Highlights
- Morgan Stanley forecasts 1 billion humanoid robots generating $4.7 trillion annually by 2050.
- Projected humanoid robot production could dramatically increase demand for rare earth elements, particularly NdPr magnets.
- The report suggests significant market potential but cautions investors to treat the projection as a scenario requiring further validation.
Morgan Stanley (opens in a new tab) has ignited fierce debate with its latest global insight report projecting humanoid robotics as a future multi-trillion-dollar industry with seismic implications for critical mineral markets—particularly rare earth elements (REEs). The report estimates that by 2050, the world could host 1 billion humanoid units generating $4.7 trillion in annual revenue. If true, the required material inputs for these robots, especially neodymium-praseodymium (NdPr) magnets, would more than double current global magnet demand.
According to Morgan Stanley, each humanoid could require up to 1.3 kg of NdPr, pushing incremental NdPr demand to +167% of 2030 levels by 2050. By then, total humanoid-induced demand across all covered critical minerals (NdPr, Li, Ni, Co, Cu, graphite) could total $50–$120 billion annually. The report upgrades long-term NdPr pricing from $135/kg to $209/kg (ex-VAT).
Compelling Vision, Questionable Certainty
Morgan Stanley’s thesis is bold, backed by a globally integrated analyst team. Their commodity modeling draws on credible sources such as Wood Mackenzie (opens in a new tab) and the IEA (opens in a new tab), and the humanoid forecasts stem from cross-sector AI and robotics work led by Adam Jonas, Head of Global Auto and & Shared Mobility Service (opens in a new tab).
Yet, the scale and certainty of the predictions raise serious questions for Rare Earth Exchanges (REEx).
First, we’ll refer to the question as “assumption creep.” The humanoid forecast rests on a chain of unproven assumptions that need more validation. For example, massive AI-driven robot adoption, minimal substitution, and unchanged mineral intensity per unit. Is it realistic to assume Tesla-like robots will be mass-deployed by 2035, especially with no operating supply chains in place? REEx suggests this needs serious vetting.
Second, the China exposure dynamic is left unresolved. The report rightly notes China’s 74% share of NdPr refined supply and >90% share in heavy rare earths. But it underplays the geopolitical risk of a humanoid-centered West dependent on China for critical inputs. Even MP Materials still ships concentrate to China for finishing. Next comes substation blind spots.
While the report acknowledges iron nitride (FeN) and cerium substitution R&D, it dismisses long-term displacement risk. FeN magnets could radically alter the NdPr trajectory, especially if geopolitical factors push Western governments to mandate REE-free supply chains. Finally, the S&P Global data cited in the report shows average mine lead times now stretch to 17.8 years. Even assuming all of Wood Mackenzie’s 25kt “incentive supply” comes online, Morgan Stanley concedes that projected NdPr deficits grow from 1% (2039) to 35% (2050). Where is the path to close that gap?
Implications for Investors
Based on this humanoid thesis, Morgan Stanley has upgraded Lynas (LYC.AX) and Iluka (ILU.AX), pointing to outsized EBITDA sensitivities to NdPr prices. The report also includes a 34-company global watchlist for humanoid-material exposure. Still, REEx cautions that price elasticity, permitting drag, and political constraints could derail timelines.
Investors should treat Morgan Stanley’s humanoid-driven mineral demand narrative not as a forecast, but as a scenario—one of several possible futures. The scenario is directionally credible but quantitatively aggressive.
REEx Bottom Line
The humanoid robot market could indeed become a transformative source of incremental demand for rare earths, especially NdPr. However, the timeline, technology adoption curve, and geopolitical assumptions warrant far more scrutiny. For now, this remains a high-stakes, high-risk hypothesis that should be stress-tested, not blindly priced into rare earth equities.
Source: Morgan Stanley Global Insight Report, “Humanoids and Global Materials” (opens in a new tab)
Lead Authors: Rahul Anand ( (opens in a new tab)Australia), Adam Jonas (opens in a new tab) (U.S.), Sheng Zhong (opens in a new tab) (Asia), and Amy Gower (opens in a new tab) (UK)
To discuss this topic visit the REEx Forum. (opens in a new tab)
Leave a Reply