This week, we’re asking the central question of the energy transition: How fast are we going?
Clean energy is bringing in $2 trillion of investment annually. Wind and solar now account for the vast majority of new electricity capacity globally. And we may already be at “peak trade” of fossil fuels.
And yet, when we look at the share of renewables in final energy consumption, they’re increasing only incrementally around the world. At the current linear pace, meaningful decarbonization may take decades longer than needed.
This week, we’re joined by one of the most prolific and respected market analysts for a 360-degree view of how the transition is playing out.
In this episode, Michael Cembalest, chairman of market and investment strategy for J.P. Morgan Asset & Wealth Management and author of the influential “Eye on the Market” newsletter, walks us through his annual energy analysis — a comprehensive assessment that covers dozens of economic and financial trends in energy.
This conversation has a little bit of everything. Cembalest walks us through his “scorpion bowl chart” that shows the gap between clean energy deployment and impact. We also dig into the transmission crisis, rising costs, equipment delays, and tariffs. And we cover the data center challenge, Europe’s economic challenges, and China’s rise.
But it’s not all gloomy. Cembalest also identifies where real progress is happening, and what a realistic path forward might look like.
Credits: Co-hosted by Stephen Lacey, Jigar Shah, and Katherine Hamilton. Produced and edited by Stephen Lacey. Original music and engineering by Sean Marquand.
Open Circuit is supported by Kraken, the only proven, AI-powered operating system for utilities. Learn how Kraken helps unlock excellent customer experiences, increased innovation and reduced operational costs at kraken.tech.
Open Circuit is brought to you by On.Energy. As one of the fastest-growing battery storage IPPs, On.Energy delivers turnkey resiliency solutions for utilities and enterprise customers. Whether you’re managing data centers or local grids, we help bring storage to your fleet. Learn more at on.energy.
Transcript
Stephen Lacey: Do you guys remember when I asked you if you had panic bought anything after the tariffs were announced?
Jigar Shah: Oh gosh, what did you do?
Stephen Lacey: I finally caved. I broke down and purchased multiple toilets and like a hundred rolls of toilet paper and paper towels.
Jigar Shah: I’m more worried about if you bought more elk meat.
Katherine Hamilton: I am finding the shelves are already starting to become depleted. I went on Amazon to get a headset, just a cheap headset for one of my kids, and the only ones you could get were used. And let me just tell you that the one thing you do not want to buy is a used headset.
Jigar Shah: A bidet. You should get a bidet, then you don’t need toilet paper.
Stephen Lacey: My wife did buy one of those Japanese toilet seats. Well, this week we brought on one of the most respected market analysts to basically tell me if it was an irrational decision or not. Michael, what do you make of my toilet hoarding?
Michael Cembalest: Let’s see. If you buy it on Amazon, you’ll have to see whether or not Bezos decides to disclose what the Chinese tariff would be.
Stephen Lacey: What a way to celebrate Trump’s first 100 days. From Latitude Media, this is Open Circuit. This week we’re asking the central question of the energy transition, how fast are we going? Wind and solar now account for the vast majority of new electricity capacity globally, and yet when we look at the share of renewables and final energy consumption, they’re increasing only incrementally around the world. At the current linear pace, meaningful decarbonization will take decades longer than needed. This week we’re joined by one of the most prolific and respected market analysts for a top-to-bottom view of how it’s all playing out. What happens when transition ambitions hit economic reality? Our conversation with Michael Cembalest is coming right up.
I’m Stephen Lacey, executive editor at Latitude Media. Welcome. I’m joined by my co-host, Jigar Shah, as a clean energy investor and former director of the DOE’s Loan Programs Office. Jigar, you announced a big thing this week.
Jigar Shah: Yeah, Jonathan Silver and I have decided to set up a new firm called Multiplier. We’re helping companies part-time. We all have other stuff going on, but helping companies get better exits.
Stephen Lacey: He’s a former director of the Loan Programs Office as well. What is it like when former directors of that office get into a room? Do you guys get commemorative jackets and sit around smoking pipes, talking about treasury curves and liquidity spreads?
Jigar Shah: No, we’re mostly just talking about Senate hearings that made us uncomfortable.
Stephen Lacey: Katherine Hamilton is the chair of 38 North Solutions. Katherine, how are you? You were at the BNEF Summit this week, right?
Katherine Hamilton: We are. It’s great to watch Jigar, and I was so excited about his announcement. Last time Jonathan Silver and I started trying to work together a few years ago, we were going to do a SPAC and I’m really glad we didn’t. I’m really glad instead that he is teaming up with Jigar on this.
Stephen Lacey: You don’t have to put yourself on the SPAC shame wall.
Jigar Shah: Nope.
Stephen Lacey: You went to Jigar’s panel with Neil Chatterjee. Did you heckle Jigar from the crowd?
Katherine Hamilton: No, just Neil.
Stephen Lacey: So our guest is Michael Cembalest, Chairman of Market and Investment Strategy for J.P. Morgan Asset and Wealth Management, where he writes the influential eye on the market newsletter, a superbly crafted analysis on economic and market trends. We’re so happy to have him here. And every year Michael writes a comprehensive energy report, which is required reading for anyone trying to understand the complexities of global energy markets. His latest report titled Heliocentrism: Objects may be further away than they appear, offers a clear-eyed view of the state of energy and decarbonization, and he is here to help us understand it and talk through it. Michael, welcome.
Michael Cembalest: Thank you.
Stephen Lacey: Really good to have you here. I read that Jamie Dimon once called you, “One of our firm’s great thinkers.” Do you put that on your business card?
Michael Cembalest: Yeah, I don’t have business cards anymore. It’s a digital world. I don’t even carry a wallet.
Stephen Lacey: I’m actually curious, before we dig into the data, you say in this report that you hope that government statistics agencies will continue to provide data under the Trump administration. Are you seeing any data decay right now?
