Hedge fund tycoon David Einhorn: Gold will significantly outperform the Nasdaq in the next 3-5 years

David Einhorn, founder of Greenlight Capital, expects that in the next three to five years, gold will significantly outperform the Nasdaq, with a potentially substantial gap.

In a recent interview with Morgan Stanley, David Einhorn issued a warning that large-cap tech stocks on the Nasdaq are facing a fundamental shift from the "asset-light monopoly" business model to a capital-intensive competitive landscape, while the long-term bullish narrative for gold continues to strengthen.

Einhorn referred to this statement as a “bold prediction.” He stated, “With the persistent loss of control over U.S. fiscal policy, the dollar’s status as a reserve currency is being eroded. Coupled with the global trend of 'de-dollarization,' the long-term strategic value of gold is rising systematically.”

Meanwhile, he believes the AI capital expenditure wave is sowing the seeds for long-term profit risks among tech giants.

This view is directly reflected in Greenlight Capital’s portfolio structure. Einhorn revealed that gold is currently the fund’s largest macro long position, which has been held consistently since 2008. He emphasized:

As long as the federal budget remains uncontrolled, gold is my friend.

Einhorn’s bullish case for gold is rooted in his long-term concerns about global fiscal and monetary order.

He believes that the fiscal deficit issue in major economies—especially the United States—remains far from under control. He called out the U.S. Treasury Secretary’s target of a 3% deficit-to-GDP ratio, saying:

I don’t know how they plan to achieve that goal; at present, progress appears minimal.

On the monetary system level, Einhorn pointed out that the West’s freezing of Russia’s foreign exchange reserves was a landmark signal, prompting many countries to reassess the safety of holding dollar assets. He said:

The point of reserves is to have access to them when needed. If that’s not possible, can they really be considered reserves?

He added that after the parabolic rally in gold seen between January and February this year, a short-term consolidation may follow. Einhorn noted:

It wouldn’t surprise me at all if gold does nothing for a year or even a year and a half.

He emphasized that this is a long-term, multi-year outlook. Within five years, gold should digest speculative peaks and resume its upward trend.

Einhorn’s bearish outlook for the Nasdaq is fundamentally based on a sober analysis of the current AI capital expenditure boom.

He believes that current tech sector profits are artificially inflated; depreciation lag will gradually erode earnings in the future.

He used memory chips as an example to illustrate the mechanism: If memory prices rise fivefold, suppliers see profits surge, but hyperscale cloud providers capitalize the high-priced equipment purchases. This means no expense is recorded for the current period, and depreciation pressure is deferred into the future.

Einhorn explained:

I’ve seen an estimate showing that, at the current scale of AI infrastructure buildout, by 2033 depreciation costs will equal the current profit of hyperscale cloud businesses—assuming those businesses continue to grow at their historical rate.

This means that by then, profitability will depend entirely on incremental returns from AI, which will face intense competitive pressure.

On a more fundamental level, large tech companies are losing the business model foundations that support high valuations. Einhorn stressed:

They are shifting from asset-light monopoly businesses to capital-intensive, highly competitive operations. In capital-intensive industries, capital chases returns, and any excess returns are ultimately competed away. When this happens, Nasdaq valuations will be re-rated.

On the long-term industrial structure of AI, Einhorn holds a view contrary to the mainstream market narrative: The value created by AI will ultimately flow mainly to users rather than providers.

He pointed out that the AI sector lacks the "moats" necessary to sustain monopolistic profits over the long term. Unlike the network effects of platforms like Facebook, AI user experiences are highly personalized, and users do not enhance value for each other through network effects.

Unlike the scale economies of traditional software, AI is a capital-intensive business where marginal costs do not approach zero. Einhorn noted:

If ten AI providers can all offer roughly equivalent services and users can switch at low cost anytime, then providers can hardly capture much profit—pricing will converge towards marginal cost.

Einhorn has not directly positioned Greenlight Capital’s investments in the widely recognized AI theme stocks. He stated that none of these stocks are “cheap; everyone is watching them.”

Greenlight Capital’s strategy is to own companies that are already fundamentally attractive and also positioned to benefit long-term from AI-led cost reduction and efficiency gains.

He gave an example: health insurance giant Centene (CNC) has substantial AI application potential in data processing and document automation. He commented:

Even if the AI upside doesn’t materialize, I like this stock for other reasons.