Gold’s Old Pricing Rules Are Broken: What Drives It Now

Gold has more than doubled in three years, climbing from roughly $1,970 per ounce in mid-2023 to above $4,200 by late September 2026. It did this during one of the most aggressive real-yield environments in decades, the exact conditions that should have crushed a non-yielding asset.

That is the paradox worth sitting with. The rules investors have used to price gold for 40 years have quietly stopped working, and the forces that replaced them look structural rather than cyclical. Two analysts saw this early: Andrew McGuire, a gold market analyst, and Danielle DiMartino Booth, macro strategist and CEO of QI Research, laid out the case in July 2023 when spot gold sat near $1,970. The years since have validated the direction of their argument.

This piece gives you a framework for separating the durable drivers of the current gold market from the ones carrying counterparty and policy risk. The goal is a more informed view of where physical precious metals fit in a portfolio built for the monetary landscape actually in front of you, not the one the old models assume.

Why gold stopped responding to real yields

For four decades, the pricing logic was simple. Real yields (the return on government bonds after inflation) moved inversely to gold. When real yields rose, holding a metal that pays no interest became more expensive in opportunity-cost terms, and gold fell. When real yields dropped, gold rallied. This inverse relationship was the anchor of nearly every institutional gold valuation model.

The inverse relationship between real rates and gold held so consistently for so long that most institutional models treated it as a near-law; the deeper mechanics of why real rates and gold diverged in this cycle reveal which structural forces are durable and which are not.

That anchor has come loose. Gold surged from approximately $1,970/oz to $4,259/oz by 27 September 2026, more than doubling while real yields stayed high. Three mechanisms overrode the old signal.

1. Geopolitical safe-haven demand decoupled from economic variables. The World Gold Council’s 2026 outlook explicitly names tense geopolitics as a primary driver of gold’s performance. Buyers are hedging systemic and political risk, not calibrating against bond yields. This lends institutional weight to the dynamic McGuire and DiMartino Booth flagged in 2023.
2. Central-bank buying is strategic, not yield-reactive. Official-sector purchases are driven by currency diversification and sanctions-risk concerns that do not respond to short-term rate moves. When a central bank buys gold to reduce dollar exposure, the real yield on Treasuries is beside the point.
3. Investment demand reached a scale that swamps the yield signal entirely. In 2025, total investment demand hit 2,175 tonnes, up 84% year-on-year, overtaking jewellery as the single largest demand category for the first time. ETF inflows alone reached 801 tonnes.

World Gold Council demand data confirms the scale of the structural shift: total 2025 investment demand reached 2,175 tonnes, up 84% year-on-year, with ETF inflows alone accounting for 801 tonnes, making investment the single largest demand category for the first time.

That 84% surge is the number that reframes everything. The yield-to-gold relationship has not merely weakened; it has been structurally overwhelmed by buyers whose motivation has nothing to do with the opportunity cost of holding a non-yielding asset.

Underpinning this is a repricing of credit risk that McGuire and DiMartino Booth used to anchor their thesis.

The credit-risk signal SoftBank issued high-yield debt at 9.875%, a level the original analysis cited as evidence that credit markets have begun pricing risk at intensities unseen in generations. As counterparty risk in financial instruments climbs, the appeal of an asset with no counterparty at all strengthens.

For investors still pricing gold against real yields, the takeaway is direct: the framework itself has changed. Understanding why is the prerequisite to judging whether current prices reflect genuine repricing or speculative froth.

The gap between paper claims and physical metal

Here is a ratio that forces a recalculation. Gold’s paper markets trade 600-800 tonnes daily against physical delivery of only 3-5 tonnes. For every tonne of metal that actually changes hands, hundreds of tonnes of claims trade above it.

According to Andrew McGuire’s structural data from July 2023, London’s gold market held an estimated 9,000 tonnes including ETF-owned metal, with only around 1,200 tonnes available as genuine free float for trading. The rest is spoken for. That leverage of paper claims over deliverable metal is the core tension in the gold market.