Michael Cembalest: Yeah, sure. There’s a couple of BEA and BELS websites that we used to go to, and when you log into them now you get a denial of service message or sometimes you’ll get, “This activity has been suspended for budget reasons.” There’s two levels of badness. One level of badness is they just decided to cut that particular function because there’s not a lot of users like me that use that particular data. The worst level is they have actively decided to which data sources to stop updating because they would provide some potential negative commentary with respect to the administration policies. It’s too early to tell which one they are. Both are bad.
Stephen Lacey: Do you feel like this is going to be really damaging for just your work and thinking through economic indicators?
Michael Cembalest: It’s not going to be great. More data is always better. Look what happens in China. There’s so little confidence in some of the officially released statistics that people derive their own… Bloomberg does this. They derive their own China indicators based on electricity consumption and port traffic and thing. Then they triangulate to what they think the Chinese economy is actually doing. There’s a lot of really important data that comes out of the EIA, and based on the budget cuts that are being proposed, we can’t tell at this point how much of that will get sustained. What’s going to happen to all the fantastic reports that come out of NREL and NBNL. I’m kind of holding my breath.
The Scorpion Bowl chart: measuring transition progress
Stephen Lacey: Okay, let’s get into the report. The subtitle, as I said, is objects may be further away than they appear, suggesting there’s this disconnect between perception and reality of how the energy transition is playing out. And I want to start with a chart that you call the scorpion bowl chart. Why do you call this the scorpion bowl chart? What’s in it and what does it tell us about the current pace of the transition?
Michael Cembalest: Sure. In the early 1980s, I went to college in the Boston area and there’s the bar called the Hong Kong Bar in Harvard Square. They served a thing called the scorpion bowl, and it was terrifying. It had everything in it, rum, gin, vodka, tequila, fruit punch. It had had everything in it. And for whatever reason I was thinking of it when I put together this chart that we monitor the speed of the transition, which looks at the renewable share of final energy consumption, because if you unravel it it’s got everything in there, right? Any efforts to decarbonize transportation, industrial production, biofuels, rooftop solar, deep geothermal. No matter what the widgets are, blue hydrogen, any of the colors of hydrogen, it’s in there. It’s a useful measure to track how the transition is going.
Jigar Shah: And sounds like a tasty drink.
Michael Cembalest: Well, it depends. It tasted good and then the next morning it wasn’t so great. So if you look at that chart, Europe’s the global leader here and it’s rising at about 0.6% a year and the US is running at about half that pace. When you look at the chart, it sends a different message than the other statistics that you cited, which are also in the piece, which is about the explosion of solar power and the fact that in a couple of years it’ll probably represent three quarters of new electricity capacity initiatives globally.
Part of the executive summary is, “Well, how can we reconcile these two things.” Where on the one hand you look to the left and you see this giant explosion of solar power and batteries and stuff like that. And then you look to the right at this other chart at the overall transition is moving much more slowly. Part of the purpose of this paper is to do the energy math to reconcile those two things because both true at the same time.
Jigar Shah: Well, the one question I had was when it comes to investment, one of the big challenges you have in a transition is, yes, it takes a long time to actually see a transition. But the other piece of it is what investment decisions are people making now? I think it’s very obvious the GE Vernova has been crystal clear that they’re not building a new natural gas turbine manufacturing plant. They’re just debottlenecking the one in South Carolina because they don’t believe that they’ll have persistent demand for new gas turbines way into the future to pay for that. And you see similar decisions being made by the auto companies around investments in new engine plants or things like that because we peaked in internal combustion engine car sales in 2019. As part of this, can both things be true? Which is the transition takes a long time and investment decisions that are quite profound are being made right now that will have big impacts around the future?
Michael Cembalest: I agree with that. We also have to make investment decisions, which can be treacherous. A lot of these businesses, some of them are highly leveraged. A lot of them survive, live and die based on combination of tax policy, purchase of voluntary carbon credits. The track record for investors in renewables, if you strip out the SPAC year of 2021, has been quite difficult. Then if you look underneath the hood, there’s five or six different renewable energy indices. Their operating margins are pretty low compared to a lot of the other sectors that we invest in. We have to be really careful when investing in this renewable transition because a lot of times the opportunities, that’s just not that profitable.
Jigar Shah: But just to push that one point more time, if you look at the net cashflow or free cashflow from all of the fracking companies from 2009 to 2020, it was atrocious, right?
Michael Cembalest: Yes.
Jigar Shah: You had like Bill Ackman talking about… This was a Ponzi scheme in 2014, 2015. It wasn’t until Wall Street took over all those companies in the COVID downturn, brought capital discipline to them and then the price of oil went up during the Ukraine conflict that people were like, “Oh, we could print money with this thing.”
Michael Cembalest: Yeah, the lack of capital discipline that you’re citing was probably the worst example of a decade of lost money. Worse than the casino build out, worse than the fiber build out, and worse than any kind of 10-year period in the capital intensive airline industry. It was incredible to see. But with a dose of capital discipline, the free cash flow statistics for the industry that survived are much better now. It remains to be seen whether or not those potential dynamics exist for a lot of the renewable companies. I sat with a company a couple of weeks ago that’s working on a certain kind of sustainable aviation fuels. A jet fuel costs about $2.5 a gallon.
I said, “How much is United willing to pay you for your sustainable aviation fuel?” It was another 25 cents, 2.75, that’s it. Their costs are $5 a gallon. This company somehow has to be profitable through a combination of 45, you add the letter of tax credits, and the sale of voluntary carbon credits to tech companies that want to use them against their scope 3 emissions. That’s a terrifying business model for somebody that’s in my seat. That’s what we have to work through. I need to see companies that are getting the majority of their revenues from their core businesses rather than these other sources.
