What Stays the Same in Climate?

We can't predict the future, but what can we count on to remain the same?

Nov 3, 2023

In 1903, Henry Ford hired a Detroit lawyer named Horace Rackham to draw up paperwork that eventually incorporated the Ford Motor Company.

Rackham was struck by Ford and the idea of the motor vehicle. He borrowed $5,000 from a few friends and a local bank to buy 50 shares of the company. The bank president that loaned him part of that $5,000 told Rackham, "The horse is here to stay, but the automobile is a novelty - a fad."

Rackham's investment was worth $15,000,000 - $275,000,000 in today's dollars - when he cashed out in the late 1920s.

The story, while anecdotal, illustrates an important truth about forecasting and investing. Significant events, the ones likely to make the most return, are those you usually can't predict.

It's probably not surprising to you that I believe climate tech and the energy transition will undergo a lot of change in the next few decades. We can't predict them, though we'll undoubtedly give it a go.

So, why do we still forecast, and how can we at least attempt to do it differently or better

We stink at forecasting, and we’ll never get better.

John Galbraith once said:

"There are two kinds of forecasters: those who don't know, and those who don't know they don't know."

Forecasting, especially the type we can profit from, must, by definition, differ from the future. If it didn't, it would be consensus. The more our prediction differs from the present:

the more likely it is to diverge from the current consensus forecast
the greater the profit would be if it's right
the harder it will be to believe and thus act on it

Not only do good forecasts need to diverge by a certain magnitude in the right direction, but the timing needs to be correct, too. Early is wrong.

People Aren’t Particles

Investing is more art than science.

In science, there are theories, and there are laws. Take physics, for example. Richard Feynman once said, "Electrons always do what they're supposed to do. They never rebel. They never go on strike. They never innovate. They never behave against their interests."

None of these things are true about people, and people comprise the economy.

Thus, models and forecasts can be directionally accurate, especially when assumptions aren't violated. But, they can't always be accurate - especially at inflection points where they'd be most valuable.

We should flip forecasting on its head.

For years, forecasting has been the study of predicting what will change. The internet and social media ensure that won’t change.

No one gets clicks or likes by predicting things will stay the same.

If you wrote this Thread, "My 2024 climate tech prediction is we'll continue on a path similar to 2023," you'll get precisely 0 likes. Not too different than my social media now.

But, how would this one perform: "2024 is the year where Tesla sells more cars than Ford and GM combined, the US approves its first nuclear power plant 30 years, and OPEC sets emissions goals."

You'd get more likes than an Elon meme on X being juiced by Russian robots. Yet, you'd still be wrong.

But what if the first Thread is how we should actually approach forecasting? Instead of thinking about what will change, what are the truths we can always count on?

Jeff Bezos often said Amazon made decisions knowing the customer always wanted things faster, cheaper, and with better selection. In fact, he loved the framework so much that he used it when he testified in front of Congress:

The customer is always beautifully, wonderfully unsatisfied.
What's going to change in the next 10 years?…And that is a very interesting question; it's a very common one….I almost never get the question: 'What's not going to change in the next 10 years?' And I submit to you that that second question is actually the more important of the two -- because you can build a business strategy around the things that are stable in time.

What's going to stay the same in climate tech?

The customer will always want things cheaper and better. Let's start with where Bezos left off. The winning formula in all things climate tech will be more affordable and better and sometimes both. Wind and solar are marginally cheaper than oil and gas, but their variability leaves the better part with much to be desired. Electric vehicles are better, but they still need to be cheaper.
Transitions take time. Energy historian Vaclav Smil calculates past energy transitions have taken 50–75 years to ripple through societies. The transition from whale oil to wood to petrol fuels was slow. Most English coal pits opened between 1540 and 1640. Yet for all this, coal had only reached 5% of the global market by 1840. In North America, coal didn't overtake wood until as late as 1884 – even as crude oil became more important.
Red tape exists. Proponents of nuclear and Marc Andreessen will say the only thing holding it back is government regulation. It exists for a good reason, but point taken. We could build faster without it. But red tape also exists in the form of interconnection queues, offshore wind permits, and solar customs. We must learn to work within the current structure or elect new leaders willing to innovate.
Commodities are volatile. The knock on electric vehicles and solar during the last few years was the material cost. Commodities like polysilicon and lithium are subject to wild swings in supply and demand. This is nothing new in the energy world, as oil and gas are subject to similar shocks due to the supply being concentrated in just a few countries.
Technology investment drives down costs. Innovation investment accelerates learning curves. Once that happens, manufacturing scale drives rapid cost declines. The same thing has happened in technology more broadly throughout history. Chips, memory, and broadband have all increased exponentially for decades, and several energy technologies will do the same. Picking which ones would be a prediction.
The world will use more power.
The US and Europe are electrifying their economies at record rates. An influx of electric vehicles and heat pumps combined with growing demand for data centers will increase our consumption in the developed world.
Emerging economies will also contribute to this trend. The middle class in places like India, China, and Africa continues modernizing their lives thanks to newfound productivity and prosperity. Each of these economies will install an increasing number of things like air conditioners and other things we take for granted.

For any secular trend, there are likely 15-20 subcomponents like these that won't change as the world around it does, and I'm confident the energy transition is no different. So, what did I miss? Or do you see more power in predicting the future than I do? I'm curious to hear what you think.