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

RE: Continuous and Perpetual contracts



PureBytes Links

Trading Reference Links

Being an end-of-day long term trader, Continuous back-adjust current
at actual-market-prices are the only contracts I use in all of my
systems.

I have looked at using the Perpetual style contract in the past to
track the market when bumping into the CFTC limits.  These size-bumps
occur when the contracts approach termination and change status.  I
don't use Perpetuals because I've never understood how the Perpetuals
could be of value as was reported for large traders.  Now when I'm
forced to move some of the position forward, I just follow the trade
and adjust the signal values using the difference between the current
and future contract close.  This difference is reported as an offset
in my daily broker faxes.

If someone has found a simpler way of doing this with Perpetuals, I
would be interested in understanding it.

My rollover timing is different than what has been reported.  My
approach is to follow the money in the contracts and stay with the
crowd so size stays under my signals.  This means that in the
physicals, tracking the Open Interest has served me well.  In the
cash settled markets, Volume is a better indicator.  In addition, for
the currencies, the floor will tell you ahead of time when the next
contract is going to be the front month.  In the Bonds and Notes,
this is always the close prior to first notice day.  This year in the
bonds, May 28th was roll day.

Constructing contracts the way I will trade the markets has made
trading less stressful because roll-over timing is anticipated and it
matches what was tested.  It is also good for the broker because he
gets to knows what I want with little explanations necessary going
forward.  He is then able to work the size in when it appears without
becoming the market.

Another feature that wasn't mentioned about Continuous contracts was
their ability to keep trade signals alive during rollovers.  Often
the pattern or market information that originally put you into the
original signal won't be there in the next contract or if is there,
it most often isn't in the same place.  If it goes away or moves
because actual contracts are being used, the trade will require
manual calculations or abandonment.  Giving up good trades isn't
necessary with Continuous contracts.

The only disadvantages I've found to using Continuous contracts is
the daily maintenance and system test result adjustments necessary
when using a static file name.  Even though my daily data maintenance
is almost totally automated, it is still a chore that must be done
everyday in addition to the actual collection of downloaded data.  As
for the adjustments, the number of rollovers must be accounted for in
all system reporting before the results can be accepted.  This
adjustment process will be automated before long and this too will be
a non-event.


Hope this helps,
Roger...



-----Original Message-----
From: Timothy Morge [mailto:tmorge@xxxxxxxxxxxxxxx]
Sent: Thursday, June 10, 1999 12:18 PM
To: Jim; Chuck LeBeau; omega-list@xxxxxxxxxx
Subject: Re: Continuous and Perpetual contracts


Omega folks:

Reading Mark Johnson's post, I see one piece of confusion of
terms. Let me re-state what I use in my trading:

Besides actual data from the exchanges, tick by tick...

When trading bonds, I use a synthetic contract that takes
the front month's open, high, low and close until the last
trading day of the month prior to the delivery month and
then switch to the 'next front month' for prices. So for
bonds, I looked at the June 1999 day only bond prices until
the first trading day in June of 1999. On the first trading
day of June, I began using the prices of the September 1999
day only bond prices. I paste these together as one long
'continuous' contract and use it for for calculating
confluences of levels [old highs, old lows, old trend lines,
retracements, projections, etc]. I use this method because
it uses 'actual' prices and for these purposes, 'real'
prices work best, even though the prices are from 'older'
actual front month contracts.

When charting weekly and monthly data, I often use a
different style of continuous contracts [and Mark Johnson
would call it perpetual contracts]. depending on the
commodity or financial, these synthetic price series chart
'more true' than adding 'actual' front month contracts
together, as in the first example.

I can't comment on their use in systems trading--I trade
from charts.

I hope this hasn't further confused the issue.

Best,

Tim Morge