Most investors are taught to value companies based on growth.

Growth is easy to measure. Revenue growth, earnings growth, subscriber growth. These metrics are visible, quantifiable, and widely reported.

But growth alone does not determine enterprise value.

Capacity does.

The macro applications of this frame — Billboard, SXSW Festival, the Golden Globes — are visible in the transaction record. But scarcity thinking applies equally at the smaller deal level, where the scarcity is less obvious and therefore more mispriced.

BuzzAngle Music is the clearest example. In 2018, BuzzAngle was not a growth story. Revenue was modest. Distribution was limited. By conventional growth metrics, it was a small data company in a niche market. But evaluated through a scarcity lens, it was one of very few platforms in the world capable of challenging Nielsen's monopoly on music consumption analytics. That capability — the methodology, the institutional client relationships, the daily data infrastructure — could not be replicated quickly. It had been built over years through specific technical choices and specific industry relationships that were not available to a new entrant.

PMC did not invest in BuzzAngle's growth. It invested in BuzzAngle's scarcity — the specific and non-replicable capability that, combined with Rolling Stone's editorial authority, created a competitive instrument that the growth metrics could not surface. The Rolling Stone Charts, launched shortly after the investment, were the first expression of that scarcity being activated. The eventual Billboard acquisition was the thesis playing out at scale.

The same pattern appears in the Long Beach Surgical acquisition. The center's scarcity was not obvious from the outside — it looked like a small physician-owned facility with modest reimbursement rates. But inside the diligence, the scarce asset was clear: a credentialed surgical team with embedded physician relationships built over years, a multi-specialty expansion opportunity that independent centers in the region could not access without institutional affiliation, and a geography with structural demand that no new entrant could satisfy quickly. The value was not in the growth trajectory. It was in the capacity that competitors could not easily replicate — and that SCA affiliation could immediately unlock.

Scarcity thinking at the deal level requires the same discipline as at the portfolio level: ignore the growth narrative, find the constraint, evaluate whether the constraint is durable, and price accordingly. The market consistently undervalues constraint control because constraint control is not visible in the metrics that analysts report.

When we acquired Billboard, the decision was not driven by its short-term revenue trajectory. It was driven by its structural position.

Billboard controls authority in music. It defines legitimacy through its charts. That authority cannot be manufactured. It cannot be replicated quickly. It exists because of decades of accumulated trust and institutional recognition.

The same principle applied to SXSW Festival.

SXSW Festival is not valuable because of its quarterly growth rate. It is valuable because it occupies a singular position at the intersection of technology, culture, and innovation. Its capacity is fixed. Its authority is scarce.

The Golden Globes represents another form of scarcity. It controls a cultural moment tied to a specific point in the awards calendar. That moment cannot be expanded. It cannot be duplicated. It regulates access to attention during one of the most globally recognized entertainment events of the year.

In each case, scarcity — not growth — determined long-term value.

This principle applies equally to physical infrastructure.

Artificial intelligence is not constrained by demand. Demand for compute is effectively unlimited.

AI is constrained by capacity.

NVIDIA controls scarce GPU compute capacity.

TSMC controls scarce advanced semiconductor fabrication capacity.

SK Hynix and Micron control scarce high-bandwidth memory capacity.

These companies do not simply participate in AI growth. They regulate how fast AI can grow.

They control throughput.

Throughput control is the foundation of structural economic leverage.

Traditional valuation frameworks such as PEG ratio assume growth reflects opportunity.

In capacity-constrained systems, growth reflects production limits.

When supply is constrained, reported growth understates structural power.

This creates persistent mispricing.

The market rewards visible growth while underestimating invisible constraint control.

This is why infrastructure controllers often appear expensive before their structural advantage becomes obvious.

My capital allocation decisions prioritize control over scarce capacity.

This applies across both media and technology.

In media, scarce authority creates durable enterprise value.

In AI infrastructure, scarce compute, fabrication, and memory capacity create durable enterprise value.

This is why companies such as NVIDIA, TSMC, Micron, and SK Hynix represent structurally advantaged positions. They control infrastructure that others depend on but cannot easily replicate. They regulate system expansion. Over time, these controllers capture disproportionate economic value relative to participants.