The last COFER print landed. Dollar share ticked up. Headlines wrote themselves: 'De-dollarization is dead.' 'The greenback strikes back.'
Here's the data. The tick is real. The interpretation is lazy.
I spent the last week cross-referencing the IMF's Currency Composition of Official Foreign Exchange Reserves (COFER) release against central bank gold purchase data from the World Gold Council. The result is a picture that contradicts both the doom-porn and the triumphalist narratives. The short-term uptick in dollar share is a function of valuation math, not policy preference. The long-term slide remains intact. And the only asset class that is pricing this correctly is gold.
Trust the hash, not the headline.
Context: The Methodological Trap
Before we dive into the numbers, we need to address a methodological landmine. COFER data is reported in nominal terms. This is not a minor technical detail; it is the lens through which every conclusion must be filtered.
When the dollar index (DXY) rallies, the value of existing dollar-denominated reserves rises in the total pool. This inflates the dollar's share without a single central bank buying a single US Treasury. It's a mark-to-market artifact. A valuation bump. Not a vote of confidence.
Conversely, when the dollar weakens, the share can decline even if central banks are actively accumulating dollars. The denominator shrinks. This is the core of the 'reserve share paradox.'
From my 2017 ICO ledger audit days, I learned a critical lesson: raw data lies unless you trace the flow beneath it. The same applies here. We cannot look at the headline share number. We have to look at the flow behavior. We have to isolate the valuation effect from the active allocation effect.
Based on my audit experience, the distinction between 'active accumulation' and 'passive valuation' is the single most important variable in interpreting this data. The recent uptick is almost certainly the former, not the latter.
The dollar was strong in 2025. It remains relatively strong. That alone explains the share increase. Meanwhile, the World Gold Council data shows central banks have been net buyers of gold for over a decade. That is not a valuation artifact. That is a deliberate, structural portfolio shift.
The narrative that 'the dollar is back' is a misreading of a quarterly snapshot. The trend is the truth. The snapshot is just noise.
Core: The On-Chain Equivalent of Reserve Rebalancing
Let me translate this into a framework I understand. Think of the global reserve system as a massive, slow-moving liquidity pool. Central banks are the largest whales. Their allocation decisions are akin to a treasury team rebalancing between USDC, USDT, and a basket of other stable assets.
In the crypto world, we track stablecoin flows to understand market conviction. If USDT's market cap rises relative to USDC, is that because of a fundamental shift in trust, or is it because of a specific yield differential or regulatory news? Usually, it's the latter. The same logic applies to reserve currencies.
The data we have is clear on the long-term trend:
- Dollar share of global reserves has declined from over 70% in 2000 to roughly 57-58% today, despite periodic upticks.
- Central bank gold purchases have surged from ~400 tonnes annually in 2010 to over 1,000 tonnes annually in 2022-2024.
- The 'weaponization' of the dollar (sanctions, frozen reserves) post-2022 has accelerated the search for non-sovereign alternatives.
The recent uptick is a cyclical blip within a structural downtrend. It's the equivalent of seeing a short-term bounce in a token's price on high volume and declaring the bear market over. You have to check the wallet clustering. You have to look at the holder distribution. You have to see where the real accumulation is happening.
The real accumulation is happening in gold vaults. Not in US Treasury custodial accounts.
Let's dig into the micro-structure. The IMF's own data, when adjusted for valuation effects, shows that the 'active' dollar diversification has been steady. The uptick we see is a DXY function. In Q3 2025, DXY rallied nearly 4%. That mechanically increases the dollar share by approximately 0.3-0.5 percentage points. The reported uptick was within that range. It's math, not policy.
Meanwhile, look at the gold data. It's not just Russia and China. Central banks in Eastern Europe, the Middle East, and even some Western allies have been quietly adding gold. This is not a coordinated anti-dollar conspiracy. It is a coordinated risk-management response to a unipolar monetary system that has become overtly political.
The 'dual-track' behavior of central banks is the key insight here. On one hand, they hold dollars for transactional efficiency and liquidity. On the other, they buy gold for long-term safety and independence. They are running a barbell strategy. Short-term operational needs are met by USD. Long-term reserve security is met by gold.
This is the same logic as a DeFi treasury holding a large stablecoin position for yield while simultaneously buying ETH for protocol alignment. It's not a contradiction. It's a hedge.
The Fiscal Elephant: Why the Slide Continues
The most critical factor that the mainstream analysis ignores is the US fiscal trajectory. The dollar's reserve status is fundamentally backed by US Treasury bonds. But the demand for those bonds is being eroded by the supply of those bonds.
The US is running a structural deficit of 6-7% of GDP. Interest payments on the national debt now exceed defense spending. This is a death spiral in slow motion. The math is simple: if debt grows faster than GDP indefinitely, the creditworthiness of the issuer erodes.
