Perpetual Data

[chris@xxxxxxxx (Chris Norrie)]

> tell me how any other method can reflect the actual market more accurately?
> do the home work and get back, and i dont mean read (some loosers) swaggerts
> comments in some magazine.

Mark or anyone else, educate me here.

Let's say:
   - I'm trading the Yen and on the 25th of the month preceeding expiration
     I switch to the next contract.
   - I rollover by exiting the old contract on the close and entering the
     new contract on the close.
   - I build a continuous contract that stitches together the data in the
     same contract months I am trading.
   - My EL system actually exits and re-enters according to my rollover
     rules (thus the contract spread does not show up as a profit or loss).

Ignoring slippage, etc, do I not have a continuous contract and an EL system
that works with it in an appropriate manner to mimic past performance.
By past performance, I mean showing me the prices and dates for recent trades
I have made plus historical trades that I did not make because I wasn't
trading the system then.

The answer is probably NO.  The reason being that my system references
past bars to make an entry/exit decision.  At rollover time, there is a
discontinuity between past and future bars!

Now let's say my system does no ratio calculations.  No calculation depends
on the absolute value of prices.  In other words, my system would perfrom
the same if my entire data series had a constant added to it.

So now when I stitch together my continuous contract, I match the closes
of the rollover bars.  Thus, bars that are not part of the current contract
are offset by a constant.  However, as discussed, this does not affect my
system.

Are my profits and loses from the trades now representative of what really
transpired in my account (ignoring slippage).  Not quite, but closer.  If
I have a moving average, for example, immediately after rollover time the
moving average will be calucualted on the prior contract's adjusted bars.
However, if I'm trading with only one data set, the moving average will be
based upon the current contract.  This may/will give me a different moving
average result.

So the rollover must either be phased in over the period of my moving average
OR, instead of trading the current contract, I trade my continuous contract
(whose bars for today will be the current contract exactly).

Given the nature of my system and approach, is this not the continuous
contract superior to the perpetual contract?

I hear your argument that the perpetual contract is flatter and therefore
harder to make a profit.  However, flatness means less volatility which
affects the sizes of swings between draw-downs and draw-ups in addition to
other measurements I may be making.  Therefore, I am getting a different
picture of what I may expect and have seen in the past in terms of account
activity.

The question is, why would I want to abandon my approach for the perpetual
contract?  What do I not understand here?

Chris Norrie