Weather Trading (prediction, forecasting): Rain to Hurricanes

Anyone have any idea how to trade weather as close to "spot" trading (like FX, metals, cryptos, oil...) as possible?

I started with the weather prediction before moving to FX. To me they are somewhat similar... Just had problem "trading" (making money) the weather back then... don't know if that have changed much, at least for "spot trading". It was rather complicated since they did not allow to make "bets".. It was more for insurance purposes for real business that needed insurance/hedge, if I remember correctly.. So i dropped the idea.. and not long after, went to FX...

Also please check related posts of mine:

1. https://www.energyexch.com/showthrea...8#post12481258
2. https://www.energyexch.com/showthrea...7#post12468297

So would really appreciate to links to some website where one could trade (bet - they try to avoid this word/activities as much as possible, it seems) weather.

I found them before but there were some complications back then... But as far as I know there is no SPOT trading for weather directly, right?
Yeah, we can trade weather indirectly by all sorts of way since like 20%? of (USA) economy is based on Weather? But I am interested in some more direct way...
Like trading/betting/hedging on the rain/sun etc... change.. vs prediction...

BONUS: And even like... prediction path of Hurricanes even... or its effect....

I did some research and found two potential avenues for trading weather derivatives.

1. CME Globex
The CME offers CDD and HDD derivatives for 10 cities. I'm not 100% clear how to trade this as a retail trader. I believe they can be traded through Interactive Brokers, but even so I'm not clear on what kind of liquidity these contracts have and what kind of fees would be involved in trading them.

2. TradeWX
TradeWX is a broker that allows you to trade weather derivatives. They're by CXmarkets which is a subsidiary of Cantor Fitzgerald and regulated by the CFTC so they should be legit. Honestly though, this is the first time I've ever heard of them so I'd still be cautious.

As far as indirectly trading the weather goes, I think certain commodity futures are our best choice. Nat Gas, Orange Juice, and Soybeans are all possible choices.

Auricforecas, I'd be interested in working on a strategy with you if this is something you'd like to pursue. Meterology is a subject I've always enjoyed and my username EF5 is actually a tornado reference. I was really impressed by your model's accuracy on Dorian, from the very beginning you called the turn and knew it wasn't going to impact the US.

I somewhat avoid futures etc... due to much different taxation and reporting, but if there would be no better choice... Also, the second link you provided doesn't seem to work due to GEO restrictions... "The Amazon CloudFront distribution is configured to block access from your country.".. But no problem there, if there would be "the juice", could find a legal way.. The focus now is the juice and profitability... The closer I came in the past was "weather insurance" but they didn't want people betting on weather.. as I explained... So still looking for some more tradable website/broker.. to trade weather directly.. without the need to have business behind it (that depends on it directly on certain date/quarter).

As for indirect trading.. no doubt.. "OJ" aka Orange Juice... seems to be the most well known example.. specially due to the movie Trading Places (1983)...
The problem with indirect trading is that one have to factor in certain other parameters... Would prefer to have some more direct way of trading weather..

Yeah, it seems I might have "the thing" with the hurricanes but I haven't modeled many... ussualy when I personaly needed to know.. When I visited some country in USA that was under evacuation process (Houston)... "had" to know what was going on.. That many years ago... let's say more then a 10... people would know/google when that was But it is very exhausting... Let's just see the last example.. Yeah, my initial model matched what came at the end, more or less.. but I still tracked i carefully and I was aware of the consequences if someone would rely to much on it (postpone evac or something), so I didn't dare to publish the actual line.. but I had it made... I was really hoping/expecting it to turn before Bahamas.. I was tracking the route constantly.. all the news.. I must add something, when there was pretty much certain, on NYT path for example, that the hit will be less severe for FL than initially predicted... many news channels/press still posted doomsday scenarios... Well yeah, it is better to be prepared, but some really overreacted even when there was pretty much clear that some areas are almost certainly out of the question even by worse modelled lines...

So at the moment I do predict/model weather but primarly when I need to know personally... OR if there is a real challenege that is hard to guess.. or certain "window of chance"... I love to work withing probability models/windows.. and compare my results.. For example I did it for elections of 2016... And Ebola pandemic... And this latest was the hurricane... But I do not do it acivetly because I haven't found a way to profit from it.. or find better potential thane existing markets.. fx, metals, energy...

But yes, if you can find a way for us to make money or at least contribute somehow, to timestamp somewhere, be evaluated and maybe become part of some group... do let me know... I could model a few things for you in advance, for you to make a test of my abilities for example...

I must conclude that it is very much pretty much very similar to me... There are many things that are similar but also some markets are (partly) specific.. so it does help to know the "ballpark"... but sometimes it is best to not know anything at all For example, my modelling became worse when I added too many variables that seemed to be important.. (talking about fx for example)... removed them later..

Ok that would be somewhat "pulp fiction" on the topic, from me..

Still finding a way to make trades/money from the weather prediction Btw the other thing would be, but that is a big money game.. to offer insurances... Meaning, selling rain protection etc.. Like this for example https://wetterheld.com/beta/ (haven't actually tested), but to get you idea...

I agree completely. I don't like making indirect bets because more can go wrong.

What software are you using for your models?

