Welcome to the free edition of The Diff. This newsletter goes out to 33,641 readers, up 212 since last week. In this issue:

• NextEra Energy's Green Growth

• Specialized Cloud

• Bailouts

• Another Look at Inflation

• Trucking Recession

• Belt and Road Alternative

• Diff Jobs

Also: the Diff team is expanding! Scroll down to the Diff Jobs section at the end for more details.

NextEra Energy's Green Growth

It's been about half a century since electricity generation in the US could be characterized as a high-growth business, and about a century since utilities counted as serious growth stocks. But any boom that creates fixed assets will eventually be followed by a boomlet when those assets reach the end of their useful lives, and that's been happening: coal power plants will make up 85% of the electric capacity being retired in the US this year, and have been a high share of retirements for a while. And a decent share of the capacity that replaces them is coming from renewables.

Which is one reason that the fairly stodgy, regulated business of electric utilities—an industry where a substantial share of revenues are tied to not just regulated prices but regulated returns on equity—has produced some impressive returns over the last decade. NextEra Energy, the largest single producer of renewable energy in the world, has generated annualized returns of 19.9% over the last decade, compared to 13.4% for the S&P 500, for a total return of 514% compared to 253% for the broader market. Not bad for an industry where demand, in terms of BTUs consumed across all energy sources, is up 0.5% total over the same period. (The utility ETF XLU underperformed the S&P, with a 10.4% return over this period. And its single largest constituent, at 15.8% of assets, is none other than NextEra.)

NextEra earns 58% of its adjusted earnings from its regulated business, and 44% from its wholesale unit, Energy Resources (these add up to over 100% because of corporate costs not attributed to either unit). These are very different companies. The regulated business consists of Florida Power & Light, and the recently acquired (and now fully merged) Gulf Power Company. These businesses serve 5.7 million customers in Florida, with a power mix that skews low-but-no-zero carbon: 70% natural gas, 22% nuclear, and 6% solar. This business has been getting cleaner over time; they shut down their last coal power plant on January 1st, 2021, and if you're particularly aggressive about ESG you're welcome to watch this video of them blowing it up.

That unit has grown in part because they've been able to drive costs down over time. Their investor day presentation has a nice slide showing that thirty years ago, their operations and maintenance costs per megawatt-hour were 4% above the industry average, and they're now 66% lower:

They've also benefited from the tailwind of favorable demographics: Florida's population has grown by 15.9% since 2010, the fourth fastest state in percentage terms and the second fastest in absolute terms. But the residential utility business is in one sense forcefully commoditized by regulations. The quality of service is predefined, the areas in which a company can operate are predetermined, and they're allowed to set rates within a range defined by their returns on equity. Cheap power, renewable power, or ideally cheap renewable power might be a tiny concern on the margin for people choosing which state to move to, but in general there's deliberately little flexibility for regulated utilities to achieve extraordinary returns. (There is some. At the company's investor day, the CEO took a moment to expand the company's addressable market by noting that as a Florida resident he doesn't pay state income taxes and that that's an option for the audience, too.) And these businesses are valued accordingly, so expansion through acquisition usually means paying the market price for this monopoly status.

A more interesting piece of NextEra's business is the wholesale power segment, where they're generally putting together projects for specific long-term commercial customers. They might sign a twenty-year deal to supply power at a particular price to a corporate customer, and then build and operate the project. This is NextEra's Energy Resources unit, which has very different economics and a very different mix: it gets 66% of generation from wind, 23% from nuclear, and 8% from solar.

In one sense, this business is also commoditized: "Our electrons are no better than the next guy's electron," as the company's CEO recently put it. But a Tegus expert call talks about how they aren't quite as commoditized as they look: NextEra actually bids a bit higher than competitors, and cites its long-term record for launching projects on time, operating them well, and not selling them to third parties who might mismanage them.

For a conventional power source, that might not be an especially big deal, but for renewables it's very important, because renewable power is intermittent. Energy Resources is 52% of the overall company's capital expenditures, and within that $8.4bn of spending on the wholesale business, $3.8bn is wind, $2.0bn is solar, and $304m is for batteries. One of the costs of non-nuclear clean energy is its unpredictability, and this can take two forms:

1. Redundancy and power storage, or

2. Not getting the power you were promised because it was overcast or the wind didn't blow that day.

There is demand from corporate buyers for clean energy; many of them have long-term commitments to reach net zero, and the lifecycle of NextEra's contracts can be long enough to overlap with those goals. But companies also demand reliable power.

