Operands that are not represented in format descriptions by value mnemonics are represented by one or more uppercase character strings that show exactly how the operand should be written. These character strings are called coded values, and the operands for which they are written are called coded value operands.

A coded value operand results in either a specific value parameter or a specific sequence of executable instructions.

If a positional operand can be written as any one of two or more coded values, all possible coded values are listed in a format description and are separated by vertical strokes indicating that only one of the values is to be used.