AI spending is surging, but when does it create value? Morgan Stanley’s ROIIC lesson for India

The global artificial intelligence investment boom is pushing technology companies to spend heavily on data centres, chips, servers and networking infrastructure. But as free cash flow comes under pressure, the more important question for investors is whether each additional dollar invested generates enough returns to create value.

A recent report by Morgan Stanley Investment Management's Counterpoint Global examines this through Return on Incremental Invested Capital (ROIIC), a measure designed to assess the return generated by new investment. The framework could offer a useful lens for investors assessing technology companies, including those in India's IT sector, as AI-related spending increases.

What is ROIIC?

ROIIC measures the change in net operating profit after tax (NOPAT) generated by additional investment in a business. Morgan Stanley calculates it using a rolling three-year period, with a one-year lag between investment and the resulting NOPAT.

The report argues that returns above a company's weighted average cost of capital (WACC) indicate that incremental investment is creating value.

This distinction is important during an AI investment cycle. A company can report weaker or even negative free cash flow because it is spending aggressively on future capacity, without that necessarily indicating that the investment is destroying value.

MUST READ: Anthropic says Claude now leads 26% of its AI R&D work: How fast is AI development accelerating?

Hyperscalers offer an example

Morgan Stanley analysed Amazon, Alphabet, Microsoft, Meta Platforms and Oracle. Their combined ROIIC peaked at nearly 40% in the first quarter of 2026, compared with an estimated WACC of about 8%.

The report expects combined ROIIC to decline to around 23% in the third quarter of 2027, before recovering to approximately 35% by 2030. Importantly, the projected returns remain above the estimated cost of capital throughout the forecast period.

At the same time, the five companies' combined free cash flow is expected to fall sharply. Morgan Stanley estimates that FCF could bottom at around -$265 billion in Q3 2027, before recovering to nearly $505 billion by 2030.

What it means for investors

The report's central message is that cash-flow pressure has to be assessed alongside the returns generated by investment. AI infrastructure spending can depress FCF initially because companies spend before the additional capacity is fully monetised.

Amazon CEO Andy Jassy similarly noted that the returns from investments in land, power, buildings, chips, servers and networking equipment become attractive after they are deployed, while early-year FCF can remain under pressure when capex grows faster than revenue.

For Indian investors watching the country's technology sector, the framework highlights metrics beyond revenue growth and margins. As companies invest in AI capabilities, ROIIC, ROIC, capital expenditure and free cash flow can help assess whether higher investment is translating into sufficient economic returns.

Morgan Stanley ultimately identifies the key test as whether sales and EBIT growth can keep pace with rising investment—a consideration that becomes increasingly relevant as the AI spending cycle expands.

ALSO READ: AI-driven drug discovery race: Anthropic quietly opens biology laboratory in major push