about the system's performance as a whole; the system's performance is not a linear function of its elements' performance

I'm not of the opinion that backtesting is useless, but I think that it shouldn't be overemphasized, and its results have to be interpreted critically, considering its limitations. More often than not backtesting results will be invalidated in the future due to the incorrect backtesting, than due to the system's obsolescence.
Quote from pem0608:

...
What I think would settle the argument once and for all is to use historical data to try to figure this out. I do not mean back testing a fib based system, what I would propose is to take historcal charts from 100 years ago, 75 years ago, or even 20 years ago and compare them to charts today. The goal would be to see if most retracements did indeed stop at or around .328, .618, etc.
...
Comments appreciated, I don't need to get shredded though.....if my idea is dumb, just say so. Not looking for a debate......:)
More...
 
How does it work. Well to put it simply, one has to learn what a "test" is. To go a step further, one has to learn to distinguish between two conditions.

1. Test and then price "takes out" a price point, pivot, or fib.

2. Test and failure to "take out" a price point, pivot or fib

For me the rule is simple. Price tests and "takes out" a price point, pivot or fib when it CLOSES above or below that point.

Conversely, price tests and "fails" when it cannot CLOSE above or below a price point, pivot or fib.

After that it is a matter of determining a stop loss and an exit strategy.

I have already done my homework, but I suggest to those who want to make some money, that it is a matter of HOW you operate the system. How you write the rules.

I have used this analogy before. It is a matter of skill and experience. If you put a tool (pliers, screwdriver, etc) in the hands of an unskilled person, you don't have much. If you put a tool in the hands of a skilled mechanic, you get a much different result. Same is true of Fibs.

Good luck Folks
Steve
 
I've used Fibs for a while now, but usually in conjunction with pivots. It almost looks like magic when you look at a chart at the end of the day and overlay it with Fibs. The problem is: they work until they don't work-- just like most everything else in trading. I find myself using them less and less. I never use them to enter a trade any more, but find them to be a good way to exit. Several times I have gotten top print . If a stock is getting a little toppy, getting out within .5 to .10 of the next level will get most of it off the table.
 
Steve46,

What I'm thinking is that a system can perform better, synergistically, than separately its entries and exits can. It could also probably perform worse than its elements taken separately.

I liked your tools analogy, but I'd like to add to it that there are different levels of skills, based not only on experience and education, but also based on their user's other qualities: creativity, imagination, problem solving skills, etc.. Quite often a skilled and experienced laborer lacks the capability to further improve over time, because he lacks other qualities.
Quote from steve46:

...
After that it is a matter of determining a stop loss and an exit strategy.
...
More...
 
You must log in or register to reply here.
Back
Top