On Carlota Perez's "Technology Revolutions and Financial Capital"

Notes on the tech cult classic

A new experiment — you can now also experience this piece as a GPT-Essay

---

Is there a pattern behind the hype, crashes, and comebacks in technology cycles?

About a year ago, I read Carlota Perez’s famous 2002 book “Technology Revolutions and Financial Capital”. It’s a cult classic and a fan favorite of Fred/USV, Chris/a16z, etc. After seeing it referenced enough times, I figured I should go straight to the source. What follows are my notes and reflections on the read.

Essentially Perez’s framework suggests:

1. There are certain technological big bangs that create a predictable ~50 year technology cycle of installation (technology being created/refined) and deployment (technology being widely adopted, used, and causing societal changes)

2. The midpoint of the cycle is a turning point in the form of a financial crisis. This is to relieve the pressure from the natural overhype that occurs during the installation phase, and to start getting rid of existing structures to make room for new things in the deployment phase.

3. This has happened multiple times during history (we’re in the fifth era)

As with all theories, it’s a good mental model but can be hard to falsify.

I think Perez is right that revolutionary technologies often spark natural exuberance that can get ahead of itself. It echoes other theories like Minsky cycle, Dalio’s Big Debt Crises, and Soro’s theory of reflexivity — all highlighting the cause-and-effect that leads to over speculation and a collapse. Perez’s unique contribution is exploring the boom-bust in relationship with technology revolutions versus just a financial bubble in isolation.

Technology Revolutions

Perez defines a “revolution” as a constellation of interconnected technologies that reinforce one another. For example:

• The “Age of Oil, Automobiles, and Mass Production” combined standardized manufacturing (Taylor process), cheap energy (oil), and new products (cars).

• The current “Information Age” links integrated circuits → microprocessors → software → the internet → and now AI.

I like this part of the theory because I don’t think technologies develop in a vacuum and do build on top of one another.

A new technology that solves a previous problem creates a new set of problems and opportunities. And then people find clever ways to combine different technologies to create brand new experiences that weren’t possible before.

You also see how progress in one technology accelerates the development of complementary ones. The iPhone is a great example. The core product helped spur new technology development in cameras, smaller chips, screens, batteries, charging cables, etc. The r&d efforts for complements gets justified.

Nuances in Installation

Perez breaks the installation phase into two parts:

1. “Irruption” (~10 years) where key parts of the technology are developed, infra is deployed, and complementary services are formed.

2. “Frenzy” (~10 years) where capital floods in, hype takes over, and speculation outpaces reality.

During the irruption phase, progress is driven by financial capital (risk seeking capital) and more radical pioneers (outside of incumbents). Notably it’s done with small amounts of capital.

This tracks for me — breakthrough startups are usually small groups of people with constraints creating new things as a result of new technologies.

A few other ideas from Perez that stood out to me:

• Cultural shifts start early. As the tech develops, behavior begins to change — often on the fringes. I interpret this as niche communities trying new things. Think car enthusiasts before highways, computer hobbyists before mass adoption, or early mobile users. This mirrors the “early adopters” in Everett Rogers’ diffusion of innovation curve.

• Each era invents new financial tools. Mass production needed credit. Railroads used joint-stock companies. The steel era birthed investment banking. Today, you could argue venture capital is the key innovation fueling the current wave. Perhaps tokens with crypto. This one is pretty interesting to think about.

• Overestimation of the diffusion of the technology. In the frenzy phase, people assume the tech will be everywhere overnight. Canals on every river. Dot-coms for everything. They might be right long term but they get the timing wrong. They also often miss what the tech is uniquely good at — or try to retrofit it into the current paradigm instead of building new ones (skeuomorphism)

Nuances in Deployment

Perez also divides the deployment phase into two parts:

1. “Synergy” where innovation starts happening and things really start coming together

2. “Maturity” where the last new products enter and opportunities start to diminish

This is when production capital takes hold, and a focus on managerial and high ROI driven business decisions (vs. a more spray-and-pray approach in financial capital). Unlike the installation phase, where financial capital takes big swings, production capital bets on what’s already working. It’s the golden age: technology moves from the fringe to the mainstream and starts impacting everyday life. I also interpret it as going from “early majority” into the “late majority”.

