Reinvestment Composition

In the AI Story

The framework matters because R&D as percentage of revenue — the standard measure of technology company reinvestment — is silent on composition. A company spending 20% of revenue on R&D may be deploying capital productively or destructively depending on whether the spending is concentrated in code-writing or in ecosystem-building. The aggregate number is the same; the value implications are opposite.

The decomposition maps onto the moat hierarchy. Investment at the code layer faces declining returns as AI commoditizes the function. Investment at the data layer generates compounding returns because data assets accumulate and become more valuable over time. Investment at the integration layer generates durable returns because each new integration deepens platform embeddedness. Investment at the trust layer generates defensive returns by expanding the addressable market to regulated industries. Investment at the judgment layer — the capacity to identify what should be built — generates the most speculative but potentially highest returns of all.

The framework applies directly to Edo Segal's decision in The Orange Pill to keep his full team and retrain rather than capture productivity gains through headcount reduction. Damodaran's assessment: the immediate cost is measurable (incremental labor expense); the return is harder to measure but real (institutional knowledge, customer insight, judgment). If the return exceeds cost of capital, value is created; if not, the retention is sentimentality dressed as strategy. The judgment is empirical, not philosophical — it depends on whether retained employees actually produce the higher-value output that justifies their cost.

The historical pattern is instructive. Microsoft survived the internet transition not by investing more in desktop software but by reallocating to cloud infrastructure. Apple survived the smartphone transition not by investing more in Macs but by reallocating to mobile hardware, software, and services. Companies that failed — Nokia, BlackBerry, Kodak, Blockbuster — continued investing in depreciating capabilities, either through inability to see the transition or through organizational inertia. The same pattern is playing out in 2026: companies reallocating from code to ecosystem will create value; those continuing to invest primarily in code-writing will destroy it.

Origin

Damodaran has written extensively on reinvestment and value creation across decades, with the AI-specific application appearing in his 2024-2026 commentary on enterprise software companies.

Key Ideas

Composition over total. R&D as percentage of revenue is silent on composition; what matters is where the spending is directed.

AI changes the return calculation. Code-writing investment faces declining returns; ecosystem investment faces increasing returns.

Mapping onto the moat hierarchy. Each layer of the hierarchy has different return dynamics under AI disruption.

Historical pattern is consistent. Survivors of past transitions reallocated capital from depreciating to appreciating capabilities.