Hi all !
I am a Junior Power trader (prop only) in a small european boutique ! I trade power in different countries on a short term basis (days and weeks). On this time frame, weather is king ! I have basis in weather understanding but I can't properly explain a move from one forecast run to another - which seems normal as super computers run half a day for each run...
I look everywhere for other forecasters than ECMWF and GFS, other sources that do their own forecasts based on the original inputs of EC/GFS etc. in order to tale a position when one is very different from another. However, it seems more like a random process than anything else to predict the next move of weather forecast (especially the wind...)
What do you guys think ? Is this a good way to trade the short term power ?
Happy to share on my little exp if you have any question !!
Hey doudounovic, what a lonely thread. I'm here since nobody responded ...so maybe one of these discussions will help:
I hope those threads give you a bit more insight.
weather is a commodity - ecmwf is the best, GFS is a POS. Honestly weather wont give you and advantage, everyone is the same and everyshop has the same vendors - find a different way to give your self an edge
Hi nattyphizz ! Yes I totally agree with you... Everybody is given the same info ! But if you trade the very short term, only weather changes creates moves !
I try to evaluate the risk, like wind is very high, I think next move has higher prob to see the wind decreasing... But nothing really reliable you know !
It seems a good strategy. Considering also Samuelson effect for longer maturity and you are talking about short term trading Power is classic example in stochastic processes examples with mean reversion and jumps, thus it comes quite hard to predict future prices
run what you think load is on 30 year normals, 10 year normals, forecast, forecast +/- 10% change and try to evaluate risk/reward. for ex a super simple trade if load comes in 10% higher maybe you think price increases $5 but if its 10% lower than forecast maybe you might model a price decrease of $20. All depends on where you are in the stack. I agree wind is huge now and very hard to predict - just use the forecast and try to evaluate probabilities
When one is modeling load based on 30 yr normals does that work if you include temperatures at several nodes? Or will that introduce errors will that under/over estimate load due to how weather patterns manifest?
I would focus more on generation/transmission outgaes coupled with weather. Also learn what goes into each model and the differences in MB height. Ex: PJM late Oct 19, gas getting hammered, no cong, hr collapsing to 10, 2nd week traded 28 w/ intra day trading 26, outages for the 2nd week was peak generation outages at ~40GW, downside on $28 is probably $25....upside $40 if colder weather comes in....models run a few days later the 2nd week is going for $46 with load now forecasted at 110GW w/ 40GW of gen out...very simple example but looking at load/outages/pipeline conditions then throw weather on top maybe be easier than saying I think the 12Z GFS is going to be colder...just my 2 cents.
Autem corporis asperiores aut officiis vel iste aliquid. Quia voluptas excepturi saepe explicabo molestiae.
Sint et ut dolores in ipsum quia. Qui tenetur quas maiores ex. Eum aut et eum illum est. Nesciunt vel reprehenderit non qui molestias eum laboriosam.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...