Some nuances here I liked:

• The distinction between financial capital — which is used to speculate and try things out — compared to production capital — where ROI is expected. Anyone who has gone through roadmapping or quarterly planning will understand this

• Maturity happens because firms start consolidating, m&a, trying new things as growth stalls

• The next irruption phase starts at the maturity phase of the previous one — people and financiers are looking for new places of growth and willing to experiment more

History lesson across the ages

I really enjoyed just learning about the history of some of these cycles. Below are the cycles the book focused on (and I’m complimenting directly with Kondratiev waves that Perez’s work also draws from)

1. Industrial Revolution (1770s to 1830s): Cotton mill big bang —> water powered industry leads to canals, roads, water wheels etc. Partnerships / factory systems.

2. Steam Age (1830s to 1870s): Steam engine big bang —> railroads, iron and coal mining, new towns. Joint stock corporations.

3. Steel & Electricity (1875 to 1920s): Bessemer Process big bang —> heavy engineering, chemicals, shipping, etc. Taylorism & scientific management.

4. Oil + Mass Production (1920s to 1970s): Model T big bang —> oil, mass production, automobiles. Fordism, credit.

5. Information + Telecommunication (1970s to ??): Microprocessor big bang —> Computers, phones, internet, social media. VC model.

Where are we today?

According to Perez, the turning point was both ‘01 and 08 and she thinks we’re just hitting the maturity part of the deployment phase / still waiting for the big bang for the new paradigm.

This might surprise some folks. Many seem to be investing as if we’re already in the installation phase of something new.

I personally think we’re later in the current cycle and more in the maturity phase. Big tech is big, products reach maturity really fast, m&a is happening, and r&d innovation def seems sustaining. See, it’s very easy to just say without evidence or ability to falsify! The problem with theories!

The feeling is hard to ignore that we’re probably squeezing the last bits from software cycles, and while there will definitely be more software, it may be harder to generate (investment style) returns from it. Distribution advantages favor incumbents. But perhaps transformers or something else is the new big bang?

To truly stand out I think we’ll need more disruptive technologies, cultural shifts, and weirder undeveloped markets to take advantage of (and I think AI, crypto, bio, and new devices will play a big part here)

What do I think of the book and Perez’s Theory? Critiques and Takeaways

I generally disliked the book, but I came to appreciate the theory. I found the book to be difficult to read, too academic for me, and only by supplementing with other analysis did I come to better appreciate it.

Another critique, and this is true for most business theories, is that it’s not really falsifiable and very very easy to cherrypick technology innovations into a broad overarching themes. My favorite theories have something that emerges bottoms up and Perez’s can at times feel too tops down “correlation is causation” kind of thinking.

I can why though her work is catnip and justification for venture investors — you can very easily use this to justify your own existence (either in the installation or deployment phase).

Cynicism aside, it’s a useful lens on how technological and financial cycles intertwine. And one that can help you take a longer-term view. A few larger takeaways for me:

• There is cyclicality and S-curve in technology development and adoption, and its interlinked with financial cycles.

• Exuberance / bubbles can be good because they lead to infra development. It’s also okay to exit / take profits when things are exuberant, and reposition for longer term after things blow up. You’re not “bearish” by taking profits, you are being realistic on its timeline, and then re-entering at a time when there is more durable compounding.

• Financial capital (making a series of bets for tail end hits) is what drives the early phase. And production capital (predictable managerial style investments) is what drives the latter phase. You can look at individual behavior to see which stage we might be in.

• Knowing which part of the curve a technology is in is helpful to know where value may accrue

  • In installation, it’ll be all the NEW things that are really changing the paradigm

  • In deployment, it’s going to be the compounding winners, and who will get much bigger than most expect

• Maturity leads to boredom and a desire for step function ROIs. This leads people and financiers to the edges and to innovate more which causes the next wave.

• Nothing happens in a vacuum and there are clusters of technology / services that build on each other (solve each others problems). This creates a lot of second order effects, and the different technologies are often combined to create new products and problems.

Whether or not you buy the full theory, I have to admit Perez’s work is a useful tool to zoom out and use to expand our horizon of possibilities.