Today’s Espresso Index

“Compute creates intelligence. Learning creates velocity. Conversion creates wealth.”

In 1986, The Economist introduced the Big Mac Index to make purchasing power parity understandable through the price of a familiar product across countries. The Velocity Edge's Espresso Index applies a similarly intuitive idea to a radically different economic question.

The Big Mac Index compares the purchasing power of currencies. The Espresso Index seeks to compare the productive power of intelligence across economies. Its starting point is the emergence of the token economy: a world in which accelerated computing, GPU architectures, optimized acceleration libraries and increasingly capable AI models are reducing the cost of producing intelligence while expanding the speed at which machines and organizations can learn.

But the true revolution is not simply faster computation. It is the compounding of learning. Accelerated computing reduces the cost of experimentation. More experimentation generates data. Data improves models and operational decisions. Better models enable more autonomous actions. Those actions generate additional feedback, potentially accelerating the next learning cycle.

1. 01 Learning becomes self-reinforcing

   The most consequential feature of the token economy is that learning can compound. Countries that combine compute, research, industrial applications and capital can generate a recursive advantage: more deployment → more feedback → better models → cheaper execution → more deployment. This could widen economic differences even as foundational AI models become more widespread. signal: The recursive loop, not raw compute, is the durable moat.

2. 02 A rotation from producers to deployers

   The United States retains the strongest combined position in frontier development and private capital. China possesses exceptional potential to convert intelligence into physical output. South Korea and Taiwan occupy indispensable semiconductor layers, while Europe's opportunity lies in connecting its engineering base to accessible intelligence. The financial consequence is a possible rotation from countries producing intelligence toward those deploying it most productively. signal: That rotation is not yet established—it is what future releases should test.

Five Key Takeaways For Investors

learning velocity over installed scale

• The next AI premium is learning velocity — look beyond installed compute toward the speed at which organizations turn inference into better decisions and recurring revenue.
• Distinguish AI producers from AI converters — the U.S. dominates capital formation; China is strong in industrial deployment. Different sources of return.
• Watch the second tier — South Korea, Japan, Taiwan and leading European industrials can capture value through semiconductors, automation and specialized systems.
• Europe needs an industrial conversion strategy — Germany, France, the Netherlands, Spain and Italy gain if energy, capital, data and procurement operate as a more integrated market.
• Earnings diffusion is the decisive catalyst — with two-thirds of forecast U.S. Q3 growth concentrated in tech, watch for verified margin and ROIC improvements in AI-adopting industries.