Stochastics
are probably the most popular indicator used in futures trading.
We have recently issued a special
report on stochastics. In it, we interviewed top experts
on stochastics, including George Lane. What we found is a
sharp diversity of opinion. Some of them thought stochastics
are fantastic. George Lane calls it "damn near infallible".
Others thought it was pretty much useless.
My own take is that it is a very useful
indicator but not in the way that most people use it
Most
people believe that they should look to buy a commodity when
stochastics dip to below 20 and sell when it goes above 80.
In general, you should do exactly the opposite. A move
in stochastics to above 80 usually means that the market is
trending strongly and should be bought. Take a look at a chart
book and check this out. We discuss this in more detail in
our special issue.
Also, you can construct profitable trading
systems out of stochastics. A simple trend following system
can be created out of crossovers of the stochastics line over
the 50 level. Buy when it crosses to above 50 and sell when
it crosses back below 50.
We tested varying lengths of stochastics to see what kind of profitability we could generate.
The classic length for stochastics is 14. However, we found that using a crossover of the 50 line by the 14 day K line was not profitable.
The good news is that using a K length of anything over 25 was profitable! The highest profitability came from K lines around 40 bars long. However, you also have to add the filter that the price must be higher than five bars previously.
Of course, this simplistic system was often chopped up in sideways markets but made a ton of money when a trend started. However, even the chops were less than most systems and the equity curve was actually quite smooth.
Having said all that, I really don't
use stochastics because I find the signals to be too inaccurate.
I'm going to do some additional work, based on our interviews,
to see if I can create a more robust use for stochastics.
Do
you use stochastics? Do you find it useful or dangerous? Send
me some email
and I will print the best comments on this page.
From
Leonardo:
I
have been a futures trader for 23 years. I trade for the long
term; and have found (statistically) that the 2-3 day breakouts
in the direction of the main trend work best when 14 period
stoch. is 70% or better, or 30% or lower. The same concept
is true in weekly and monthly charts, although there are precious
few of them. You could just trade breakouts when you have
these extremes and you would be roaringly profitable!
RETURN