Sticking points: tariffs, rising costs, and equipment shortages
Stephen Lacey: Let’s walk through some of the sticking points. There’s a few that I want to get to. Let’s see if we can get to them. There’s the slow pace of electrification, transmission, equipment shortages, and then rising costs across the board. Katherine, did any of these sticking points stick out to you?
Katherine Hamilton: Yeah, interestingly, I’m the executive director of the Thermal Battery Alliance. This is a group of companies that are doing thermal batteries that produce heat for industrialization and in all kinds of industrial processes. And the whole goal of these folks is to take electricity when it’s the cheapest. So when you’re overproducing wind or solar, store it, and then produce heat and they can keep it for as long as they need. These are long-duration storage technologies. Then provide heat cheaper ostensibly than natural gas to industrial customers. I’m looking at the sticking points and I realize electrification is not going as fast as it could, but there are things that are waiting in the wings that I think we’re at a very close point of scale.
It’ll take a little bit. It’ll take more investment. It’ll take more policy for them to scale. But I think things like thermal batteries, geothermal… There’s some things out there that are percolating along, and I’d love to hear what Michael thinks about not just where we are now, but what does it look like going forward for electrification?
Michael Cembalest: In principle, I like the industrial heat idea a lot. An enormous amount of industrial energy use is used for things below 200 degrees centigrade. People have this vision that it’s all at the 800 to 1,000 degree level, but it’s not. There have been different studies, but something between a half to, let’s say, three quarters of all industrial energy consumption is taking place at lower temperatures, which can be provided in a number of different ways. But the issue I have with the premise that you cited is… Look at ERCOT. They don’t have capacity payments for batteries. We looked at some investments in this space and I said, “Well, wait a minute, the arbitrage looks great. But what happens when everybody else starts doing the arbitrage? It’s going to disappear pretty quickly.” I’ve seen this in commodities in terms of the slope of the curves. I’ve seen this in commercial real estate.
Yes, there are places in the country right now which have negative power prices. For example, in the Midwest. I’ve always been mystified why data centers don’t locate there. But the more people start doing these things where they’re taking advantage of lower overnight power prices, some of those arbitrage opportunities are going to get arbitraged away. Yeah, I think it’s one of the best premises of all the things that are waiting in the wings to use this kind of industrial heat storage and then later deployment, as long as your battery round trip efficiency is above 85%.
Jigar Shah: The one thing I’d say though just to follow on that question is I hear what you’re saying, which is that the structural transition is going more slowly than people are necessarily hyping. But one of the things I’m curious about is once the cost of things actually comes down, then when an external shock to the system occurs, you can imagine things happening faster. Like when you think about Pakistan and just how terrible the coal plant was that the Chinese built and didn’t work and whatever, and then they had an electricity crisis, they got 22 gigawatts of solar installed without anyone really even noticing until GIS mapping found it. And now the cost of solar panels in Pakistan are cheaper than plywood.
Michael Cembalest: Yeah, the good news is all about the speed of the decarbonization of the existing grid. Almost everything I see that falls into the category of a faster transition has to do with the existing grid, and the stuff that’s moving at a snail’s pace is using these existing grid for more applications. It’s essentially home heating and industrial energy use, and to some extent the vehicles and shipping and aviation and things like that.
Jigar Shah: Right. But I guess what I’m saying a little bit differently is that I think that the people that are running our energy system right now are not as thoughtful as they used to be, and that can lead to external shocks to the system, right? That could lead to far longer times after a hurricane for people to reconnect, right? The president just denied FEMA assistance to Arkansas. They just denied FEMA assistance to North Carolina. If some of that happens and people are without power for an extra week, you could see a huge transition to solar plus battery storage because Generac has eight month waiting lists. Right? You’re in this weird spot right now where in some ways the fact that these transmission lines are getting demonized, like rain belt, they should just build the damn transmission line, right? Means that some of these shocks, not just in the United States but around the world, are going to lead to these bursts of deployment.
Michael Cembalest: Yes, and certainly the doubling of gas combined. The CCGT plant cost per kilowatt over the last couple of years will also accelerate some of that. My approach has generally been… When I started doing this 15 years ago, I was overwhelmed with all of these hockey stick forecasts from BNEF and McKinsey and Vinod Khosla on cellulosic ethanol, and I just decided at some point I’m going to stop looking at all these hockey stick projections and I’m going to focus more on the near term of what we can see happening, because in some ways it was more informative about where we were going next. I’m perfectly willing to assume that we’ll start seeing this stuff and we’ll be actively monitoring it.
But you know what? Rooftop solar, at the same time that those trends are taking place, you have more and more states moving away from fixed compensation for midday solar, moving to time of day pricing, right? That’s going to extend your payback period in states like California and New Mexico. That’s another trigger point where right now states in the Northeast don’t really have to worry about how much they’re paying, rooftop solar customers for midday solar, but they one day will, once it gets to be a large enough part of the grid. Look, California is already talking. Severin Borenstein writes about this. California is already debating whether or not to means test electricity bills based on income.
Stephen Lacey: Okay, let’s talk about two other bottlenecks, equipment shortages and transmission. I know, Katherine, that you wanted to talk about transmission quickly. So what jumped out at you about transmission?
Katherine Hamilton: Yeah, being here at this Bloomberg New Energy and Finance Summit, there’s been a lot of conversation about transmission, specifically this Champlain-Hudson Power Express, which is the line that Québec-Hydro is bringing down, otherwise known as CHPE. They first announced it in 2010, and it is finally going to be energized by 2026. It’s about 75% complete, and it’s supposed to, when it’s done, provide 20% of the electricity to New York City, which is amazing. That is a long time, from 2010 to 2026, to build a transmission line. Michael, what can we do? What do you have to say about that?
Stephen Lacey: Right. And you say that transmission is stuck in a rut here in the US.
