Published on

The Curious Case of Gold - Part 3: What Now?

Authors

Gold is currently trading at $5,172 as I write this - March 7, 2026. A few weeks after the most violent precious metals crash since 1980, the metal has climbed back above $5,000 with a composure that would seem implausible if you'd watched what happened on January 31.

Silver is at $80.62. Still well below its $121 peak. The gold-to-silver ratio sits at 62.8 - elevated, with silver lagging, suggesting that the speculative premium it carried in January has not fully returned.

So we're back. Or something close to it. Which raises the obvious question: what now?

This is the third and final part of this series. Part 1 covered the rise and Part 2 covered the crash. This one tries to do something harder: figure out what any of it means, what to actually watch going forward, and whether gold at $5,000 is a bubble nearing its end or a structural repricing that still has further to run.

I'll give you my honest view. But first, let me show you where the smart money stands.

Table of contents

Where the banks stand

After the crash, you might have expected Wall Street to quietly lower its targets and move on. The opposite happened.

Major Bank Gold Price Targets: End-2026 (USD/oz)

JP Morgan raised its year-end 2026 target to $6,300 on February 2 - the Monday immediately after the crash. Wells Fargo lifted its target to $6,100–6,300. UBS's base case is $5,400, with a bull scenario at $7,200. Goldman Sachs sits at $4,900. Deutsche Bank at $4,950. Bank of America and ANZ at $5,000. The average of the major bank targets is around $5,180.

These are not perma-bull boutique gold dealers. These are the same institutions that manage trillions of dollars of assets and have sophisticated risk management teams. When JP Morgan raises its gold target after the biggest crash in forty years, it is because its model - driven by central bank buying volumes, ETF flow projections, and dollar trajectory assumptions - still produces a bullish output.

The underlying rationale, from JP Morgan's own research, is instructive: they project roughly 585 tonnes of quarterly demand from central banks and investors combined throughout 2026. Their model says around 350 tonnes per quarter is the threshold for gold prices to rise. Above that, every additional 100 tonnes is worth approximately 2% more per quarter. At 585 tonnes, the direction of travel, by their own arithmetic, is still up.

But I want to be careful here. The banks were also bullish before the crash. In late January, some of those same targets looked like they might be hit in weeks rather than by year-end. They were wrong about timing and trajectory even if they are right about direction. Target and thesis are different things.

The bubble case

Let me steelman the bear case properly, because it deserves steelmanning.

The single most important thing that happened in January 2026 was that precious metals markets started behaving like meme stocks. The silver RSI touched 93.86. The gold-to-silver ratio compressed to 46 - its lowest since 2011. Silver gained 68% in a single month. These are not normal market dynamics. These are what happens when momentum and leverage interact without friction, when retail FOMO collides with institutional trend-following algorithms, and when a fundamentally sound thesis gets used to justify prices that have temporarily separated from the underlying fundamentals.

The bubble case says: gold at $5,000 still contains a meaningful speculative premium above its fair value. Even accepting all the structural drivers - de-dollarisation, central bank buying, fiscal deficits, dollar weakness - gold's intrinsic value is lower than its market price. The January peak was not $5,600 worth of structural repricing. It was $4,000 worth of structural repricing and $1,600 worth of leveraged momentum.

There is also a scenario where the Warsh appointment turns out to be genuinely consequential. If the next Fed Chair oversees a period of tighter monetary policy than markets expect - if real yields stay elevated, if the dollar stabilises or recovers, if the "debasement trade" thesis proves premature - then the fundamental case for gold weakens. Not collapses, but weakens. A 15–20% correction from current levels ($5,062) would bring gold to $4,050–4,300. That is painful if you bought at $5,500.

And there's a geopolitical wildcard that most gold bulls underweight: resolution. If the Russia-Ukraine war reaches a durable settlement. If US-Iran nuclear talks succeed. If US-China trade tensions de-escalate meaningfully. Each of those would remove a piece of the risk premium that gold currently carries.

The bubble case is not that gold goes back to $2,000. It is that gold consolidates in the $4,000–4,500 range for a year or two while the structural forces slowly reassert themselves. That is the most bearish credible scenario.

The repricing case

The repricing case is simpler and, I think, more compelling.

Gold at $273 in 2000. Gold at $1,050 in 2015. Gold at $1,800 in 2020. Gold at $2,600 in January 2025. Gold at $5,062 today. Each of those levels looked extreme when first breached. Each became the new normal.

What is different about this particular cycle is that the source of demand has changed permanently. In the 2000s, gold's rally was driven primarily by retail investors and commodity funds. In the 2010s, ETF demand was the dominant force. In the current cycle, the dominant marginal buyer is sovereign central banks - purchasing gold as a geopolitical act, not a financial trade. Central banks don't care about quarterly earnings or valuation multiples. They are buying gold because they have decided, as a matter of state policy, that they want fewer dollars and more physical metal. That decision does not reverse on a bad RSI reading.

