Follow-Up: When the AI Bubble Deflates, Gold Remembers What Money Is

Rafi Farber's measurement of the AI mania against gold ounces, combined with the sharp reversal in AI-heavy Asian markets, strengthens the case that capital fleeing a deflating technology bubble will ultimately reinforce the monetary role of gold and the cash-flow power of the mining equities that produce it.

In the previous analysis, we examined the growing signs of capitulation inside the AI trade: widening credit spreads on hyperscalers, the violent unwinding of momentum positioning, heavy selling in memory names, and the forced reduction of crowded long technology exposure. We argued that the capital leaving that trade would eventually seek less crowded, cash-generative homes—and that senior gold producers already generating billions in free cash flow at $4,000 gold stood as one of the clearest candidates. A complementary perspective now sharpens that thesis further. In a recent discussion, Rafi Farber of the Endgame Investor examined the AI bubble through a monetary lens that most equity analysts still neglect: he measured it in ounces of gold.

Measuring Bubbles in Real Money

Farber’s central observation is straightforward and difficult to dismiss. When the peak of the late-1990s semiconductor-heavy technology bubble is expressed in gold ounces, it reached approximately 4.75 ounces. The recent peak of the AI-driven semiconductor complex registered closer to 3.5 ounces. By this measure, the dot-com excess was larger. The AI advance was faster and more concentrated, but it did not surpass the earlier mania when priced in the one monetary asset that cannot be printed. The same comparison expressed in silver ounces yields a similar conclusion: the earlier bubble was bigger. What matters more for the present is the direction of travel. From its early-2026 peak near 3.5 gold ounces, the semiconductor index has already declined to roughly 2.87 ounces. The deflation of the AI valuation extreme, measured against gold, is underway. This framing is not an academic exercise. Bubbles financed by expanding credit always look most impressive when measured in the same expanding credit. Only when they are measured against an unexpandable monetary asset does their true scale—and their subsequent contraction—become visible.

The Global Footprint of the Reversal

Farber points to a broader geographic pattern that reinforces the sense of a peaking cycle. Stock markets with heavy exposure to AI-related and technology companies have begun to reverse sharply:

• Japan’s Nikkei, after a powerful advance, has retreated from its highs.

• South Korea’s market, driven in significant part by AI-linked heavyweights, has triggered circuit breakers and suffered abrupt declines.

• Indonesia’s index has fallen hard from its recent peak, retracing toward longer-term trend lines last tested during previous crises.

• Taiwan, the critical node in advanced semiconductor manufacturing, has rolled over from its highs.

• Even Israel’s market, which advanced strongly despite years of conflict, has begun to show its first meaningful decline from the recent peak.

These are not isolated corrections in peripheral markets. They are concentrated in the very jurisdictions whose equity indices became most leveraged to the AI narrative. When the leadership of a global theme begins to fail simultaneously across multiple countries, the probability rises that the move reflects something larger than local profit-taking.

Historical Rhyme: Bubbles, Bailouts, and Gold

Farber draws an explicit parallel to the aftermath of the dot-com collapse. The bursting of that bubble did not simply produce lower equity prices. It produced financial stress, policy responses, and, critically, the end of gold’s two-decade bear market. Gold bottomed in 2001 and began the secular advance that, despite cyclical corrections, has defined the subsequent quarter-century. The mechanism is familiar. Leveraged speculation ends. Balance sheets come under pressure. Central banks respond with liquidity and lower real rates. In a system already carrying elevated sovereign debt and strained fiscal positions, each successive round of intervention risks further eroding confidence in the currency itself. Gold, as the monetary asset that stands outside the credit system, has historically been the beneficiary of that erosion. Farber’s current thesis extends this logic. He views the AI bubble’s deflation as a potential catalyst for renewed stress in the financial system, followed by the inevitable policy response. In his framing, that response may prove more difficult to contain than previous episodes because the underlying monetary foundation is already more fragile. The result, in his view, is not merely another cyclical rally in gold but a deeper recognition of gold’s role as money.

Connecting the Threads to Gold Equities

This monetary analysis complements the market-structure observations made earlier. The AI trade is experiencing the classic late-stage sequence: extreme positioning, credit-market skepticism, momentum failure, and forced de-leveraging. At the same time, the valuation excess, when measured in gold, is already contracting. Meanwhile, the gold mining sector continues to operate in a different economic reality. Senior producers are converting prevailing gold prices into substantial free cash flow. Newmont’s recent quarters have demonstrated the scale of that cash generation even after the metal’s own correction from its highs. The equities themselves, however, remain depressed relative to those cash flows—reflecting both the broader risk-off tone and the sector’s chronic under-ownership. If Farber’s sequence proves directionally correct—bubble deflation, financial stress, policy response, renewed focus on monetary integrity—then the capital currently exiting crowded AI exposures will eventually confront a limited set of alternatives that offer both scarcity and cash generation. Physical gold is one. The equities of the companies that mine it, currently available at valuations that price in considerable skepticism, are another.

A Correction Within a Larger Cycle

Farber also notes that gold’s roughly 30 percent decline from its recent peak resembles, in magnitude, the mid-cycle correction of 2011–2015. That earlier drawdown proved temporary within a longer secular advance. Whether the current correction follows a similar path will depend on the interaction between real yields, central-bank demand, geopolitical risk, and the broader confidence in fiat systems.What is already observable is the divergence in treatment: the AI complex is being forced to defend its capital intensity and balance-sheet expansion, while gold producers are quietly harvesting wide margins and returning capital to shareholders. Markets that ignore such divergences for extended periods often correct them with equal intensity once the dominant narrative loses its grip.

Conclusion

The AI bubble, measured in gold ounces, never exceeded the scale of its dot-com predecessor and is now deflating. AI-exposed equity markets across Asia are reversing. Credit markets have begun to question the balance-sheet implications of the infrastructure build-out. Positioning has been forcibly cleaned.These developments do not guarantee an immediate surge in gold or gold equities. They do, however, strengthen the structural case that capital leaving a deflating technology mania will ultimately reinforce the monetary attributes of gold. And they leave the mining sector—already generating record cash flows at current prices—positioned as one of the more coherent destinations for that capital once the rotation gathers force. The previous article argued that the capitulation in AI was the best thing that could happen to gold stocks. Farber’s analysis supplies the monetary logic that explains why. When bubbles measured in credit begin to contract against gold, the metal—and the companies that produce it—tend to reassert their relevance. The process is rarely orderly. It is, however, recurring. And it appears to be underway once again.

Disclaimer: 

This article is for informational and educational purposes only. It does not constitute investment advice, a recommendation to buy or sell any securities, or a prediction of future market performance. Views expressed by independent commentators are their own. Gold, silver, and mining equities involve substantial risk of loss. Readers should conduct their own due diligence and consult qualified professional advisors. Past performance is not indicative of future results.