|
PureBytes Links
Trading Reference Links
|
I have tested various strategies using opt f, secure f, etc but have
never been
able to overcome the shortcomings of the shift in f over time. Fixed
ratio
appears to be a bit more caitious and I also like the fact that as
account equity
grows each contract must still earn the same "increase amount" to
increase the
position size, unlike in opt where once you're trading hundred lots a
few ticks
can mean a substantial position size increase. However, I am still
trying to
find a way to use fixed ratio with pyramidding. Most of what I do is
always in
the market - long or short. On a single contracts the systems do quite
well.
Like most systems the majority of the money is made during the large
moves. What
I am trying to do is find a way to incorporate fixed ratio without
having to
treat each additional trade like a different system. A simplified
example is the
crude chart below. One could set aside capital for the 2nd, 3rd, etc
system but
who is to say this pattern won't repeat itself 10 times? Also, the time
the
system spends in a pyramid mode is 10% so to set aside capital when the
future
max position size is unknown and the system is only in such a mode a
small
percentage of the time doesn't seem to be the best use of capital.
Testing does
show increased drawdowns with pyramidding but the avg yearly p&l / worst
case
drawdown ratio improves considerably. As a prop trader this is probably
my most
important indicator. I can't go through a 2year drawdown without
getting my
capital cut severely and I also like to know how long historically it
will take
me to work myself out of a drawdown period. In the past I have traded
mys
systems based on dividing capital / combined worst case drawdown x 2. I
then
increase or decrease my contracts allowing myself to drawdown 30% of
capital.
This approach does allow me to increasedecrease size over time but I
feel this is
not the best approach to the matter.
x x
x x
x x 3rd buy
2nd buy x xx x x
x x x x x x x x
x x x
x initial buy
Kevin243@xxxxxxx wrote:
> I tried that with stocks and got burned badly. You are dollar cost
averaging
> up. You can turn a series of winning trades into major losers. I
much more
> prefer "Run and Gun", where you just keep moving to your next initial
buy
> signal. That way you are always buying in at a presumed bottom. The
profits
> realized or leverage via unrealized, are reinvested into the next
signal
> without dollar cost averaging up.
>
> Granted, there are some instruments like crude that have had a long
run.
> That's also why I use a more sensitive system that trades more
frequently. I
> would try a system that gives more buys and sells along the way thus
> controlling risk. The worst alternative would be a system that has
many add
> on trades and then goes flat "after" a major retrace. This will kill
you if
> you are highly leveraged. (Been there, done that)
>
> I do use fixed ratio on one of my least profitable systems ie. many
short
> lived trades, and this keeps the losses down. The inefficiency of the
system
> is more than made up by the leverage.
>
> Would you rather have a system that tweaks a few more points out with
add-on
> trades, or a system that gets fewer points, but leverages up the
points you
> do earn.
>
> I'll take the latter system any day - my preference.
>
> You are certainly getting to the heart of the matter when you talk
about
> combinations of trading systems and leverage. Only extensive system
testing
> combined with money management will answer your question fully.
>
> I would guess that the first trade is likely to risk the least
drawdown,
> thus the fixed ratio method would have you bet the most. The second
trade,
> likely to have a bigger drawdown risk, thus less capital would be
bet. Bet
> size is related to estimated loss/drawdown. So the bet size in
relation to
> add-on trades would be related to, for example, delta=1/2 drawdown for
each
> add-on trade based upon the drawdown character of each add-on trade.
I
> really believe that each add-on trade should be treated as a separate
trading
> system, including money management for each. I believe Ralph Vince
also
> takes that approach for optimal f. BTW, his books are worth reading
even if
> you stay with fixed ratio money management.
>
> Happy Testing
>
> Kevin Campbell
>
> In a message dated 12/7/99 12:50:17 PM Central Standard Time,
> polar@xxxxxxxxxx writes:
>
> > I have been playing around with some Ryan Jones fixed ratio method.
I
> > was wondering how one would go about implementing this strategy
with
> > systems that allow pyramidding. For example, your buy condition is
met
> > and you buy 1 contract. The next day your buy condition is no
longer
> > true but your sell condition is also not true so remain long 1.
The
> > third day the buy condition becomes true again so a 2nd contract is
> > bought. The easiest way (but also most time consuming way) is to
treat
> > each pyramid signal as an independent system. But this leaves
valuable
> > capital on the sideline and it also assumes that the maximum number
of
> > pyramid contracts in the past will not be exceeded. Any ideas?
> >
> > Brian
> >
|