The implication follows from the number. Paper gold markets are primarily a financial instrument market, not a metal market. Most investors participating through futures or unallocated accounts hold a financial claim on a counterparty, not bullion in a vault.

A concrete illustration of how that plays out: the original analysis cited Jane Street’s large silver position, which was fully liquidated and contributed to a sharp price decline. Speculative paper-market participants can be forced to sell during margin-call events, creating openings for physical holders who are not exposed to that mechanism. Asian physical markets, by contrast, showed persistent premiums at the time, averaging around $10/oz for gold and roughly 13% for silver.

The paper gold system rests on a structural assumption that most participants will never simultaneously demand physical delivery, but as sovereign buyers accumulate allocated metal and ETF inflows consume deliverable float, that assumption is being quietly stress-tested in real time.

One important caveat. The London free-float and daily-delivery figures above are from July 2023 and may have shifted materially since. Treat them as an illustrative baseline for the market’s structure rather than current values; updated figures for London vault free float were not recoverable from available research.

What Basel III changes for gold investors

Basel III, the global bank capital regime, draws a hard line between allocated physical gold and unallocated gold claims. Allocated metal held on a bank’s balance sheet carries a zero risk weighting, treated as favourably as cash. Unallocated positions do not.

The Net Stable Funding Ratio (NSFR) provisions are the pressure point. They make it more expensive for banks to fund longer-dated unallocated gold positions with short-term liabilities, raising the cost of the traditional bullion-banking model. Independent analysts argue this constrains banks’ capacity to build large unallocated books.

Here is why that matters to you. The regulation nudges institutions toward physically backed, fully collateralised products and away from synthetic exposure. That is a structural tailwind for physical gold ownership, written into the rulebook, not a passing cyclical swing.

Central banks as structural buyers: what the data confirms and what it does not

The scale of official-sector accumulation settles one question decisively. Central banks bought a record 1,092.4 tonnes in 2024, then 863.3 tonnes in 2025. Even after that pullback, 2025 ran at roughly double the pre-2022 annual norm of 400-500 tonnes. The 2023 thesis that sovereigns were re-rating gold as a strategic reserve asset is confirmed by the numbers that followed.

The named buyer list tells a specific story about motivation.

• National Bank of Poland: 102 tonnes (largest buyer, second consecutive year)
• National Bank of Kazakhstan: 57 tonnes
• Central Bank of Brazil: 43 tonnes
• State Oil Fund of Azerbaijan: 38 tonnes
• Central Bank of Turkey: 27 tonnes
• People’s Bank of China: 27 tonnes
• Czech National Bank: 20 tonnes

According to World Gold Council survey data, the motivations cluster around de-dollarisation, sanctions-risk hedging, and inflation protection. These are structural concerns, independent of where real yields sit in any given quarter. The strongest demand is coming from sovereign entities explicitly managing their exposure to dollar-denominated reserve systems, and that motivation does not fade when rates rise.

Reserve diversification motivations vary significantly across the buyer list: Poland’s accumulation reflects NATO-era balance-sheet prudence, while Brazil and Azerbaijan are managing commodity-export revenue recycling alongside dollar-exposure reduction, producing different sensitivity profiles to any future normalisation of geopolitical risk.

The original analysis illustrated the underlying logic with a single object: a British pound coin from 1930, once redeemable at face value, now estimated to be worth around £900 in gold-equivalent purchasing power. That is the fiat-erosion thesis that drives sovereign accumulation, expressed in one coin.

Now the genuine qualification. This is not a structurally accelerating trend.

The 21% drop from 2024 to 2025, and the WGC’s projected further easing to around 850 tonnes in 2026, means the correct read is precise rather than simply bullish. Directionally the thesis holds; the magnitude is moderating. Central-bank buying provides a floor of price-insensitive demand, but that floor is settling, not rising.

Sovereignty, cash yields, and the structural case for physical ownership

The logic that drives central banks scales down to the individual portfolio. Sovereigns buy gold to protect national purchasing power against currency debasement and geopolitical risk. An individual holding physical metal is applying the same reasoning at a personal level: sovereignty over an asset that carries no counterparty.

