|
PureBytes Links
Trading Reference Links
|
Hi Joe & all,
Gee, I actually thought I'd stumbled upon something new - I've been looking
at this multiple MA crossover phenomenon for a couple months now. My
comments are based on preliminary observations, but here's what I've noticed
that appears to warrant further investigation:
J.F. - "There was an interesting article in Stock and Commodities back in
Feb 98. The title "Multipule Moving Avg. by Daryl Guppy. For short term
traders he used the averages of 3, 5, 8, 10, 12, 15 day EMA. Long term, 35,
40, 45, 50, 60 EMA."
D.C. - I'd been looking at a combination of 21, 34, 55, and 89 period EMA
regardless of trading timeframe (like Steve Karnish, I'm just crazy about
those Fib numbers). This seemed to yield better results on intraday charts
(much better using all session data). Just for kicks, today I looked at a
daily chart of the NASDAQ composite and tried a combination of simple moving
averages using the same period inputs. The simple MA combo seems to be more
accurate than the more sensitive EMA. But I think I use it differently than
the author in TASC...
J.F. - "How to use it, fairly simple. When one group of EMA would bunch up
and start to cross the other group, you would know to go short or long."
D.C. - Does this imply that if the MA crossover occurs in an upward
direction that it's time to start looking for a long entry? From what I've
observed, using my EMA combination, prices would be at new swing highs; once
the 21, 34, and 55 EMAs have all moved above the 89 EMA, I see prices
correct a bit and then hit one or two new highs before turning down for a
playable short (this minor correction should last for several periods before
going to a new high). Occasionally, a third high will be hit before a
worthwhile short play. But the short play is always more profitable than
going long on the crossover (of course, the opposite is true when the EMA
crossover happens in a downward direction). So I'm looking to fade the MA
crossover, as opposed to following along in it's direction. Of course, in a
parabolic move up, one could wait for hell to freeze over before finding a
correction and rally setup to fade the crossover. So there'd be a lot of
missed profit by not playing the MA direction - but since those parabolic
moves like to correct sharply, they're not something I'm inclined to jump on
anyway.
I'm attaching two charts of the NASDAQ with the simple and exponential MA
combos to illustrate the difference. In a sense, some combination of MA
crossovers would seem to be a decent leading indicator - in a contrarian
way. No, I'm not trying to reheat that argument about leading/lagging
indicators - please, let's not go there again! If it's something that works
for you, then great - use it. If not, then don't. I'm just looking for
possibilities here. All I know is that when I see a crossover now, I start
to salivate... : )
Regards,
Dennis C.
Attachment Converted: "f:\eudora\attach\EMA.gif"
Attachment Converted: "f:\eudora\attach\SMA.gif"
|