|
PureBytes Links
Trading Reference Links
|
Al,
I think you need to take Neal up on that challenge (Of which at the
moment I can't remember. But it had something to do with trading with
other ratios.) We could construct a whole new sequence that approximates
the Fibonacci sequence but yet gives us a ratio close to 0.618 in the
limit but yet slightly different. Since Neal has revealed to us the true
meaning of trading with Fibonacci ratios, we can now exploit has
challenge.
Just kidding Neal.
Harley
Al Unger wrote:
> Harley's post raises a couple of thoughts. First, the number 0.618 is
> not a
> Fibonacci number, but rather the number to which the ratio of
> successive
> Fibonacci numbers converges as the series expands. As such, it is an
> approximation by definition. Secondly, exchange quotes are quantized
> in
> tick increments, which means the typical three decimal "Fib" ratio has
> more
> significant digits than the prices allow one to use. As such, 0.625
> or
> 0.600 are as good as 0.618 most of the time. The difference is even
> less
> important if one is using time rather than price, as it is a rare
> occurrence
> when (0.625 * some number of days) produces a result in days notably
> different from 0.618 multiplied by the same value. It would seem a
> significant variance wouldn't occur unless the time interval were
> minutes or
> possibly even seconds. Whether Fib numbers or the ratios involving
> them are
> useful in trading may be open to debate, but requiring the exact
> number
> 0.618 for market analysis is not necessary or practical.
>
> ------------------
> Harley Meyer wrote:
>
> <Let me throw a fly in the ointment of Fibonacci here. I would say
> that
> <most folks would not argue with me if we said that a fib retracement
> of
> 0.618 is approximately 2/3 and .382 is approximately 1/3. Likewise
> 0.61803999 is approximately 0.618. So do I place my trade or support
> line at the fib number or at the other fractions. Since they are very
> close.
>
> The point that I am trying to make is that in mathematics everything
> is
> defined to have a very specific meaning. For example 0, 1, 1, 3, 5, 8,
>
> ... is defined to be the fibonacci sequence. 0, 2, 4, 6, 8, ... is NOT
>
> the fibonacci sequence, by definition. The "NOT" that I am using is
> the
> 'logical' not that turns a true statement into a false statement and
> vice versa.
> (I hope you can see where this is going by now.)
>
> So if the retracement was actually 0.61803999999999999... which isn't
> a
> fibonacci number then the retracement isn't fibonacci. Some ratios
> that
> are observed in nature that are fibonacci are exact, not
> approximations.
>
> Hopefully this takes us back to trading using Fibonacci techniques is
> more of a social construct than being one with nature. Let alone by
> definition it isn't even Fibonacci, since an approximation of a
> Fibonacci number is Not a fibonacci number.
>
> <Just something to think about.
> <
> <Harley
|