NextEra can make a plausible claim to provide this. They acquired an analytics company a decade and a half ago, both to get better at predicting short-term changes in supply and demand and to identify the best sites for renewables. Finding a good site is a combination of natural resources (wind and sunlight), grid interconnections, and cost; it's the kind of thing that's hard to simultaneously optimize for without direct experience, which gives the business some scale advantages. That's not the only scale advantage they have: they noted that they're #5 on the list of US companies with the highest annual capital expenditures, so they're generally their suppliers' most important customers. That's good for costs, and for reliability, too; if GE has trouble producing all the wind turbines they've promised customers this quarter, NextEra is probably not the one who gets a late shipment.

They also have the dubious advantage of operating in an area with plenty of bad weather, including chronic problems like frequent lightning strikes (compare them to the rest of the US on this map) or salt in the air that corrodes equipment. They also deal with hurricanes. So few other US residential power providers have as many irritating operational problems. Economically, some of that is a wash: their return on investment is an input into rates, so higher operating costs from their location ultimately get added to the customer's bill. But they had experience with intermittent power before they were ever a renewables company, so they had a head start.

NextEra has one other advantage: they've done financial engineering to get more efficient access to capital. One tradeoff with growth companies, which makes it hard to design their incentive programs, is that in the short term the market cares about growth and in the long run it cares about the return on investment generated by all that growth. Since long-term returns aren't knowable in advance, the incentive is often to grow a bit faster than the numbers really justify. Some investors own the stock for cash flows from its existing assets, others are long because they're excited about future ones, and it's hard to balance those interests. But one tool is the OpCo/YieldCo split: the company hives off some of its long-lasting assets into an entity with lower growth and more predictable dividends, and lets investors own that one if they want to take less risk. This setup was a trend ($, WSJ) a few years ago, and the NextEra version is still going strong (15.5% annualized returns in the eight years since its IPO).

The company has absolutely enormous ambitions, because transitioning the grid completely to renewables is about $2 trillion. And it's a long process—they note that none of their existing gas assets are expected to be shut down for economic reasons before they're physically worn out. Another way to frame that cost is that it's 125 years of NextEra's capital expenditures. They note that renewables are an inflation hedge, because they don't require much incremental raw material to build—but they're also exposed to inflation because the initial investment does require lots of raw materials. They've set a goal to have near zero emissions without the use of offsets by 2045 ($, WSJ). So what the company ends up illustrating is the massive scale of investment required for a low-carbon economy. Even the US's biggest utility is only making a small dent in overall emissions in any given year, and the US uses only 17% of the world's energy. But the company's performance also shows that effective financial engineering and real engineering can make those investments viable, and still produce a great return.

Disclosure: I have a financial relationship with Tegus. You can sign up for a free trial here. No position in NEE.

Share

A Word From Our Sponsors

Here's a dirty secret: part of equity research consists of being one of the world's best-paid data-entry professionals. It's a pain—and a rite of passage—to build a financial model by painstakingly transcribing information from 10-Qs, 10-Ks, presentations, and transcripts. Or, at least, it was: Daloopa uses machine learning and human validation to automatically parse financial statements and other disclosures, creating a continuously-updated, detailed, and accurate model.

If you've ever fired up Excel at 8pm and realized you'll be doing ctrl-c alt-tab alt-e-es-v until well past midnight, you owe it to yourself to check this out.

Elsewhere

Specialized Cloud

One seeming paradox of cloud computing is that the economics get better at massive scale, but the way to achieve massive scale is to find ways to sell to specific niche use cases. Smaller cloud computing platforms have been particularly focused on this model, since their biggest comparative disadvantage is in selling generic services at commodity prices against larger competitors. So one element of Oracle's strategy in its Cerner acquisition is to offer healthcare-specific cloud services. One part of Google's strategy is to offer supply chain services, and they've already done a lot with retail. So it's partly an effort to outflank Amazon in two directions: by offering services that Amazon's competitors might be reluctant to buy from Amazon itself, and by doing things that weaken Amazon's core business.

Disclosure: long AMZN.

Bailouts

Goldman Sachs is working on a deal to acquire assets from crypto lender Celsius, which has halted withdrawals and is considering bankruptcy. This is not a vote of confidence in the long-term viability of the crypto lending business—they're looking into buying loans, not the lending business itself—but it does imply that they either view the crypto selloff as temporary or suspect that crypto lenders have been hurt more by liquidity problems than by making bad loans.