Well, not to get the focus there.. I must say that almost "any" competitive software (many of them are free) is good... I think people focus too much on some tools, software... I am much more focused on setting the proper initial conditions, set the proper (co)relation... using "right" parameters and ponder them correctly... and using historic chart, different sources etc... to get the "extremes" estimations etc..

Some models are just "lost in the woods"... I have seen different models for different markets/industries and it seem... they overcomplicated to an extent that they no longer knew what they are doing... They sure got the time series right, curve fitting and many beautiful statistical conclusion but they forget what the MAIN force is behind something... One of the best example I have seen is inflation prediction by the FED... It is wonderful and very scientific until you ask some wrong question about some main parameter(s) (that is proven to drive the economy for example, by formulas!) and they say that it is already too complicated so they didn't include that one So the key is in the assumptions or initial conditions Also a good example that I use as an example is assistant system of certain car manufacturers that try to predict when the driver is tired and gives him warning.. It is a very good case how they overcomplicated in the beginning (with camera recognition etc...) but later found out that they could predict this from steering wheel/pedals... (alone), for example... Similar to markets... People using 100+ indicators that might not be (that) important at all, in the longrun, might be more true some of the time however.. It is, I believe, similar in weather, but much less than say markets... So if I get the REAL juice anywhere.. it is here.. knowing what to watch for and what (co)relates... Would leave the "basic" computation from then on.. to others/computers

But back to your question... I obviously rely heavily on what is already available from some weather "companies"... so I use that as a base.. and I try check myself what model line or "spaghetii" have the highest probability... and what the alternative route would be possible etc... So once I get that "donor"... something that already have probability distribution for example.. I use pretty much standard (stat) tools to determine the potential outcome... I use rather "industry specific" tools from one industry, so If I give a name, one could somewhat narrow me down much easier, so would prefer not to say But I think it is not vital, at this point, specially because we are dealing with many pre-calculated data/models already (in the case of weather)... one just had to use/cross-check them... might not even need so much comp/model power for that...

I mean it depends how far/deep one needs to go, you know how they say... 10x the work for every additional decimal point But for the ballpark.. one might not need a lot of data.. much less than some market present Of course with more data, the better precision, high-res matrix etc.. but the general idea can be applied rather fast... if using proper things..

I can give you example from other markets... Some experts that are a little BEYOND.. would need like 2-3 numbers to give you and initial estimate... that is true is financials.. some very good fin people check like 2-3 numbers in some company.. to get the picture of its STATE... also in eurusd for example... I know a guy that could model the market from NEWSPAPER chart and cross-checking with some 2-3 parameters... to be sure... And on and on... So for weather, hurricanes...
I tend to focus on the "whole pictures"... instead of computing micromatrix or something But that might change if I would be trading this market... At which point, I would be using as much tools/data as possible... and to fine-tune some matrix.. but we are talking about big money now

But I must tell you that my "tools" are basically the same either If I'm modelling FX, metals, energy, weather, hurricanes.. Well yeah, there are some specific and it is good to be "in" to get some ballpark/feel (at other times it would be better to not know anything outside the chart, so to not know what is impossible ... But the CORE is the same, at least for me... I must tell you one more thing.. I sadly come to the conclusion that many traders are not using anything from math/stat... but some "jargon" garbage I found that more scientific guys are rarely interested in trading.. and traders are rarely interested in science

For Dorian for example... I've done some fast rather deep research of all models posted, cross-checked with some parameters I think are the most important... and somewhat made my best guess but with putting ballpark/historic/other_models... into consideration... and then... live-tracking and fine-tuning, if needed.... Because the main question was the rather "ballpark" trajectory... It would be much different to microposition it however... but in the same direction

And P.S... Some are surprised what all can be done with "simple" Excel... not to mention Matlab...

That's an interesting point you make. Its normal to see physics and computer science PhDs in the markets, but someone with a background in meteorology could do really well too. I can imagine modeling a hurricane's path probably isn't too dissimilar to predicting asset prices. Its all roughly the same approach, just very different inputs.

Of course... What is hardest is to find bridge between the hard science guys and traders... Most of the time, at least in my experience... those are of very different personalities... It might be best not to confuse the two.. Just use those and unite them together (or even better, their findings)...

Anyway, if you know how to make profits modelling weather/hurricanes... Please let me know... I mean, preferably with direct SPOT trading, or if no batter, CFDs etc...
Yeah, we know about indirect weather trading... But would love to have it some more directly.. without some insurance contracts etc... At the time FX seemed much better but things might have changed... Atm FX is completely different... I mean when I started trading FX was "the shit".. real juices were flowing... But now it has been rather slow, almost completely stall at times/weeks.. or jumping about sone nonsense price... So I am also looking for alternative markets to trade... Did so now with CO2... Weather is of particular interest to me... still... So, should I find better market to trade... It does not have to be a market as we know it... It could be any form of making money with it... Posting/sending data/analysis/prediction somewhere. You get the idea.. Weather prediction->money.. that is the only criteria... How to make money with it and what would be requirement to enter, because "channels" might be much tougher if there is not a SPOT market where everyone could enter....

It looked like TradeWX was basically the only option to trade weather directly.

Tnx for the research... Yeah, seems to be same thing for direct weather trading... If you find anything else... Please keep us in the loop.
There are some ways to indirectly trade it... since much of the economy/stocks are related to it but I haven't found anything attractive enough yet or at least not more attractive than the current existing/other markets...