The result type is the same data type as that of the numeric input argument.
If the input argument is defined as a DECIMAL/NUMERIC with a precision less than 38, the return DECIMAL/NUMERIC value will have its precision increased by 1. For example, if DECIMAL(6,4) is passed in, it will be increased and returned as a DECIMAL(7,4). If the precision is 38, the scale will be reduced by 1 unless the scale is 0. For example, a DECIMAL(38,38) results in a return data type of DECIMAL(38,37).
For information on default data type formats, see Teradata Vantage™ - Data Types and Literals, B035-1143.