The balance sheet nobody elected
Consider an emerging economy that exports copper and imports fuel, with a central bank sitting on roughly $40 billion in foreign exchange reserves. How that pile is divided up comes down to some fairly hard-nosed practical considerations: the country's external debt obligations, how many months of imports the reserves need to cover, and which assets can actually be sold in a hurry when things go wrong. That last point tends to matter most. A reserve that looks good on paper but finds no buyers in a crisis is, for intervention purposes, close to worthless.
Foreign exchange reserves are assets held by a central bank in currencies other than its own, kept primarily for market intervention. When the exchange rate comes under pressure, the central bank sells foreign currency to buy its own, providing support. Liquidity is not a nice-to-have in this context. It is the whole point.
The dollar's structural grip
The US dollar has dominated global reserves for decades, typically accounting for somewhere between 55 and 65 percent of reported holdings. Inertia plays a role, but it is not the main story. The US Treasury market is simply the deepest and most liquid bond market on earth. A central bank can sell a very large position in Treasuries without moving the price against itself in any meaningful way, which is a genuinely rare property. From there, a self-reinforcing logic takes hold: commodities get invoiced in dollars, importers need dollars to pay for them, and governments want dollar reserves on hand partly just to signal credibility to the markets watching them.
The euro is the second-largest reserve currency, accounting for roughly 18 to 22 percent of global holdings. It benefits from deep European sovereign bond markets and the eurozone's considerable trade footprint. There is, however, a complication. The eurozone has no single Treasury, so a reserve manager holding euro assets must decide whose debt to buy, an extra layer of judgment that simply does not arise with dollar holdings.
The four questions every reserve manager asks
When a central bank allocates new reserves, four questions tend to structure the conversation.
First, what currency is the country's external debt denominated in? This sets a floor: the central bank needs enough of that currency to service obligations without being forced into the market at the worst possible moment. Second, what currency do the country's main trading partners use? The answer shapes how much of that currency is useful to hold as a buffer against trade disruptions. Third, how does the reserve currency move relative to the domestic one? A currency that tends to appreciate precisely when the domestic currency is falling offers genuine insurance rather than just a different label on the same risk. Fourth, what yield and duration options does the reserve currency offer? A well-developed yield curve lets the reserve manager earn a return while keeping enough liquidity in reserve, a balance that matters more than it might sound when you are managing tens of billions of dollars.
None of these questions has a clean, permanent answer. They get revisited as trade patterns shift and as debt structures evolve.
What people get wrong about gold and the IMF's SDR
Two reserve assets attract a disproportionate amount of confused commentary. Gold remains on central bank balance sheets around the world, but it is not a currency, generates no yield, and bears no direct relationship to trade flows. Central banks hold it as a hedge against systemic currency crises, not as something they expect to deploy in routine intervention. The Special Drawing Right, the IMF's composite unit based on a basket of five currencies, is not directly spendable in currency markets at all. It has to be converted into one of its constituent currencies before it can actually be used, which limits its practical role somewhat (though it does serve as a useful unit of account in official transactions).
The slow drift toward diversification
Reserve compositions do shift. They just do it slowly. The dollar's share has fallen from above 70 percent in the late 1990s to the 55-to-65 percent range seen more recently, with the gap filled partly by the euro, partly by the renminbi, and partly by what analysts call non-traditional reserve currencies, the Australian and Canadian dollars chief among them.
The renminbi's rise is real but modest so far. China's capital account remains only partially open, its central bank does not operate with the kind of independent inflation-fighting mandate that reserve managers find reassuring, and its sovereign bond market, while growing, is not yet in the same league as the Treasury market for liquidity. Whether that changes is one of the more consequential open questions in international finance.
Ultimately, a central bank's reserve portfolio is a statement about which stress scenario it considers most likely and which currencies it trusts to hold their usefulness when that scenario arrives. Watching where the actual flows go tends to be more informative than listening to the official commentary about them.