Wind, Sun, and CO₂: The Weather Factor in carbon trading

[MUSIC] Welcome to Carbon Trading Chronicles, the podcast that untangles the complexities and potentials of emissions trading. Whether you're an industry expert or simply curious about how emissions trading supports the energy transition, this podcast is your platform to join the conversation. [MUSIC] The market is back and so are we welcome to episode one season three of the Carbon Trading Chronicles. My name is Stefan Feichtinger. I'm your host here with Virtus Environmental Finance. With me today, as always, luckily is Reham, our co-host, also Virtus. Welcome. Good to be back. And honestly, time flies by. It's been three years doing this. Three years since we are in season three. Today our guest is Eric Stein. He's the lead trading meteorologist for Eon Energy. Markets welcome. >> Thank you very much for having me here. It's a big honor. I've been following your podcast probably since season one. I'm a big fan of your podcast. >> You were one of those that clicked although, you know, when there were just three people. >> Yes, absolutely. No, I really found it always very refreshing that this format to get, but not really next year. >> And we're really excited about today because you're a trader and motorologist or you were a trader and motorologist, you're now the lead motorologist. But you off energy and gas and other things like that, right? >> I'm lead trading, motorologist on the trading floor. But in the past, I also traded, for example, US net gas, which is a very weather-driven product. And of course, the connection to carbon is there. >> But before we do that, we have that nice little thing in this podcast, which is, you have to decide if your bullish bearish on the carbon market. And I know you're not trading carbon, but I want to know are you bullish bearish? >> Yes, I'm bullish in the long term, right? And the reason for me is very straightforward. I think the renewable new build will continue to grow even though it's slowing here and there. But to meet the equation is if you have renewables more intermittent energy, you need balancing energy. Batteries cannot cover it, so in the end you will burn fossil fuels to some extent. And to me, also the interesting part is that the energy demand will also be growing. So we will always have a bit of a red race between renewable new builds and conventional. >> Rehamb, bullish bearish. >> Well, in the short term, I'm cautiously bullish on the longer term I'm bullish. As that, for September, I do have three reasons for this. >> Okay. >> I have the compliance, but I wouldn't be necessarily super bullish on the compliance, because overall, we believe that most of the compliance entities have covered for the benefit. But then you have the leading event of the month more or less it is the expiry. So the options expire on the 24th of September. And I think there is a bit of action that is going on there. And there is a play out of a trade. >> So how about you? >> And then we broke out, yeah. I'm ultra bullish. >> Ultra bullish, definitely. >> I've never been as bullish. I think since we started this topic. >> What's your ceiling? >> Oh, what's my ceiling? Well, I'm not that bullish then. >> Okay. >> I think when I hear forecasts of 120 or 150 next year, I'm like, well, you've got to be realistic nevertheless. But I think we really have a nice setup. It depends, however, on a lot of factors. So fundamentally speaking, we recently also put that out. We changed our forecast. We think there's somewhere around 190 million tons of shortage next year. That depends really on two big factors. >> Yes. >> And other factors that play a role. But really the two variables that can mess it up. Number one is power emissions. >> Yes. >> And power demand that the whole hedging side of things. And number two, three allocation. >> Exactly. >> Right. And the first one is really the one that we're going to speak a lot. Excuse me. Now with Eric. And why do we even speak about the weather? As we are moving towards more renewables, the times when there is a lot of wind and solar and others really matter for the carbon price. That's really when you see the things move a lot. And also through hedging, because as people start hedging closer to the actual event, if there is a change in the way that moves the price a lot, we've seen that in November last year. >> Exactly. >> When a few weeks before seem to us, that's the kind of price at the timeline when people start pricing. Is that correct? >> Oh, I think of course it depends on the time horizon, the lead time that you forecast. It goes, for example, the Donkelflout events. It days with low solar, low wind, and central Europe. We call the Donkelflout. The Germans got to work for everything, isn't it? Which is basically days with extremely low solar. We are talking less than 5% wind and solar of the actual capacity. >> Right. >> And these days, they are to exactly forecast them. You need to have a lead time of 70 hours max. That's something otherwise you cannot really put to the estimate. >> You see that it gets tight. >> You see that it's getting tight. We saw this last year. And we introduced something like a Donkelflout watch, and a Donkelflout warning. What did we do? Donkelflout is related to certain weather patterns in regimes. And while you cannot really forecast a Donkelflout, a 10 days ahead. That's impossible. You can forecast the average weather pattern around that time of the year. And there's also a climatology of the Donkelflout. So there is a