Precision & Rounding Standards
Financial accuracy is not just about getting the right answer; it's about retaining precision through every step of an equation. Here is how we prevent the "Penny-Gap" error.
The Floating Point Problem
Most standard calculators use "floating-point" math (IEEE 754), which attempts to trade accuracy for speed. This often results in microscopic errors that compound over time.
Standard JavaScript Error
0.1 + 0.2 = 0.30000000000000004In a billion-dollar payroll run, this error creates significant liability.
Commonrule uses a custom Arbitrary Precision Engine that treats numbers as decimal strings rather than binary approximations, ensuring that 0.1 + 0.2 equals exactly 0.3.
Our Rounding Standards
- 1
Half-Up (Standard)
Used for general currency display. 0.5 rounds to the nearest neighbor away from zero.
- 2
Banker's Rounding (Half-Even)
Used for bulk payroll aggregations to minimize statistical bias. 0.5 rounds to the nearest even number.
- 3
Floor / Ceiling
Strictly used when legal statutes dictate "at least" or "no more than" thresholds (e.g., minimum wage compliance).
Verification Matrix
| Operation | Standard Calculator | Commonrule Engine | Variance |
|---|---|---|---|
| 0.1 + 0.2 | 0.30000000000000004 | 0.3 | Low |
| Tax on $1,050.55 at 8.25% | $86.670375 (Display: $86.67) | $86.670375 (Stored) | Critical (Banking) |