Welcome to the TFC Commodity Trading Forum.
Please feel welcome to join in on these informative ongoing discussions about trading futures and commodities.

The Trading Forum is intended for the open discussion of commodities trading. The management of this Forum does not agree or disagree with the ideas exchanged, and does not exert editorial control over the message posted herein. Read and post at your own risk. The risk of loss in trading or commodities can be substantial. We discourage the use of this Forum to promote trading that is acknowledged to be risky. Please note: many links from the Forum lead to pages on other web sites. We cannot take responsibility for nor endorse the information presented on those sites.

TFC Commodity Trading Forum

Re: IMF: 'Decisive action' needed for euro/Hal

Hi Lee. Predictons for NG use are based on degree days; specifically "cooling degree days" or CDD, which is in turn based on the difference between the average temperature of a day and 65F. Its a pretty crude measure, and does not take into account explicitly the efficiency of cooling systems, the insulation of dwellings or structures, the details of the ambient weather (humidity, cloud cover and wind) nor the choice to turn on the AC made by individuals or systems. As a predictor, nevertheless, it seems to be pretty good, as the predictions of net injection generally do not deviate too far from the EIA numbers. So a good prediction of net injection comes down to a good forecast of CDD. Therin lies the possibility of error, and the possibility of an edge for traders. This is why using a reliable, accurate weather forecast can help make a trader money. As I understand it, commodities traders do employ meteorologists to do this work, but each one has preferences and knowledge which creates their preferences for data and models, and their forecast. Each will use a different approach which creates a divergence in forecasts of net injections. I generally use a multi-model approach and also, extensively, use an ensemble which is composite of 2 models run 20 times each, initialized with either perturbed intial conditions or internal physics, then averaged. So far, its working well enough although at times i've had to wait a few days or weeks to make a profit. As for your paper calls, my gut is that we will indeed see some hurricane action in the gulf this summer and that its this sort of rather unpredictable event which can turn out to be the real money maker. But that's just a bet based on pretty new science.
Cheers, CJ

Messages In This Thread

Re: IMF: 'Decisive action' needed for euro/Hal