Following a remarkable five-year bull run that peaked around $5,400 in January 2026, gold has since corrected to approximately $4,600 against the backdrop of conflict in Iran. Analysts at J.P. Morgan argue that gold at $4,600 is expensive with justification, though the primary buyers have yet to enter the market.

On May 7, a research report from J.P. Morgan's Asia-Pacific team, issued after a meeting with John Reade, Chief Market Strategist at the World Gold Council (WGC), presented a straightforward conclusion: by any quantitative model, gold is expensive, but that does not mean it is mispriced.

According to historical models, the fair value of gold should lie between $1,000 and $1,900. The current price represents the largest premium level since 1971. However, being expensive does not equate to being wrong. This significant surge—a 188% return in U.S. dollar terms over five years—has been entirely driven by physical demand from Asia and purchases by emerging market central banks, particularly undisclosed buying from China. In contrast, Western institutional investors have been net sellers over the past few years.

In the near term, gold may continue to consolidate around $4,600, with further downside risk from momentum-driven liquidations. However, medium-term catalysts are real and currently underestimated by the market. J.P. Morgan views gold at $4,600 as expensive insurance. But considering the risks it hedges, the premium might be worth paying.

The bank believes the next major structural wave supporting gold prices—a "century-scale reallocation" by Western pension and insurance funds away from the faltering 60/40 stock-bond portfolio into gold, coupled with a market repricing of "stagflation"—has not yet genuinely begun.

**Models Fail Across the Board: Expensive, But Not Necessarily Wrong**

To understand the current pricing logic for gold, one must first acknowledge a fact: all traditional models point in the same direction—gold is significantly overvalued.

The WGC's GLTER framework estimates long-term gold returns as U.S. CPI plus 2-3%, implying a fair value around $1,900. Furthermore, the stable two-decade relationship (2001-2021) between gold and U.S. real yields has completely broken down since 2021. Had this relationship held, gold's price today would be near $1,000, not $4,600.

So, what are investors actually paying for?

J.P. Morgan agrees with the WGC's framework: the current premium reflects two factors.

First, an insurance premium. Against a backdrop of public questioning of the Federal Reserve's independence, record peacetime fiscal deficits, ongoing Middle East conflicts impacting energy markets, and accelerated de-dollarization by emerging market central banks, gold's high premium is a rational pricing response to an unusually unstable environment. As long as this instability persists, the premium will likely remain.

Second, the models themselves have become ineffective. For the past three decades, government bonds were the hedge for equity portfolios. This assumption is now crumbling. When the correlation between bonds and stocks turns positive during crises—meaning they fall together instead of hedging each other—the value of gold as an alternative diversifier shifts from tactical to structural. The 60/40 portfolio is increasingly seen as outdated by major asset allocators, and the logical beneficiary is gold.

This is the most underestimated core insight: gold currently constitutes only about 2% of global investable portfolios. Decades of portfolio optimization research by the WGC consistently finds that the risk-adjusted optimal allocation is between 5% and 10%.

J.P. Morgan illustrates the scale: global pension and insurance assets total approximately $80 trillion. If the average gold allocation rises from 2% to just 3%, this 1 percentage point reallocation would require roughly 5,000 tonnes of incremental demand—surpassing the annual global gold supply of about 4,500 tonnes. This reallocation has not yet begun.

**Bull Market Built by Asia, West Remains Absent**

The report emphasizes what is perhaps the most counterintuitive fact in the gold market: between 2021 and 2024, Western institutional investors were net sellers of gold via ETFs, even as the price doubled.

While global ETF flows turned positive in 2025 with net inflows of about 800 tonnes, this buying was late, primarily momentum-driven rather than conviction-based, and has partially reversed during the subsequent correction.

The real drivers of this bull market have been a different set of buyers:

* Physical bar and coin demand, primarily from Asian retail investors, now stands at about 1,500 tonnes annually, significantly higher than the historical average of around 1,000 tonnes, representing the most persistent and stable component of demand. * China and India together account for roughly half of global gold demand. China's demand composition has shifted from being predominantly jewelry-based (about 75% of demand pre-2020) to having investment demand exceed jewelry consumption. * Emerging market central bank purchases: between 2021 and 2025, central bank buying averaged about 225 tonnes per quarter, roughly double the pace of the 2016-2020 period.

J.P. Morgan states that central bank buying is a significant demand driver, but official figures severely underestimate the actual scale. An increasing amount of central bank purchases are not reported to the IMF and are classified as "undisclosed" buying.

In Q1 2026, reported central bank demand was weak, but the WGC and Metals Focus estimate total official sector purchases at around 245 tonnes, implying substantial activity from undisclosed buyers.

Meanwhile, J.P. Morgan believes that the reallocation from bonds to gold by Western pensions, insurance companies, and real money portfolios will be the engine for the next structural upswing. This process has not yet started on a large scale.

**Two Emerging Buyer Types Overlooked by Mainstream Models**

J.P. Morgan identifies two emerging sources of gold demand not captured by any standard supply-demand framework, currently hidden within the "over-the-counter and other" residual category used by analysts when data cannot be reconciled.

**Chinese Insurance Companies**

In early 2025, China's top ten insurance companies received regulatory approval to allocate up to 1% of their Assets Under Management to physical gold. Based on their size at the time, 1% of AUM equated to approximately 200 tonnes of gold. These institutions executed some pilot trades in 2025 but largely remained on the sidelines, waiting for a satisfactory price correction that never materialized. Buying began in 2026 but remains well below the approved quota.

The significance of this dynamic lies in the fact that the 1% ceiling is almost certainly just a starting point; a 5% allocation would imply demand scaling up by an order of magnitude. Furthermore, as these institutions begin disclosing their holdings, this demand will be reclassified from the OTC residual to visible investment demand, potentially triggering a repricing of demand data.

**Tether**

Tether, the issuer of the world's largest dollar stablecoin, USDT, with a market capitalization of around $200 billion, is one of the largest global holders of non-sovereign U.S. Treasury bills.

In 2025, Tether purchased approximately 100 tonnes of gold to back its USDT reserves. Tether's management has publicly stated concerns about structural issues with the U.S. dollar and views gold as a more enduring reserve asset.

In an annual market of about 4,500 tonnes, a single-year purchase of 100 tonnes represents a buyer of significant size. This demand is also currently hidden in the OTC residual, invisible to consensus models.

**Next Catalysts: Stagflation Trade and Western 'Century Reallocation'**

J.P. Morgan suggests that while Asia has built a solid floor for gold prices, the future upside will be determined by the West.

First is the stagflation trade. The Middle East conflict points towards higher inflation and lower growth, a macroeconomic environment historically most favorable for gold. Investors have not yet positioned themselves adequately for this scenario. A key signal to watch is when the rolling correlation between gold and equities turns persistently negative, indicating the stagflation trade is formally underway.

Simultaneously, the net managed money position on Comex has been significantly reduced in 2026, even with gold near historic highs. This suggests the current rally is driven by structural buyers rather than tactical funds. Western momentum traders have not yet built positions for a stagflation trade.

Second is the structural reallocation by Western institutions. With gold constituting only about 2% of global investable portfolios versus an optimal 5-10% for risk-adjusted returns, the potential is vast. A mere 1 percentage point increase in the average allocation from 2% to 3% across the $80 trillion in global pension and insurance assets would require about 5,000 tonnes of incremental demand—exceeding total annual global supply. This scale of reallocation has not yet begun.