Michael Cembalest: Yes. By the way, again, when I started doing this a few years ago, of all the data that was the hardest to find, good hard data on the growth of trend of gigawatt miles of grid expansion is the hardest thing to find. There’s not a lot of good sources for it, and we finally have cobbled together a few of them. And when you look at that chart, instead of rising, it’s going down, right? A few years ago, the miles added to the grid were much higher than they were. It’s been steadily declining at a time when people think it should be rising, so it’s going in the wrong direction. This gets into 100 years of US industrial history and the use of eminent domain to build the interstate highway system, fiber optic grid, the railways, civil and military aviation installations, benefited from substantial federal eminent domain and legislation at the congressional and state and local level to make it happen.
Also, a world where there were a lot fewer environmental impact statements, which is one of the ironies here is that a lot of environmental protections have been used to block these renewable transmission lines from being built. This is public information. We have the Anschutz family in Wyoming. They own some wind farms, and they’ve been trying to build this Northwest Passage power line to the California-Nevada border. They put a shovel in the ground last year. It’s year 18. 80% of the projects on public lands, and Obama, quote, unquote, “Fast-tracked it.” And that’s how long it took for them on the 20% of the non-public lands to sneak their way through and get the thing built.
Now, Chris Wright has said publicly that he plans to focus on this intently and that they will be able to make some headway here. But I have attended transmission line breakout sessions at energy conferences for the last 11 years. So I’ll believe it when I see it, because there really is a very complex maze of federal, state, and local things that have to be overcome to build out the grid.
Jigar Shah: But this is also where competence matters a lot, right? Part of the problem with the Transwestern Pipeline with philanthropists, which I’ve gotten to know Phil pretty well through that project, is that it’s going to come in at 4.5 cents kilowatt-hour. It’s not going to be 1.80 cents a kilowatt-hour. Right? And the question becomes, in a market-based system, how do you pay for that? It’s very obvious that if you rate based that line, it would reduce costs for everybody across the entire grid and pay for itself in three or four years, right? The same is, for instance, if you build a transmission line from Boston to North Carolina in federal waters, it would cost $20 billion.
It would link Neapol, New York ISO, and PJM together, and Neapol customers alone from that line would save $4 billion a year, right? Because their wholesale prices are roughly double what is it in PJM in New York ISO. But Lord Almighty, who actually does the cost allocation, right? How do you get Neapol to pay for all of it, right?
Michael Cembalest: Yeah.
Jigar Shah: We’re in this weird spot where it’s like…
Michael Cembalest: I know.
Jigar Shah: … “I don’t know how the markets work.”
Michael Cembalest: I’ve heard Ernie Moniz talk about this. He does a lot of work with J.P. Morgan as part of his futures initiative. It’s amazing to see a guy as brilliant as Ernie say… Yeah, he would sit in a room with three people and he just couldn’t get any of them to agree to do anything in their own collective interest. That’s why I think you’re going to need… You know how the Mountain Valley Pipeline thing got built? Not everybody’s in support of that, but that took explicit legislation that was written to prevent anybody from interfering with it.
Last year there was a big hydro recovery in terms of capacity factors. But two years ago it was a super warm year and the hydro capacity factors globally plummeted. What’s unclear is in a bad hydro year, whether or not that pipeline will be subordinated to Canadian hydro consumption, I don’t think they have a senior claim on hydro generation. My gut feel is that New York City would end up being subordinate to whatever local consumption of hydro was taking place in a weak hydro year. But we’ll see.
Katherine Hamilton: Yeah, ironically, when the Inflation Reduction Act was passed, the one piece that fell out was a tax credit for transmission, which was heartbreaking because Chairman Manchin said, “We don’t need this.” And I think it was because of some utilities had got to him, but I’ve been working with SOO Green for years, and that’s a merchant line, and it’s supposed to connect Iowa and Illinois all on rail rights-of-way, highway rights-of-way. It’s not about permitting and citing, that is not the issue. The issue is capital cost and trying to figure out how do we defray it, and certainly an investment tax credit would’ve made a huge amount of difference in a project like that. We have to be able to figure out how do we get something like that done? How do we unlock it? How do we… It’s going to be more expensive because it’s underground, and yet they’ll be able to deliver more electrons through the pipe than they would overhead a lot faster. I don’t know, I do think we need some more policy and some more thought about how do we get these big things done.
Michael Cembalest: Yeah, the government with respect to energy doesn’t seem to be very good at normal NPV analysis that business people do all the time, where you make an upfront investment, it costs you money, and then you reap benefits. Interestingly, there are parts of the government which do this real well. Right now the CMS is publishing all sorts of reports on how providing GLP drugs solely for obesity purposes might not pay off, because based on the projected costs of the drugs, they won’t yield enough long-term benefits in terms of their use in Medicaid and Medicare. There’s parts of the government which are pretty good at NPV analysis, but I don’t see it applied often enough in the energy space.
Stephen Lacey: Coming out of COVID and into the trade war era, we’ve seen supply chain disorientation, rising costs for everyone. Your report highlights transformer delivery times extending from many weeks, four to six weeks, to multiple years. There are rising costs for wind and solar PPAs. Meanwhile, the cost of building gas is climbing 3x according to NextEra’s CEO. Talk to me a little bit about how equipment shortages and rising costs are playing into the development of both renewables, storage, and gas.
Michael Cembalest: Yeah, it’s one of the things we try to do each year. I take a lot of potshots at Lazard, because I think levelized cost of energy is the cocktail napkin of energy math, right? Because…
Jigar Shah: Peter Orszag deserves it.
Michael Cembalest: I think for 15 years or so, they were publishing LCOE measures which weren’t really fairly measuring intermittent costs for wind and solar against gas because they weren’t including any kind of cost of grid stabilization, backup, thermal power, battery storage, et cetera. They’ve added in some kind of token, non-disclosed calculation, but me and a lot of other people like Paul Joskow at MIT are just dubious of the concept. But right now it’s particularly difficult even if you have models which adjust for all those things to nail down the relative cost of power.