maximum of, there is a time in the year when the risk is. >> Which is November. >> To mid-December, roughly. >> Okay. >> But it could be also October. >> It could be October. There's also the chance that you have a Donkelflout today in high winter, January or February. But it's more common ironically in November. Actually, Donkelflout risks in January or February would be even worse because then the average temperatures would easily go up. >> It's not too low. >> Right. So the Donkelflout is always. >> But couldn't you. I mean, so February last year wasn't that also to some extent a Donkelflout? >> Yeah, it has to be really bad for the whole world. >> It was close. But in February, the length of day is already in our more than it is in the beginning of December. So yes. But really the bad time or the time in the year where you have a culmination of events over the climatology of the last 50 years or so is November and December. >> Right. And then you. That's really just before. But what about the long term? I mean, aren't people like yourself trying to forecast now or the winter or. >> Yeah. >> Is there anything that you would say you can. >> Yeah, of course. I think most utilities, Hatchfons, banks, they do have weather teams, right? >> That don't try to look out for them. >> Yeah, some Hatchfons actually build around weather teams, like Citadel. They had the old cumulus guys taken over. I guess their weather team must be your mongous. The normal regular weather teams and utilities and banks, they're of course trying to assess whether the scenario is for the upcoming winter or summer season, for example. And if we look at the Hatchfons positions from last year, I think I remember was something like three weeks or so before this Dungal-Flaude event happened. You saw on the carbon side the length of the Hatchfons growing and it could be a coincidence. But it could also be that. >> Maybe that's what. >> No, I think there was already some forecastability in there because three weeks is what we call sub-seasonal timescale and there are actually credible models with reasonably good confidence if your question is. is course enough, if you have two detailed questions you cannot do weather forecast beyond seven or eight days. But if you want to have, okay, I need an idea about the potential weather pattern regime, what is the normal frequency and probability of such a pattern. And then when this probability moves up, you increase the risk for potential Dungal-Flaude. And if you are hatching your position then you start probably first buying out of the money options to cover that. And then the option market trickles down into the spot market. And then you see the curve moving up with the spot. >> And I think this. >> And you see that you see to start three weeks before it's a bit slower and then the curve must speed up, right? Where you see those in the market? >> So, calendar week three and four is the classic period for sub-seasonal forecasts. >> So assuming that there would be, let's say, expectations of Dungal-Flaude or something, we should expect the hedging to start for it somewhere in October. And this should trickle down to some price action as we get in the winter, right? >> I guess all players, traders, decision makers are human beings, right? And we have this massive Dungal. It's pretty much a mother of Dungal-Flaude events, I would add last year. If you look back into the climatology, it was quite unusual, actually. >> Right, but that's how likely is it? I mean, as an analyst, that's what we have all the time, right? We see what happened last year and we're always guided by, you know, recentifiers. >> That's what it's called, the probability now is. >> I mean, I don't know, it's moving up overall, but it's still more of an outlier, right? >> Yes and no, if you go back over seven tiers of climatology, this is how we have the data, the historic data sets really available from, for example, the ECMWF. >> One of the big European. >> Yeah, the European weather agency. >> Agents, yeah. >>. you want. Then the probability is the same as last year, right? Yes, it's not really high, but. But nevertheless, number one, the impact becomes higher. The more renewables you have, then don't go flour to becomes from a kind of noisy, low likelihood, low impact, or high likelihood low impact. High likelihood low impact. High likelihood low impact becomes more and more to low likelihood high impact. And that's when it really matters. And this is when it really matters. And if you know, and then as I mentioned, there is always a bit of a memory effect. We had it last year, so people have it well in their back, in the back of their hats. And they do not want to be caught on the wrong side if it should happen this year. And the longer the last donkel flout is in the past, people will probably be coming less aware of it again. I think one point maybe that is connected to this, but it's a different event. It was the water levels in the Rhinane, like a couple of years ago. I think it was two years ago around the energy crisis. And then like this year also, like we've seen the same effect, or like the same pattern happening in the water levels. And I think the market hasn't reacted. Yeah, but the market hasn't reacted to it as much as it has reacted before. But so then 22 maybe you can actually shine some light on that, because 22 was very interesting. You had low availability of French nukes. Must be what was that? I had to, yeah. Yeah, yeah. Hydro was flattened completely, right? And then there was a complete