Technology Revolutions and Financial Capital

The Cyclical Structure of Technology Revolutions and Financial Capital

Both technology revolutions and financial capital develop in strongly cyclical patterns; the two reinforce each other, and their relationship shifts from phase to phase. Counting from the First Industrial Revolution, technology revolutions have run through five iterations, and we are now entering the opportunity of a new AI cycle. Within every cycle, financial capital has encouraged and diffused innovation — and earned returns beyond ordinary profit for doing so.

More important, the separation of financial capital from production capital helps new innovation mutate in nonlinear ways. Precisely because financial capital exists, the winners of the previous cycle do not always capture the next cycle’s innovations, while financial capital finds it easier — and is more willing — to back what looks, by the old cycle’s standards, like “peripheral innovation,” and is therefore positioned to catch the birth of the new cycle’s killer artifact. The killer artifact — the category-defining product — is the core of every technology revolution: its appearance opens the cycle, and its penetration determines which phase the cycle is in.

Within a single technology-revolution cycle, different kinds of financial capital play markedly different roles at different phases. The Irruption phase is dominated by venture capital, production-expansion capital, and the introduction and iteration of financial instruments, and its hallmark is the appearance of the killer artifact. The Frenzy phase is led by speculative capital, debt capital, and grey capital, as the market whips itself into a mania around a single theme. In the Synergy phase all six kinds of financial capital participate, harvesting the killer artifact’s rapid growth and backing the adaptive innovations that grow alongside it. By the Maturity phase, production-expansion capital, debt capital, and grey capital dominate, earning steady returns in adaptive-innovation categories while scouting the next cycle’s opportunities.

Zooming in on a single cycle, the penetration curve traces a classic S-shape: the killer artifact’s appearance opens the Irruption phase; during the Frenzy, financial capital gradually decouples from production capital and the bubble bursts at the turning point; thereafter financial and production capital re-couple through the Synergy and Maturity phases. The first half of the cycle (the Installation period) is driven by financial capital, the second half (the Deployment period) by production capital.

On a longer horizon, five technology revolutions have succeeded one another: the First Industrial Revolution of 1771 (the Spinning Jenny), the steam-and-railways revolution of 1829, the steel-and-electricity revolution of 1873, the oil, automobile, and mass-production revolution of 1918, and the information-technology revolution of 1974 (with the PC’s mainstreaming, epitomized by Windows 95, as its landmark) — followed by the new cycle opening around 2022. Each cycle’s diffusion lasts roughly half a century (58, 43, 45, 74, and 40+ years), and the idle financial capital accumulated in one cycle is precisely the fuel that ignites the next revolution.

The First Three Technology Revolutions: Financial Capital Struggles to Capture Value

In the first three technology revolutions, textiles, steel, and the electrical industry created the greatest value, yet financial capital struggled to capture it. A technology revolution revolves around the appearance of its killer artifact; in essence, new technology drives a step-change in industrial productivity: the first two revolutions opened with the textile machine and standardized production, while the third erupted with the mass production of steel — new means and modes of production taking hold and lifting productivity dramatically.

Across these three cycles, financial capital was still at an early stage of its development and depended heavily on industrial capital: it had reached visible scale, but its core function remained enabling industry; it had not yet become an independent industry of its own. Only at the end of the third technology revolution did financial capital, hatched out of industrial capital, begin to acquire real independence.

The first two technology revolutions: financial capital as an aid to production efficiency

The driving force of the first two technology waves was raising productivity in cotton textiles. The cost of spinning 100 lb of cotton in Britain fell from £2.1 in 1780 to £0.1 in 1830 — down 94% in 50 years. The worldwide consequence of collapsing production costs was a surge in the value of cotton-goods exports: Britain’s cotton-textile exports totaled a mere £260,000 in 1780 and reached £41.05 million by 1830, compounding at roughly 10–20% a year through 1780–1810 and beyond. In this process, the bank capital that provided credit to the industry profited most.

The third technology revolution: steel and electricity capture the most value; financial capital reaches initial scale

The third technology revolution formed a three-layer structure with electricity as the energy infrastructure, steel as the killer artifact, and heavy industry as the derivative application. The rapid fall in steel prices marked the revolution’s irruption: in relative-price terms, steel fell from 2.8 in 1870 to about 1.1 around 1885, then held near 1.0 through the Frenzy, turning, and Synergy–Maturity phases, ending at 0.9 in 1920.

Financial capital in this era had reached initial scale but remained inseparable from industrial capital. The sales of the major American and German electrical companies from 1893 to 1913 show the industry’s order-of-magnitude expansion: over two decades, Westinghouse’s sales grew from £1.0M to £8.0M and General Electric’s from £4.24M to £17.84M, while Germany’s Siemens and AEG reached £23.65M and £22.65M. Over the same period, US investment in industry (net of inflation) rose from $2.7B in 1879 to $20.8B in 1914.