As you mentioned, you’re getting skyrocketing costs for transformer equipment. In the PPI report in the US, the producer price index has about 47 items in it. The number one increasing item over the last three years has been anything related to the grid, transformer equipment and things like that. At the same time, you have these skyrocketing costs for the gas turbines. There’s an organization called LevelTen Energy which tracks the PPA prices for wind and solar. Those have been going up. There’s just been a generalized increase in everywhere you look, so it’s difficult right now to say, “I’m going to put a post in the ground and tell you exactly what the cheapest grid would look like based on levelized cost estimates in the moment.”
We’ll have to see. The markets work pretty well. The decisions that the big utilities and independent power producers make over the next few years will tell you a lot. I thought the decision to resuscitate some of the closed but not decommissioned nuclear plants… There’s only a couple of them where that’s even feasible. That told you a lot about the demand for power, because they wouldn’t be doing that at the cost they’re doing it if it was that easy to build behind-the-meter wind, solar, and battery.
Katherine Hamilton: Yeah, it’s super interesting because it also depends on your goal. What are you trying to get out at the other end? I work with a microgrid company that said, “A few years ago our clients were coming to us because they wanted to reduce greenhouse gas emissions. They had all these sustainability goals.” Then that evolved to more resilience, reliability, flexibility, especially in California with all these communities and public safety shutoffs, and now they said, “We are getting orders just because people need electrons. They just need electricity. That is the entire goal.” Yeah, sure, they love sustainability and they want resilience, but they need power more than anything and they need it fast.
Michael Cembalest: Yeah, I would say, and as a general answer to Stephen’s question, this import substitution approach that the administration is using of which tariffs are just one part, has a poor track record in places where it’s been deployed. If you look at… The tariffs on capital goods imports are just as high as they are on consumer goods imports. So if the administration, when they claim that they’re primary focused on excess consumption and they’re trying to not mess up intermediate business and supply chains, I don’t see any evidence of that because I see plenty of tariffs and other kinds of costs for the exact kind of capital goods that feed into the energy infrastructure.
Jigar Shah: Yeah, the one thing I would push back on though, Michael, is I think that the LevelTen data that you’re referring to, a lot of the reason why solar and wind have gone up in cost, and they really should go up in cost, is because they’ve added battery storage. A lot of the $30 per megawatt hour stuff had no battery storage, which I would suggest was suboptimal for grid interconnection because you were suboptimally using the interconnection point that you were given. And one of the arbitrages that I think is going to be fascinating over the next six years because of what you just said on the trade tariff piece is, I think, you’re going to find that Brookfield and NextEra are going to find that the most profitable thing for them to do is actually just repower their old solar sites that have 13% efficiency panels with 23% efficiency panels. Add four to five hours of battery storage and use the exact same interconnection point, but double the capacity factor from 25% to 50%.
Michael Cembalest: Yeah, I agree. Look, at the end of the day, all of those solar statistics that you’re citing and that we cite in the executive summary of the piece, economics are driving that way more than the existence of the PTC or ITC in isolation. Now, we haven’t talked about it yet. The administration, I think, is going to look to the energy bill as a big pay for… The only big numbers that I see them coming up with are gutting the energy bill and about $600 billion of Medicaid cuts to pay for extension of the tax cuts and no taxes on tips, no taxes on overtime, 300 billion of Homeland Security spending. We’re looking at various iterations of the reconciliation bill, but most of them involve clawbacks to the energy bill. We get the sense at this point that most of the clawbacks are going to be focused on EV credits and infrastructure. But with this group of people, who knows.
Katherine Hamilton: Yeah, and interestingly, Michael, we find with Congress with the Inflation Reduction Act that there’s this like comorbidity factor with the tariffs, because you’re going to do it both, you’re going to crush it in two different ways. So we’re finding Republicans starting to push back a little bit more on the Inflation Reduction Act. But I noted that we are almost 7% through Trump’s presidential term, which is lower than any tariff he has put on anything so far. It would just be interesting to hear from you. I know here at the Bloomberg Summit, everybody’s like, “Oh, it’s going to settle down. Don’t worry. Everything will be okay.” Because I think people don’t want to contemplate this constant up and down and not knowing what the tariff structure is going to be. I’d love to hear your take on just where this is headed.
Michael Cembalest: I wish I knew. I did a webcast a couple of weeks ago where I talked about this and I started…
Stephen Lacey: This was the redacted report that you’re talking [inaudible]?
Michael Cembalest: Yes, it was the one where I had the penguins on. I brought some penguins as the trade representatives from the McDonald and Heard Islands.
Katherine Hamilton: Brilliant.
Michael Cembalest: I started the webcast by saying… I can imagine two states of the world, and let’s use Vietnam as an example. Vietnam’s a country that has modestly higher tariffs on the US than the US has in exchange and has some local non-tariff barriers, particularly as it relates to things like agriculture. I can imagine the administration saying, “All of this that what we’re doing is really just designed to open the playing field for American companies to do business in Vietnam. And if we can get them to dismantle a lot of their tariff and non-tariff barriers, we’ll do the same and life goes on, mission accomplished. We’ve done what we needed to do.” In that case, the trade deficit with Vietnam will still be there because of relative cost of production, relative cost of labor, and all sorts of macroeconomic things.
Scenario B is that’s not what they’re looking to do. They don’t really just want a level playing field, they want to eliminate the trade deficit with Vietnam, which is a much harder hurdle to accomplish because you would need gargantuan tariff levels to offset all of those underlying economic drivers which are creating the trade deficit in the first place. Depending upon the day, the week, the hour and which particular cabinet member is talking, they’re pointing either to A or to B. And my fear is whenever the markets go down, they tell you it’s A. And when the markets go back up, they tell you it’s B. Right now, I think the markets are still underpricing the risk that they are really trying to achieve a long-term economic rebalancing, which they believe would repatriate manufacturing and production back to the United States. I think that’s going to be just a very, very, very difficult and costly road.
