XAG

The code for silver under ISO 4217, the standard that assigns a three letter code to each currency. Codes beginning with X are reserved for units that are not any country's money, and the precious metals sit in that range: XAG for silver, XAU for gold, with platinum and palladium alongside them. The two letters after the X follow the Latin name, argentum. The unit the code names is one troy ounce, the weight precious metals are measured in, which is heavier than the ounce used for ordinary goods and is not interchangeable with it.

Because the code occupies a currency's position, silver is written as a pair, XAG against the US dollar being the common form, and the number is read as any pair is read: how much of the second unit one of the first costs, here dollars per troy ounce. The dealing conventions, though, follow the metal rather than the currency market. Contract sizes are stated in ounces rather than in units of a base currency, silver is conventionally quoted to a finer decimal than gold because its price per ounce is so much smaller, and the value of one increment is the ounces the contract covers multiplied by the size of that increment.

Carrying a currency code does not make silver a currency, and the difference surfaces in financing. No policy rate is set on an ounce of metal, so the financing on a metal position held past the daily cut off derives from the dollar interest rate on one side and the metal's own lease rate on the other rather than from two central banks. A rate decision therefore reaches a metal position through a different and slower route than it reaches a currency pair.

What silver actually is remains genuinely disputed. A large share of the demand for it is industrial rather than monetary, which places it in two categories at once: it has tracked gold closely through episodes driven by real interest rates and the dollar, and tracked industrial activity when that was what moved. The ratio of the gold price to the silver price is watched as a convention for describing that tension, and it describes where the two have been rather than measuring where either is going. The practical consequence is that silver's realised volatility has generally run above gold's, and a correlation between the two computed over a long window can be absent in the window that matters.