Financial capital’s participation in the electrical industry already foreshadowed its later ability to capture industrial value independently: behind Edison’s financing stood two financial houses, Drexel and Morgan; Thomson-Houston was funded by Boston capitalists; and the merger of the two into General Electric was engineered by Henry Villard — a man of both finance and railroads.

The Fourth Technology Revolution: Oil, the Automobile, and Mass Production

The fourth technology revolution formed a cycle with oil as the energy infrastructure, the automobile industry as the core killer artifact, and mass consumption and applications (aircraft, automobiles, and the like) as derivative applications. Infrastructure, killer artifact, and applications realize their value in sequence, each in its own phase of the cycle.

The killer artifact captures the most value within the cycle, while derivative applications live longer: automobiles captured the greatest value during the Synergy phase, with the stock rising 8-fold — the measure of a killer artifact’s worth within its cycle. Upper-layer applications, by contrast, stay vigorous after the cycle ends; their growth early in the next cycle can even exceed that of the Synergy-phase killer artifact.

Infrastructure companies capture the new cycle’s value first

Standard Oil, the energy infrastructure, saw its stock rise about 400% during the “eve of irruption” (1892–1909), before automobiles were even in volume production — infrastructure companies always capture the new cycle’s value first. Through the Irruption phase its stock climbed another 200% or so, punctuated by a crash on the antitrust breakup of 1911 before resuming its rise.

The killer artifact captures the most value in the Synergy phase

General Motors’ share-price trajectory maps the killer-artifact company’s performance across the whole cycle: up about 200% to $5 in the Frenzy (1920s); down 35% in the turning phase (1930s), touching $0.4 at the low; up about 800% in the Synergy phase (1940s–1950s), from around $2 to $18; then, in Maturity (1960s–1970s), a spike to $40 followed by a 39% retreat to $11 by the late 1970s.

Derivative applications capture value in Maturity and early in the next phase

The derivative-application companies of the Nifty Fifty — IBM, Coca-Cola, Merck, GE, and Disney among them — first rose about 200% in the fourth revolution’s Maturity phase (peaking at 3.06 around 1972), then gave back 65% in the bear market; but from the mid-1970s to the late 1990s — early in the next phase — they compounded roughly 6,900%. The growth momentum of derivative applications carried into the next cycle, even outrunning the Synergy-phase killer artifact.

Auto production and cycle milestones

Seen through output, US auto production went from about 0.4 million units around 1908 to 5.4 million in the Frenzy, back down to 3.1 million in the turning phase, then 8.2 million and 13 million through the Synergy phase, finally reaching 22.8 million in 1974. Penetration crossed 5% during the Irruption phase — precisely the signal of the automobile switching from slow growth onto an exponential track.

The Fifth Technology Revolution: The Killer Artifact Takes Most of the Profit

The fifth technology revolution formed a cycle with the microprocessor as hardware infrastructure, the PC and the iPhone as core killer artifacts, and internet applications as the upper layer — and for the first time, financial capital could capture value across both private and public markets. The fifth revolution offered the largest opportunity of any technology revolution, and its killer artifacts the largest multiples: Microsoft and Apple each grew roughly 600-fold, in the Frenzy and Synergy phases respectively — which for the PC and for mobile were precisely their own irruption windows. The application layer grew fast too, but still a shade behind the killer artifacts.

Infrastructure: growth concentrated in the first half of the cycle

Intel, the emblem of hardware infrastructure, saw its growth concentrated in the cycle’s first half: up about 350% in the Irruption phase (1972–1987) and about 5,900% in the Frenzy (1987–2000); after the bubble burst in 2000 it fell 75% through the turning phase, added only about 275% in the Synergy phase (2003–2020), and gave back about 50% in the Maturity stage after 2020. The infrastructure company’s window for capturing value opens visibly earlier — and closes sooner — than the killer artifact’s.

Killer artifacts: Microsoft and Apple capture the most value in the Frenzy and Synergy phases

As the killer artifacts of the PC and the mobile internet, Microsoft and Apple captured value exactly when their platforms erupted: Microsoft rose about 59,900% in the Frenzy (1987–2000), retreated 50% in the turning phase, added another 1,325% in the Synergy phase, and slowed to 32% in Maturity; Apple rose 100% in the Irruption and 625% in the Frenzy, fell 80% in the turning phase, then — riding the mobile internet the iPhone opened — rose about 64,900% in the Synergy phase (2003–2021), easing to 12% in Maturity.

Derivative applications: the most value captured in the Synergy phase

The derivative applications, represented by Google, Meta, and Amazon, rose about 4,500% in the Frenzy (1997–2000), crashed 92% in the turning phase, then compounded roughly 37,330% across the long Synergy phase (2002–2021), and still logged a 25% gain into Maturity. Solid fundamentals should keep them supported through the Maturity phase.