Data centers and AI growth
Stephen Lacey: Let’s turn to data centers now. Katherine, you shared that anecdote about the microgrid company saying that people are coming to them because there’s this mad scramble for power, and so much of that is coming from data centers, many of which have a big AI component to them. What do you make of these crazy projections, Michael, of coming data center capacity? And do you think that the economic value of AI is going to actually justify the expansion we’re seeing? I want to get your view on data center expansion broadly, and then we can talk about the energy component.
Michael Cembalest: The hyperscalers are spending hundreds of billions of dollars a year on this AI infrastructure. We will eventually need to see a massive uptake of inference models by the corporate sector to pay for all of that. And I’ve written over the last year a few pieces, drawing on some analysis that was done by Sequoia Capital, suggesting that we are going to need to see a few hundred billion a year of corporate adoption of AI to be willing to pay for all this. And we’re right now in the zone where you’re still getting the hyperscalers spending. I think we’ve got another 18 to 24 months of that continued spend without the proof statement of adoption on the corporate side, because there’s so much momentum behind it. But within 18 to 24 months, I think people are going to focus less on what the hyperscalers are doing and what companies like J.P. Morgan are doing.
Now, we have 300 AI projects or so going on within the bank, fraud detection, issues around customers, and settlement, and custody and things like that. But the jury is out about the magnitude of the productivity benefits they will generate and how much we’d be willing to pay for them. There’s lots of interesting anecdotes at this point. Klarna said they’ve saved 40% of their cost by reducing their data center employee requirements because of the chatbots and CoPilot has made coding much more efficient. It’s a cloud of anecdotes right now as it relates to the real productivity benefits associated with AI adoption. And we’re trying as hard as we can to put our finger on the pulse to see what’s actually going to happen. But the data’s very squishy at this point, for lack of a better word. Now, that brings us to the… I’m assuming your next question is about all of these energy consumption forecasts?
Stephen Lacey: Yes. What do you make of them?
Michael Cembalest: Well, first, if you go back to 2007, the last time there was this skyrocketing perception of electricity demand. You roll the clock forward and US electricity consumption is unchanged versus what it was in 2007, for reasons of efficiencies and things like that. Now, this time around you’re looking at not just the data centers but electrification of some degree of home heating through heat pumps and also electrification vehicles. So it does seem like there’s going to be some grid growth. But there’s a huge gap between some of the really ambitious forecasts and some of the lower forecasts, which have different implications for what kind of grid build that you’d need. From the ’50s to the ’90s, we had rapid growth in electricity consumption in the United States in generation, but that was an era when it was gas, nuclear, and large gigawatt level plants. And this time around we’d be trying to do it with smaller renewable and batteries and things like that.
I’m going to take the under, which my sense is that the cost of doing this is going to be so prohibitively high that it reverse forces certain kind of engineering efficiencies into AI energy consumption that will lower their practical demands on the grid. Because, for example, I don’t think that Talen Energy-Amazon deal will be done in the future. A lot of my colleagues at the bank don’t like it when I say that because they think it’s perfectly within Talen’s rights to do what they did. I just think the public optics are very difficult and it’s going to be difficult to do stuff like that. As I mentioned, you can count on one hand the number of nuclear plants that can be recommissioned that haven’t really been actively dismantled yet.
So that’s not going anywhere. TerraPower, 2035. Small modular reactors, I don’t know if that’s ever really going to happen. I’m skeptical about SMRs for reasons we can discuss. When you look at all of those pieces, I think it’s going to put pressure on the AI industry to figure out much more efficient ways. Look, DeepSeek, there’s still a lot we don’t know. But DeepSeek showed you that you could generate at a lower cost and lower energy consumption AI results that were not that different from some of the Gemini models they’re competing against.
Katherine Hamilton: Yeah, it was super interesting yesterday, the president of Schneider Electric on a panel said that they’re really focused on making data centers more efficient, citing them where they’re distributed energy resources available or easy to install where there’s good water management. And this is like getting to what Jigar talks about all the time, they are building DERs and that helps get interconnection lowered to three years rather than forever and doing virtual power plants. They’re thinking about that, “How do we get these things on where it’s cheaper and you’re not waiting a decade for a gas turbine to show up?”
Michael Cembalest: Yeah, it’s amazing when you look at the map of negative power prices, not just in the United States, but other countries. There are opportunities there in terms of taking advantage of that for the data center industry.
Jigar Shah: Yeah, there’s a couple of pieces here. One is that I think it’s important to separate data center growth from AI growth. I think data center growth is going to continue unadulterated. There’s a lot of people who continue to save pictures on their account and keep paying Google an extra 3.99 a month, and those folks need data centers. That’s going to double data center consumption from about 170 terawatt hours, I think, to 340 terawatt hours by 2030. Then I think I totally agree with you, AI is going to be a lot less than people are saying it’s going to be in terms of power consumption. I’ve been sticking to my number of 25,000 megawatts of new load growth by 2030. I think that number is coming closer into view as people downrate stuff. But the one thing I would say is a megatrend that I think is going to happen is it is very clear to me that Waymo is going to succeed, and especially under this administration.
And I think when you think about what that means, Waymo is collecting ginormous amounts of data to be able to drive every one of their cars and to figure that out they already have much higher market share than Lyft in San Francisco, and they’re launching in DC soon. I think as these asset utilization breakthroughs occur, I think there’s this nexus between AI, those asset breakthroughs, right? And the finance behind them. That then provides that killer use case because, as you know, automobiles are notoriously terrible for asset utilization.
Michael Cembalest: They are. For that thesis to play out, the cost of doing so has to be able to be retrofitted in the usage of those assets.
