Yes, the COT report should be taken into consideration when trading a spread. 30 years ago the COT report didn't exist, the market would make an educated guess re composition of open interest. Now it is all there, neatly compiled for the market to see, It might be added that the COT works to the detriment of the speculative content of the market, e.g. funds. Anyways, it is one tool that should be taken into consideration with other influences. The markets have learned to anticipate how and when funds will manage their positions e.g. Goldman Roll. Saying that, a large spec long doesn't guarantee that a spread will move towards carry. There have been contracts squeezed by funds, most recently oats (which was predictable). From personal experience, 20 years ago i often held significant long spec positions prior to first delivery day, but I wouldn't replicate those trades now simply because of the COT report.

Why do "large traders" and "commercial hedge traders" seem to go opposite directions The problem with the Commitment of Trader reports are their accuracy. Back in 2008 the CFTC reclassified Vitol from being a commercial hedger to non-commerical speculator. The result? An 10% shift in the report. Yes reclassifying one company resulted …

SMCJB 

The problem with the Commitment of Trader reports are their accuracy. Back in 2008 the CFTC reclassified Vitol from being a commercial hedger to non-commerical speculator. The result? An 10% shift in the report. Yes reclassifying one company resulted in a 10% change in the COT reports! Makes you wonder about how so many other companies are classified. Reuters:- Big CFTC data revision raises oil traders' eyebrows

IMHO, if you have a good price-based model with confirmation, it will pick up on seasonal spread moves. You won't absolutely "time" a bottom or top - but nobody can do that anyway.

PropArb 

IMHO, if you have a good price-based model with confirmation, it will pick up on seasonal spread moves.

Just curious what you mean by this? I think of seasonal trades as having a (very?) significant 'time based' component which a pure price based model doesn't have.

ZS U-X spread traded -59 over night

SMCJB 

ZS U-X spread traded -59 over night

and as tight as -53.75 on the sell off

The bean spread that is going to get interesting over the next month is the N/X. funds long 130k N, half the open interest. I think they are going to take delivery of zero. Could get ugly if they roll.

Another bean thread that may be interesting. Some discussion of the U-X spread

Reminiscences of a Bean Trader or Why These Ain't Yo Daddy's Beans No-Mo SMCJB not really. i was never a spread trader pre se, except that i would trade bond calendar spreads in the pit, mainly because it was there and the transactions costs weren't; and and i could get the edge. in other words, i could put on a lot …

Very nice seasonal movement coming at sugar

SMCJB 

Just curious what you mean by this? I think of seasonal trades as having a (very?) significant 'time based' component which a pure price based model doesn't have.

My personal opinion is that terms like "seasonals" and " mean reversion" are descriptive terms or definitions that do not necessarily have to be a unique trading style or strategy. Most believe it has to be that way, but I am not.

A price based model that uses a well designed technical entry signal and relies upon historical trading ranges to set profit targets and stop-loss levels will pick up on "seasonal" moves in commodities like grains, energy, and softs just fine speaking from personal experience. Just my 2 cents.