Laying the fifth revolution’s five phases side by side makes the relay structure — infrastructure → killer artifact → upper-layer applications — plain to see:

Further Reflections

Reflection one: technology evolves along infrastructure → killer artifact → derivative applications

Within a single technology cycle, technology consistently evolves along the path of infrastructure → killer artifact → derivative applications, and financial capital must rotate its sub-sector themes accordingly to capture the most value in each phase. The third revolution ran on electricity as energy infrastructure, steel as the killer artifact, and heavy industry as derivative applications; the fourth on oil as energy infrastructure, the automobile as the killer artifact, and durable consumer goods, communications, and defense manufacturing as derivative applications; the fifth on the microprocessor as hardware infrastructure, the computer and the smartphone as killer artifacts, and PC- and mobile-internet applications as the upper layer.

Meanwhile, across successive technology cycles, the value captured by the killer-artifact company keeps growing, and financial capital’s toolkit for capturing it keeps expanding. Financial-capital instruments have developed alongside the cycles: from a single function — bank credit — at the start, to financial capital partly hatching out of industrial capital in the third revolution, to the growth of public-market equity investment in the fourth, and the full flowering of private- and public-market equity investment in the fifth. Every added instrument marks another increment in financial capital’s ability to capture industrial value; and because each new technology cycle is larger than the last, the later the cycle, the more value financial capital captures — the fourth revolution’s Synergy-phase killer artifact (the automobile) grew 800%, while the fifth’s (Apple) grew 64,900%. Within one cycle, the investment theme also differs by phase and must stay aligned with the cycle’s progress.

Reflection two: falling production costs and killer-artifact penetration above 5% mark a new revolution’s irruption

Two signals identify the irruption of a new generation of technology revolution. The first is a rapid fall in the production cost of the means of production, which marks that the material basis for the new revolution is in place: in the first two revolutions, Britain’s cotton-spinning cost fell 94% over 50 years (Exhibit 4); in the third, steel’s relative price plunged during the Irruption phase (Exhibit 6); in the fourth, the crude-oil price index likewise fell severalfold across the 1870s–1890s — the collapse in oil production costs paved the way for the automobile.

The second is the killer artifact appearing and its penetration crossing 5%: once a cycle’s killer artifact breaks through that threshold, scale becomes possible and the category flips from slow growth onto an exponential track. In the third revolution, steel’s share of output crossing 5% marked the eruption of heavy industry; in the fourth, automobile penetration crossing 5% marked the eruption of assembly-line manufacturing (Exhibit 12); in the fifth, PC penetration crossing 5% marked the eruption of the information industry — the Intel 4004 of 1971 held just 2,250 transistors, versus 1.2 million by 1989; after the IBM PC launched in 1981 and the Apple Macintosh in 1984, computer sales climbed an exponential path, and MS-DOS, the World Wide Web, Mosaic, and Internet Explorer in turn pushed the cycle into its Frenzy.

Reflection three: Installation vs. Deployment — is the killer artifact settled?

The deepest difference between the Installation and Deployment periods is whether the killer artifact has been settled — and the investment categories that capture the most value differ accordingly. Installation-period strategy: put 80% into the previous cycle’s application companies for steady growth and 20% into the coming cycle’s infrastructure — the fourth revolution’s derivative-application companies kept performing through Maturity and the next cycle’s Installation period (Exhibit 11), while Intel, the fifth revolution’s infrastructure standard-bearer, completed most of its rise precisely during Installation, i.e. the Irruption and Frenzy phases (Exhibit 13). Deployment-period strategy: back the category definer’s growth without wavering — General Motors’ 800% rise in the Synergy phase (Exhibit 10) and Apple’s 64,900% (Exhibit 14) are both footnotes to the category definer in Deployment.

In short: as a new cycle opens, the best allocation pairs the previous cycle’s mature application companies with the current cycle’s technology innovators, combining stability with growth; in the Frenzy, invest in the private market in the companies most likely to become the category definer in their track; and in the Synergy and Maturity phases, own the category definer and its derivative-application companies.

Reflection four: implications for the current stage

AI, the central theme of a brand-new technology cycle, must likewise pass from infrastructure to killer artifact to applications — and it is still early in the Irruption phase: the AI stack remains in its infrastructure-building stage, with compute, data, and energy as the three pillars; prototypes of the killer artifact have appeared, but the product form that triggers a rapid penetration breakout is still awaited. The next milestone is killer-artifact penetration crossing 5%; the milestone after that is the emergence of the representative company in the killer-artifact race — most likely a new company building a new product, not an extension of a mature company from the previous cycle.

What this implies today: own the internet cycle’s application companies plus the AI cycle’s infrastructure. On one side, invest in the previous technology cycle’s application companies (the Magnificent 7) for dependable growth — after rising 37,330% through the Synergy phase, Google, Meta, Amazon and peers still logged a 25% gain in Maturity on solid fundamentals (Exhibit 15). On the other, position in the new cycle’s infrastructure themes: compute (Nvidia), data (Google, Meta, and other companies with edge data-collection reach), and new energy (fusion, batteries, and other new energy technologies).