Consider the scale: since 2022, central banks have absorbed 3,232 tonnes of gold. At current prices, that is roughly $530 billion of structural demand that did not exist in prior cycles. The physical gold market is not big enough to absorb that volume without moving prices - and it has moved them.

The repricing case also notes something subtle: the IMF's own data shows that the dollar's share of global foreign exchange reserves fell from 64.69% in early 2017 to 56.92% in Q3 2025. That is a 7.77 percentage point decline over eight years. In COFER terms, the researchers at Robin J. Brooks' Substack called the Q2 2025 quarterly drop "a 2.5 standard deviation move" - as close as that data ever gets to an earthquake. This is not a conspiracy theory or a BRICS press release. This is the IMF's own survey of what 149 central banks are doing with their reserves. The direction is clear and it has been clear for years.

If the dollar's reserve share continues declining - and the structural forces (US fiscal position, geopolitical fractures, BRICS expansion, weaponisation of sanctions) all push in that direction - then gold's role as the alternative reserve asset is not a narrative. It is a function.

The repricing case does not require everything to go right for gold. It only requires that the structural forces that have been building for fifteen years continue to build. That is the lower bar.

The dollar question

Everything in this story ultimately comes back to one question: is the dollar genuinely in structural decline as a reserve currency, or are recent trends a temporary wobble in a system that remains intact?

This is a genuinely hard question, and I want to resist the temptation to give you a confident answer when the evidence is mixed.

On one hand: the dollar's reserve share has fallen from over 70% in 2001 to below 57% today. The BRICS "Unit" currency pilot is live and pegged to gold. China holds $683 billion in US Treasuries, down from $1.3 trillion in 2013. Russia's reserves were frozen by executive action in 2022, demonstrating that dollar reserves carry genuine political risk. The US national debt is approaching $36 trillion with annual interest payments over $1 trillion. These are not small things.

On the other hand: the IMF's exchange-rate-adjusted COFER data shows that much of the apparent dollar decline in 2025 was mechanical - driven by the dollar's own weakness rather than by active portfolio shifts. When you strip out the valuation effect, the actual reduction in dollar allocations was around 0.12 percentage points in Q2 - far less dramatic than the headline number. The renminbi, frequently cited as the main beneficiary of de-dollarisation, actually holds just 1.93% of global reserves - unchanged for years. There is no credible alternative to the dollar that is deep, liquid, and politically neutral enough to replace it at scale. Even gold, which is nobody's liability and can't be devalued, is not a settlement currency. You can't pay for oil with gold bars.

So where does that leave us? Probably here: the dollar is not going to be dethroned in the next five years. But it is going to be held in declining proportions by more and more of the world's central banks, replaced by a basket of alternatives that includes other currencies, non-traditional holdings, and - increasingly - gold. That slow, structural shift does not require a crisis or a collapse. It is already happening. And as long as it continues, it is a tailwind for gold.

What to actually watch

If you want to track this story as it develops, here are the things that will actually move the needle - not the things that get the most airtime.

Central bank buying pace. The single most important structural number. The World Gold Council publishes quarterly demand data with a lag. The question for 2026 is whether purchases hold above 750 tonnes annually - well above the 2010–2021 average of 473, and a sign that the structural shift is intact. If buying falls below 500 tonnes, the price floor weakens.

The Fed's actual path, not its projections. Markets currently price in two rate cuts for 2026. If economic data weakens faster than expected - rising bankruptcy rates, cooling employment - the Fed may cut three or four times instead. Every additional cut is incrementally bullish for gold. If the economy stays resilient and Warsh takes a hawkish line, fewer cuts are possible. Watch the real yield on ten-year TIPS: when it falls, gold tends to rise; when it rises, gold faces headwinds.

The dollar index (DXY). Gold is priced in dollars. When the DXY falls, gold rises in dollar terms - sometimes regardless of underlying fundamentals. The DXY fell more than 9% in 2025, its largest annual drop in years. If it stabilises or recovers in 2026, that removes one tail tailwind from the gold price.

Chinese ETF and futures positioning. The January crash was partly a Chinese leverage story. Watch the net long position in Chinese gold and silver futures - when it is at extreme highs, the market is vulnerable to a fast unwind. When it is normalised, the underlying physical demand story reasserts.

US fiscal trajectory. The US government currently spends more on interest payments than on defence. If that ratio worsens - if deficits expand, if bond markets start pricing in higher long-term yields to compensate for default risk - gold's role as a fiscal hedge becomes more acute. Watch the 30-year Treasury yield and the CDS spread on US sovereign debt (still tiny, but no longer zero).

Geopolitical de-escalation. This is the most underpriced risk for gold bulls. A genuine resolution to the Ukraine conflict, a US-Iran nuclear deal, or a meaningful US-China trade de-escalation would each remove meaningful geopolitical risk premium from gold. None of those outcomes is likely in 2026, but they are not impossible.

Silver specifically

Silver's position as we sit here in February 2026 is more complicated than gold's.