setup. Yeah, and on top of the water levels were super low. Yeah, this is this is also quite an interesting one. The actually the outflow of the Rhinane is dominated by an annual cycle. In the winter, it's precipitation. In the spring, it's snow melt. And the summer is how high is the snow line moving? So if the snow line in a less warm summer, the snow line stays lower and the glacier survives. And then every year the glaciers in the Alpine are losing mass. And so it's actually the glacial melt off at the end of the summer that keeping river levels up higher. And if that potential melt off is shrinking because the actual mass, the parent mass is shrinking, then the risk for low rivers if we should have a dry autumn or relatively try a late summer, then the risk for very low river levels in Rhinane and Rhone and E-Zair and all these rivers is increasing. Yeah. So how do you see it currently? If you compare it to 2022, at least now, French nuclear availability to me looks quite healthy. And then hydro levels, I believe, are also-- Yeah, we had after a very dry Q2. We had a relatively wet July. And even though there was not massive glacial melt off available, the most reservoirs have been replenished. We're seeing a relatively wet episode delivering. And so in the end, I would say that risk is not one of the extreme years. But one needs to have an eye on it because it builds very quickly. You have four dry weeks. Does it have an impact on the extent of the dunkelflutter or is in different areas? Of course, the flood is more northern Germany. Yeah, if you have a dunkelflutter, you let everything run that you have. So even small river, run of river plans will be running. And of course, if then river levels are low, you have even-- then the situation is even worse. If there are some moth-bolt petroleum plant is jumping in, then A to profit from the high prices, but mainly to secure the stability of the grid. And I mean, in one-- you have renewables that increase every year, maybe slowing down a little bit lately. But overall, that should reduce the need for fossil burn. But on the other hand, as you said, that also makes-- if there is an event where there's not a lot of wind, that makes the impact much stronger. I think what is even more important is-- especially in solar, we are now past 100 gigawatt-mark solar capacity. And this is not actually looking at all these solar panels that you can buy for your balcony from the discounters, right? What becomes pressing are the flanks. You have to balance the flanks. You are losing with every minute or gaining with every minute. Gigawatt's of electricity into the grid. You have to keep the grid stable. And yeah, what can you do? You have to either have very flexible nukes or very flexible gas power plants or very flexible batteries. Yeah. And-- And absence of the batteries, like more or less, the most reliable option here is the gas. And that's the case with gas, I guess, also in general, right? Especially since the energy crisis we have seen that were, before we wouldn't have to care much about the weather in China. But then last year, for example, maybe you can speak a bit about that if you remember. Last year, I remember there was a lot of bullish bets on the gas side from hedge funds and others on LNG and on TTF because of that. I think it never fully played out so much. I think the idea was that the weather that demand is going to be very strong and the weather is quite extreme in China and others. And then a lot of people got burned. But yeah, I have to say summer weather was actually very hot in China. It was in northern China, in South Korea, in Japan. We had one of the hottest summers specifically in Japan. But this is really showing how complex things are. Everybody was expecting the NG market to be probably not everybody was expecting, but there was-- But that's exactly what an idea there. But that was the idea. Because as you say, things are so complex. And people need to simplify it, right? Because you cannot trade everything. And everything at once. And with this idea goes. And the interesting thing is when the market gets crowded on one of these ideas and it doesn't fully play out, then things get interesting in the market. And the trouble is that the things-- why things are not fully playing out is political risk and weather risk. And that's the thing with carbon as well where I wanted to get to because we are right now seeing. Maybe it's not crowded, right? But it's starting to be more of a crowded. Everybody basically expects the market to be tight. I mean, we're all Polish here also, right? For the first time. And I think there is a good reason for that with the fundamentals that we're seeing. Nevertheless, if you're imagining a very, very mild winter, then all of a sudden, the tightness is not so tight anymore. And then I want-- so I want a little bit-- also the position-- if you are a hedge fund and you're building a big long position right now, it's always the question, how long term it is, right? Everybody wants to make you tell you that's for very long term. But is it really the long term? Or do you need to make money now in Q4? And if you do, then you probably need exit liquidity. In other words, if the price does go up and all of a sudden, it gets really cold or another dunkelflow, that great, because now you can exit and you don't crash the price. But what if it gets really mild? And yes, the whole year of 26 overall is tight. But let's just assume industrials are not going to massively buy in winter yet, because