The purchasing-power anchor A British pound coin from 1930, once redeemable at par, is estimated to be worth roughly £900 in gold-equivalent value today. That is the scale of a century of fiat erosion set against gold’s preservation of value, as cited in the original analysis.

Three sources of demand look durable rather than cyclical.

1. Sovereign accumulation driven by de-dollarisation and sanctions-risk hedging that does not respond to yield levels.
2. Credit-risk repricing that raises the appeal of a counterparty-free asset as financial-instrument risk climbs.
3. The Basel III regulatory shift tilting institutional preference toward allocated, physically backed gold.

The cash-yield counterpoint

There is a genuine argument against gold right now, and it lives in money-market funds. U.S. money-market fund assets stood at $7.94 trillion for the week ended 23 September 2026, according to the Investment Company Institute, with taxable funds yielding around 3.9% at year-end 2025. In a high-rate world, that is a rational place to hold cash.

The catch is policy risk. Those yields have already compressed from roughly 5% in July 2023 to 3.9% by end-2025. If rates fall rapidly, as they did during the COVID-19 response, that yield can collapse toward zero while inflation and currency risk remain fully intact. That trajectory is precisely the sensitivity the original analysis warned about, and it argues for treating cash and gold as complements rather than substitutes.

Distinguishing the structural thesis from the product pitch

Retail physical demand is not niche behaviour. Bar-and-coin demand hit 420 tonnes in Q4 2025 alone, a 12-year quarterly high, signalling that individual and institutional physical buying is genuinely elevated.

But there is a distinction worth holding firmly. The original analysis treated major investment banks recommending gold to retail clients as a separate matter from the long-term structural case for holding physical metal. A brokerage sells products and narratives; the fundamental thesis stands on its own.

The value of understanding the structural argument is that it lets you interrogate any recommendation. When a bank pitches a gold product, the question becomes concrete: does this deliver physical ownership, or paper exposure sold as its equivalent? Those are not the same asset, and only one of them removes counterparty risk.

Gold investment vehicle selection carries structural consequences that go beyond fee differentials: the choice between allocated physical, gold ETFs, and unallocated accounts determines whether a portfolio holds the asset itself or a financial instrument whose value is contingent on a counterparty chain that may behave differently from spot gold during a stress event.

Where the structural gold thesis goes from here

At over $4,200 per ounce, gold is no longer pricing in a modest structural premium. It is pricing in a sustained structural realignment. That makes the sensible posture one of watching specific variables rather than settling on a verdict.

Thesis confirmation signals:

• Sustained central-bank buying above the pre-2022 norm of 400-500 tonnes
• Continued dominance of investment demand over jewellery demand
• Persistent physical premiums in Asian markets
• Basel III-driven migration from unallocated to allocated gold

Thesis deterioration signals:

• Rapid normalisation of geopolitical risk
• A sustained return of the gold-to-real-yield inverse correlation
• Central-bank purchases falling meaningfully below the pre-2022 norm
• A reversal of ETF inflows

The gold-real-yield relationship deserves particular attention. If real yields stay elevated while the non-yield drivers normalise, gold could face a period of price vulnerability. The decoupling is not necessarily permanent, and the original framework acknowledged this implicitly by focusing on structural demand rather than price prediction.

The scale of the repricing The LBMA PM gold price set 53 all-time highs during 2025, with full-year demand including OTC reaching a record of approximately 5,002 tonnes.

The near-term benchmark to watch is the WGC’s projected 850 tonnes of central-bank demand for 2026, measured against that 2025 record baseline. The structural shift is identifiable and partially confirmed by data. The disciplined move is to monitor the variables above as the numbers arrive, rather than treating either the bullish or bearish narrative as settled.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions.

Past performance does not guarantee future results. Financial projections are subject to market conditions and various risk factors. These statements are speculative and subject to change based on market developments.