[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Exponential moving average



PureBytes Links

Trading Reference Links

On Sep 30,  2:08pm, Carroll Slemaker wrote:
> Keep in mind that a MA is a low-(frequency)-pass filter.  That is, it passes
> through most variations of longer duration than the MA span but removes most
> (or all) of shorter-period variations.  To use one for revealing a cycle,
> therefore, one should start with a MA having a span of, perhaps, a quarter
> of the length of the cycle of interest and compare its result with that of
> successively longer-period MAs. 

Interesting point.  As a short term trader, I might be holding positions
on average about 3 to 5 days, or 1 to 10 days nin/max.  If my "cycle"
of choice is 4, than 1/4 of that is 1 day. Does that mean that my
short term average is simply yesterday's close (or yesterday's change
from the previous close).  Does this imply that it won't be worthwhile
short-term trading with only daily closing data?

> What you want is the shortest-period MA
> which produces the desired degree of smoothing (removal of short-period
> variations) withOUT significantly reducing the size of the desired cycle.
> 
> You can take this one step further by also calculating a MA with span equal
> to the cycle of interest
> and then displaying the difference - short period
> MA minus long period MA.  This gives you a rough band-pass filter.

As Bob Fulks mentioned, the problem with using a simple average
whose length is that of the cycle is that it will average to zero (note
in a rising market, with a cycle imposed on top of it, the average
would work out to the upward bias).  EMA's won't do that, but they do
attenuate the signal, and they're late.

Take a look at this chart:
http://www.intrepid.com/~gary/ma.gif

It shows a sine wave with a 10 period cycle (in red).  The simple
moving average is in green, and is zero.  The 10 period EMA (using a
conventional weighting of 2/(length + 1)) is shown in blue.  The EMA
lags by 1 to 2 bars (about 15% of the cycle length in this case).

Interestingly, if we traded the simple MA, by trading on the close,
after a cross of the MA, we'd buy at 0.6 and sell at -0.6.  If we
traded the EMA the same way, we'd buy at 0.0 and sell at 0.0.  Neither
method looks too good.

If this example were changed to use a much shorter cycle of, say, 4 bars,
probably the best we could hope for is the system homes in on the fact
that the period is repeating, and is sufficiently out of phase to exactly
hit the tops and bottoms.