maybe they don't see the shortage. Maybe they think the price is too high as it relative to much. Maybe they're waiting for better times. And the weather is mild. Who's going to be the backholder? Who's going to be the person who's actually buying the stuff from you that now you want to sell? If you want to get out of your position. Maybe that doesn't crash the price, but that could lead to another kind of a boring face. That's the thing, because I think you're going to have the people that get burned basically in the positions. It might not be a major crash on the market. But overall, I wouldn't say that even if next year would have mild temperatures, or even if next year has like lower demand from industry, that this is going to basically delete-- Delete the bullish story. It's very hard to write. It's a bullish story, because-- It's really hard to delete the bullish story. Yeah, let's say the-- let's say the revival has a materialized. Let's say the power emissions are lower than-- and instead of the shortage that you have now at 140 million or slightly above 100 million, you'd have something around 80 million. This is how many times, like this double what we have for this year. So in general, that creates some pressure, but maybe not that much. Exactly. So let's say instead, I don't know, what would be the forecast for next year? We have a forecast that is going above 80 for sure. So instead of maybe tackling or attacking the 90, then it would be comfortable trading comfortable between 80 and 90. And this would be the sideways. And then it would depend on how-- So that's exactly because our base forecast, I think it goes to 100 or edges off from there. But that's exactly the point. And also, the final benchmarks are not out yet. So in our forecast, that's really bullish because we expect one of the biggest drops in free location. I think that we've ever seen in this market. And that will affect industry that I think to the largest extent is not really aware of it, or they're aware of it. But they don't have the cash right now to worry too much or the products to trade this. Or-- Well, first of all, the numbers get published. in Q4. - Yeah. Q4 most probably because the numbers were published the time before and Q4 as well. So they'd be aware of these numbers. >> They will because then it comes through the associations and the will tell them. >> Exactly. So like until everyone is aware of the numbers, probably would be into 2026 because we're speaking of here and the holiday season is kicking in. So people would be aware of this. Then they basically assess the impact. >> Until they buy actually, right? That could spill, until they exactly they assessed the impact. >> Yeah, by the end part. >> The winter. >> And after the winter might be when, I mean, that's really where it comes down to the weather. So you have the ECMW, the regular physical models. >> The regular physical models, yeah. >> And now you have the A.A. models. >> Right. >> And they're gaining a lot of traction and people start thinking they are better than the human one in some aspects. >> But there are a few problems in there. Number one, first research already showed that machine learning models are not so good in forecasting extreme events. >> Right. >> Because they're kind of smoothing the past a little bit, right? So they might even be better in forecasting a hurricane track. But in forecasting a heavy rainfall event in Europe or a drought or an event, there seem to be just a very fresh paper out there. It seem to be not so good. That's by the way also our experience. >> You have immediate feedback. >> ECMWF published the latest version of their ensemble model in July. And now we have already three months of daily work experience with that. And this is so exciting about this. >> But for us, yeah, about the job and what it's doing to the markets. And it's so interesting I set to my colleague. We are, I'm 20 years in the business. And now this AI model thingy is kind of leveling the playing field again. Because my know how is also not more than with this one model more than two months, like everybody else's. >> Right. >> Whereas with the old traditional models, I could bring in 20 years. >> Now the AI part is part of the base. >> Yes, the AI is of the ensemble or whatever. >> Yes, yes, they come all together. They're coming from the same sources. >> And in your experience is the market usually pricing in just ECMW based case. >> Yeah, in 20 years ago it was like almost only GFS, America, the podcast model. And the ECMWF developed so much better than the GFS that it came to gold standard. And now everybody is jumping on the machine learning models and the whole field is re-shuffled again. >> Right. >> But the market is pricing in usually those. >> Yeah, yes. >> If you see a big move in the AI, FS model or in the machine learning model, then the market will react right away. >> And then the market will react right away. >> You mentioned earlier that there was a forecast that got published to end of August. And now it is sort of like the word on the sheet in terms of the weather forecast for the next few months. I would like to ask you, did this forecast have deviation from the previous ones? Is it showing some extreme events actually for the winter? Because what we're trying to know in here or trying to understand in here, what are the scenarios that they are pricing in for the winter? >> Yeah, I would actually even say that the market at the moment is probably not looking at forecast model scenarios. Of course, the