The fundamental supply case is strong. Silver has run a structural supply deficit for multiple years - meaning industrial and investor demand has consistently exceeded mine supply and recycling. That deficit is not closing: solar photovoltaic installation is accelerating globally, electric vehicle production continues growing, and AI data centre infrastructure demands silver-intensive components. Each solar panel uses roughly 20 grams of silver. The world is installing panels at an accelerating rate.

But silver's January 2026 price action revealed how much of its premium had become speculative rather than fundamental. The RSI of 93.86, the comparison to GameStop, the 68% gain in a single month - none of that reflects physical supply and demand. It reflects a momentum trade that got extremely crowded and then unwound violently.

The question for silver in 2026 is whether it can re-establish itself as a fundamentals story rather than a momentum story. The conditions for that are: physical demand staying strong, the speculative overhang continuing to clear, and gold maintaining its level (which gives silver a floor through the gold-silver ratio). Several analysts have identified $100 as the next major target - achievable, but probably not before a meaningful period of consolidation in the $75–90 range.

One thing is worth stating plainly: at $80/oz, silver is still extraordinarily expensive by historical standards. As recently as early 2025 it was at $30. The structural case remains intact, but anyone buying silver at these levels is accepting significant volatility risk in exchange for exposure to a genuine supply-demand story.

An honest take

I'm not a financial advisor, and none of this is investment advice. But you didn't read 10,000 words across three parts of this series for me to end with a shrug.

Here is what I actually think.

Gold at $5,062 is not a bubble in the sense that requires a crash back to $2,500 to resolve. The structural forces are real. Central banks buying 800–1,000 tonnes a year is not a meme. The US fiscal position is not going to improve in any scenario I can construct. De-dollarisation is slow but it is happening. Gold's role as the world's neutral, counterparty-free reserve asset is not a narrative - it is a function that has existed for five thousand years and has never been permanently displaced.

At the same time, $5,062 after a 21% crash from $5,600 is not obviously cheap. The January peak contained speculative froth. That froth has partly deflated but not completely. Gold could easily consolidate at $4,500–5,000 for most of 2026 while the leverage overhang continues to clear and the market rebuilds on a more sustainable base.

What I am confident about is this: the direction of travel is not going to reverse because of one Fed Chair appointment. The forces that produced the 2025 rally - fifteen years of de-dollarisation, central bank gold buying at historic rates, US fiscal deterioration, geopolitical fracturing, and declining trust in Western financial institutions - are not going to resolve themselves in 2026. They are structural, they are slow-moving, and they are accumulating.

If anything, the January crash illustrated that this is not a smooth escalator to the moon. It is a structurally rising asset class with violent episodes of speculative excess and brutal corrections. The investors who did well through the 2025 rally were the ones who bought in 2022 or 2023 and held. The investors who got destroyed in January were the ones who piled in at $5,400 with leverage in the final weeks of a parabola.

The lesson is not "gold is a bubble." The lesson is that timing parabolas is a mug's game, and the structural thesis does not require you to be at the peak.

Gold as mirror

There is one final thing worth saying, and it is not really about price targets.

Gold does not produce anything. It does not innovate. It does not cure diseases or build infrastructure or create employment. It just sits there, in vaults, gleaming. And yet it has been money - or money-adjacent - for every coherent civilisation in recorded human history. No government has permanently suppressed its price. No paper money has outlasted it.

When gold rises, it is usually telling you something unpleasant about the world. In 1979, it was telling you about inflation, OPEC, and the erosion of post-war certainties. In 2008, it was telling you about financial system fragility. In 2020, it was telling you about pandemic-era money printing. In 2025, it was telling you about fiscal excess, geopolitical fracture, and a slow global loss of confidence in the dollar-centred financial architecture.

What gold has told us, over the last two years of extraordinary price movement, is that a significant portion of the world - not just gold bugs, but central banks and institutional investors who have no ideological skin in the game - has decided that the existing monetary order is less reliable than it was. That physical metal, which you can hold and store and move without anyone's permission, is worth more than it was when trust in the system was higher.

That is not a comfortable message. It is also not a prediction of collapse - most of the dystopian gold forecasts, from $20,000 to $100,000, depend on scenarios that would destroy the value of gold along with everything else. But it is a message worth taking seriously.

Gold at $5,000 is not the world ending. It is the world hedging.

This is the final part of the series. Part 1 covered the rise. Part 2 covered the crash. If you have thoughts, I'd love to hear them.

Sources: JP Morgan Global Commodities Research; Goldman Sachs Metals Research; UBS Wealth Management; Wells Fargo Investment Institute; IMF COFER Data (Q3 2025); Federal Reserve International Role of the USD (2025 Edition); World Gold Council Gold Demand Trends Full Year 2025; Robin J. Brooks Substack; State Street Global Advisors Gold 2026 Outlook; SSGA; Metals Focus 2026 Outlook; FXStreet; Fast Company; CNBC Markets.