|
PureBytes Links
Trading Reference Links
|
Hi Yuki
Tks for your considered and lucid reply. I am
a momentum trader and I guess your reply
confirms that momentum - its start, fade, finish, direction (or
change of), and velocity can be measured with many different styles
of ruler.
What has been useful in applying the "Yuki's Double Dip Day" nuance
to my momentum indicator is that it now changes colour in
confirmation on a "Yuki Double Dip Day". I don't even have to check
the ADX. I hate excess work.
Over 60% of the time I'm in on a `double dip day', having received an
earlier momentum entry signal the day before.
However, because I scale in and out of positions this is great for
helping confirm a second position, and its size, on the following
day.
Another benefit is confirmation of trend strength when I have an open
position (the double dip appearing sometimes within a trend, not only
at the start) and I am considering stoploss levels for the following
day this. I have confidence that the following day will be a positive.
One final observation ? Backtesting a reverse double dip (both +&-DI
up), returns not one occurrence on 5 years of data on this index.
My brain feels as tho I have been double double dipping for apples in
a barrel all day so I'll leave investigating that for another time.
Tks again
Regds
Gerard
--- In amibroker@xxxxxxxxxxxxxxx, Yuki Taga <yukitaga@xxx> wrote:
>
> Hi Gerard,
>
> The example you showed is during a pretty darn good uptrend. ADX is
> one of the better indicators ever created, IMHO, and as you have
> seen, when a strong uptrend pulls back, the odds are very strongly
in
> favor of a resumption of that trend.
>
> The period of time you posted shows a *really* strong ADX(14), the
> lowest dip just barely getting under 30. This in itself is a very
> strong measure of trend strength.
>
> What you are seeing on the "double dip" days appears to me to be
much
> more indicative of a long entry the closer +DI has come to touching
> -DI. When they are far apart, mean reversion would indicate a
higher
> probability of that knock-out correction you want to enter on.
Again,
> this is in keeping with the idea that a trend takes a breather, but
> doesn't really reverse -- until it's not a trend anymore. You have
> to be wrong at the end, but you are right the rest of the time.
>
> The spikes on -DI in a strong trend, like you posted, are entry
> signals -- when they reverse. The only problem is making sure you
> have a real spike. You seem to have a good momentum indicator (in
> black) to go with this. Notice that on the minor -DI spikes and
> reversals, that black level is still rather high, telling you that
> you're not really in a prime entry spot at that time. But when your
> momentum gauge goes below zero, all the -DI spike-reversals seem to
> show good entries. I'm not surprised, since they also correspond to
> a good convergence of +DI and -DI.
>
> Other periods of ADX are useful as well ...
>
> Yuki
>
> Friday, June 2, 2006, 10:15:37 AM, you wrote:
>
> GC> Hi all
>
> GC> I noticed recently that often +DI & -DI both DECREASE on the day
> GC> a possible entry is signalled, by price action or another
indicator.
> GC> (See the attachment)
> GC> This is not always the case as shown here but it appears to be
much more
> GC> than just a coincidence.
> GC> Anyone have any insights?
>
> GC> It does happen on other occaisions ? Could this perhaps be used
as an
> GC> indication of trend strength, usable when holding a position?
>
> GC> I am aware that one of the ADX nuances is to look for an entry
when the
> GC> ADX is above both the +DI and ?DI.
>
|