models that you were mentioning, they have a fixed schedule. There is one model appearing on the first of every month, next on the fifth and then next on the 11th. So you have these kind of time steps that can be market moving, especially at this time of the year. >> I think that the scenarios for this winter are more built about questions like, do we have a Laninia or not strong Laninia events? This is what I love so much about working with traders because traders are next to meteorologists the most weather savvy people I know. I can say there is potentially a Laninia developing and most traders instantaneously understand what I'm talking about. They know exactly in the back of the head on, this means this and this. So yeah, obviously we are in my team, we are also shaping our ideas for the winter which I cannot share here. But that's one of the things that people are looking at. >> When do you think people start really looking at what happens in the winter? When is the market, we said this three weeks before, is that the same here? Do you think three weeks for weeks that we start seeing impacts because that's also trickling through to novel models? >> I think the impacts already started. >> I think that usually the first requests from the traders to the meteorologist for the upcoming winter are popping up in as early as August. >> Really? >> Then of course we have, I would say, the seasonal weather models. Usually you might have to wait till October, till October version to get kind of a scenario with a good confidence. Of course you get scenarios out of the latest weather models. You get scenarios out of that. This time it was maybe so far a bit disappointing because the latest update was not so much different from the August update. So that is. >> But when it is different than anything. >> Yeah, but when it is different things are. >> We should start feeling the impact. >> Yeah, yeah, yeah, you start feeling the impact. >> Yeah, but the good example actually was El Nino in 2015. The El Nino was relatively early agreed on to be one of the strongest El Nino's in this millennium, which was actually true. And strong El Nino's create very warm wet windy patterns in Europe. You could actually see how the coal price already started to drop in October. >> Really? >> I'm not sure whether the carbon price back in the day was affected. >> Yeah, I guess it was very low. >> It was very low. >> I guess it was 456. >> Yes, it was very, very low already. But I could imagine that if such a situation would pop up again. >> And it actually played out in 2015. >> It played out like a charm. >> Yeah. >> And one last thing on my side actually, it is connected to gas and pricing because you mentioned coal. And then now we see that gas prices are trending lower. They are not trending higher. Like with some deviations here and there. But in general, look at the trend. It is quite on the lower end of things. Do you think the gas pricing as of now is connected to this weather event or it's a combination between the supply, the expected supply of LNG, expected pipe supply as well as the weather events? >> Yeah, I mean gas and weather is one of the original partners in terms of trading. I mean, it started back in the day with heating demand. And today it is via the LNG kind of a global market if you want. So if you have very hot weather in Japan or in the past, or let's put it the other way around because if you have very cold weather in Japan and you have very cold weather in the US as well, which can happen. But you have mild weather at the same time in Europe. Your spot cargo will probably divert and deliver those markets who are more promising in terms of spot prices. And this is already happening since quite a number of years and it's of course becoming more and more interesting. >> Speaking of infrastructure though, I think we have a question from our last guest. >> Exactly. >> And then we're going to ask you a question for the next guest. >> Yeah. >> What was the question from last? >> So the question was left from Oliver. He said or he asked, do you think we can find a way to make the EU competitive again? Maybe we can just make it very short on that one before we wrap up. >> I think we should improve things where we are already good, which is education, standardization and science. I think the EU is actually gaining ground again in research. But if we are kind of sticking together and cooperating more together, I think we can be very competitive. >> I think for God who said, it's a classical thought philosophy of first principle thinking if you want to know how to do something, think about how not to do it. How can we be the most in competitive that we can possibly be? And then we know what are the obstacles maybe that we need to remove. What's your question though for the next guest? >> The next question for my guest will be actually whether this person thinks that whether the route is will finally take off or not. >> That's going to be a tough one. >> Great, well. >> That's a good one. And Eric, I must say, you are the first person who shared this ray of sunshine in this podcast about the future of Europe. So thanks for that. >> Yeah, that's good. >> Well, thank you very much, Eric, for joining. Thank you as well, Reham. And thank you, everybody, for watching of yet another exciting episode of the Carbon Trading Corp. [MUSIC] This podcast is brought to you by Virtus Environmental Finance. Join us every month for new episodes. Don't forget to subscribe, rate, and share. Stay tuned for our next episode.