Central banks are not stupid. They read the same Congressional Budget Office reports I do. They see the debt-to-GDP ratio heading toward 150%+. They see the political gridlock on fiscal reform. They are making a rational decision to diversify away from an asset whose issuer is on an unsustainable path.
Yields don't lie. The 'exorbitant privilege' of the dollar is being priced out by the 'exorbitant duty' of US fiscal profligacy.
The recent uptick in dollar share is a temporary reprieve driven by high interest rates. But high interest rates are the very mechanism that worsens the fiscal situation. The higher the rates go to support the dollar, the faster the debt accumulates, and the sooner the long-term slide accelerates. It's a self-defeating policy.
The market is starting to price this. The gold price rally to all-time highs is not a fear trade. It is a structural repricing of reserve asset safety. Gold has no counterparty risk. It has no political risk. It has no yield, but it also has no liability.
In a world where the issuer of the world's reserve currency is running a Ponzi-like debt scheme, the asset with no issuer becomes the ultimate store of value. This is not a conspiracy theory. It's a balance sheet analysis.
Contrarian Angle: The 'Safe Haven' Misconception
Here's where I diverge from the consensus, both the crypto maximalist and the gold bug camps.
The common narrative is that central banks are buying gold because they are bearish on the US economy or expecting a crash. That is incorrect. They are buying gold because they are bearish on the US fiscal policy, not the economy. The US economy is still the strongest in the G10. That's why the dollar share is ticking up. But the fiscal path is unsustainable, which is why they are hedging with gold.
This is a far more nuanced and dangerous situation. It means the dollar can remain strong in the short term while its reserve status erodes. It means the US can have a strong economy and a weak currency in the long run. It's a divorce of economic performance from monetary credibility.
The second misconception is that this is a zero-sum game where the Yuan or the Euro will automatically replace the dollar. The data doesn't support that. The Euro's share is stagnant. The Yuan's share is growing but from a tiny base (around 2.5-3%). The real beneficiary of de-dollarization is not another fiat currency. It is gold.
This is the 'no-alternative' problem. Central banks want to diversify away from the dollar, but they don't trust each other's currencies either. So they default to the one asset that has been money for 5,000 years. Gold is the neutral, apolitical, sovereign-free option.
This creates a structural bid under gold that is independent of the economic cycle. It is a policy-driven bid. And policy-driven bids are sticky. They don't reverse quickly. If a central bank starts buying gold, it's very hard to stop because the political rationale (hedging against US sanctions) doesn't disappear when the price goes up. It only strengthens.
Therefore, the recent dollar uptick is not a signal to fade gold. It is a signal that the cycle is still intact. The dollar is strong because of rates. Gold is strong because of risk. Both can be true. The market is paying for safety with one hand and for yield with the other.
The On-Chain Corollary: Stablecoin Dominance and the BTC Signal
Let me bring this back to my native habitat: on-chain data. We can see this same dynamic playing out in the crypto market structure.
Look at the flows between USDT, USDC, and BTC. When there is a risk-off event, we see stablecoin inflows to exchanges. When there is a risk-on event, we see stablecoin outflows to buy BTC or ETH. The stablecoin is the transactional layer. BTC is the reserve asset.
This is exactly what central banks are doing. They are holding USD (stablecoins) for transactions and gold (BTC) for reserves. The recent ETF flows show a similar pattern. Institutional money is buying BTC as a long-term treasury reserve asset, not as a short-term trading vehicle.
The correlation between BTC and gold is not accidental. They are both responding to the same macro signal: the debasement of fiat currencies. They are both non-sovereign, decentralized (in their own ways), and supply-capped.
BTC is the 'gold 2.0' for the digital age. It has the same properties but with a verifiable, auditable supply. For a data detective like me, the ability to verify the exact supply schedule on a public ledger is a massive advantage over gold, where official holdings can be opaque.
Chaos is just data waiting for the right query. The data is clear: the world is moving to a two-asset reserve system. One for liquidity (USD), one for safety (Gold/BTC). The current COFER uptick is the liquidity leg. The gold purchases are the safety leg. Don't confuse the two.
Takeaway: The Signal to Track
The next COFER print is due in June. The World Gold Council data comes out monthly. I will be watching one specific metric: the velocity of gold purchases.
If we see another month of 80+ tonnes of central bank buying while DXY holds above 100, it confirms the structural trend. It confirms that central banks are using the dollar's strength to buy gold at better prices. It is a distribution event for the dollar and an accumulation event for gold.
If, however, we see gold purchases slow down significantly, it might indicate that the high price is deterring buyers. That would be a warning sign for the gold bull thesis.
My base case is that the buying continues. The geopolitical incentives are too strong. The fiscal trajectory is too clear. The dollar's short-term bounce is a gift to long-term diversifiers.
Stop guessing. Start querying.
The blocks remember. The vaults do too.