To be clear, a company that runs into liquidity problems but has fundamentally valuable assets has still made a colossal mistake, and they end up paying two penalties: first, they lose all their equity and have to liquidate, and second, whoever buys during the liquidation makes money. It's clear that the crypto lending market got overheated, and focused more on providing liquidity to speculators than on any other useful task. While that's an indictment of the business model, it's not an indictment of crypto itself.

This Diff post from a year ago looks at how crypto lending analogizes to emerging market bubbles: when asset prices are rising, leverage to bet on further asset price increases crowds out investment in improving their earning power. One difference I'd note now that the bubble has deflated a bit: emerging market recoveries can be slow when those markets raise outside money on the condition that they adopt economic austerity and engage in structural reforms that may not be appropriate.¹ Since this concept can't be easily imported to crypto, the collapses are sharper and more painful, but the recovery can start right away.

Another Look at Inflation

This BIS piece is a good statistical exploration of what high- and low-inflation regimes look like. One difference is that the higher inflation gets, the more interconnected it is. When inflation is low, high gas prices or wage increases don't tend to percolate through the rest of the economy; they just mean that some money gets redistributed between energy consumers and energy producers, or between wage earners and companies. But the higher inflation gets, the more everything moves together; gas inflation leads workers to ask for cost-of-living increases, those cost-of-living increases lead to higher prices for consumer goods, that leads to more spending on real assets like real estate over financial ones, and the process feeds on itself.

While inflation levels are high right now, policymakers spent a decade-plus being over-worried about inflation compared to other concerns, and just a year or two more recently being too blithe about it. But it's good to remember why inflation was such a salient concern in 2008 or in the post-crisis period. Once it's something people worry about and talk about, it's a self-sustaining phenomenon.

Trucking Recession

Between high diesel prices and the end of the great retail overstocking, trucking companies, especially smaller ones, have been shutting down at a record-breaking pace. This is partly a response to the strong trucking cycle earlier in pandemic:

New trucking fleets poured into the market to profit from these sky-high rates. From July 2020 to now, almost 195,000 new carriers have entered the market, according to Vise of FTR. About 70% of these new carriers were just one truck. The previous record 23-month period saw just 86,000 new carriers.

Some of this will lead to consolidation: bigger trucking companies have more long-term contracts, so they'll survive (and some independent operators will join—or rejoin—them). But it's a classic case where the supply grows on a lag and demand can shift faster than that. And this has a macro impact as well, since trucking is such a common job.

Belt and Road Alternative

The G-7 has released a global infrastructure investment plan with the US planning to "mobilize" $600bn through 2027 ($, WSJ), through some combination of government lending and private sector activity. Like the Belt and Road initiative, it will be difficult to put accurate numbers on this since projects that were already going to happen will want to tie themselves to it.

In one sense, this would have been a cheaper project and a better idea when 10-year treasurys were yielding under 1%. On the other hand, real yields are deeply negative for the moment, and in the long run you can view inflation as an issue of the supply and demand for real goods and services—more infrastructure increases demand in the short term but increases supply over time, so it's easier to manage inflation if there's more production happening in more parts of the world.

Diff Jobs

Diff Jobs is our service matching readers to job opportunities in the Diff network. We work with a range of companies, mostly venture-funded, with an emphasis on software and fintech but with breadth beyond that.

If you're interested in pursuing a role, please reach out—if there's a potential match, we start with an introductory call to see if we have a good fit, and then a more in-depth discussion of what you've worked on. (Depending on the role, this can focus on work or side projects.) Diff Jobs is free for job applicants. Some of our current open roles:

• The Diff is looking for an associate to support Diff Jobs, newsletter growth, research, and other areas of the business. (Remote, Austin a plus)

• An edtech company is looking for multiple roles (remote, US ideal), including data-literate product leaders and software engineers.

• A new fintech startup is looking for people with deep insurance industry experience and an interest in crypto. (US, remote)

• An API-first bank operating in developing markets is looking for payments experts and people experienced with Finacle Core. (Remote)

• An alternative data business is looking for multiple roles, including Python developers with web scraping experience and financial analysts who are fluent in data. (US remote, NYC a plus).

Find a Role With Diff Jobs

Hire With Diff Jobs