DECIMAL vs FLOAT vs MONEY

DECIMAL, FLOAT, and MONEY in SQL Server

The main difference is how accurately numbers are stored.

1. DECIMAL(p,s)

Stores exact numeric values with configurable precision and scale.

• p = total number of digits

• s = digits after the decimal point

DECIMAL(10,2)

This allows:

• 10 total digits

• 2 decimal digits

• Maximum positive value: 99999999.99

Example:

DECLARE @Price DECIMAL(10,2) = 1250.75;

Use it for:

• Prices

• Tax calculations

• Account balances

• Interest rates

• Financial reports

NUMERIC and DECIMAL are synonyms in SQL Server.

2. FLOAT(n)

Stores an approximate floating-point number.

DECLARE @Measurement FLOAT = 10.123456789;

It supports very large or very small values but may introduce tiny rounding differences.

DECLARE @Value FLOAT = 0.1 + 0.2;

SELECT @Value;

The internal value may be approximately:

0.30000000000000004

Use it for:

• Scientific calculations

• Measurements

• Engineering data

• Statistical calculations

Avoid it for:

• Currency

• Exact equality comparisons

• Financial calculations

FLOAT(1–24) uses approximately 7 digits of precision, while FLOAT(25–53) uses approximately 15 digits. FLOAT without a value defaults to FLOAT(53).

3. MONEY

Stores an exact monetary value with exactly four decimal places.

DECLARE @Amount MONEY = 1250.7567;

Range:

-922,337,203,685,477.5808
to
 922,337,203,685,477.5807

Use it for simple monetary values when four decimal places are sufficient.

However, MONEY is usually avoided for complex financial calculations because its fixed scale and calculation behavior can cause rounding issues, especially during multiplication and division.

For most financial systems, prefer:

DECIMAL(19,4)

Example

CREATE TABLE Product
(
    Price       DECIMAL(18,2), -- Exact price
    Weight      FLOAT,         -- Approximate measurement
    LegacyPrice MONEY          -- Fixed four-decimal monetary value
);

Key interview points

• DECIMAL stores exact values with configurable precision and scale.

• FLOAT stores approximate values and can introduce rounding differences.

• MONEY is exact but has a fixed scale of four decimal places.

• Use DECIMAL for money, tax, banking and accounting calculations.

• Use FLOAT for scientific or measurement data where approximation is acceptable.

• Do not compare FLOAT values using exact equality when precision matters.

• MONEY does not store a currency symbol or currency type; store the currency code separately.

• In most modern financial applications, DECIMAL(19,4) is safer and more flexible than MONEY.