Why Intel Can't Beat TSMC (And What It Reveals About Every Business
The first of the 7 Powers
“A monopoly business gets stronger as it gets bigger: the fixed costs of creating a product (engineering, management, office space) can be spread out over ever greater quantities of sales.” — Peter Thiel
Two companies go head-to-head competing for market share in the digital streaming industry. Each has excellent management, brilliant content, and state-of-the-art facilities, but one company has something the other doesn’t—numbers.
Company A has 20 million subscribers and Company B has 100 million. If both companies spend $100 million on creating original content, it’ll cost Company A $5 per subscriber, but Company B only $1. Basic maths states Company A is going to have a hard time competing against Company B because it costs them five times more to make the same product. What’s Company A to do?
The obvious solution would be to cut the price of their subscription fee, ensuring it’s materially lower than Company B’s. However, Company B, being the astute strategists they are, know this play and drop their prices to match. This folly favours Company B — with lower fixed costs and higher margins protecting them from competitive onslaught. Company A, with higher fixed costs and lower margins, feels the squeeze and retreats gratefully back to the trench it came from.
This is Scale Economies at work.
Scale Economies create a moat not just through size, but through the prohibitive costs a challenger must pay to reach the incumbent’s efficiency.
This is the first deep-dive in the 7 Powers series. Hamilton Helmer’s landmark book on how businesses use strategy and value creation to create consistent differential returns — something he calls Power. In the introductory essay, I explained why value creation alone isn’t a moat. Now let’s examine the first Power that separates good businesses from great ones.
What Are Scale Economies?
“A business in which the per-unit cost declines as production volume increases.” — Hamilton Helmer
Scale Economies is fundamentally about structure — not operational efficiency. When Netflix decided to move away from leasing film studio properties to creating originals, they solidified an element of their business model that had long had variable costs attached to it. If the film studios set a price for a property lease, there wasn’t a lot Netflix could do — but by creating originals, they now set the budget. If they wanted to spend $100 million on an original, that’s what they’d spend, and the more their subscriber base grew, the lower their costs were per original.
This is what Helmer means by “per-unit cost declines as production volume increases.”
The Different Types
Scale Economies can emerge through multiple mechanisms:
Fixed costs: High upfront capital expenditure (factories, equipment, R&D) spread over increasing volume. Netflix originals, semiconductor fabs, software platforms.
Volume/area relationships: Utilising production area to increase volume results in lower per-volume costs. Refineries and warehouses.
Distribution network density: As distribution networks grow denser with more customers per area, delivery costs decrease through more economical routes. Amazon Prime delivery.
Purchasing power: Large-scale buyers negotiate better pricing than smaller companies. Walmart, Costco, supermarkets.
Learning economies: Knowledge that reduces costs or improves quality, tied to production levels. Experience compounds over volume.
Benefits and Barriers
Benefits are more common in business: pricing power, cost reduction, lower operating costs. But barriers — those that prevent competitors from engaging in tit-for-tat tactics — are less so.
Benefit: The conditions that created the Power must materially grow cash flow through increased prices, reduced costs, or lower capital expenditure.
Barrier: As well as persistently increasing cash flow, there must be a barrier that prevents competitors from engaging in value-destroying arbitrage.
Barrier conditions deserve your utmost attention. Scale Economies can only exist when the cost advantage is material enough to change competitive behaviour — when competitors purposefully decide not to engage because it’s economically inefficient to do so.
Case Study: Taiwan Semiconductor Manufacturing Company (TSMC)
TSMC produces approximately 60% of the world’s semiconductor chips and over 90% of the most advanced chips. Their semiconductors can be found in modern cars, industrial sensors, and power management chips. Advanced chips — requiring more processing power — are found in iPhones, MacBooks, and AI processors.
TSMC isn’t a chip designer: Apple, Nvidia, AMD, et al. design their own chips, and depending on the complexity of manufacturing, that chip will almost certainly go to TSMC, where they’ve built an insurmountable moat. This is called a “foundry” business — they’re a fabrication plant for hire.
Their operating margins are consistently 40–45%, while their nearest competitor Samsung’s foundry business margins sit at 10–15%. Intel, who entered this market in 2021, has negative margins. Over two decades, these margins haven’t compressed. They haven’t been eaten away. TSMC is able to consistently create that determining factor of genuine Power: persistent differential returns.
The fixed cost base
TSMC’s fabrication plants cost between $20–30 billion to build. They’re filled with dozens of ultraviolet lithography machines costing in excess of $150 million each, run by thousands of people with PhDs, spending $3–4 billion annually on R&D developing the next generation of manufacturing processes.
That’s an eye-watering amount of capital expenditure. They have 12–13 operating fabs across Taiwan and Japan, with each fab having multiple phases, meaning the total number of operating fab phases is closer to 50.
And there’s more to come. TSMC is currently expanding to Arizona, where they plan to build 3–6 new fabs, two advanced packaging facilities, and an R&D centre totalling an estimated $165 billion. There are plans for 2–3 more in Japan, one in Germany, and one more back home.
Imagine trying to compete with that.
The volume advantage
TSMC can invest that kind of capital because they have the volume. TSMC produces 13x more wafers than Intel and 50x more than Samsung’s advanced foundry nodes. Intel and Samsung simply don’t have the volume that would enable them to invest in the kind of infrastructure and research needed to close the gap on TSMC.
What creates a barrier your competitors can’t cross?
You’ve seen how TSMC built their fortress. But here’s what separates genuine Power from temporary advantage: the barrier must be insurmountable. In the full essay, you’ll get:
The complete TSMC breakdown: Why Intel can’t catch up despite 60 years of experience and unlimited capital
Surplus Leader Margin explained: The simple calculation that reveals whether any business has genuine Power
The diagnostic framework: Three questions to determine if a company has Scale Economies or just operational leverage
When Scale is a trap: The red flags showing when chasing growth destroys value
What separates TSMC from everyone else: Why most industries will never achieve this Power
This is the foundation for the entire 7 Powers series. Subscribe to get access to this essay plus the full breakdown of each Power, complete with frameworks for analysing any business.
The barrier: Why Intel can’t catch up
