|
PureBytes Links
Trading Reference Links
|
Mark Johnson wrote:
>
> Good old CSI certainly believes their Perpetual
> Contracts are lots "better" than continuous contracts.
> I think I saved the issue of their newsletter that
> discussed why -- I'll look for it tonight. Something
> about "stationary" as I recall.
>
> Prices in a perpetual contract are a weighted average
> of two actual prices from two actual contracts.
> For example, the two actual contracts might be June 1998
> Crude and Dec 1998 Crude. The "weights" are arranged
> so that the nearer contract is given more weight.
> Examples:
>
> For 02Jan98, perpetual = ((105/106) * CL98M) + ((1/106) * CL98Z)
> For 15Mar98, perpetual = ((53/106) * CL98M) + ((53/106) * CL98Z)
> For 01Jun98, perpetual = ((1/106) * CL98M) + ((105/106) * CL98Z)
>
> This has the "advantage" that perpetual prices don't
> get "distorted" by rollovers. Look at a continuous contract
> for S&P500, they will tell you. In 1982 the continuous
> contract prices are around 120 .... but the futures themselves
> traded around 300. What a "lie" the continuous contract
> is telling. And plenty of commodities have continuous
> contracts that INCLUDE NEGATIVE NUMBERS! Horrors!
>
> Algorithms based on the absolute level of prices, might do
> better with perpetual contracts.
>
> Me? I use continuous contracts. Exclusively.
>
> Mark Johnson
> --
>
Something I've wondered about. If here in Jun., I want to consider
relatively longterm support/resistance values (where the mkt turned
previously), is it best to look at the nearby contract at that time
(say, Mar.), or where Jul. was trading back then.
For support/resistance it seems you have to go to the actual contracts
not continuous contracts. Even then, if I look at where Jul corn was 6
months ago, back then it included 6 months of storage, so support then
might not be exactly where it would be today. (Of course a lot of
things could have changed, so support and resistance are only
approximate numbers at least for longterm, right?)
Maybe this is an advantage of perpetual? The carry or interest or
whatever is approx. constant because a perpetual effectively keeps a
constant time from expiration by varying the weights of the contracts.
So particularly for currencies, for example, a perpetual might give the
best support/resistance values (assuming you look also at the current
perpetual contract to do today's analysis). However, for commodities
with discontinuities like old crop/new crop, I remember CSI warning that
perpetuals were not as desirable here.
Other than sup/res I think i lean toward continuous, not perpetual - tho
I am far from a n expert.
Conrad Bowers
|