|
PureBytes Links
Trading Reference Links
|
RTs,
I, like many others am daily watching a number of indicators unrelated to my
system, for possible
reversal patterns. Out of 25-30 oscillators say a third produces divergences
every day, whether
these consist of price moving higher and oscillator lower or chart patterns
(lower low for price,
double bottom for an oscillator etc). Are these divergences tradable? I did
a lot of backtesting (historical
daily data, lots of it) and:
yes! Very! And this is the problem, LOL. Because it's mostly BULLISH
divergences that are tradable. And no wonder, as for most of the time
markets (major market averages) have been going up, bear markets
being an exception. If bearish divergences are much less/untradable at all,
then doesn't it make you
suspicious. I mean, a system results can be improved not by a divergence,
but by the bullish nature
of the market. Which makes these divergences random, might just as well try
flipping coins to go long:-((
So, what can be done about this?
(1) if (IF) we do assume that markets will keep repeating their at least
general characteristics (this case: going
higher) - and, BTW, isn't this the first assumption of TA? - with occasional
and short-lived bears, then
there's little problem using long signals of even random character, if these
backtest to be very
profitable. I mean, hey, we're in a bear market for just a year and in a
little while the market will
revert to her original course. Using random bullish signals in a bear could
be costly, but (xx) stops are in place
and rallies do occur + all the usual stuff and procedures, + money from
shorting etc. - so trading bullish divergences would mean maybe less money
in general at this time,
but still money. And when the bear is finally over, would pay big.
Obviously if we OTOH get a bear market for 10-15 years (+recession... war...
famine... earthquakes... you name it) then, less money (but see {xx} above).
A lot would depend on how long we actually plan to be in the business.
(2) Stick with more down-to-earth regularities, even though they mean much
less money overall than trading divergences. It's also easier
psychologically, cos trading divergences (+oscillators in general) means
picking bottoms, etc.
On the other hand: if markets prove themselves to be very bullish (proof:
these bullish divergences work; bearish don't) then isn't this the first
regularity to come to terms with? Or do we quarrel with the market?
OTOH to the previous OTOH:-)) - but they prove themselves to be very bullish
in the very long haul. Can't be otherwise, if DJIA goes from 100 to 11000. I
always suspected long backtesting periods to be a double-edged weapon :-((
Ideas, comments please?
Bullishly,
Yarroll
PS. Continuing on this note, one would be very suspicious of most long
signals generated by TA indicators:-((
OTOH, every short strategy, properly back-tested, would pick some very
non-random regularities. "Don't confuse brains for the bull market", ROTFL.
PPS. If you don't check up on oscillators unrelated to your systems, please
visualize what Im talking about:
pull up a chart of DJIA, Naz etc. and overlay it with ROC-5, CCI-5 (periods
would matter). Divergences are everywhere, LOL.
To unsubscribe from this group, send an email to:
realtraders-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|