Jigar Shah: Totally.
Michael Cembalest: That’s the piece that we’re still waiting to see. Some of the early Waymo models we’re actually quite expensive to operate based on how their prototypes were built. By the way, here’s some fun fact for you on autonomous vehicles. Of the many DOGE cuts that have taken place to the government, shockingly, and you can draw your own conclusions in the National Transportation Safety Administration, the cuts fell disproportionately on the department that oversees the safety and operation of autonomous vehicles.
Jigar Shah: Oh yeah. And the offices…
Michael Cembalest: You can draw your own conclusions.
Jigar Shah: And the offices that denied Andreessen Horowitz’s company’s cryptocurrency banking licenses.
The state of play in Europe and China
Stephen Lacey: I want to take a geographic approach now and talk about Europe and China. I know these are really heady subjects, but I want to see if I can lump them together here. Europe, of course, now gets 50% of its electricity consumption showing, as you say, what’s technically possible with grid decarbonization. But energy prices in Europe are two to four times US levels. Manufacturing output is declining substantially in energy-intensive sectors. There are some real challenges in Europe right now, and I want to unpack what those challenges are and what it means for Europe’s economic outlook. Then I want to get your read on what the story of China is. Obviously, a country that dominates renewable energy and battery storage supply chains. It is decarbonizing final energy consumption actually faster than the US, but slower than Europe. And I want you to characterize the state of China’s energy transition as well. Let’s take those together.
Michael Cembalest: Okay. Well, on Europe, when you come home and there’s a broken window, the dog gets blamed. Now maybe he was involved or maybe he wasn’t. The renewable transition tends to get blamed for a lot of ills in Europe, some of which are accurate and defensible and some of which aren’t. Europe is undergoing deindustrialization, particularly in Germany and in the UK, that have a lot of factors involved. A lot of which is the competitiveness of Chinese vehicle exports. Some of which has to do with the cost of imported natural gas. Some of which have to do with the cost of duplicated electricity generation capacity in the [inaudible] transition in Europe, part of the decision by Germany to shutter all of its nuclear plants, increasing its exposure on imported Russian pipeline gas, US LNG. So there’s a lot of stuff going on. And just like the power outage in Spain this week, renewables will immediately get the blame. Whether or not they deserve, it’s a different question. If you want to talk about the grid inertia issues, we can do that.
Europe is paying a price. Broadly speaking that its governments and people have been supportive of, right? The Carbon Border Adjustment Mechanism is unique in the world in that regard. Europe is, as a society… Look, compare their gasoline taxes versus the US, right? As a society, they have more committed to paying an economic price for the energy transition than the US has ever been and probably will ever be. And there’s a cost to that. I thought we would start seeing more fiscal federalism in Europe than we have been, right? It’s a monetary union but not a fiscal one. That could have offset some of the costs here, but we’ll see.
And now you have Germany actually pushing to finally do some increased fiscal spending, which could offset some of the blow. But I think the example of Europe, to most people, will be that you can achieve a greater penetration of renewables on the grid and accelerate the transition, but there’ll be an economic cost for doing so.
Jigar Shah: Yeah, it really just comes down to figuring out what you’re transitioning to. I think they stayed on feed-in tariffs way too long. It cost the German people 200 billion euros more than it should have as a result. But now you look at the UK for instance, where they have this contracts for different structure for offshore wind, and yet they’re paying record prices for wholesale market prices because of the LNG gas imports that they have right now. I don’t know why they keep paying these massive windfalls to the offshore wind folks when wholesale market prices are high. Some of these market mechanisms just need to be adjusted because they’re not working.
Michael Cembalest: Yes, and it’s weird because you could understand why those market mechanisms were needed in the inception of some of these things.
Jigar Shah: Yeah.
Michael Cembalest: But it was like they never contemplated how they would exit them. And then you have these vested interests that are going to fight you tooth and nail before they give them up.
Katherine Hamilton: There are a lot of folks talking about how renewables somehow took down the grid in Spain and Portugal. We don’t know yet exactly what happened, but they do have a real structural issue, which is they have 30 gigawatts of solar and they only have a gigawatt of batteries. By comparison, the UK has 13 gigawatts of solar and 6 gigawatts of batteries. Laurent Segalen sent me some of this information, which I was really grateful for. They also only have one interconnection. It’s a three gigawatt interconnection with France. Part of what Europe desperately needs to do. And this happened with France three years ago where they were about to go down, but their interconnections were so strong, they had like 20 gigawatts of interconnections that they were able to stay up.
I think that is part and parcel of what needs to happen there. It’s like what we’ve done with our organized markets and interconnections in the US. They have to be able to use solar and batteries. They have to install batteries. They can do it as long as they have the right interconnections.
Michael Cembalest: Regardless of what ends up being determined here, and it could easily have been nothing to do with solar and some kind of equipment failure on protective equipment and circuit breakers and stuff like that. There’ll be some fairly modest cost, infrastructure improvements, grid forming inverters, and stuff like that. For everybody who thinks that the solution here is to go back to a gas-dominated grid, that’s not what Spain’s going to do, and that would certainly not be the cheapest solution for them. Making some modest grid investments that deal with issues around frequency regulation and low voltage conditions is where they’re going to go here. I think this will be a useful episode for them to deal with in other countries, but it doesn’t really change the viability, I think, of electricity grids with high contributions for renewable energy. I don’t think it does that.
Stephen Lacey: So what do you think the story of China is?
Michael Cembalest: Well, when you look at China, you have to make a decision whether or not your chart’s going to be as a percentage of something or in megajoules, because as a percentage of energy consumption, China really looks like it’s making a lot of progress. The problem is that in absolute terms, they’re just consuming more of everything, right? More renewables, more coal, more gas, more oil. Now, I think the pace of that is going to start to slow along with the Chinese growth. Chinese CO2 emissions haven’t come down, right? In absolute terms, they’re still rising and China has effectively been investing in a lot of renewables which are meeting incremental demand growth on the grid. I think that’s the best way to look at it.
Jigar Shah: Which I think your report basically says is true globally, that renewables basically helps with growth. It has a hard time with baseload. I think the real interesting thing with China is they have 30 nuclear plants under construction, and when those 30 nuclear come into service, the question is does their coal consumption go down? A lot of people are saying that it will, including the Chinese government is saying that it will, that that’s what they’re waiting for. And if that happens, then the question is does that then send a clear signal to India and other people that scaling up nuclear is the way to go to offset coal?
Michael Cembalest: Yeah, whenever I go to energy conferences, there’s always so many panels on nuclear. And there’s tons of excitement now about modular reactors, thorium reactors, fast breeders, and low-high gas temperature reactors.
Jigar Shah: Are we talking about family planning or nuclear design?
Michael Cembalest: Nuclear. I have this very simple chart that looks at both the cost and time to build of nuclear plants in the emerging world and in the developed world. And, of course, the costs are multiples higher and the building timeframes are double. Diagnosing those differences is probably one of the most important tasks ahead of the developed world if they want to pursue this nuclear renaissance, which you keep hearing about, but which in practice is not really happening. A lot of people conjecture on it like, “How much of it is just lower cost of labor? How much of it is regulatory? How much of it is lower cost of concrete? How much of it is other oversight and procedural issues?” I don’t really know. I will tell you something. Yeah.
Jigar Shah: All of which is interesting, but my point is just simply that even if you don’t believe that the West can build nuclear, if China successfully builds 30 nuclear plants, which I think we all believe that they will do, the question becomes whether that dramatically changes the trajectory of their carbon emissions.
Michael Cembalest: I think it probably would. I think at that point they would continue to use renewables to meet incremental demand growth and that you would get them canceling out a lot of coal that’s certainly used for residential heating in China with some of those legacy systems that have been around for 100 years. I think it would. The thing about China to remember is something like 40 to 50% of its critical mineral industry is unregulated, right? Think about a regulated critical mining and refining operation in China, and now imagine an unregulated one. And all the pictures you’ve seen of fluorescently colored Chinese rivers. Part of the speed of the Chinese energy transition can be directly attributed to regulatory issues and things like that, which is the polar opposite of what you have in the United States. Where it would probably take 12 to 15 years to get the permit to develop a lithium mine in terms of both production and refining.
Stephen Lacey: I’m going to tie this all together and get your take on how you read this phase of the energy transition. I’ve seen some writing about your report and people label it energy realism. Let me give you an example. Earlier this year, Jeffrey Currie wrote this report called The New Joule Order from Carlisle. He argued that we’ve reached peak fossil fuel trade and that local security… Will Trump decarbonization as a primary driver of the energy transition because we’re in this era of trade wars, supply chain disruptions.
Rather than thinking about it as the net-zero era or the era of decarbonization, the energy security will be their primary driver. And those localized development of supply chains and projects will ultimately help benefit renewables and batteries along with a suite of other technologies. How do you frame out the way you describe the macro framing for this report and what era we are in energy transition?
Michael Cembalest: Well, there’s certainly been a shock to the system when compared to the energy bill just from three years ago. The number one complaint… We have more CEO clients in our private bank than anybody, and the number one complaint from those CEOs is the violent pendulum swings in US capital policy, tax policy as it relates to things like this. It’s very difficult to plan. Yeah, the pendulum is swinging right now. And it does make sense to think about energy independence, certainly when you saw what Europe went through. But I think people need to be also realistic about the duration of US natural gas consumption, which is on a consumption to probable reserves basis.
The US is not in as good shape as Qatar, Turkmenistan, and Russia and places like that. This natural gas boom is not going to last forever. So at some point, energy security is not just going to be a reflection of your potential to generate fossil fuels, but is going to have to deal with other kinds of electricity and power supply as well. Which is why the energy secretary is putting energy behind things like geothermal, because I think he understands that there’s a lifespan associated with natural gas that is not infinite.
Stephen Lacey: Michael Cembalest is Chairman of Market and Investment Strategy for J.P. Morgan Asset and Wealth Management. Michael, this was a fantastic conversation. I really love your writing, your analysis, and this report in particular. Thank you for coming on the show.
Michael Cembalest: You’re welcome. It was a lot of fun.
Stephen Lacey: Make sure to check out Michael’s Eye on the Market newsletter. We’re going to provide a link to his recent energy report in the show notes and to Eye on the Market as well. Katherine, good to see you. You’re going back to the BNEF conference now?
Katherine Hamilton: I am, yes. And thank you so much, Michael. I am just going to commend your reports to everybody, because not only are you smart and use data, but you also have a good sense of humor.
Stephen Lacey: Yes. Your wordplay is fantastic. Jigar, great to see you. You’re going back to the conference as well?
Jigar Shah: I think so, but mostly I’m just going to turn on the lights so that I’m not so dark in this hotel room.
Stephen Lacey: Because you can only hear Jigar. He’s been looking like he’s in the witness protection program with a dark background and a light blinding his face. Anyway, thanks all. This was great.
Open Circuit is produced by Latitude Media. Jigar Shah and Katherine Hamilton are my co-hosts. The show is edited by me, Stephen Lacey. Sean Marquand is our technical director and mixes the show. He also wrote our theme song. Anne Bailey is our senior podcast editor. Latitude Media is supported by Prelude Ventures. Prelude backs visionaries, accelerating climate innovation that will reshape the global economy. Learn more at preludeventures.com for more in-depth reporting on the topics we cover on this show. Of course, go over to latitudemedia.com. You can subscribe to our newsletters and give this show a rating and review wherever you get podcasts. We ask it every time, but it is hugely helpful and we thank you very much. We’ll see you next week.
